Space Grant Consortium - University of Wisconsin - Green Bay

Space Grant Consortium 

wisconsin 

UNIVERSITY OF WISCONSIN-GREEN BAY 

Proceedings of the 19th Annual 

Wisconsin Space Conference 

The Cassini Encounter 

with the Gem of the 

Solar System 

August 13-14, 2009 

Milwaukee School of Engineering 

Milwaukee, Wisconsin

For information about the programs of the Wisconsin Space Grant Consortium, 

contact the Program Office or any of the following individuals: 

Wisconsin Space Grant Consortium 

University of Wisconsin-Green Bay 

2420 Nicolet Drive 

Green Bay, WI 54311-7001 

Tel: (920)465-2108; Fax: (920)465-2376 

www.uwgb.edu/wsgc 

Director 

R. Aileen Yingst 


(920)465-2327; yingsta@uwgb.edu 

Program Manager/Associate Director for Outreach 

Sharon D. Brandt 


(920)465-2941; brandts@uwgb.edu 

Chair, WSGC Advisory Council and Institutional Representative 

Karen Valley 

Wisconsin Department of Transportation 

(608)266-8166; karen.valley@dot.state.wi.us 

WSGC Associate Director for Scholarships/Fellowships 

Steven Dutch 


(920)465-2246; dutchs@uwgb.edu 

WSGC Associate Director for Student Satellite Programs 

William Farrow 


(414)277-2241; farroww@msoe.edu 

WSGC Associate Director for Research Infrastructure 

Gubbi R. Sudhakaran 

University of Wisconsin-La Crosse 

(608)785-8431; sudhakar.gubb@uwlax.edu 

WSGC Associate Director for Higher Education 

John Borg 

Marquette University 

(414)288-7519; john.borg@marquette.edu 

WSGC Associate Director for Special Initiatives 

Thomas Bray 


(414)277-7416; brayt@msoe.edu 

WSGC Associate Director for Industry Program 

Eric Rice 

Orbital Technologies Corporation 

(608)827-5000; ricee@orbitec.com

THE CASSINI ENCOUNTER WITH 

THE GEM OF THE SOLAR SYSTEM 

19 th Annual Wisconsin Space Conference 

August 13-14, 2009 

Host: Milwaukee School of Engineering 

Milwaukee, Wisconsin 

Edited by: R. Aileen Yingst, Director, Wisconsin Space Grant Consortium 

Sharon D. Brandt, Program Manager, Wisconsin Space Grant Consortium 

Martin Rudd, Assistant Professor of Chemistry, UW-Fox Valley 

Nadejda Kaltcheva, Department of Physics & Astronomy, UW-Oshkosh 

Cover by: Jonathan Carlson, Student Assistant, Wisconsin Space Grant Consortium 

Layout by: Sue Weiler, Office Coordinator, Wisconsin Space Grant Consortium 

Published by: Wisconsin Space Grant Consortium 

University of Wisconsin-Green Bay

�� 

�� 

�� 

�� 

� 

��5�� 

� 

�� 

� 

��

Th he proceedinngs 

of our 199th 

Annual WWisconsin 

SSpace 

Confeerence 

comess 

in a 

mo omentous yeear. 

In 20099 

we celebrated 

the moost 

amazing, , most increedible 

thi ing that humman 

beings hhave 

ever doone 

— 40 yeears 

ago hummans 

for thee 

first 

tim me broke thee 

gravitationnal 

pull of thheir 

own plannet 

and livedd 

on anotherr 

one. 

Th his year hhas 

seen wwonderful, 

meaningful, , and sommetimes 

excciting 

celebbrations 

of th his event! UUnfortunatelyy, 

at the samme 

time, we aalso 

were suubjected 

thiss 

year 

to rerruns 

and new w productioons 

of slicklyy-produced 

television sppecials 

"sugggesting" 

thaat 

the 

entiree 

Moon land ding was a hooax. 

We live in n a time whhere 

technoloogy 

has madde 

it easier thhan 

ever to be fooled — and 

paraddoxically, 

wh here less andd 

less of a prremium 

seemms 

to be put on the criticcal 

thinking skills 

that wwould 

best protect p us fr from being ffooled. 

As eeducators 

(thhose 

of us wwho 

are) wee 

talk 

aboutt 

the importa ance of Scieence, 

Technoology, 

Enginneering 

and Mathematiccs 

(STEM) ffields. 

As wwe 

all know, however, leearning 

youur 

multiplicaation 

tables sso 

you can ppass 

a test iis 

not 

why STEM is so o important. . No, it's beecause 

to truuly 

be able to understand 

the ruless 

that 

underrlie 

the work kings of the universe — that is, phyysics, 

chemistry, 

mathemmatics, 

and sso 

on 

— yoou 

must have e critical thinnking 

skills. You must bbe 

able to appproach 

a prooblem 

from mmany 

anglees, 

embrace e multiple, sometimes conflicting hypothesess, 

apply creative 

soluttions, 

perseevere 

throug gh dead endss 

and failed experimentts 

and your own fatiguee 

and frustraation, 

and mmost 

importa antly, to let tthe 

universe teach you, rrather 

than trrying 

to molld 

the univerrse 

to 

what you want it t to be. Youu 

must not oonly 

be objecctive, 

but bee 

willing to be convinceed 

by 

superrior 

evidence e, especiallyy 

when that eevidence 

poinnts 

in a direcction 

you doon't 

want to ggo. 

This is why 

STEM fiields 

are releevant 

to all ccitizens, 

andd 

why spacee 

fields especcially 

are reelevant. 

Spac ce and aerosspace 

represeent 

the very edge of the envelope, annd 

thus the ffields 

where 

these critic cal thinking skills are mmost 

essentiall. 

This is whhy 

the work published heere 

is 

relevant 

to Wisco onsin and thhe 

country — because byy 

the very nature 

of whaat 

we are strriving 

for inn 

aerospace pursuits, p thee 

very best iss 

demanded of us. We cannot 

be lazzy 

in our thinnking 

or ouur 

work; we cannot c settlee, 

we cannot "make do." We must exxcel. 

Acknnowledgeme 

ents 

I have bee en privilegedd 

to review the work puublished 

heree, 

where I seee 

that excellence 

manifested. 

I'm grateful g to eeach 

person who made the 19th AAnnual 

Confe ference and these 

Proceeedings 

poss sible. The WWisconsin 

Sppace 

Grant CConsortium 

especially eextends 

thannks 

to 

our hhost, 

the Mil lwaukee Schhool 

of Enginneering, 

andd 

WSGC Exxecutive 

Commmittee 

memmbers 

Tom Bray, Will liam Farroww, 

and especcially 

Loretta 

Krenitskyy 

for all thheir 

assistancce 

in 

runniing 

a beautif ful conferennce. 

We alsoo 

thank Dr. Ellis Minerr, 

who serveed 

as the Science 

Manaager 

for the NASA N Cassini-Huygenss 

Saturn Orbbiter 

and Titaan 

Probe for 12 years (ammong 

his mmany 

accolad des) for giviing 

keynote presentationns: 

The Casssini 

Encountter 

with the Gem 

of thee 

Solar System 

I & II. Dr. Minerr 

spoke to mmembers 

off 

the 1997 WWisconsin 

SSpace 

Confference 

at MS SOE on the Cassini Mission 

before llaunch 

and tthis 

brought us full-circle. 

Our ssession 

mod derators are all volunteeers 

and we aare 

grateful for their geenerous 

efforrts 

in 

ensurring 

our ses ssions run ssmoothly. 

Finally, 

as aalways, 

we wwould 

like to thank alll 

the 

scienntists, 

engine eers, studentss, 

educators and others, wwho 

contribuuted 

papers to this volumme. 

Forwward! 

Pr reface 

R. Aiileen 

Yingst, 

Ph.D. 

Direcctor 

iii

Student Programs 


• Undergraduate Scholarship 

• Undergraduate Research 

• Graduate Fellowship 

• Dr. Laurel Salton Clark Memorial Graduate Fellowship 

• University Sounding Rocket Team Competition 

• Student High-Altitude Balloon Launch 

• Student High-Altitude Balloon Payload 

• Student High-Altitude Balloon Instrument Development 

• Industry Member Internships 

• NASA ESMD Internships 

• NASA Academy Leadership Internships 

• NASA Centers/JPL Internships 

• NASA Reduced/Gravity Team Launches 

• Relevant Student Travel 

(see detailed descriptions on next page) 

Research 

The Research Infrastructure Program provides 

Research Seed Grant Awards to faculty and staff from 

WSGC Member and Affi liate Member colleges and 

universities to support individuals interested in starting 

or enhancing space- or aerospace-related research 

program(s). 

Higher Education 

The Higher Education Incentives Program is a seedgrant 

program inviting proposals for innovative, 

value-added, higher education teaching/training 

projects related to space science, space engineering, 

and other space- or aerospace-related disciplines. The 

Student Satellite Program including Balloon and Rocket 

programs is also administered under this program. 

Industry Program 

The WSGC Industry Program is designed to meet the 

needs of Wisconsin Industry member institutions in 

multiple ways including: 

1) the Industry Member Internships (listed under 

students above), 

2) the Industry/Academic Research Seed Program 

designed to provide funding and open an avenue for 

member academia and industry researchers to work 

together on a space-related project, and 

3) the Industrial Education and Training Program 

designed to provide funding for industry staff members 

to keep up-to-date in NASA-relevant fi elds. 

Programs for 2009 

v 

Aerospace Outreach Program 

The Aerospace Outreach Program provides grant 

monies to promote outreach programs and projects that 

disseminate aerospace and space-related information to 

the general public, and support the development and 

implementation of aerospace and space-related curricula 

in Wisconsin classrooms. In addition, this program 

supports NASA-trained educators in teacher training 

programs. 

Special Initiatives 

The Special Initiatives Program is designed to provide 

planning grants and program supplement grants 

for ongoing or new programs which have space or 

aerospace content and are intended to encourage, 

attract, and retain under-represented groups, especially 

women, minorities and the developmentally challenged, 

in careers in space- or aerospace-related fields. 


The Wisconsin Space Conference is an annual conference 

featuring presentations of students, faculty, K-12 

educators and others who have received grants from 

WSGC over the past year. The Conference allows all to 

share their work with others interested in Space. It also 

includes keynote addresses, and the announcement of 

award recipients for the next year. 

Regional Consortia 

WSGC is a founding member of the Great Midwest 

Regional Space Grant Consortia. The Consortia consists 

of eight members, all Space Grants from Midwest and 

Great Lakes States. 

Communications 

WSGC web site www.uwgb.edu/wsgc provides 

information about WSGC, its members and programs, 

and links to NASA and other sites. 

Contact Us 


University of Wisconsin–Green Bay 

2420 Nicolet Drive, ES 301 

Green Bay, Wisconsin 54311-7001 

Phone: (920) 465-2108 

Fax: (920) 465-2376 

E-mail: wsgc@uwgb.edu 

Website: www.uwgb.edu/wsgc


Undergraduate Scholarship Program 

Supports outstanding undergraduate students pursuing 

aerospace, space science, or other space-related studies 

or research. 

Undergraduate Research Awards 

Supports qualifi ed students to create and implement 

a small research study of their own design during the 

summer or academic year that is directly related to 

their interests and career objectives in space science, 

aerospace, or space-related studies. 

Graduate Fellowships 

Support outstanding graduate students pursuing 

aerospace, space science, or other interdisciplinary 

space-related graduate research. 

Dr. Laurel Salton Clark Memorial 

Graduate Fellowship 

In honor of Dr. Clark, Columbia Space Shuttle 

astronaut and resident of Wisconsin, this award 

supports a graduate student pursuing studies in the 

fi elds of environmental or life sciences, whose research 

has an aerospace component. 

University Sounding Rocket Team 

Competition 

Provides an opportunity and funding for student teams 

to design and fl y a rocket that excels at a specifi c goal 

that is changed annually. 

High School Sounding Rocket Team 

Competition 

For high school students. This program is in its initial 

stages. It mimics the university competition. 

Student High-Altitude Balloon 

Instrument Development 

Students participate in this instrument development 

program through engineering or science teams. 

Working models created by the students will be fl own 

on high-altitude balloons. 

Student Programs for 2009 

vi 

Student High-Altitude Balloon Payload/ 

Launch Program 

The Elijah Project is a high-altitude balloon program in 

which science and engineering students work in integrated 

science and engineering teams, to design, construct, 

launch, recover and analyze data from a high-altitude 

balloon payload. These balloons travel up to 100,000 ft., 

considered “the edge of space.” Selected students will 

join either a launch team or a payload design team. 

Industry Member Internships 

Supports student internships in space science or 

engineering for the summer or academic year at WSGC 

Industry members co-sponsored by WSGC and Industry 

partners. 

NASA ESMD Internships 

Supports student internships at NASA centers or WSGC 

industry members that tie into NASA’s Exploration 

Systems Mission Directorates. 

NASA Academy Leadership Internships 

This summer internship program at NASA Centers 

promotes leadership internships for college juniors, seniors 

and fi rst-year graduate students and is co-sponsored by 

participating state Space Grant Consortia. 

NASA Centers/JPL Internships 

Supports WSGC students for research internships at 

NASA Centers or JPL. 

NASA Reduced Gravity Program 

Operated by the NASA Johnson Space Center, this 

program provides the unique “weightless” environment 

of space fl ight for test and training purposes. WSGC 

student teams submit reduced gravity experiments to 

NASA and, if selected, get to perform their experiments 

during a weightless environment fl ight with the support 

of WSGC. 

Relevant Student Travel 

Supports student travel to present their WSGC-funded 

research.

Preface 

19th Annual Conference 

TABLE OF CONTENTS 

Part 1: Student Satellite Program: High Altitude Balloon 

Balloon Launch Team: 

Eric Deering, Milwaukee School of Engineering 

Brian Nguyen, Milwaukee School of Engineering 

Nathan Sward, Milwaukee School of Engineering 

Balloon Payload Team: 

Antonio Castillo, University of Wisconsin-Madison 

Brittany Hauser, Milwaukee School of Engineering 

Zachary Parsons, Milwaukee School of Engineering 

Chris Reichard, Milwaukee School of Engineering 

Victorialynn Salas, Marquette University 

Max Witte, Milwaukee School of Engineering 

Part 2: Student Satellite Program: Rocket Design Competition 

1 st Place – Non Engineering 

Schmitt Triggers 

Brad Hartl, University of Wisconsin-La Crosse 

Jacob Wardon, University of Wisconsin-La Crosse 

Steven Welter, Western Technical College 

Mark Witte, University of Wisconsin-La Crosse 

1 st Place – Engineering 

Team Narwhal, Milwaukee School of Engineering 

Neal Bitter, Team Leader 

Adam Harden 

Chelsey Jelinski 

2 nd Place - Engineering 

Team Drew and Crew - Milwaukee School of Engineering 

Drew Falkenburg, Team Leader 

Wes Larrabee 

Caleb Varner 

3rd Place - Engineering 

Rocky Mountain Miners - Milwaukee School of Engineering 

Brian Mortensen, Team Leader 

Ryan May 

Jon Neujahr 

vii

Part 3: NASA Reduced Gravity Programs 

Part 4: Engineering 

Repose Angles of Lunar Mare Simulants in Microgravity, Isa Fritz, Erin Martin, 

Samantha Kreppel, Brad Frye, and Caitlin Pennington, Undergraduate Students, 

Carthage College 

Dynamics Characterization of the Electron Beam Freeform Fabrication System 

Matthew Kallerud, Undergraduate Student, Milwaukee School of Engineering 

Analysis of Circulation Properties in Wake Vortices, Vanessa Peterson, 

Undergraduate Student, University of Wisconsin-Madison 

Validation of Novel Rigid Body Frictional Contact Algorithms using Tracked 

Vehicle, Simulation: A Stepping Stone for Billion Body Dynamics, Justin 

Madsen, Graduate Student, University of Wisconsin-Madison 

Modeling and Optimization of a Two Stage Mixed Gas Joule-Thomson 

Cryocooler, Harrison Skye, Graduate Student, University of Wisconsin-Madison 

ECLSS Vapor Compression Distillation System and Brine Processing, Cheryl 

Perich, Undergraduate Student, Marquette University 

Phaeton Mast Dynamics Mechanical Systems, Adam Harden, Undergraduate 

Student, Milwaukee School of Engineering 

Risk Mitigation Study for ISIM Systems Engineering, Code 443, Aaron Olson, 


Part 5: Atmospheric Science 

Part 6: Astronomy 

Is Lake Superior a Significant Source of Atmospheric Carbon Dioxide?, Val 

Bennington, Graduate Student, University of Wisconsin-Madison 

Realizing a Better Hydrostatic Response in NWP with MODIS Products, Jordan 

Gerth, Undergraduate Student, University of Wisconsin-Madison 

A Comparative Study of Type IIb Supernovae, Bradley Rentz, Cyrus Vandrevala 

and Michael Heim, Undergraduate Students, Marquette University 

Gravitational Heating and Evolution of Galaxy Groups, Melania Riabokin, 


viii

Part 7: Physics 

Understanding the Evolution of Supernova Progenitors, Cyrus Vandrevala and 

Bradley Rentz, Undergraduate Students, Marquette University 

Computational Fluid Dynamical Model of a Cyclone Separator in Microgravity, 

Brad Frye, Undergraduate Student, Department of Physics, Carthage College 

Simulation of Fast Magnetic Reconnection using a Two-Fluid Model of 

Collisionless Pair Plasma without Anomalous Resistivity, E. Alec Johnson, 

Graduate Student, Department of Mathematics, University of Wisconsin-Madison 

Part 8: Biological Science 

Part 9: Chemistry 

Reflection and Refraction of Vortex Rings, Kerry Kuehn, Associate Professor, 

Department of Physical Sciences, Wisconsin Lutheran College 

Understanding Risk Determinants of Chagas Disease in Peri-urban Peru, 

Megan Christenson, Graduate Student, University of Wisconsin-Madison 

Astronaut Advanced Life Support: Engineering Extra Terrestrial Extremophile 

Plants, Dan Hawk, Undergraduate Student, College of Menominee Nation 

Toxic Offgassing Analysis at Marshall Space Flight Center, Kenion Blakeman, 

Undergraduate Student, Department of Chemistry, Carthage College 

Preparation and Characterization of Platinized Electrodes for Use in Oxygen 

Extraction from Lunar Regolith, Nathan Wong, Undergraduate Student, 

University of Wisconsin-Madison 

New Initiatives in the Project on Fossilization via Silicification, Vera Kolb, 

Professor, Department of Chemistry, University of Wisconsin-Parkside and 

Patrick Liesch, Department of Entomology, University of Wisconsin-Madison 

Part 10: Education and Public Outreach 

Combining Writing Across the Curriculum Strategies in Community-Based 

Programs to Teach Core Scientific Concepts, James Kramer, Executive Director, 

Simpson Street Free Press 

Educators’ Aerospace Workshop for the 21 st Century, Jake W. Blake, 8 th Grade 

Science Teacher, Brookwood Middle School 

Teaching the Teachers, Coggin Heeringa, Director, Crossroads at Big Creek 

Environmental Learning Preserve 

ix

EAA Women Soar – You Soar, Lee Siudzinski, Manager, Education 

Relationships, Experimental Aircraft Association (EAA) 

EAA Space Week 2008, Lee Siudzinski, Manager, Education Relationships, 

Experimental Aircraft Association (EAA) and Chrissy Paape, Vice President, 

Space Explorers, Inc. 

Space Travel Simplified – Part 1, Bradley J. Staats, President, Spaceflight 

Fundamentals, LLC 

Expanding GIS Across the Curriculum, Jennifer Johanson, Associate Professor, 

Physical Science Department, Alverno College 

Astronautics Course, Harald Schenk, Associate Lecturer, Physics/Astronomy 

Department, University of Wisconsin-Sheboygan 

New Directions in Astrobiology Education, Vera Kolb, Professor, Department of 

Chemistry, University of Wisconsin-Parkside 

Providing High School Students with Earth Imaging Tools, Thomas Jeffery, 

Assistant Professor, Geography and Geology Department, University of 

Wisconsin-Whitewater 

Appendix A: 19 th Annual Conference 2009 Program 

x


Part One 

Student Satellite Program 

High Altitude Balloon

Elijah High Altitude Balloon Project Launch Team 

Eric Deering from Milwaukee School of Engineering; Brian Nguyen from Milwaukee School of 

Engineering; and Nathan Sward from Milwaukee School of Engineering. 

Please note: Due to the deadline for this report, it will primarily be about the first launch of 2009. The 

responsibilities of the Launch Team extend through the summer and the following academic year, and 

accommodate both the Payload Team and the Instrumentation Team. Multiple future launches are 

planned, therefore, which are not covered in this report. 

Members and Roles 

Because the Launch Team has only three members this year, each member is involved with all aspects of 

the project. However, Eric Deering served as the primary liaison in communications with WSGC and 

StratoStar. All members bring some experience to the table, since Brian Nguyen and Nathan Sward 

were on the Payload Team last year, and Eric Deering was on the HALO II project. 

Equipment 

Overview 

The 2009 Launch Team is again using the StratoSat complete flight package, by StratoStar Systems. This 

system was first used by the 2008 team and was found to be an excellent tool. The system contains 

everything necessary to plan, launch, and remotely track the payload. The following components are 

included: 

• Command pod with GPS and data 

transmitter 

• Three payload pods 

• Two 900 Mhz Modules and antennae 

• StratoStar software, which interacts 

with Microsoft MapPoint 

• Battery chargers for the pods 

• Parachute 

• Cables and tethering 

Figure 1: StratoSat System 

In addition to the StratoStar package, the launch team acquired high altitude weather balloons, helium, 

and the payload to be carried, which was supplied by the payload team. 

Balloons 

The Elijah launch team began the season using a 1200 gram balloon that was left over from last season. 

After witnessing the performance of the first balloon and after talking with the company who sells the 

balloons (Kaymont), the decision was made to purchase two more 1200 gram balloons for future flights. 

1

From some of the information Kaymont had on their website regarding balloon performance we were 

able to get a better idea of how our balloons should be performing. The 1200 gram balloons we 

purchased had an average ascent rate of 1050 feet per minute with an average bursting altitude of a 

little over 30 kilometers. This now gave us some ground statistics which would allow us to better predict 

landing locations for future launches. 

Command and Payload Pods 

The StratoSat system comes with one command pod and three payload pods. The command pod is 

responsible for flight data acquisition such as position, altitude, speed, and heading while the payload 

pods can be set up with other sensors to collect data such as temperature and pressure. The payload 

pods can communicate wirelessly with the command pod which sends all of the information gathered by 

both the command pod and the payload pod back to the tracking software. Once tracking is complete 

the data can be simply exported into an excel file and analyzed even without the recovery of the system. 

Equipment Testing 

Since the tracking of our balloons was to be done with StartoStar Systems StratoSat setup, we began by 

installing Microsoft MapPoint and the StratoStar software. The new StratoStar software conveniently 

interfaced with Microsoft MapPoint which allowed for easy geographical location of the balloon. Before 

our first launch we took all of the StratoSat equipment outside to verify it was working properly. After 

overcoming a few obstacles with the USB to serial converters, we were able to locate the command pod 

on our computers. After tracking one of our team members driving around Milwaukee we concluded 

that the equipment was working properly. 

Our team quickly noticed that the command pods transmitter needed to be directionally positioned to 

the antenna receiver. For example, when the transmitter was pointed away from the antenna, the GPS 

transmitter on the command pod could not communicate with the receiver. While this remains an issue 

once the balloon is on the ground, this problem is diminished when the balloon is high above the GPS 

receiver. 

High Altitude Launch Opportunity (HALO) II 

The Wisconsin Space Grant Balloon Launch Team began its annual duties in the 2009-2010 season by 

participating in the High Altitude Launch Opportunity (HALO II) this spring. The HALO II project consisted 

of the simultaneous launching of 15 high altitude balloons from 15 universities in 9 different states. The 

purpose of the HALO II launch was to set up a high altitude communications network along with 

collecting temperature, humidity, pressure, altitude, CO2, rate, and radiation data. By having each 

payload transmit a radio frequency, the different balloons could link together through the radio 

transmissions and share data. Taylor University in Indiana served as the mission control for the project 

and through the communication network created, could receive data in real time. This eliminated the 

need for a sometimes unsuccessful recovery of the balloon. 

2

Our experience with the HALO II launch 

went well and was a good starting point 

for the rest of the season. We were able 

to experience some of the basic 

procedures such as caring for and filling 

the balloon along with practicing with the 

software and the payload unit that was 

sent to us from StratoStar. While the 

procedures of the launch and actually 

launching the balloon went well, tracking 

and recovering the balloon was a different 

story. Up to about half way through the 

balloons flight, the tracking software was 

working well and was transmitting GPS 

points back to our software. However, 

shortly after the balloon reached Figure 2: HALO II Network Coverage 

approximately 100,000ft we stopped 

receiving GPS and data packages. After driving around for about four hours trying to pick up a signal on 

the payload, our team was forced to abort the recovery mission. 

A few weeks ago our team contacted Jason Krueger from StratoStar to try and figure out the reason for 

the GPS failure. His conclusion was that the payloads aerodynamic design caused it to twist and turn, 

eventually becoming tangled in the cord, causing the antenna to break off. We were not the only team 

to have trouble with tracking and data transmission so hopefully these problems will be corrected for 

the next team that decides to partake in the continuing HALO project. 

For more information on the HALO II launch you can go to the following website 

http://www.nearspacenetwork.com/group/halo2project 

Launch Planning 

Balloon Choice 

As stated above our team chose to use 1200 gram balloons for our launches. 1200 gram balloons are 

relatively small so they allow us to maximize our use of helium while still providing ample lift for any 

payload the team may choose to send up. 1200 gram balloons also have similar maximum height and 

ascent rate characteristics when compared with larger balloons. 

Location Choice 

An optimal location for a launch is dependent upon 1) the ideal location of landing, 2) weather 

conditions, and 3) risk factors in the area. The ideal location of a landing would be in a rural location, 

away from power lines and heavy traffic. Further, it would be someplace that is not heavily wooded, 

since this makes retrieval difficult. The location for which we aimed for our first launch was near 

Monroe, Wisconsin. This is a rural, sparsely populated area, with less trees and hills than Western 

Wisconsin. 

3

Prediction Software 

To be able to accurately predict where a high altitude balloon will land is next to impossible unless you 

are able to consider all of the variables that come into play. Fortunately, flight prediction software 

based online at (www.nearspaceventures.com) is able to consider all of these variables and put them to 

use. This software takes into account the weather predictions for the area in question and plots take 

off, landing, and balloon burst points on a map. 

In days leading up to the launch, flight prediction software was used to predict the course of the flight 

using weather forecast details from the closest weather station and details about ascent and descent 

rate which we obtained from the balloon manufacturer. This tool was first used to predict the flight 

path from Boscobel, WI. The software predicted that the payload would land near Dubuque, Iowa, 

which was unacceptable. Mt. Horeb, WI was finally decided on as the most convenient site that would 

result in a landing location around Monroe, WI. 

For the second launch the same prediction software from (www.nearspaceventures.com) was used to 

again try and find an optimal landing zone as discussed above. Below is a figure showing the results of 

the prediction for the second launch. 

Figure 3: Flight Prediction for Launch 2 

Launch Day 

Launch 

On Saturday July 18 th , the launch team and 4 members of the payload team headed to Mt. Horeb to 

execute the first launch of the 2009-2010 season. The first task for making our first launch successful 

was finding a safe place to launch the balloon. By using the satellite view on Google maps we were able 

to find a nearby football field. The launch site worked well due to the fact that it had no trees or power 

lines in the immediate area and the site was elevated relative to the surroundings. 

After finding the site we brought out all of our gear and laid out everything we would need on the grass. 

A tarp was laid on the ground so the balloon could be filled without causing any damage to it. Rubber 

gloves were worn by the launch team members when handling the balloon to avoid getting skin oil on 

the balloon which can hamper the balloons expansion at high altitudes. The balloon was then filled 

4

which had been a problem in the past due to a check valve in the connection fitting to the helium tank. 

After unsuccessfully looking for a new tank fitting without a check valve, the decision was made to 

simply drill out the check valve to provide better air flow. This decision proved to be a success as the fill 

time of the balloon went from about 2 hours to 15 minutes. While the fitting now doesn’t have a check 

valve it is still safe since the flow of helium can be regulated by the tanks 

valve. 

At approximately 11:00AM the balloon and the payloads underwent a final 

check along with turning the GPS system on and checking to see how it was 

performing. A few minutes later the balloon was released and instantly we 

knew something wasn’t right. The balloon slowly began ascending and then 

started descending. The members present rushed to grab the balloon and it 

was brought back to the launch pad. An oversight had been made while 

reading the scale which measured the amount of lift the balloon was 

providing. Since the wind was blowing the balloon sideways while the lift 

was being measured, the wind added on to the amount of lift significantly. 

After holding the balloon steady we were able to get the lift up to around 

6lbs, about 3/2 the weight of our payload. 

The balloon was released a second time, this time with much better results. 

The members present watch the balloon rise away and then proceeded to 

pick up all of the gear and 

begin tracking the balloon. 

Figure 5: GPS Generated Track of Balloon Flight 

Figure 4: Balloon and Payload 

Tracking 

Possibly the most fun part of being a member of the 

balloon launch team is tracking the balloons. 

Following the balloon on our laptops, our convoy of 

three vehicles drove around back roads trying to stay 

close to the balloon. We had set up two laptops in 

separate cars, both tracking the balloon. This reduced 

the risk of losing the balloon due to computer 

problems. We tracked the balloon from Mt. Horeb, 

WI to just north of Monroe, WI. The StratoSat 

tracking system preformed well and got us to the 

landing point. 

5

Local Search 

While we were right under the balloon when it came down we were unable to see it. This began a long 

process of trying to find the balloon in a wooded area. We knew the balloon was close but we were 

unable to obtain a final landing point from the tracking system. Due to the directional nature of the 

antenna and transmitter as discussed above, it is very have to obtain tracking information once the 

balloon is on the ground. Once the search area was narrowed down using our GPS systems, we began 

to look in a limited area for the balloon. After talking with the landowner, he volunteered to help us try 

and find the balloon. Following about four hours of searching the woods with trucks, an ATV, and on 

foot, we had to call it a day. We were not able to find the unit even though the system had worked 

perfectly. 

Ideas for Future Launches 

The largest problem with the current system is the directional nature of the antenna and transmitter. 

Once the balloon is on the ground the only way to find it is to get a final landing point and then go to 

that point and search the area. Sometimes the final point is obtained while the balloon is still in the air, 

allowing the balloon to drift an unpredictable distance before landing. A solution to this came in the 

form of a radio tracking system that the launch team had purchased previously. The system was in need 

of a new tracking collar and so one was purchased and the system was tested. With an ideal range of up 

to seven miles the new secondary tracking system seemed ideal for short range tracking. 

Another idea to try and locate the payload once the balloon was on the ground came in the form of a 

loud beeper. This beeper made a sound similar to that of a truck back up alarm. This would allow us to 

audibly locate the payload from well over 100 yard away. 

Lastly, since we had determined the parachute had not opened in our first launch due to our high 

decent rate data, we wanted to find a way to more consistently have the parachute open. We 

determined that making the cord from the top of the parachute to the balloon slightly longer would give 

the parachute more time to fill with air after the balloon had popped. 

Second Launch 

Armed with our new secondary tracking system and the loud beeper we were confident we would 

retrieve the payload after our second launch. After running the prediction software each day the 

previous week, Mt. Horeb was again determined as an ideal launch site. The prediction for this flight 

(Figure 3 ) told us the balloon would travel north east, landing in south central Wisconsin. After getting 

everything unpacked some weather began to roll in, threatening our ability to launch. With still an hour 

before the storm hit we decided to launch the balloon. This time the balloon was carrying the payload 

team’s experiments, a secondary radio tracking device, the StratoStar command pod, and a very loud 

beeper. 

After tracking the balloon for a little over an hour the balloon hovered over the Horicon marsh for 30 

minutes at an approximate altitude of 25 kilometers. While the balloon remained in the same position 

horizontally, it slowly climbed in altitude where it finally popped at around 30 kilometers. The balloon 

again began traveling northeast during its descent. Staying just ahead of the looming storm we tracked 

the balloon across highway 41 where it landed just east of Lomira. 

6

Figure 6: 3-D GPS Generated Track of Balloon Flight 

Using our secondary tracking device we were able to locate a general area in which we thought the 

balloon had landed. Heading in on foot with the radio tracking device we continued in the direction of 

the strongest signal until we heard a noise. It was the beeper on the payload which seemed to be just 

beyond a patch of woods. When we arrived at the payload we were both excited to have found the 

system, and disappointed to see that it was stuck 60 feet up in a tree. With no means of retrieving the 

payload from the tree we were forced to temporarily abandon our payload. 

Conclusion 

The launch team has identified some problems with the system of launching and retrieving the balloons 

but has also come up with solutions to many of those problems. The solution to finding the balloon 

after it was on the ground came in two separate systems. The radio collar tracking system and the loud 

beeper that was placed on the payload for easier short range recovery. The radio tracking system led us 

to the payload where we were then able to hear the alarm and locate the payload in a tree. Also, 

adding a few feet to the tether between the balloon and the parachute solved the problem of the chute 

not opening. 

These additional systems have greatly increased the percentage of balloon recovery. While our first 

launch recovery was unsuccessful even with the StratoSat system performing well, our team was able to 

come up with secondary systems to aid in the recovery of the balloons. With multiple launches in the 

team’s future, we plan to further refine the system of launching and recovering high altitude balloons to 

make it easier on teams in the future. 

7

2009 Elijah Summer Payload Team 





Victoria Salas, Marquette University 


Project Synopsis 

The Elijah Payload Project is an opportunity for science and engineering students to research, 

design, develop, and implement their own experiments. A high altitude weather balloon is 

launched into the stratosphere with the students’ designed payload attached. There are no limits 

other than weight as to what can be included in this payload; the students are encouraged to 

develop their own ideas over the course of the 10-week program. Through weekly meetings and 

brainstorming sessions, four separate experiments were designed and constructed. The first 

payload was lost and the second has not been launched as of the writing of this paper, but what 

follows is a detailed summary of the projects of the 2009 Elijah Payload Team. 

Seed Germination 

Abstract The purpose of the watermelon seed experiment is to determine if the near space 

environment at 100,000 feet has an effect on the seeds. This was tested by first growing the seeds 

and then comparing the growth of the experimental seeds to the growth of the control seeds. 

Background Everything on earth is exposed to small amounts of natural radiation (Health 

Physics Society, 2001). Exposing seeds to higher amounts of radiation can cause inhibited 

sprouting, slow seedling growth, reduced plant fertility, and chromosome alterations. 

Watermelon seeds were chosen because they have a fast germination of three to seven days and 

do not require a special environment for growth. 

Methodology Four bottles with approximately 15 seeds in each bottle were created. Two bottles 

with seeds were sent up with the balloon; one insulated and the other not insulated. The other 

two bottles were used as controls; one staying at MSOE and the other traveling with the 

experimental seeds until the launch. Since the payload was not recovered, the seeds were tested 

in the lab with a bell jar and dry ice. 

The three packs of seeds purchased are Sugar Baby Watermelon Seeds, packed in 4g packages 

by the Livingston Seed Co. All three packages of seeds were labeled 2009 Run B, sell by 10/09. 

The seeds were planted in a 72 cell seed insert with Stein seed starter mix. The seed tray was 

placed in an 11” x 22” base tray and covered with a two inch tall humidity dome. The seed tray 

was then placed on a Hydrofarm heated germination station that uses 120V and 17W. Two 13W 

Ott lights, model # OTL13BPB, were placed above the seeds to provide light and approximately 

three ml of distilled water has been given to each seed daily. 

9

Results The seeds tested in the lab were planted on 08/05/09 and have not germinated yet. A 

photo of the setup for planting and the reference plants can be seen in Figure 1. Control seeds 

were planted on 07/27/09 for reference and are growing well. Four experimental groups were 

tested in the lab with fifteen seeds in each new group; control seeds, seeds exposed to the dry ice 

for an hour, seeds exposed to the vacuum for an hour, and seeds exposed to both the dry ice and 

the vacuum for an hour 

each. 

Figure 1: Watermelon seed germination container 

Conclusion Due to the fact that we could not retrieve the payload a conclusive decision about 

this experiments success or failure could not be made at this time. 

References "Answer to Question #1280 Submitted to "Ask the Experts"." Health Physics Society 

Radiation Effects — Biological Effects of Radiation 18 OCT 2001 Web.10 June 2009. 

. 

Coffee Taste 

Abstract The coffee experiment was based on how the atmosphere at 100,000 feet would affect 

the taste of the coffee in a near space environment, with interest developing from how 

microgravity affects the taste of food for astronauts. Four samples of coffee beans were prepared 

and the difference in taste were compared with the sample being subjected to extreme cold 

having the best taste and the control sample having the worst taste. 

Background Motivation for this experiment developed with the team members interest in how 

different types of foods and beverages taste differently in space. This was later narrowed down to 

how the flavor of coffee will be affected once placed into near space conditions. The genuine 

10

taste of coffee and the fresh aroma are commonplace here on earth but becomes a luxury up in 

space for those serving on a space mission or living aboard the international space station. Food 

as well as coffee is known to taste different in space according to the astronauts conducting their 

missions. Most of this difference can be attributed to physical effects faced by the astronaut’s 

body once in microgravity conditions such as nasal congestion due to upper-body swelling which 

reduces the sense of smell causing a reduction in taste. Since an experiment could not be 

performed in space, an examination of how the near space conditions could play a role in altering 

the flavor of the coffee became the central part of this experiment. Coffee freshness is affected 

by three major elements: oxygen, moisture, and temperature. Oxygen is the primary enemy that 

reduces the flavor of the coffee. This resulting loss of flavor is due to oxidation, which attacks 

the aromatic volatile compounds within the coffee beans. Moisture will degrade the flavor when 

it is absorbed into the coffee bean by depleting the flavorful oils. Moisture and oxygen also link 

in with temperature, because the cooler temperatures cause water vapor to condensate, coffee 

beans will experience excess moisture formation furthering the reduction of the flavor and during 

warmer temperatures there is a higher thermal energy that spurs on the staling effect which 

increases the solubility of any oxygen that is present. Thus coffee beans should only be subjected 

to these temperature extremes at a minimum, usually once is alright. 

Methodology A basic coffee was purchased from Starbucks, whole bean Komodo Dragon 

Blend, and used for this experiment so a better understanding of the effect on the flavor could be 

established. The experiment consisted of four bottles of coffee beans, two of which were sent up 

with the payload and the other two remained on the ground. One container was insulated from 

the cosmic radiation and part of the extreme weather and the other was left out so it can be 

affected by the extreme conditions. Of the two samples left on the ground one was left in the 

science laboratory and the other was carried along to the launch. The payload was launched on 

the high altitude balloon, however it 

was unable to be recovered. 

Therefore ground analogs of the 

effects due to the pressure and 

temperature drop were performed in 

the laboratory in order to gain a 

prediction of the results being 

explored in the near space 

environment. One sample of coffee 

was left as a control, one was 

subjected to a temperature of 

negative thirty degrees, one was 

subjected to a pressure drop, and the 

last sample was subjected to the 

temperature drop of negative thirty 

degrees as well as the pressure drop. 

To simulate the drop in pressure a 

bell jar was used and to cool the 

samples to negative thirty degrees, dry ice was Figure 2: Example of Coffee Survey 

placed in a cooler. Each of the samples being 

placed in a different temperature and pressure were subjected to these differences for exactly one 

11

hour as we assumed the flight in the atmosphere would be approximately this amount of time. 

After the appropriate tests the coffee beans were grounded in four different grinders to prevent 

contamination of the samples. The samples were brewed just using a separate filter for each, with 

hot water being poured through it. A double-blind test was performed by surveying the public, to 

evaluate the flavor difference of the samples and not telling the test facilitator which coffee was 

which. A survey asking a few questions and a rating of the coffee is shown in Figure 2. 

Results The coffee beans were tested in a ground analog experiment due to the failure to find the 

original payload. The coffee beans were prepared as described in the methodology and served to 

eight random people with in MSOE’s Fluid Power Institute. Info gathered from the survey 

showed all but one person likes coffee and the majority preferred a dark roast coffee which 

coincides with the type of coffee used for this experiment. The rubric for the double blind 

experiment was as follows: coffee A was the cold and pressure, coffee B was only subjected to 

the cold, coffee C was only subjected to pressure, and coffee D was the control. The results of 

the test based on the differences in taste obtained from the survey are shown in Table 1. 

1 2 3 4 

Coffee A 1 1 4 2 

Coffee B 5 3 0 0 

Coffee C 2 3 3 0 

Coffee D 2 1 0 5 

Table 1: Results organized by frequency of the rank for each of the four samples 

Table 1 shows that the test subjects preferred the taste of coffee B which was subjected to the 

cold the most and preferred coffee D, the control, the least out of the four samples. 

Conclusion We are still optimistic of finding the original payload in order to perform a more 

exact analysis to gain better results. However based on the results from the ground analog the 

beans subjected to the cold turned out the best which can be based on the cold keeping the beans 

the freshest, however with repeated cycles of the extreme cold and room temperature the results 

are expected to be different with more of a flavor loss being experienced due to excess moisture 

buildup. 

References Romanoff, Jim. "When it comes to living in space it's a matter of taste." 

Scientific American 10 Mar 2009 Web.17 June 2009. 

. 

"Storage & Packaging." The QARR coffee. 2006. 17 June 2009 

. 

Watch Functionality 

Abstract The watch functionality experiment is designed to study the effectiveness of different 

types of watches when exposed to low temperatures and pressure. The expected results include 

the glass/plastic covers either cracking of breaking off due to the lower pressure, and a slower 

watch due to the lower temperature. 

12

Background The idea behind the watches experiment started with Atomic Watches. Atomic 

watches are the most accurate watch as of current because every day they receive a signal that 

adjusts the hands to read the exact time. Atomic watches are essentially quartz watches, only 

they automatically adjust themselves every day. 

The team wanted to experiment 

with the radio signal that the watches 

receive daily. Would the altitude 

have any effect on the watches ability 

to receive the signal properly? This 

was to be our initial experiment. It 

then became apparent that the 

Atomic watches only receive their 

signal twice a day at most, one of 

which is around 2am. In order to 

fully study how altitude affects the 

radio signal we would need a 

constant stream. However the only 

receiver we could obtain was one 

from Europe which uses a different 

frequency for their Atomic watches 

than in the US. 

This led the team to the current 

experiment of using basic quartz 

watches of different values and 

comparing how they are affected by 

reduced pressure, temperature, and 

the miniscule amounts of radiation 

they are to be exposed to. 

Methodology The setup for this 

experiment is fairly simple. The watches are mounted on a single foam core board using Velcro 

straps. (As seen in Figure 3) A video camera is setup a small distance away and will be 

recording the watches the entire time during the flight. To ensure that the camcorder records the 

entire flight an external battery pack will be used along with a large enough SIM card to hold the 

video file. A basic LED from a small flashlight is mounted with a disperser to provide light 

inside the payload. A control setup was also constructed to record at the exact same time as the 

payload. 

Results 

Figure 3: Set up of the watches and camera inside the payload 

13

Experimental Results There are two specific results that are to be expected from this 

experiment. The first is how the glass/plastic covers of the watches will either crack or pop out 

from their holders due to the lower pressure. Standard watches are designed to withstand the 

reduced pressure inside of airplanes, which is usually around 2000 feet. The high altitude 

balloon is expected to reach at most 100,000 feet. The difference in pressure at the altitudes is 

immense and therefore a result is expected. The second is that the clocks will slow down or 

possibly stop due to the decreased temperature. Most basic materials contract a small amount 

when exposed to cold temperatures. Due to this the gears in the watch will become smaller and 

more rigged thus reducing the speed in which they take time. If the gears shrink small enough 

they will no longer mesh with each other resulting in the hands not moving at all. 

Testing Results The watch experiment was pressure testing using a vacuum bell jar. It is 

unknown how low the pressure dropped to however nothing happened to the watches and the 

camera operated normally. 

Conclusion Conclusions for this experiment are still pending; the expected launch is scheduled 

for 08/09/09 

Sound Transmission 

Abstract This experiment deals with sound as it travels through air. The qualities of the air will 

affect the manner with which the sound travels. Using basic electronics equipment and 

MATLAB for analysis, we hope to see these factors effect sound, namely its speed as it moves 

through air. We have identified two factors that will be relevant: colder temperature, which will 

slow the speed of sound, and reduced pressure, which will increase it. We hope to 

experimentally verify these theories. 

Background This particular experiment was based on the McNeese LA-ACES Group Sound 

Experiment (McNeese) done in 2007. Our goal was to measure the speed of sound relative to its 

altitude. We planned on attempting to replicate their findings, in order to compare their results to 

ours. We were very curious about what would happen to sound when subjected to the extreme 

cold and high altitude pressures provided by the stratosphere. 

Methodology A simple 4000 hertz buzzer was used to broadcast a continuous tone throughout 

the experiment. This tone was picked up by two microphones a fixed distance apart. Two noise 

cancelling mics were used to cut down on the wind noise sure to be encountered during the 

flight. The sound data was recorded in stereo onto an SD card by a digital voice recorder with a 

high sampling rate. The basic setup shown in Figure 4: 

Buzzer 

Fixed 

Distance 

Mic 1 

14 

Mic 2 

DVR

Figure 4: Sound Apparatus 

An MP3 reading program was downloaded for MATLAB, 

storing the data from each mic separately. The sound 

appears as a sine wave. In a program we wrote, the data was 

grouped, peaks were found, and the time separation between 

these peaks determined. Knowing the amount of time it 

took the sound to cover the fixed distance, speed can be 

easily calculated. An apparatus was designed to hold the 

buzzer and mics stationary. The mics had to be securely 

fixed, and positioned in a way that would not block or 

interfere with the sound in any way. Figure 5 shows a CAD 

model of the designed and constructed setup for the 

experiment, with the mics and buzzer included: 

The wires from the mics to the DVR and from the buzzer to 

the 9V battery will run inside our insulated foam box, so the 

conditions of the stratosphere don’t interfere with 

functionality or battery life. The holder was constructed out 

of wood in MSOE’s machine shop. 

Mic 2 

Mic 1 

Buzzer 

Results With the setup constructed as shown and the 

program written, several tests were run under different 

Figure 5: Experimental Apparatus 

conditions. One such test was made at 23.3°C in dry air. The 

speed returned was 374.85 meters per second. This is fairly close to the accepted value for the 

speed of sound under standard conditions, 343 meters per second. The difference can be 

accounted for by the difficulty of getting the microphones an exact distance apart; even with 

calipers, there is bound to be some difference. The plot generated by the test, for one 

millisecond of recording, is shown in Figure 6. The mic nearest the buzzer is shown with greater 

amplitude, and the farther mic is shown with lesser amplitude: 

15

A series of tests were done in a cooler with dry ice to examine the effects of temperature on 

Figure 6: Normal Conditions 

sound. The program output a speed of 340.77 

meters per second at a temperature of -28.1°C. 

This is significantly slower than the control test, and in the plot, attenuation can be seen, shown 

in 

0.4 

0.3 

0.2 

0.1 

0 

-0.1 

-0.2 

-0.3 

Figure 7: 

-0.4 

5 5.0001 5.0002 5.0003 5.0004 5.0005 5.0006 5.0007 5.0008 5.0009 5.001 

Figure 7: Cold Conditions 

In the control, the peaks of the far mic (lower peaks) are slightly behind those of the near mic 

(higher peaks). In the cold test, the far mic peaks are shifted to the right. The slower speed 

16

agrees with the accepted model for sound transmission. Tests were also done in a bell jar to 

analyze the effects of air pressure on 

sound. Figure 8 shows the tests being 

done: 

The program output a speed of 454.38 

meters per second. Again, note the 

attenuation from the control case, shown 

below in Figure 9: 

0.5 

0.4 

0.3 

0.2 

0.1 

0 

-0.1 

-0.2 

-0.3 

-0.4 

-0.5 

5 5.0001 5.0002 5.0003 5.0004 5.0005 5.0006 5.0007 5.0008 5.0009 5.001 

Figure 9: Low Pressure Conditions 

This time, the low peaks are shifted to the left. The calculated increase in speed also agrees with 

the accepted system for analyzing sound transmission. 

A second experimental apparatus of identical dimensions that returns the same numbers as the 

flight apparatus was constructed. This setup will run under normal, dry air conditions during the 

flight. This will serve as the control and basis for comparison of the flight data. 

Table 2 shows additional results from the cold tests: 

17 

Figure 8: Bell Jar Testing

380 

360 

340 

320 

23.3 

Temp. (Celsius) Sound Speed (Meters Per Second) 

Control 23.3 374.85 

Trial 1 -18.1 358.22 

Trial 2 -19.6 361.71 

Trial 3 -22.9 360.23 

Trial 4 -27.2 351.63 

Trial 5 -28.1 340.77 

Table 2: Cold Test Results 

Figure 10 shows these results plotted. Note the obvious downward trend: 

-18.1 

-19.6 

-22.9 

Figure 10: Temperature and Speed 

-27.2 

Table 3 

shows additional results of the pressure tests and demonstrates the consistency: 

Trial 1 

Sound Speed (Meters Per Second) 

454.38 

Trial 2 468.56 

Trial 3 416.5 

Table 3: Pressure Test Results 

Conclusion As of now, the launch has not yet occurred. With the ground tests as indicators, we 

would expect to see significant differences in the way sound travels relative to altitude, 

according to the effects of cold and pressure. It would be interesting to see how the combined 

conditions of both colder air and lower pressure would interact in manipulating sound. With a 

working system in place to compare how quickly sound travels in specific conditions, we hope to 

have the chance to analyze flight data eventually. 

References Townsend, Tate, et al. "McNeese LA-ACES Group Sound Experiment 2007." 

Columbia Scientific Balloon Facility. 10 June 2009 

18 

-28.1


Part Two 

Student Satellite Program 

Rocket Design Competition

First Place, Non-Engineering: Schmitt Triggers 

2009 WSGC Intercollegiate Rocket Competition 

Brad Hartl, Jacob Wardon, Steve Welter, Mark Witte, Michael LeDocq 

University of Wisconsin-La Crosse & Western Technical College 

Synopsis 

Plan of action. Prior to making any major design plans or ideas, a precise plan of action 

was created for attacking this problem. The first step was to find out what the ideal shape of the 

dart and booster stage should be. There were a few specific restrictions that had to be accounted 

for, such as motor and altimeter size, but beyond that, the options were wide open. The first thing 

considered was the drag equation (equation 1), which is the force of air resistance on the rocket 

and ultimately the most important concern. 

(1) 

The first two variables of the equation are ρ, which represents the air density, and u, which 

represents the velocity at which the object is traveling at. Both of these variables can not be 

controlled directly by the shape of the rocket. The last two variables, CD (coefficient of drag) and 

A (cross-sectional surface area), can be controlled by the shape of the rocket. In order to decrease 

the air resistance on the rocket, both of these terms should be minimized. The cross-sectional 

surface area is basically the surface area you would see, as if you were looking at the rocket from 

above it. Since there was a four inch required diameter on the lower stage, there really wasn’t 

much control over this. However, the width of the fins would contribute slightly and thus their 

thickness and length was kept to a minimum. 

The drag coefficient is a measure of how smoothly and easily air can pass around the object. The 

more streamlined the object, the less turbulent flow there will be, and thus the less resistance. 

The profile shape with the lowest drag coefficient is a tear drop shape. [4] Thus, an ideal shape 

was found to model the rocket after. While not directly related to the drag equation, the next 

most important variable to calculate was the ideal masses of the rocket stages. Newton’s second 

law states that 

Net Forces = Mass x Acceleration (2) 

When the motor is firing, it is the primary force on the rocket, thus it would be ideal to keep 

mass minimized in order to allow the acceleration to be maximized. Integrating acceleration 

twice will give distance. Hence, the less mass during takeoff, the higher the rocket will go. 

However, after the motor has shut off, the dominant force term is the air drag. In this case, the 

rocket will actually be decelerating from its maximum velocity. Now, the mass should be 

maximized in order to reduce the deceleration of the rocket. In conclusion, the mass of the rocket 

will have to be precisely calculated and will need to account for both of these time frames. 

Now that there is an understanding of what the rocket should ideally look like, it is now time to 

find the appropriate parts in order to build a structure that resembled that shape. Obviously 

certain design problems would be met, and only so much of the rocket could be built as 

specified. One important part of this process was watching where the weight was being 

distributed throughout the rocket. Ultimately the main concern was with where the center of 

1

mass (CM) was along the rocket. The CM is a mathematical representation of all of the mass of 

the object, such that the total mass of the rocket could be placed as a point mass at the CM. 

Along with this, the center of pressure (CP) was of concern. The CP in generic terms, is the point 

where half of the profile surface area is above the point and half is below. Thus for a cylinder, it 

would be exactly in the middle. However, if you added a set of fins to one end, the CP would be 

shifted towards the fins. 

The reason why the CP and CM are so important is that the distance between them is directly 

proportional to how stable the rocket is during flight. The CP is related to where the net air 

resistance will occur on the rocket if it tilts off its velocity axis. The CM is related to the inertial 

momentum that the rocket carries. In order to keep the nose of the rocket as close to the velocity 

axis as possible, it is ideal to have the CM as close to the nose as possible, and the CP as close to 

the tail as possible. It is recommended that there be at least one to two body diameters between 

the CM and the CP. [1] Thus, as parts were being placed for the rocket, a close watch had to be 

kept in order to ensure that these two points stayed appropriately apart from each other. 

The last part of design was integrating electronics and other accessories to the rocket. This 

involved finding appropriate altimeters, parachutes, etc. while at the same time adhering to the 

above specifications. The final step in the plan of action was to run a practice launch. This would 

give us an excellent idea of the rocket’s performance and would remove a lot of the guess work 

and calculations involved in its predicted altitude and acceleration. 

Wiki. In order to facilitate the communication of information throughout the team, a very 

reliable system was needed. Because of the extremely busy schedules of everyone on the team, 

having weekly meetings was deemed impossible to plan for and make happen. It was thus 

decided to use an online wiki. Through www.pbwiki.com, the team was able to create a personal 

small wiki that each member could post information in the form of web pages, links to sites, or 

actual files they found useful so that the entire team could read them. The wiki also served as a 

basis for important information regarding upcoming meetings and events. Most importantly, the 

wiki housed the specifications for the rocket as it progressed from a rough sketch, to a CAD 

design, to a parts list, and finally, a completed rocket. After the launch, the wiki will be available 

for public viewing. 

Design Features of Rocket 

Aerodynamics. 

Dart. The only constraint that existed was that the inner diameter had to be at least 29mm 

wide in order to accommodate the WSGC altimeter in it. The ideal shape of the nose cone was 

determined by running various simulations on RockSim. Because of the relatively low velocities 

of the rocket, an elliptical or parabolic shape would perform just as well as a more complicated 

Haack series. The length to diameter ratio was also found to not make a very large effect. Thus, a 

4:1 ratio was chosen as ideal. The length of the body section should be kept to a minimum. 

Obviously this is taking away from the tear drop shape and only adding resistance to the shape of 

the dart. 

The ideal shape for the tail cone would be to match the tail end of a tear drop. Simulations 

showed that a roughly 6:1 length to diameter ratio provided the least amount of resistance. Most 

2

of the basis for fin designs came from the Handbook of Model Rocketry [1]. It suggests that the 

most aerodynamic fin shape is the “clipped delta.” The ideal dimensions of this being a root cord 

length of 2D, a tip cord length of D, a span of 2D, a leading edge cord of D, and a thickness of 

0.2D (where D is the diameter of the body of the rocket). The ideal profile would be a 

teardrop/airfoil. The fins should be located as far back on the rocket as possible, as to provide the 

most CP benefit. Obviously, a minimum of three fins should be used. 

Booster. There were three constraints that had to be considered when designing the 

booster stage. The first and most important was the four inch diameter that was required by the 

competition. Another constraint was the length of the motor. The booster had to be long enough 

to fit the entire motor in it. The last concern was the transition area. The booster had to have the 

ability to accept a tail cone from the dart stage. The Haack series determines the shape that gives 

the least amount of drag for the transition cone is to be a raindrop [5]. Thus, this would be the 

ideal shape of the cone. The value of C used in the Haack to find the ideal length of the cone is 0 

[5]. A length to diameter ratio of roughly 5:1 was found to be the most ideal. However, RockSim 

simulations showed that the additional weight of the cone added reduced the positive benefits of 

the cone. Thus, a 1:1 ratio was used for the transition cone. 

The length of the body section should be kept to a minimum. This section is only increasing drag 

by taking away from the tear drop shape. The tail cone shape came from trying to maintain the 

tear drop shape as much as possible. Ideally, the cone should be as long as possible, but it was 

limited to the length of the motor as to not add unnecessary weight. Because of the motor that 

needed to be placed in the tail cone, it would have to end flat instead of coming to a point, as 

would be ideal. The design of the booster fins was nearly identical to those of the dart, with a 

couple of differences. The location of the fins was chose such that they extended past the end the 

rocket. Thinking ahead, this would help keep the CP towards the tail end of the rocket. However, 

this then required that small triangles be cut from the inside corners to prevent the motors hot 

gasses form damaging he fins. Lastly, the fin size was also scaled as to provide appropriate CP 

assistance. The air flow inside of the four inch diameter is very turbulent [1], hence it is 

suggested to have fins that extend past this area. For this rocket, the fins extend a whole body 

diameter past the perimeter of the four inch diameter body. 

Structure 

Dart. The ideal mass of the dart was quite difficult to calculate as described earlier. A 

known booster stage mass had to be used and was plugged into the methods described later on. 

The small diameter of the body limited the usable shapes for the cone. The nose cone was 

purchased through Public Missiles Ltd., as a solid composite cone. However, the material was 

carefully drilled out, so that a large brass weight could be placed in the cone. Brass was chosen 

as the weight, because it is more dense than steel and there was easy, free access to it. This extra 

mass was used to keep the CM closer to the nose of the rocket. Phenolic tubing was used for the 

body of the dart stage. This kind of tubing had the lowest density for a reasonable price. 

Additionally, the strength likely exceeds the needs for the rocket. 

A custom wooden bulkhead was used to separate the electronics bay from the parachute bay and 

to give the parachute a solid attachment point. The bulkhead was kept in place by using a dowel 

that was placed perpendicularly through the body and bulkhead. It was friction fit in place. 

3

Additionally, two holes needed to be drilled vertically through it in order to allow the wires from 

the altimeter to enter into the parachute bay. An eye bolt was installed vertically through the 

bulkhead to provide the parachute cord with an attachment point. The last modification made 

was a small narrow slot cut in the top of the bulkhead. With the use of a long screwdriver, this 

allows the rocketeer to orient the bulkhead so that the holes matched up and the wooden dowel 

could be placed through the part. 

A hollow fiberglass tail cone was purchased through Wildman Rocketry. A hollow cone was 

chosen to help keep the CM towards the front of the dart and also provide a location for the 

transmitter to sit in. The fins were designed to be constructed with 1/16” G-10 fiberglass. 

However, Wildman Rocketry shipped the order to the wrong address. Due to shortage of time 

ABS was used as a substitute. “Clipped delta” fins were custom cut from 3/32” ABS sheet. The 

“clipped delta” shape was per recommendations by [1]. 

In order to prevent the electronics from compressing into each other during takeoff, a simple 

electronics mount was built using ABS plastic. The simplest way to protect all three devices 

(WSGC altimeter, team altimeter, transmitter) was to build a mount for the middle device (team 

altimeter). This way, the WSGC altimeter couldn’t run into the other altimeter and the 

transmitter had no danger of running into the center altimeter either. In addition to the four 

mounting holes for the altimeter to be mounted with, several air holes were created to allow air 

to freely flow into the upper area around the WSGC altimeter. The last modifications made were 

two holes for wires to pass through and an antenna mount for the transmitter’s antenna. 30 

minute epoxy was used for every connection point in the dart stage to provide necessary 

structural integrity. 

ABS altimeter mount, ABS darts fins, Bulkhead and dart body tube 

Booster. The ideal mass of the lower stage is fairly obvious, it should be as light as 

possible. This is because the mass of the lower stage is a concern only during the phase when the 

motor is firing. After motor burns out, the two stages will separate and there will no longer be a 

concern regarding the fate of the booster. As described in the Executive Summary, the rocket 

should be as light as possible for this period, as to allow the maximum acceleration of the rocket. 

The transition cone, which is possibly the most important element to the rocket, was produced 

using a custom prototyping machine at WTC. The prototyper created the hollow part using ABS 

plastic. After completion, the cone had to be manually sanded down to remove all of the 

roughness and to ensure perfect fitting. The transition cone must be able to hold the dart 

extremely firmly yet release it as soon as the motor burns out. This was accomplished by letting 

the dart essentially sit on its fins while inside the holes of the transition cone. The upper hole was 

hand sanded to match the size of the tail of the dart exactly. A second hole on the bottom of the 

cone was placed to help stabilize the dart from rocking side to side. Additionally, the transition 

cone must fit extremely firmly into the body of the booster, yet be able to easily slip off when the 

4

ejection charge from the motor fires. This was accomplished by hand sanding both the inside of 

the body and the shoulders of the cone. 

Top and bottom of transition cone, prototyper machine 

Phenolic tubing was used for the body of the booster stage. For a rational as to why this was 

chosen please see the “Body” paragraph of section 2.2.2. 

The fins were custom built by Giant Leap Rocketry using G-10 fiberglass. The fins were beveled 

on both the leading and trailing edges in order to provide better aerodynamics. A thickness of 

3/32” was used per recommendations by Giant Leap. 1/8” could have been used; however, a 

thicker material was used to err on the safer side. The fins were in the “clipped delta” shape as 

per recommendations by [1]. The G-10 fiberglass was chosen as the construction material 

because is has been widely used and tested in the HPR field. It is lighter and cheaper than almost 

any feasible metal to be used. Additionally, it can be easily epoxyed to the body of the rocket, as 

opposed to needing to weld or rivet metal fins. In the rare case that it would crack, it could also 

be repaired quickly and easily with epoxy. 

Custom G-10 fiberglass fins 

A custom tail-cone was purchased through Public Missiles Limited. The tail-cone, or boat-tail, 

that PMI sells happened to fit the rocket design perfectly. It was rough 12 inches long and went 

from a four inch diameter to a one and a half inch diameter. The only custom parts of the cone 

were the three slots that were cut for the fins to fit in. This was extremely helpful as there was 

much less work needed when trying to epoxy the fins in place and one could assume that they 

would be almost exactly straight. 

A combination of friction fitting and a simple screw and washer will hold the motor in place 

during ejection charge firing. The design for this was found in the motor retention device FAQ 

found here [2]. In order to increase the strength of the entire booster stage, the lower half of the 

transition cone was filled with a two part expanding foam. This provided exceptional rigidity to 

the already strengthened lower section that was fortified with 30 minute epoxy. The surface of 

the entire rocket was sanded down using 600 grit sand paper. Adding a paint coat to the rocket 

would only add extra unnecessary weight. Ultimately, the surface is still extremely smooth 

without a paint coating. 

5

Electronics and accessories 

Altimeter. Per recommendation by Frank Nobile, an altimeter was purchased from Adept 

Rocketry to be used in the dart stage of the rocket. At the time of purchase it was not decided 

whether the parachute would be deployed at apogee or at 750 feet above ground. Hence, a model 

was chosen with the options for dual deployment. Weighing in at only 26 grams, the ALTS2 is a 

very bare minimum altimeter. It reports maximum altitude, but not accelerometry data. This is 

acceptable because the only function really required by this device is the ability to deploy the 

parachute at the proper time. The altimeter is connected to e-matches from Wildman Rocketry. 

Black powder is then used as the propellant for ejection, as it is the generally accepted standard. 

Locating beacon. Being able to track the rocket was an extremely important objective. 

GPS systems would be ideal. However, they are also bulky and expensive, and if there is trouble 

making contact with the satellites, they can be completely useless. An alternative to GPS is radio 

tracking, similar to how one would track wildlife or hunting dogs. It works by having a very 

small radio transmitter on the rocket that transmits a radio signal. Then a radio receiver listens to 

that signal using a directional antenna. By waving the antenna in various directions, one can 

easily tell where the signal is coming from and thus where the rocket is. 

Only a few models could actually be found on the internet. The smallest and cheapest were 

found at Adept Rocketry. Weighing in at only 14 grams and a receivable range of over one mile 

on the ground, it was the obvious choice to go with. A directional antenna was also purchased 

from Adept Rocketry to go with a receiver purchased from the Ham Radio Outlet. An important 

note here is that to purchase this equipment, an FCC amateur radio license needed to be obtained. 

This involved taking an FCC test and paying a small fee. 

Booster parachute. A parachute was chosen from Public Missiles Ltd. that was 30 inches 

in diameter. This parachute was chosen because there was already going to be an order placed at 

PMI and it would save on shipping to order through them. The size was chose based on a target 

falling weight of 1 Kg. Using their weight to parachute diameter chart PDF found here [2], the 

smallest parachute that would still hold 1 Kg was chosen. Additional tests using RockSim 

ensured that the rocket would not be falling at a speed greater than 20 feet per second. 

To attach the parachute to the booster, an 11 foot section of 3/8” nylon tubular webbing was 

used. This webbing was chosen because the stretch of it would decrease the stress on the booster 

when the parachute opened up. The width of the webbing provides more strength than is really 

needed, but this allows room for error in case part of the webbing is melted during the ejection 

charge. The length of webbing was chosen to allow more stretch and also to prevent the webbing 

from ripping a hole down the side of the body tubing. [2] A parachute “slider” was also rigged 

up for the parachute to slow the opening of the parachute. The idea from this comes from BASE 

jumping parachutes, which use a similar mechanism to prevent the parachute from opening too 

quickly [3]. 

Dart parachute. This parachute was chosen in a similar fashion the booster’s parachute. 

However, because PMI only carried an 18 drogue chute and not a normal chute, an order had to 

be placed through Madcow Rocketry. This chute was of similar construction to the PMI one 

6

using rip-stop nylon cloth. To attach the parachute to the dart, fireproof cord/string was also 

purchased through Madcow rocketry. Estes model rocketry wadding will be used in order to 

protect the parachute from the heat of the ejection charge. This parachute also utilizes a small 

parachute “slider.” 

Construction of Rocket 

Obstacles. The largest obstacle in building the rocket was trying not to add too much 

weight via adhesives. A certain amount of adhesive weight was calculated into the rocket, and it 

was important to adhere to that while at the same time not compromising the strength of the 

rocket. Fitting most of the parts together was actually very easy as most of the parts were custom 

build to exact specifications. The only other problem that was encountered was the time that it 

took to complete some of the custom parts. The most notable obstacle was the time that it took to 

create the custom transition cone. That being said, every custom part was more than worth the 

money and wait. 

Completed rocket photo 

Testing 

Altimeter ejection charge ignition. The altimeter was tested using a bell jar vacuum in a 

campus lab. The altimeter was connected to a flashbulb in order to simulate an ejection charge/ematch. 

After a minimum pressure was reached, the pressure valve was slowly opened and as 

expected, the flashbulb ignited. Using a pressure sensor and an altitude vs. pressure chart, an 

approximate altitude was calculated. This calculated altitude matched closely to the altitude that 

the altimeter recorded. Thus, the altimeter was working properly. 

Practice launch. It would have been ideal to be able to practice launch the rocket at least 

once. This would give us three really important pieces of information. First, what the maximum 

altitude that the rocket would achieve. Second, what the acceleration of the rocket would be like. 

And third, it would test the structural integrity of the rocket and make sure that all of the 

mechanisms worked properly in it. 

Two attempts were make at this. One at the Bong Recreation Area on April 18 th and the second 

on April 25 th near Rochester, Minnesota. Both of these involved extensive communications to 

ensure that an unlicensed rocket could be launched and also that a motor could be purchased. 

Unfortunately both of these dates were rained out and the rocket was never practice launched. 

However, the rocket was thrown off of a high building with its parachute open in order to ensure 

that it fell at an appropriate decent rate. 

Budget 

7

Dividing up the budget. After performing a bit of searching it was found that building 

an entirely custom rocket would be more than the budget could handle. However, it could easily 

afford buying a lot of “semi custom” parts and then assembling them together. Thus, just over a 

third of the budget was designated for major structural parts for the rocket. The next part that was 

considered was electronics. The first was an altimeter that would deploy the upper stage, and the 

other electronics were tracking devices for the rockets. Both of these were deemed extremely 

critical, therefore almost half of the budget was allotted to these parts. This left over a decent 

amount of budget to use on a practice launch and various other miscellaneous parts and tools 

needed. 

Predicted Performance 

Maximum altitude. Maximum altitude: 7730 Feet 

Maximum acceleration. Maximum acceleration: 1490 Feet/Second/Second 

Acceleration versus Time: 

Methods used to calculate above. In order to make the above calculations three major 

methods were used. The first involved using RockSim to automatically calculate the above 

predictions. The second involved finding an exact analytical solution, using Mathematica, to 

solve the differential equation modeling the forces on the rocket after maximum velocity was 

achieved. Inputs to the model consisted of the rocket’s CD and its maximum velocity just after 

motor burn out. Both of these variables were taken from the RockSim program. The last 

calculation made was using MATLAB. A small program, based on an open source code, was 

written to simulate the rocket’s flight computationally. Each of the different methods produced 

slightly different flight paths. Ultimately a weighted average was taken of the three to use as the 

predicted maximum altitude and acceleration noted in sections 5.1 and 5.2. 

Post Flight Evaluation 

Verified Launch Events. 

Flight. The following is an account of the launch to the best of the team’s knowledge: 

The rocket launched from the pad extremely straight and perpendicular to the ground. The dart 

separated correctly, immediately after motor burnout. The parachute opened almost instantly 

after motor burnout. The dart disappeared out of sight less than a second after separation. The 

parachute separated from the booster body. All three fins could be seen falling separately from 

the body. A radio tracking signal was heard as parts of the booster were falling to the ground. 

After two minutes, no radio tracking signals were ever heard again. 

8

Post-flight recovery. A field search returned the following parts of the rocket: The 

booster was the first object to hit the ground and be recovered. On the actual body, the centering 

ring (which the parachute cord was attached to) was fractured in half as seen in 

figure 2a. On the inside of the body tubing, the epoxy fillet was completely missing. Part of the 

inside wall of the body tubing, which likely stuck to the part of the centering ring when it was 

broken from the rocket was also gone. The upper lip of the body tubing was also partly fractured. 

The radio transmitter retaining system was found intact, but no transmitter was recovered. 

The one intact fin slot (figure 1b) contains remnants of the external epoxy fillet used to attach the 

fin to the tail cone. The one edge of the slot (denoted with the red arrow in figure 1b) has been 

deformed and rolled out from the center of the slot. On the other side of the slot, the expanded 

foam below the surface has also been compressed. On the other side of the tail cone, a large 

section was missing between the two remaining slots (shown in figure 1a). The plastic was torn 

along the top and bottom and the expanded foam was sheared in many places. The exposed 

motor mount contained thick amounts of foam attached to it in the center and large chunks of the 

tubing itself had been chipped off near the fin attachment locations. Both of the remaining 

outside edges of the two fin slots had deformed edges, to a larger extent than the first edge 

described in the previous paragraph. 

Figure 1. (a) Top, damaged side of tail cone; (b) bottom, reverse side of tail cone. 

Only one of the three fins could be found. The fin itself was completely intact; there were no 

chips, dents, or bends in the G-10 fiberglass. The inside edge (facing down in figure 2) contains 

fragments of the motor mount tubing, of which it was epoxied to. It also has a thin layer of 

sheared expanded foam on the surfaces of it that were inside of the tail cone. The epoxy fillet, 

placed on the outside of the fin/tail cone, was intact in some places and broken off in others. The 

upper part of the transition cone was recovered completely intact cosmetically and structurally. 

However, the bottom of the transition cone was completely broke off, as visible in 

figure 2c. The break was very clean around the entire circumference of the shoulder of the cone. 

Figure 2. (a) Left, top of booster body; (b) middle, booster fin; (c) right, transition cone. 

9

The parachute was found fully intact with the full length of the parachute cord webbing used to 

attach the transition cone and booster body to it. No part of the booster body was attached to the 

webbing; however, a section of the transition cone bottom was still tied to it. As is visible in 

figure 3, the parachute traveled the farthest of all the parts. Using Google Maps and making a 

few calculations, gives an estimate of 0.52 miles for the traveled distance by the parachute from 

the launch site. The parts fell in a very linear fashion along the direction of the wind. Evidence of 

this is shown in figure 3, which shows a mapping of how the different parts fell over the marsh 

area. No sign of the dart was seen or heard after it separated from the booster stage during 

takeoff. 

Figure 3. Areal overview of recovered parts: launch site (1), booster body (2), booster fin (3), 

booster transition cone (4), booster parachute (5). 

Unverified launch events. There was no quantitative data that could be collected for the 

dart’s flight; however, it is known that it had an almost perfect takeoff. Using updated mass 

values for the dart and booster (due to small modifications made in the field) a new maximum 

altitude of 7600 feet is expected. Conservative estimates still put the dart going at least 6500 feet. 

According to internet weather sources, checked the evening of the flight, winds were estimated at 

roughly 20 mph (~30 ft/s) with gusts up to 30 mph. This means that if the parachute deployed 

and the dart fell at the calculated 20 ft/s, it could be around two miles from the launch site. If the 

parachute didn’t deploy, it would have taken just over 20 seconds to fall, giving an approximate 

radius of 1/8 th of a mile. This would place it roughly at the same location where the booster fin 

was found; an area that was scrutinized heavily in attempt to find the remaining fins. Since the 

entire debris field—from launch to the parachute—was walked and the sky was watched for the 

first few minutes after the launch, it is very unlikely that the dart’s parachute did not deploy. 

Causes of shortcomings 

Booster. The rocket likely experienced an incredible maximum acceleration of ~420 

m/s/s (43 G) and a maximum velocity of ~305 m/s (0.9 mach). Despite these incredible numbers, 

these two factors are not the likely causes for the structural failures. Had acceleration caused the 

problems, they would have likely occurred within 20 meters of leaving the ground, when the 

acceleration was maximized. Also, the fins would have likely pulled out much cleaner and not 

deformed just one single edge on each of the slots. Had the vibrations from velocity been the 

cause, the fins would have again likely been pulled out cleaner. It is important to also remember 

that the CM and CP were still intact for the booster, even after the dart separated. The booster 

itself, with an empty motor casing, was still stable by a factor of 2.5 body diameters. 

Additionally, none of the above factors explain why the parachute exited from the body just after 

separation. 

10

Thus, it is believed that the main contributor to the shortcomings was a premature ejection 

charge from the motor. This seems to best explain the events that happened during launch. The 

motor was supposed to come with an ejection charge delay that allows the rocket to slow down, 

and in most cases reach apogee, before the parachute is deployed. Hence, the rocket was 

designed to open up its parachute at speeds at or less than its terminal velocity but certainly not 

at 0.9 mach. Structurally, it was not designed to handle those forces. 

Using all of the information above, the team’s best theory of events is as follows: 

-The ejection charge fires very shortly after separation 

-The transition cone and parachute are exposed to ~305 m/s air speeds 

The cone experiences a drag force of ~375 Newtons 

The fully deployed parachute experiences a drag force of ~7500 Newtons 

The body without the transition cone experiences a drag force of ~230 Netwons 

-The parachute accelerates from the body and transition cone extremely quickly 

-The transition cone, with a shorter length of webbing, breaks off first 

-The body, with a slightly longer length of webbing, breaks apart next 

-The force of the parachute on the body turns the body sideways 

Each fin experiences a force of 660•Sinθ Newtons, where θ is its angle of attack 

The two unrecovered fins face upwind and break off 

They “pry out” part of the tail cone as they exit 

-The body rotates and the last fin is ripped from the tail cone 

Dart. Ultimately the high winds caused the dart to fall so far away. Because there was 

only one deployed parachute at apogee, the distance it would fall from the launch would be a 

function of the wind speed. The team was expecting much calmer winds (watching wind speed 

weather reports periodically during the months before the launch). Also, the fact that wind 

speeds increase as altitude increases was not accounted for. The rocket likely fell so far away, 

that the radio tracking system was pushed past its maximum use on land, which was roughly a 

mile and a half. 

Improvements to be made 

Booster. Assuming a normal ejection charge from the motor, the only modification that 

would be made to the booster would be the removal of its radio transmitter. The booster and its 

parachute are large enough that they are easily visible in the sky and in the field when on the 

ground. Another option would be to try and purchase a transmitter from a different company. 

This would only work if it broadcast at a frequency far enough away from the dart’s, such that 

they could be tracked separately. If it would be possible to not have an ejection charge on the 

motor at all, it might be worth it to have an entirely separate altimeter on the booster that would 

deploy the parachute. Thus the booster would not have to rely on the motor’s ejection charge. 

However, if the rocket was to be rebuilt to be able to withstand a deployed parachute at 

maximum velocity, many modifications would need to be made. Every part used on the tail cone, 

except the fins, failed structurally. The plastic tail cone tore, the expanded foam sheared, the 

motor mount tubing chipped apart, and the epoxy held in only some cases. Additionally, the fins 

cannot be reduced in size very much, as this causes a dramatic reduction in the CM/CP stability 

factor. Rough calculations show that the body tube, transition cone, and centering ring could be 

11

made out of fiberglass. Although, the centering ring would need to be mounted to the outer body 

tube and motor mount tubing with metal corner braces and appropriate hardware, in addition 

excessive amounts of epoxy. The fins and the tail cone however, would almost certainly need to 

be made from carbon fiber and would have to be a solid one piece custom design. 

Dart. A solution for the dart is a fairly obvious one. The use of a dual deployment 

tether/parachute system could have allowed the dart to fall much faster and could have likely 

kept it within a half mile radius of the launch. This would have added significant amounts of 

weight to the dart, but now knowing that Tripoli will launch under almost any wind conditions, it 

would obviously be worth it. A practice launch, as was attempted twice, would have also alerted 

the team of this. Lastly, as mentioned above, if there was not a radio transmitter in the booster, 

the dart could easily be tracked as it fell. If nothing else, the rough direction of its location could 

be known. 

Conclusion. Maybe the hardest part of this project is just not knowing what happened to 

the dart. With school now over, a more intensive and comprehensive search will be made. Using 

GPS, a metal detector, and all the information from this report, the team will spend one or two 

days attempting to locate the dart in the weeks ahead. 

It is the team’s hope that this report illustrates that the rocket was still designed particularly well 

and most of the missed objectives came—not from a lack of engineering—but due to random 

events out of the team’s control. Had the parachute been deployed even a few seconds after 

motor burnout, the booster would have likely came down intact and the team could have spent 

more time looking for the dart. Without the high winds and premature ejection, there is a very 

good chance that it would have been a very successful flight. Furthermore, a successful flight 

would likely have produced a maximum altitude anywhere between 50 and 100 percent higher 

than any other rocket launched that day. 

The problems that did occur, were never read about or came up in conversations with any 

experts. Furthermore, it would have been a poor use of resources to even try and account for 

such a rare occurring event. Except for the modification of the transmitter in the booster and the 

use of dual deployment in the dart, no other modifications would be made. Ultimately, the team 

still feels that that the rocket was designed extremely well. 

In addition to all of the work that the team put in, this project could not have been complete 

without the help of many people and organizations. The team would gratefully like to express out 

thanks to the following: The National Space Grant College and Fellowship Program and the 

Wisconsin Space Grand and Consortium for providing all of the funding for this project. Western 

Technical College for providing use to their custom prototyping machine at an extremely 

discounted rate. Jim Connors, a Mechanical Design Technology student from WTC, who helped 

with the operation of the custom prototyping machine. The University of Wisconsin-La Crosse 

Physics Department for providing a location to have meetings and store the rocket. 

References 

[1] Stine, G.H. 2004. Handbook of Model Rocketry. Wiley, University of Michigan. 

12

[2] Public Missiles Ltd. Frequently Asked Questions. 2007. 

. Accessed 15 March 2009. 

[3] Parachute 2009. . Accessed 25 March 2009. 

[4] Drag Coefficient 2009. . 

Accessed 26 February 2009. 

[5] Nose Cone Design 2009. . 

Accessed 28 February 2009. 

13

Student Rocket Design Competition 

Team Narwhal 

Neal Bitter, Chelsey Jelinski, Adam Harden 


Milwaukee, WI 

Abstract 

The goal of this year’s Wisconsin Space Grant Consortium Collegiate Rocket Design Contest was to develop, build, 

and fly a boosted dart rocket. Designs were constrained to using a minimum 4 inch outer diameter airframe for the 

booster and an Aerotech I435 rocket motor. The criteria for a successful flight were as follows: successful 

separation of the dart from the booster, the dart coasting to an apogee higher than the booster’s apogee, and the safe 

recovery of all rocket components. 

Team Narwhal focused on two critical design areas: the stability of the rocket and drag. Significant analysis and a 

number of design iterations yielded a configuration in which both the booster-dart stack and the dart itself were both 

aerodynamically stable. Drag was reduced by using the smallest feasible airframe size of the dart, 1.5 inches, and by 

placing boattails on both the booster and the dart. 

Three test flights were conducted prior to the competition launch. The first test flight proved that the basic design 

concept is sound. However, this flight suffered from over-stability and altitude was severely reduced. The second 

test aimed to determine the drag coefficient of the booster and dart assembly by screwing the two bodies together. 

Because the dart was unable to rotate and self-align, the flight was slightly unstable and poor quality of data was 

obtained. The third test flight featured a 1.5-inch dart and shortened booster section. Due to an insufficient venting 

of the parachute chamber, the pressure trapped in the airframe prematurely pushed the parachute out of the rocket, 

resulting in a loss of the dart. However, the booster was recovered in reusable condition. 

The competition launch was entirely successful. Launch, separation, deployment, and recovery all went without a 

hitch. The peak altitude was 3600 ft, the highest flight of the day. The actual flight was much lower than the 

predicted altitude of 5500 ft. However, post-flight analysis yielded the explanation. First, our predictions neglected 

the effects of the fillets which hold the fins in place—these significantly increased the cross sectional area of the 

rocket and boosted its drag. Secondly, our estimate of drag coefficient was too low. Finally, launch rail friction is 

known to reduce altitude by several hundred feet. 

15

Design Features 

The chief requirements of this design problem were to optimize the weight of the dart and to 

ensure a stable flight. These design efforts, among others, are detailed in this report. 

Booster A layout showing the features and components of the booster and dart is shown in 

Figure 1. 

Figure 1: Layout of Sounding Rocket, showing internal components 

All kinetic energy imparted to the booster is wasted energy as far as the final altitude of the dart 

is concerned. Thus the primary goal in the design and construction of the booster was to keep its 

weight to a minimum, allowing the transmission of a maximal fraction of the motor’s energy to 

the dart. This goal was accomplished by using clean fillets to reduce epoxy use, installing the 

smallest safe size of hardware, and minimizing the length of the body tube. The fully loaded 

weight of the booster was 4.3 lbf. 

Dart The primary design objective for the dart was to minimize drag and ensure stable flight. 

Each of these design efforts is explained in turn. 

The drag coefficient is minimized primarily by streamlining the shape of the dart. The principle 

sources of drag are friction drag and pressure drag, but since rockets are of fairly blunt shape and 

fly at very high Reynolds numbers, pressure drag is the dominant drag source. Pressure drag is 

generated by flow separation at the rear of an object. Thus the best way reduce pressure drag is 

to prevent flow separation by installing a boattail. 

The second geometric feature that affects drag is the cross sectional area of the dart. Although it 

is clear that the area should be minimized, an additional design constraint is present—a certain 

volume of components (parachute, shock chord, fire blanket, and electronics) must be contained 

within the body. At first, we built a 2.1-inch (54mm) diameter dart to ensure that all these 

components would fit. After successfully constructing and flying the 2.1-inch dart, we 

endeavored to further reduce the diameter to 1.5 inches (38mm). This diameter change reduced 

the cross sectional area by a factor of 2.0. Since drag force is directly proportional to cross 

sectional area, the diameter reduction cut the dart’s drag exactly in half! We were able to 

16

successfully fit the necessary components inside this size dart, but concluded the further 

reductions in size were not practical. 

The second major design objective for the dart was to ensure stability. Stability is of major 

importance in rocketry. As a rocket flies, all the drag forces and wind forces effectively act as 

though a single force is applied at one location—the center of pressure (CP). If the CP does not 

exactly coincide with the center of gravity (CG), then the resultant force causes rotation of the 

rocket about the CG. Three classes of flight stability depend on the relative positions of the CG 

and CP: 

1) CP is in front of CG: Rocket is unstable. 

2) CP is just behind the CG: Rocket is stable. 

3) CP is far behind CG: Rocket is over-stable. 

It is clear that case 2 is the most desirable. As a rule of thumb, a rocket with a CP that is located 

between 1 and 2 rocket diameters behind the CG falls into this category. Barrowman’s theory 

was used to approximate the CP of the rocket. This theory was submitted by James Barrowman 

as a practical means to compute a slender rocket’s center of pressure under general 

considerations (most prior theories involved detailed assumptions that made them invalid for the 

majority of applications). Avoiding such limiting assumptions, Barrowman’s theory applies to 

most slender rockets flying at subsonic speeds and low angles of attack. Barrowman found his 

theory’s predictions to be within 10% agreement of wind tunnel testing in most cases. 

The following scale drawings show the CPs and CGs of the dart. Note that the rocket falls in the 

stable range, where the CP is 1-2 calibers behind the CG. 

CG CP 

Figure 2: Scale drawing of 1.5 inch diameter dart showing CP and CG 

Entire Assembly The booster and dart cannot be designed individually; rather, they must 

operate in tandem and be designed with compatibility in mind. For the interface between the dart 

and booster, two design features are of great importance: the method of joining/separating, and 

the stability of the assembly. 

17

The joint between the dart and the booster must accomplish four tasks: 

1. Maintain alignment between the booster and dart along the axis of the dart 

2. Allow the dart to rotate about its axis of length 

3. Support the force necessary to accelerate the dart 

4. Cleanly separate with minimal friction 

Longitudinal alignment between the booster and dart is required for a straight flight. 

Misalignment would cause both flight instability and a large bending moment at the joint, which 

could fracture the airframe. To prevent misalignment, two points of contact between the booster 

and dart are required. The chosen contact points were a sliding “sleeve” fit between the transition 

and the dart’s airframe, and a concentric pin-hole fit at the base of the dart, as shown in Fig. 3. 

Sliding 

“Sleeve” 

Joint 

“Pin” Joint 

Figure 3: Method of Joining Booster and Dart 

This joint type maintains sufficient longitudinal alignment, as was verified test launches. 

The chosen joint also allows the dart to rotate relative to the booster along its longitudinal axis. 

This is important for stability. If the booster and dart are not aligned, center of pressure of the 

rocket will not be located in the predicted location. Moreover, asymmetrical forces will act on 

the dart, promoting unusual behavior. Conversely, if the dart is able to rotate, then it will 

naturally self-align in the proper orientation as the rocket flies. 

The third critical feature of the junction is that it must be able to support the inertia force of the 

dart. The predicted peak acceleration of the rocket is approximately 850 ft/s 2 . The force required 

to accelerate the dart at this rate is: 

� � � 

2.1 �� 

�� 

�� 

32.174 �� 

�� 18 

� 850 �� 

� 55 �� 

��

To accommodate this load, a double bulkhead was installed as a support for the dart. This 

bulkhead was reinforced with a 4-inch long section of coupler tubing epoxied into the booster’s 

airframe. By our estimation, the strength of the bulkhead and the shear strength of the epoxy 

applied over the area of the coupler are sufficient to carry the load. Thus this type of junction 

adequately meets the design criteria. 

The second design feature of great importance when considering the booster and dart assembly is 

stability, because these components must fly stably both together and separately. Thus two sets 

of CPs and CGs must be placed, both of which influence one another. The problem is magnified 

because the CG cannot be found until the fins are already attached (this is particularly true for 

the case of the booster, for which fin attachment adds significant weight) 

To solve this problem, the fins on the dart were slightly oversized to begin with (based on rough 

estimates). This way, the assembly could be built and then material could be cut away from the 

dart’s fins until the optimal fin area is achieved. A Matlab script which applies Barrowman’s 

equations to the rocket was written to facilitate these calculations. An algorithm was also written 

to numerically integrate over fins of non-standard shapes and compute the CP. Parameter input 

was streamlined so that various rocket configurations could be analyzed within seconds. 

Because of this streamlined process, an iterative approach could be quickly used to optimize the 

CP of the booster and dart. During these iterations, the booster fins were kept constant and the 

dart fins were modified. Once the optimal dart fin size and shape was found, the dart’s fins were 

cut down to size. Since the material removed from the dart was minimal, the CG was not 

significantly affected. 

The following figure shows the locations of the CPs and CGs of the dart + booster assembly. 

Figure 4: Drawing of booster + 1.5-inch diameter dart showing overall CP and CG 

It is noteworthy that the contribution of the booster’s fins to the CP is slightly uncertain. Because 

the fins are mounted on a boattail, the flow over a large portion of the fin area is under influence 

of the body, as shown in Figure 5. 

19 

CG 

CP

Area dramatically 

influenced by flow 

over body 

Figure 5: Model of Booster Boattail showing fin areas under significant influence of flow over the body. 

Although Barrowman’s theory accounts for the effects of fin-to-body interference, it provides no 

concession for the effects of fin-to-boattail interference. As a result, the predicted center of 

pressure is probably closer to the nose of the rocket than the actual CP is. However, this effect is 

not expected to be large enough to push the rocket into a state of instability. 

One additional stability concern was considered: As the motor burns, the center of gravity of the 

rocket assembly moves forward. It is likely that this effect will couple with the boattailinterference 

effect, resulting in a stable flight until the dart separates. 

Optimization 

There is an optimal mass of the dart for which the dart reaches the greatest altitude. In order to 

find this mass, the differential equations governing flight were derived and a MATLAB script 

was written to integrate these equations of motion. The numerical method accounted for the 

varying mass of the rocket as fuel is burned. The following chart shows the predicted maximum 

altitudes for different dart weights. 

Maximum Altitude of Dart [ft] 

6000 

5500 

5000 

4500 

4000 

3500 

3000 

2500 

2000 

0 1 2 3 4 5 

Weight of Dart [lbf] 

Figure 6: Maximum Altitude of the dart for various dart weights. The drag coefficient of the dart was 

assumed to be 0.3 and the weight of the booster was 4.3 lbf. The booster + dart assembly’s Cd was 0.35. 

20

As this plot shows, the optimal weight of the dart is 1.6 lbf. We weighted the dart to match this 

optimal value. 

One major concern about the numerical model was the accuracy of the assumed drag 

coefficients. Flight data was available for two rockets that were similar to ours in shape, so 

parameters from these rockets (such as weight, cross sectional area, and motor) were used to 

numerically predict the altitude of the rocket as a function of time. The Cd was then adjusted 

until the predicted velocity-vs-time curve matched the experimental data. For both of these 

rockets, the Cd was almost exactly 0.30. Because our rocket was similarly shaped, we chose to 

use a Cd of 0.30 to model our dart as well. A Cd of 0.35 was used for the dart + booster 

assembly. 

Anticipated Performance 

The following figures show the anticipated performance of the boosted dart under the optimal 

weighting of 1.6 lbf for the dart. 

Velocity [ft/s] 

800 

700 

600 

500 

400 

300 

200 

100 

0 

0 5 10 15 20 

Time [sec] 

Figure 7: Anticipated Velocity of Dart 

21

Altitude [ft] 

6000 

5000 

4000 

3000 

2000 

1000 

0 

0 5 10 15 20 

Time [sec] 

Figure 8: Anticipated Altitude for Dart 

Design Versions 

The following sections describe in detail the development of our competition entry. Our rocket 

went through four revisions to accomplish the following tasks: 

1) Proof of viability of concept 

2) Gather Drag Coefficient data 

3) Proof of viability of small diameter (1.5-inch) dart 

4) Competition Entry 

These four versions are discussed in the following sections. 

Version I: “Green Bean” Based on preliminary estimations, this version was designed in 

September 2009. Eventually nicknamed “Green Bean”, this iteration of the boosted dart was 

expected to serve as a proof of concept, verify the mathematical models, and gather performance 

data. In order to cut costs and reserve a larger portion of the budget for the final design and any 

contingencies, some elements of “Green Bean”, such as the dart fins, were taken from currently 

available supplies instead of purchasing their optimized equivalents. 

The dart itself was designed around a 2.1-inch (54mm) airframe. At the time, it was felt that this 

diameter would provide a good balance between low drag and ease of flight preparation (the 

loading of parachutes, etc). 

This version of the boosted dart was flown at the Midwest Power rocket launch in October, 2008. 

During testing, flight instability was discovered. Because the dart’s fins were from existing 

supplies (to avoid purchasing them), they were larger than necessary, and as a result CP of the 

22

entire rocket assembly was too far forward. In order to compensate, weight was placed in 

nosecone, around the threaded rod. This stabilized the fully assembled stack, but over-stabilized 

the dart. In spite of this, the decision was made to fly in this configuration and the launch was a 

success. However, the dart noticeably arced into the wind and considerable altitude was lost. 

Nevertheless, the basic concept was proven to be viable: the booster-dart junction provided clean 

separation and recovery devices deployed properly. 

Version II: Drag Coefficient Test Flight Because of the overstability of “Green Bean,” the 

drag coefficient of the rocket could not be determined from the Midwest Power flight. Version II 

was built as a second attempt to gather data. This version focused on the drag coefficient of the 

booster and dart while they remain in contact. 

To get the data, the booster and dart from Green Bean were simply screwed together at the joint 

between the two. Recovery devices were deployed with motor ejection only, and the flight 

computer was used only for gathering data. The flight was technically a success because the 

rocket was recovered with no damage. However, a spiraling flight path (caused by the dart’s 

inability to rotate) rendered the data useless for the purposes of finding the drag coefficient. 

Version III: 1.5-inch (38mm) Dart Following the drag coefficient test flight, it was decided that 

we should attempt to fly with a smaller diameter dart. Due to the decreased cross sectional area, 

a dart of this diameter was expected to gain about 1200 feet of altitude. 

Similar in many respects to the 2.1-inch dart, the 1.5-inch dart included a number of changes and 

improvements. The stability problem from Version I was corrected by re-evaluating the fin size. 

The booster was also rebuilt and reduced in size to minimize its weight. The airframe was also 

lengthened in order to provide the necessary volume to store a small parachute and the necessary 

fire blanket and shock cord. 

The booster was designed so that either dart (2.1-inch or 1.5-inch) could be flown 

interchangeably. Changing darts only requires that the transition section be removed and the 

alternative size of transition be riveted on. 

Unfortunately, the test flight of this design was not successful. A major malfunction occurred at 

approximately 500 feet, immediately following powered flight, and resulted in the loss of the 

rocket. The cause of the failure was insufficient venting of the parachute chamber. As the dart 

flew higher into the air, the pressure outside the rocket dropped while the pressure inside 

remained constant. The pressure difference was sufficient to push the nosecone out of the 

airframe, deploying the parachute. The parachute deployed while the rocket was moving at 

maximum velocity and was immediately torn off. The damage to the dart was irreparable. 

However, the booster was recovered in fair condition and both the fin can and transition were reuseable. 

Despite this failure, it should be noted that the flight was aerodynamically stable even 

23

though a strong crosswind was present during the launch. This verified that the CP and CG 

calculations were accurate. 

Competition Flight 

Team Narwhal’s competition launch was, in most respects, very successful. The flight was 

extremely straight and stable, in spite of the high wind speeds. Both sections separated and 

deployed according to plan, descended safely to the ground at a near-perfect rate, and were 

recovered with no damage. The only unfavorable outcome from the competition was that our 

achieved altitude (3600 ft) was drastically lower than the expected altitude (5500 ft). However, 

analysis of the flight data demonstrated that two primary factors caused this discrepancy. 

First of all, the cross sectional area of the dart that was used for altitude predictions was too 

small. The outer diameter of the dart was used to compute the cross sectional area, which failed 

to account for the area of the substantial fillets at the roots of the fins. Out post-flight analysis 

indicated that the fillets add 25% to the cross sectional area of the dart. Accounting for the actual 

cross sectional area brings the predicted altitude much closer to the actual altitude. 

Secondly, we were unable to accurately predict the drag coefficient of the rocket. For our 

prediction, we used a value of 0.3 for the dart and 0.35 for the dart + booster assembly. Analysis 

of the flight data indicated that the actual drag coefficient on the dart is 0.45. When the 

numerical simulation was re-run with the updated cross sectional area and drag coefficients, the 

predicted altitude was within 400 ft of the actual altitude. The remaining difference can be 

explained by launch rail friction. 

The results of this competition highlighted one key principle—the importance of obtaining good 

test data. At the level of sophistication of which we are capable, we were unable to accurately 

predict the flight performance. Had our attempts to take data earlier in the year been more 

successful, we would certainly have come much closer to correctly predicting the peak altitude. 

Conclusion 

The goal of this year’s Wisconsin Space Grant Consortium Collegiate Rocket Design Contest 

was to develop, build, and fly a boosted dart rocket. Our team met these goals with a solution 

that was robust yet nearly optimized. Our entry flew safely and successfully, and logged the 

highest flight of the day at 3600 ft. Although this altitude fell short of our predictions, post-flight 

analyses indicated that the cause was a poor estimate of drag coefficient. 

We feel that this competition was quite successful for team Narwhal. We enjoyed seeing our 

hard work pay off with a beautiful flight and a perfect recovery. As we designed, built, and tested 

our rocket, we learned valuable lessons about rocketry, testing, analysis, and teamwork. 

24


“Drew and Crew” 

Drew Falkenburg, Wesley Larrabee, Caleb Varner 


Milwaukee, WI 

Our Story 

A mutual friend of all of us, Steve Bothe, had competed in the WSGC’s rocket completion for 

two years prior to our entry. Not being members of the team, we watched and sometimes helped Steve 

with his rockets. We were all interested in how rockets worked and the design problem each year. With 

a basic understanding of the components and basic idea of how a rocket works we decided to start our 

own rocket team. 

At the beginning of last school year the email came out announcing the WSGC rocket design 

and build competition. Drew again talked to Caleb and Wes to make sure we all were going to go 

through with it. We decided to go for it. After getting the papers in order and attending the first 

meeting we ordered our parts. As a safety we had Steve Bothe, our self proclaimed advisor and now 

MSOE alumni, present as to make sure that one we order everything we need and two that we order the 

right sizes of everything. 

Roughly two weeks later our parts arrived. This was particularly interesting as Drew tried to 

explain why we were receiving a package for “Public Missiles Unlimited” at the MSOE Residence 

Halls. It was that very next weekend that our rocket started to take shape. Early Saturday morning 

began with a trip to ACE Hardware, which lead to a sponsorship by them, to get the miscellaneous nuts 

and bolts needed for the rocket. Within the next few hours we worked to build our rocket. We had 

made all basic design decisions prior and so our only task was to get the framework together. After a 

few bottles of epoxy and several mounds of sawdust combined with plastic shavings our rocket looked 

like a rocket. Around nine at night we called it quits and started to clean up. Unfortunately, due to our 

busy schedules we were not able to get together for several weeks. 

Realizing that the competition was a week away we decided that it was in our best interest to 

finish our rocket. We got together yet again to finish the building of our rocket and to tweak it as 

necessary. After many trips to the Science building on MSOE’s campus we had our rocket completely 

built. All it needed was some paint. It was quite a task to try to find a spot to paint our rocket in 

downtown Milwaukee, but the next day we managed. At that point our rocket was completed. And just 

in time, the next day was our design presentation. 

Our presentation went well. We had a well planned out design and conveyed our ideas 

effectively. We were also well prepared for the day of the launch. Day of all we had to do we tie our 

parachute on and wire up our electronics. Our first launch went off so well in fact that after we had our 

data analyzed we did a second launch. It did not count for the competition, but we wanted to just for 

fun. All of us enjoyed the competition so much that we will be competing again next year. 

25

Our Design 

Our rocket uses the basic laws of physics in concept and design. The booster is simply meant to 

push the dart upwards. Knowing this we wanted to design it as aerodynamically and light as possible. 

We decided to use a boattail on the booster because this makes the booster, and the rocket in its 

entirety, more aerodynamic. It does this by minimizing the disturbance as the air comes back together 

at the end of the rocket. In keeping with the light design we went with a plastic boattail instead of a 

fiberglass one. We believe that it will be strong enough. At the end of the boattail we have a threaded 

cap for motor retention. It allows for easy retention without having to deal with threaded rod through 

the boattail going to nuts holding a washer. 

The body is the required 4” body tube. We choose a phenolic airframe. This again is because we want a 

very light design and believe that a phenolic tube will still be strong enough. We cut the tube very 

short, roughly a foot; to make the booster light. More importantly however is the aerodynamics of the 

design. We want the booster to create as little drag as possible thus making it short. Simply put, drag is 

directly proportional to surface area; the more surface area the more drag. Within the body of booster 

we used as few centering rings and bulk plates as possible. This reduces weight. We will be using 

motor eject for the booster. This is because it is the simplest and is all we need for the booster. 

The nosecone for the booster is different from most rockets. Because the rocket has a dart 

coming out of it we decided to turn a plastic nosecone into a custom boattail. What this allows us to do 

is to recede the dart into the nosecone making the overlap length longer. Having a longer overlapping 

length makes the rocket less flimsy where the dart comes out. Having the rocket less flimsy reduces the 

amount of energy lost when the rocket ascends. 

The dart consists of a plastic nosecone on each end. This is for the same reason as the boattail 

on the booster. It makes the dart more aerodynamic. The airframe is made of phenolic tube same as the 

booster; this is due to the weight advantage. We made the dart three feet long. Inside the airframe we 

have two electronics bays. The judges’ is in the back end of the dart, just in front of the fins. In the 

front of the dart are our electronics. We are using Missile Works’ RRC2-mini Rocket Recovery 

Controller 1.2 This will record altitude, but more importantly it will ignite the charges the blow the 

chute after apogee. Between the two electronics bays is a parachute connected by a shock cord. In the 

front is a plastic nosecone. 

The back nosecone was converted into a slight boattail. We cut off the tip of the back nosecone, 

turning it into the slight boattail. We cut it so that the hole was just large enough for the brass tube to 

fit. The other end was secured by a bulk plate with a hole cut into the center. In the front boattail of the 

booster is a long nail, roughly ten inches long. This nail is held in the front bulk plate with epoxy. The 

brass tube slips over the nail. The rod will be used to guide the dart off the booster. This allows for a 

relatively frictionless separation between the booster and the dart. This is key in our design. 

The difficult part of designing the rocket was to ensure balance the center of pressure and center of 

gravity at specific ratios. It was calculated that for the utmost performance the center of pressure 

should be located 1.5 to 2 times the diameter behind the center of gravity. This will allow for the most 

efficient thrust and achieve the highest altitude/performance possible. The center of pressure is 

determined by the geometry of the rocket and was checked using the simulation program RockSim 8. 

The center of gravity; however, was found by locating the point on the rocket where it balanced on a 

fulcrum. 

26

In order for the rocket to achieve the best performance the Boosted Dart as a whole needed to 

have a difference in the center of gravity and center of pressure in respect to the booster. This means 

that the difference would need to be between 6 and 8 inches between the two points. A medium needed 

to be determined so that the boosted dart was set up as stated, but also have a certain difference 

between the center of pressure and gravity of the dart on its own. This is important because when the 

dart separates from the booster the momentum from the launch will act as the thrust even though the 

booster has separated. This means that the center of pressure for the dart will have to be behind center 

of gravity by around 3 inches. This was difficult as the weights in both the tip of the dart and the end of 

the booster needed to be adjusted as well as the lengths of the phenolic tubing. 

The physics behind the boosted dart is explained by the law of conservation of momentum. This 

states that the total momentum of a system is constant, meaning that the center of mass of the system 

will continue with the same velocity unless acted on by an outside source. The momentum of the dart 

will continue on long after the booster separates gaining more altitude. 

Constructing Our Rocket 

We started construction with the dart. We arbitrarily made the dart three feet because that is the 

increment the airframe came in. We decided to have Public Missiles Ltd. Pre-cut fin slots in the 

airframe, that way we knew the fins would be square. The dart would need two electronic bays, one for 

our electronics, and one for the judge’s electronics. The nosecone that would be used for the boattail 

had to be modified. Because we were going to use a nail and brass tube to stabilize the dart as it left the 

booster we had to cut off the end of the boattail so that the hole left was a snug fit for the brass tube. 

This brass tube acts as the receptor for a pin, or in this case, a nail that is on the booster. In the 

nosecone of the dart, there is a threaded rod that will be used when finding the perfect center of gravity. 

Fender washers and bolts will be used to shift the wait forward when needed. 

The booster design is unique because it’s composed of two boattails as opposed to having a nosecone 

and boattail. The front boattail was modified so that it would fit the 4 inch phenolic body tube and 

receive the 2.5 inch dart. A bulkhead with a large nail was fastened to the inside of the boattail. This 

nail will go inside of the dart and hold it in place until separation of the launch. The other boattail will 

act as a traditional boattail and receives the four inch phenolic body tube and then tappers down to 1.5 

inches where it will hold the motor housing. There is no need for electronics in the booster, so there 

will be a separation for when the motor eject separates the tube. The body tube is composed of two 

parts so that when the buttons are attached it will be easy to align when on the guide pole. 

As previously stated, the washers in the tip of the dart would be used to help find the center of gravity. 

The first step was to find the center of gravity of the boosted-dart with a booster body length of 12 

inches and a dart length of 48 inches. This showed to be a problem because the center of pressure and 

center of gravity are too close together. When the weight was shifted in dart to move the center of 

gravity, the difference in points of the dart alone too close, making it unstable. In order to relieve this 

problem, the booster body tube was kept at 12 inches and the dart body was cut down to 36 inches. By 

doing this and adding some weight to the darts tip, the boosted dart’s center of gravity and pressure 

were within an acceptable distance apart, and then after separation the center of pressure and gravity 

were at the proper distance for the dart alone The final touches on the rocket are the paint job which 

gives it a sleek and professional look. This was done by laying a solid silver coat, then taping over the 

fins so they would retain their color. From here, fishnet stockings were stretched over the pieces and 

bright orange was sprayed on. The paint scheme was selected that so when in the air the metallic silver 

would reflect the sun light and it would be easy to keep track of. The orange was used because it would 

27

e easily recovered. Bong recreation center has a very rough terrain, and the color alone may not be 

enough to find the booster and dart. In order to make this even easier to recover, panic alarms will be 

placed inside so that there will be a loud pitched noise coming from the landing site so that it will be 

easy to head in the correct direction. 

28

Flight Performance 

Table 1 Simulated and Actual altitudes 

versus time 

Time (s) Altitude 

Sim 

(ft/s/s) 

Altitude 

Act 

(ft/s/s) 

0 0 0 

0.5 93.96 61 

1 341.18 215 

1.5 657.86 523 

2 958.78 773 

2.5 1239.04 1024 

3 1500.28 1245 

3.5 1743.91 1436 

4 1971.17 1628 

4.5 2183.13 1788 

5 2380.71 1950 

5.5 2564.74 2112 

6 2735.93 2242 

6.5 2894.92 2373 

7 3042.28 2472 

7.5 3178.56 2571 

8 3304.06 2670 

8.5 3419.32 2769 

9 3524.64 2836 

9.5 3620.34 2902 

10 3756.69 2936 

11 3950 3036 

11.5 4100 3069 

12 4250 3103 

12.5 4300 3120 

13 4425 3136 

13.5 4500 3122 

14 3950 3036 

14.5 4600 

15 4625 

15.5 4657 

16 4670 

29 

Table 2 Simulates and Actual Accelerations 

versus time 

Time(s) Acceleration Acceleration 

Simulated (ft/s/s) Actual (ft/s/s) 

0 0 32.2 

0.005 954.66 1155.98 

0.015 937.03 1027.18 

0.025 919.32 953.12 

0.035 901.53 872.62 

0.045 883.67 821.1 

0.055 865.75 779.24 

0.065 847.77 750.26 

0.075 829.73 734.16 

0.085 811.61 721.28 

0.095 793.44 714.84 

0.1 784.33 705.18 

0.2 733.91 692.3 

0.3 697.22 660.1 

0.4 682.33 624.68 

0.5 670.78 576.38 

0.6 638.7 531.3 

0.7 560.7 489.44 

0.8 498.15 431.48 

0.9 434.53 386.4 

1 365.32 315.56 

1.1 203.5 35.42 

1.2 60.22 1.288 

1.3 59.98 38.64 

1.4 59.85 48.3 

1.5 58.16 48.3 

1.6 56.51 48.3 

1.7 54.92 48.3 

1.8 53.38 51.52 

1.9 51.89 51.52 

2 50.45 54.74

Performance Characteristics of Predicted and Actual 

Figure 1 Altitude of the simulated and actual flights versus time 

Figure 2 Acceleration of the simulated and actual flights versus time 

30

Discussion of Results 

Acceleration 

According to RockSim, based on the rocket geometry, weight, and motor thrust the anticipated 

acceleration was 1119 ft/sec squared. However, this is assuming perfect conditions. The actual 

acceleration was approximately 1156 ft/sec squared. This is most likely due to a loss in thrust 

provided by the motor, is due to several factors. For the actual launch, separation between the 

rocket and the launch rail is not frictionless, therefore acceleration is lost. The coefficient of 

friction on the surface of the rocket was most likely higher than RockSim predicted. This would 

create more drag and reduce acceleration. The last contribution to acceleration loss is wind. On 

the day of the launch there was a strong wind. This affects the acceleration by altering the 

rockets orientation slightly. This changes the rockets momentum vector. By changing the vector 

some acceleration is lost. It is because of these factors that the actual acceleration was different 

from the predicted value. 

Altitude 

Our predicted altitude at apogee was 4670 feet according to RockSIm, found in table 1; however, 

the actual altitude reached was 3136 feet. The altitude is related to the acceleration in the way 

that the loss of the acceleration leads to the loss of altitude. If the acceleration is lowered then the 

momentum will also be lowered. With a lower momentum the dart will not travel as high after 

separation from the booster.With less momentum the dart will not travel as high. 

Wind will also affect the altitude. Since the center of gravity leads the center of pressure, the 

rocket will turn into the wind as it travels. Ascending at an angle lowers the altitude in respect to 

the vertical. 

When simulating the rocket in RockSim, only model the rocket as a whole instead of a booster 

and a dart. This would keep the drag the same from launch till apogee instead of the case of a 

dart, where the dart has less drag thereby ascending higher. 

There is also the friction when the dart separates from the booster. In our design we tried to 

minimize this amount of friction however there will always be some. Because of this amount of 

friction some momentum is lost during the transfer between the entire rocket and the dart. These 

are the reasons for the difference between the predicted altitude and the actual altitude. 

Conclusion 

Ultimately, the boosted-dart rocket was a success in that it completed the primary goals of the 

competition. That is, to design, build, and simulate a boosted-dart rocket that would have the dart 

separate from the booster at motor burnout. The simulated values for the altitude proved to vary 

from the actual, with a percent error of 32.8%. The reasons for these differences were established 

in the discussion section of this report. The acceleration that was predicted showed to be close to 

the actual value, with only a 3.3% error. All in all, the launch and post flight analysis showed 

that results were collected and analyzed correctly. 

31

Background and Context: 


“Rocky Mountain Miners” 

Brian Mortensen, Jon Neujahr, Ryan May 


Milwaukee, WI 

The rocket was designed for the 2009 Wisconsin Space Grand Consortium collegiate 

rocket competition. The objective of the competition is to construct a boosted dart rocket and 

attain the highest altitude. The boosted dart rocket consists of two parts. The first part is the 

booster, which is powered by an I-435 motor. Secondly, the dart is carried by the booster until 

the thrust from the motor ends. At this point, the dart flies as a projectile until apogee is 

achieved. There was no experience from any of the members of the team in rocketry. 

Procedure and Methods: 

The dart had three clipped delta 

fiberglass fins evenly spaced around the 

bottom of the dart’s airframe. The 

specific shape of the fins provide a 

reduction of any unnecessary drag on 

the dart while maintaining a stable flight 

with the CP lower than the CG. 

Recovery system of the dart 

A parachute with a diameter of 18 inches, 

0.5 inch wide climbing webbing, and a 

flame retardant cloth were used for the 

recovery of the dart. The flame retardant 

cloth served as a firewall. After apogee is 

achieved, the altimeter signals an explosive 

charge to separate the dart, and the 

parachute is deployed from the rocket. 

The dart is separated right above the 

electronics storage bay as shown in Figure 

2. One end of the webbing was attached to 

the top of the electronics bay; the other end 

was attached to the bottom of the nose 

cone. 

Figure 1: Dimensions of the Dart Fins 

Figure 2: Recovery for the Dart 

Funding for competition provided by: Wisconsin Space Grant Consortium 

33 

Sweep = 3.5 in 

T.C = 2.5 in 

Span = 3.25 in

Electronics/ electronics storage 

The altimeter used in the dart was a PML 

AccuFire Staging timer. This altimeter and 

the altimeter used to record flight data were 

secured to a plexiglass sheet via zip-ties, and 

this was slid into the electronics storage bay. 

This was made out of plastic. This material 

provides a stronger airframe section than the 

phenolic tubing. A stronger electronics bay 

was necessary to protect the fragile 

altimeters. The electronic bay was capped 

on both ends by bulk plates, and the bulk 

plates were held to the electronic bay tube 

by threaded rods that ran the full length of 

the bay. 

Center of pressure/ center of gravity 

The locations for the center of gravity and the center of pressure were determined. The CP and 

CG for the booster and dart together were determined for the thrust portion of the flight. The CP 

and CG of the dart by itself was determined for the dart’s solo flight. It was necessary to have a 

CP lower than the CG in order for the rocket to have a vertically straight and stable flight. 

Distance to CG = 30.5 in. 

Distance to CP = 20.9 in. 

Results 

Figure 

and 

4: Location 

Findings: 

of the CP and CG 

for the Booster and Dart Together 

34 

I-bolt to secure 

parachute webbing 

Figure 3: Electronics Bay 

Threaded rods to 

secure bulk plate 

to the top of the 

electronics 

storage bay 

Distance to CG = 23.5 in. 

Distance to CP = 7.8 in. 

Figure 5: Location of the CP and CG 

for the Dart

The Rocky Mountain Miners’ competition flight was judged to be a successful flight. The rocket 

launched, the booster and dart separated, both parachutes deployed, and the rocket was recovered 

in flyable condition. The table and figures shown below compare the actual flight data to the 

predicted performance. 

Predicted Actual % Difference 

Maximum Acceleration (ft/s 2 ) 598.9 998.2 66.7 % 

Maximum Altitude (ft) 4524.9 2988 34.0 % 

Table 1: Comparison of Actual to Predicted Flight Performance 

Altitude (ft) 

Acceleration (ft/s^2) 

1200 

1000 

800 

600 

400 

200 

0 

Acceleration History 

-200 

0 5 10 15 

Time (s) 

Figure 6: Comparison of Actual to Predicted Flight Acceleration 

5000 

4000 

3000 

2000 

1000 

0 

-1000 

Altitude History 

0 5 10 15 20 

Figure 7: Comparison of Actual to Predicted Flight Altitude 

35 

Time (s) 

Actual 

Flight 

Predicted 

Flight 

Actual 

Flight 

Predicted 

Flight

Conclusion: 

The components for the rocket were selected in order to achieve a stable, high strength, 

and low weight design. Phenolic tubing was selected for the airframe to deliver a proper 

combination of high strength and low weight. Three clipped delta fiberglass fins were placed on 

the booster and the dart. The geometry of the fins was carefully selected to reduce excess drag 

on the rocket while maintaining a stable flight with the center of pressure below the center of 

gravity. Also, the dart is equipped with an ogive shaped nose cone on top and a boat tail on the 

bottom. They contribute to a streamlined design that reduces flow separation and drag during the 

rocket’s flight. The transition from the dart to the booster was achieved be placing the boat tail 

of in the dart into the boat tail of the booster. The dart was held in place by a centering ring in 

the booster’s boat tail. The results of the competition were well favored by the Rocky Mountain 

Miners. A successful flight resulted in a third place outcome in the overall competition with a 

maximum altitude of 2988 ft. The knowledge gained from this experience will be useful for the 

team when competing in the 2009-2010 WSGC rocket competition. 

36


Part Three 

NASA Reduced Gravity Programs

Repose Angles of Lunar Mare Simulants in Microgravity 

Isa Fritz 1 , Samantha Kreppel 1 , Kevin M Crosby 1 , Erin Martin 1 , 

Caitlin Pennington 1 , Bradley Frye 1 , Joe Monegato 1 , and Juan Agui 2 

1 Department of Physics, Carthage College, Kenosha, WI 

2 NASA Glenn Research Center, Cleveland OH 

Abstract 

Repose angles for lunar mare simulants were measured in rotating drum experiments aboard 

NASA’s microgravity aircraft, Weightless Wonder. We measured both the maximum critical angle 

of stability and the static angle of repose for simulants JSC-1A and GRC-3 as a function of 

drum rotation rate. These measurements were conducted under vacuum to simulate conditions 

of the 1/6 − g lunar environment, and under standard atmospheric pressure to examine the effects 

of interstitial gasses on inter-particle cohesivity. We find no detectable difference in repose 

behavior between simulant flow at standard atmosphere and flow in a low pressure environment 

of 10 −2 Torr. We further investigate a plausible scaling relationship for the dependence of repose 

angles on effective gravitational acceleration. The relevant scaling parameter is √ Fr where 

Fr = ω 2 R/ge f f is the Froude Number, with ω the drum rotation rate, R the drum radius, and ge f f 

is the effective gravitational acceleration acting on the simulant. We find sufficient evidence in 

the data to support the scaling hypothesis. 

Introduction 

Lunar regolith is an unconsolidated aggregation of rock, mineral, and glass that extends from the 

lunar surface to depths ranging from centimeters to several hundred meters. Regolith in the mare 

regions of the moon is several meters thick on average, while the older highland regions have 

average regolith depths on the order of 10 m. Grain size in the regolith ranges from sub-micron 

particles near the surface, to larger, millimeter sized grains below the surface. The lunar regolith 

on the lit side of the Moon is subject to continual ultraviolet radiation and micrometeoroid bombardment. 

These processes have dramatically influenced the morphology and chemical structure 

of the surface regolith, leaving the lunar surface with a significant electrostatic charge, a rough, 

jagged microstructure, and high surface energy. As such, lunar regolith represents a unique and 

important granular material whose properties are little understood. In the experiments reported 

here, we consider one particular property of the lunar regolith relevant to engineering processes 

that may one day take place on the Moon. 

1

The angle of repose of a granular material refers to the maximum angle (as measured from the 

horizontal) at which the material will form a stable heap. In practice, two angles of repose are 

commonly defined. The dynamic angle of repose measures the steady state heap slope angle obtained 

while the heap grows by continuous deposition. The static angle of repose measures the 

angle achieved by the surface of a static pile relaxing after an avalanche event. 

The angles of repose characterize critical flow properties of granular materials. In particular, 

knowledge of repose angles is essential in establishing processes and guidelines for excavation 

depths and other engineering constraints on soil processing. However, neither the static nor the 

dynamic angle of repose is a fundamental material property. Instead these angles depend on experimental 

conditions. Therefore, useful engineering data on repose angles requires that the experimental 

conditions reproduce as closely as possible the anticipated engineering environment. Of 

particular concern for lunar In Situ Resource Utilization (ISRU) applications is a measurement of 

dynamic repose angles of realistic lunar regolith simulants under vacuum conditions at 1/6 − g. 

We report here the results of an experiment to measure the range of repose angles for lunar mare 

soil simulants under variable gravitational forces and under vacuum conditions of near mTorr pressures 

as well as standard atmospheric pressure and temperature (STP). Gravity plays an integral 

role in granular flows and in the thermodynamics of granular media. Finding a relationship between 

gravity level and the angles of repose is an important step in establishing protocols and 

design specifications for civil engineering processes on the Moon. The angles of repose are key 

parameters in the characterization of the stability of heaps and piles, and their values in the vacuum 

and reduced gravity of the lunar environment are currently unknown for most materials including 

lunar regolith simulants. 

Sustained periods of microgravity are made possible by parabolic flights that offer variable gravitational 

levels for periods of up to thirty seconds. Our experiments were conducted on NASA’s 

Weightless Wonder parabolic aircraft at Johnson Space Center, and made possible by the NASA 

Systems Engineering Educational Discovery (SEED) program. 

Overview and Background 

Dynamic angles of repose are most accurately determined using a rotating drum apparatus. In 

these experiments the material of interest is slowly rotated in a drum which is outfitted with a clear 

window for observing the granular flow. The drum is rotated around its principal symmetry axis 

at a rotation rate ω. Over a range of angular velocities, the granular material will exhibit constantangle 

flow as surface particles at the top of the heap slide down the heap, mix into the contact 

layer with the rotating wall, and are brought back up to the top by the rotating wall. By varying 

the rotation rate, the range of stable repose angles can be explored. The minimum and maximum 

angles obtained in this fashion are related to the static and dynamic repose angles of interest in 

engineering applications. 

Prior studies using a rotating drum-type apparatus have examined the influence of inter-particle 

forces on repose behavior in model granular materials [Forsythe et al., 2001]. These experiments 

2

were conducted under 1 − g conditions and standard atmospheric pressure. Iron spheres of 400 

micron diameter were used as the granular material. An induced magnetic field supplied by a pair 

of Helmholtz coils was used to simulate inter-particle forces. Both the static and dynamic angles 

of repose were found to increase approximately linearly with inter-particle force. This is important 

to the extent that lunar soils exhibit strong inter- particle forces due to their intrinsic charge from 

UV bombardment. Interstitial atmospheric gasses may serve to moderate inter-particle cohesive 

forces, potentially making measurements under standard atmospheric conditions less reliable as a 

predictor of flow behavior in the lunar environment. 

Other studies have examined fine powders whose flow properties might be more similar to lunar 

soils. Castellanos et al. studied the fluidized layer in fine powder flows in a rotating drum at the 

repose angle [Castellanos et al., 2001]. This work was also performed under constant 1 − g, STP 

conditions. We have used the results of the Castellanos work to estimate the flow properties and 

design specifications for the experiment proposed here. 

Recently, investigators have studied repose behavior in rotating drum experiments under both 

hyper- and reduced-gravity conditions. Brucks et al. examined the flow behavior of particles 

under effective gravitational accelerations between 1 − g and 25 − g [Brucks et al., 2007] under 

STP conditions. Their results were consistent with the work done by Klein and White on model 

granular materials under reduced gravity [Klein et al., 1990]. 

In the analysis of Ref. [Brucks et al., 2007], in which repose behavior under hyper-gravity was 

studied, a phase diagram of flow behavior was obtained that appears to have universal applicability 

to repose measurements. Specifically, one can define the dimensionless Froude Number as the 

ratio of the centrifugal force acting on particles in the rotating drum to the effective gravitational 

force acting on the particles: 

Fr = ω2 R 

where ω is the angular rotation rate of the drum, R is the radius of the drum, and ge f f is the 

effective gravitational acceleration experienced by the particles. Values of Fr ≈ 10 −4 result in 

repose behavior across the range of ge f f explored in the work of Brucks et al. 

Regolith Simulants 

In the experiments reported here, two well-characterized lunar regolith simulants are used. JSC- 

1A, manufactured by Orbitec, Inc. [Orbitec, Inc.], is a dark, bulk mare simulant with grain sizes 

≤ 1mm. Production of GRC-3 was commissioned by NASA’s Glenn Research Center to simulate 

the geomechanical properties of lunar mare samples returned during the Apollo missions. GRC-3 

is less cohesive than JSC-1A, and has a larger average grain size. 

3 

ge f f 

(1)

Experiment Design 

The experiment consisted of three rotating drums mounted to a flight rig as illustrated in Fig. 1. 

Each drum was partially filled with a lunar regolith simulant and had its principal symmetry axis 

perpendicular to local gravity. Three video cameras, each centered on a drum viewing window, 

recorded the drums as they rotated. The rotation rates were adjusted using pulse-width modulation 

(PWM) controllers that allow precise control of rotation rates. To obtain repose behavior in 

our experiment under lunar gravity conditions, we have designed PWM-controlled, geared motor 

systems to drive the simulant drums in the 0.1-3.5 RPM regime. 

Figure 1: Experimental rig schematic and close-up of a simulant drum loaded with GRC-3 at a 30% fill. 

Each of the three drums has a video camera aligned with the drum view window to record simulant flow. The 

drums have vacuum flanged fittings and can be mounted to a diffusion pump for low pressure investigations. 

Each drum is 5.5” in length with a 3.5” inner diameter. The drums are milled from 6061 T6 

aluminum and each has a flange-mounted polycarbonate viewing window sealed by an O-ring and 

bolted circumferentially to one face of the drum. The drums were filled to 30% by volume with 

lunar soil simulants. Each drum contained one of two different lunar simulants, allowing us to 

make comparative measurements of repose angles in two different media under various pressure 

conditions. On the first flight day, JSC-1A and GRC-3 were under vacuum and the third drum 

contained GRC-3 at atmospheric pressure. Unfortunately, a pressure leak on one drum rendered 

vacuum data for GRC-3 unusable. 

For experiments on the second flight day, the first and second drums contained the lunar highlands 

simulants OB-1 and NU-LHT1. The third drum was once again under atmospheric pressure and 

filled with NU-LHT1. These simulants proved to be extremely cohesive, complicating the analysis 

of repose behavior. Analysis of the data from these simulants is ongoing and will not be discussed 

in this report. 

For experiments under vacuum, the drums were baked and pumped down prior to the flight using 

4

a diffusion pump, particle trap, and sub-micron particle filter. Pre-flight pressure measurements 

indicate that starting pressures were in the range of 10 −3 − 10 −2 Torr. We had hoped to achieve 

mTorr pressures, but trapped gasses within the simulant material made reductions in pressure below 

10 −2 Torr difficult to achieve. The drums retained sufficient vacuum for the duration of the flight. 

Given the small size of the drums, accurate, post-flight pressure measurements were not possible, 

but we estimate a diffusion rate of less than 0.5 micron (0.5 mTorr) per minute. 

Before the drums began to rotate, the simulant media lay flat in the drums due to the force of 

gravity. As the drums began to rotate at a very low rotation rate, the simulant began to move with 

the drum. At a critical value of the rotation rate, the simulant achieved a constant, non-zero slope 

with respect to the horizontal. The rotation rate of each drum was adjusted to achieve the range of 

repose angles as functions of rotation rate. Target rotation rates were calculated using the Froude 

number and the phase diagram developed in [Brucks et al., 2007]. 

Three mini-DV cameras were aligned with the viewing windows of the drums for recording the 

motion of the simulants in the drums. The video from each camera was analyzed after the flights 

using the software package imageJ [ImageJ, 2008]. The software permits a frame-by-frame analysis 

of the video stream including surface angle measurements and rotation rate determinations. We 

anticipated that the 180 measurements (thirty parabolas × two flights × three cameras) would be 

sufficient to provide statistically useful results given our experimental protocol. An accelerometer 

mounted to the rig provided time- coded information about local acceleration for use in synchronizing 

with the video from the cameras. 

Video footage from the parabolas was correlated with accelerometer data to provide effective gravity 

vectors for each parabola. These vectors define the surface planes from which flow angles were 

measured in the drums. The accelerometer data is necessary because the flight trajectories do not 

always produce effective gravitational accelerations normal to the flight deck. 

Technical Results 

Our analysis of the flight data can be broadly divided into phenomenological observations concerning 

flow regimes and quantitative measurements of surface angle behavior as a function of 

Fr. 

Phenomenology of Flow Regimes 

The behavior of the flowing granular media is a function of the rotation rate of the drum, the geometry 

of the drum, the effective gravitational acceleration acting on the drum, and the microstructure 

of the material. The primary experimental parameters in our studies are the drum rotation rate ω 

and the effective gravitational acceleration ge f f . Different microstructures are examined by studying 

a relatively cohesive simulant (JSC-1A) and a less cohesive simulant (GRC-3). The drum 

rotation rates can be varied over the range 0.1 ≤ ω ≤ 3.5 RPM. By conducting experiments 

5

on the ground and in the Weightless Wonder, we explored repose behavior for both simulants at 

ge f f /gs = 1/6, 1.0, and 2.0, where gs = 9.81m/s 2 is the surface gravitational acceleration. 

In general, we find that flow regimes for each simulant are well characterized by the Froude Number 

Fr (Eq. 1). We have identified three flow regimes in our ground and flight data. 

1. For JSC-1A, Fr < 10 −3 corresponds to an avalanching motion in which the surface angle of 

the simulant builds up to the maximum angle of stability, β and then collapses to a smaller 

value deemed the static repose angle, α through avalanching. The simulant GRC-3 demonstrates 

similar behavior for Fr < 10 −4 . 

2. JSC-1A: 10 −3 < Fr < 10 −1 , GRC-3: 10 −4 < Fr < 10 −1 corresponds to rolling motion in 

which the surface angle of the simulant reaches a constant value which is identified as the 

dynamic angle of repose, θ. 

3. Fr > 10 −1 corresponds to a regime in which centrifugal effects cause the surface of the 

simulant to have a continuously changing angle resulting in an S-shaped surface. We did not 

explore this regime in our experiments. 

The relevant surface angles measured in our experiments are shown in Fig. 2. 

Figure 2: Surface angles measured in the the experiments. In regime 1, heap angle increases until the heap 

reaches the maximum stability angle β before collapsing to the static repose angle α. In regime 2, drum 

rotation rates are sufficient to keep the surface flowing freely at constant repose angle θ. 

Surface Angle Measurements of JSC-1A and GRC-3 

While no theoretical basis for a scaling hypothesis is known to the authors, computer simulations 

of granular flow in heaping experiments suggest that Fr 1/2 is a robust scaling parameter for the 

surface angle when the average grain size in the granular media d is much smaller than the drum 

6

adius R [Orpe et al., 2001, Walton et al., 2007]. In experiments where the condition R >> d is not 

met, surface angles do not exhibit universal scaling with Fr [Walton et al., 2007]. 

In the experiments considered here, the R >> d criterion is satisfied. We anticipate that, if Fr 1/2 

is a true scaling parameter, we should expect that all surface-angle data for a given material at 

a fixed gas pressure should collapse onto a single scaling form for all gravity levels and angular 

velocities. In Fig. 3, measured surface angles for JSC-1A are plotted against the scaling parameter 

Fr 1/2 . The data includes all 1 − g experiments, a limited number of 1/6 − g experiments, and 2 − g 

measurements. The 2 − g measurements were obtained from the parabolic flights during the ascent 

portions immediately following the microgravity descent portions of the flights. Reported surface 

angles for Fr < 10 −3 are obtained by averaging the critical stability angle β and the static repose 

angle α: θ = (α + β)/2 [Liu et al., 2005]. 

Average Surface Angle (°) 

60 

55 

50 

45 

40 

35 

JSC-1A Surface Angles 

1/6 g - 0.01 Torr 

1.0 g - 760 Torr 

1.0 g - 0.01 Torr 

0 0.01 0.02 0.03 0.04 0.05 

√Fr 

Figure 3: Measured surface angles for JSC-1A. Error bars indicate variance in the measurement sets. 

Uncertainty for most lunar (1/6 − g) data is not available because each data point represents only one or two 

angle measurements. 

Our JSC-1A data is suggestive of this scaling hypothesis, but given the large uncertainties in the 

measurement of surface angles, is ultimately non-conclusive with regard to scaling. Granular flow 

is an inherently chaotic process and is particularly difficult to reproduce consistently in Regime 

1. For this reason, each data point in Fig. 3 has a large statistical uncertainty associated with it 

represented by the error bars. Each error bar is the variance of the measurement over five or more 

7

measurements for the 1 − g data. Estimating uncertainties in lunar data is difficult due to the short 

time available for each measurement at 1/6 − g. Lunar data is typically limited to one or two 

surface angle measurements per Fr value. 

Fig. 4 shows the surface angle measurements for GRC-3. Again, uncertainties for lunar data are 

quite large given the relatively few rotations available at each rotation rate. The GRC-3 simulant is 

markedly less cohesive than the JSC-1A. The relatively low cohesivity results in more consistent 

and well-defined surfaces. The uncertainty in the data is thereby reduced relative to the surface 

angle measurements for JSC-1A. From the preliminary data including apparent trends in the 1 − g 

and 1/6 − g data in Fig. 4, we conclude that universal scaling with √ Fr is plausible. While our 

data is consistent with the hypothesis, more data is necessary to draw conclusive results. 

Average Rolling Angle (θ) 

60 

55 

50 

45 

40 

GRC-3 Surface Angles 

1/6 g - 760 Torr 

1.0 g - 760 Torr 

1.0 g - 0.01 Torr 

35 

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 

√Fr 

Figure 4: Measured surface angles for GRC-3. Error bars indicate variance in the measurement sets. 

Data for both simulants suggests that the presence of atmospheric gasses does not significantly 

affect the stability of heaps. While our experiments did not achieve the near perfect vacuum of the 

lunar surface, the independence of repose angle on pressure down to 10 −2 Torr suggests that future 

experiments can reliably be carried out under atmospheric conditions. 

Conclusions 

We have examined the repose behavior of two bulk lunar mare simulants under both standard atmospheric 

and vacuum conditions at 1/6, 1.0, and 2.0 gs. We find that surface flow is characterized 

8

y the Froude Number Fr = ω 2 R/ge f f . Three flow regimes, avalanching, cascading, and centrifuging 

were observed with transitions between regimes occurring at fixed values of Fr that is material 

dependent. For JSC-1A, the critical transition between regimes 1 and 2 occurs at Fr ≈ 10 −3 . For 

GRC-3, a much less cohesive simulant than JSC-1A, regime transition occurs at Fr ≈ 10 −4 , so that 

almost all measurements of GRC-3 take place in Regime 2. 

Surface angle measurements were made in the avalanching and cascading regimes. We find no 

detectable difference in surface angle behavior with ambient gas pressure in the range 10 −2 − 10 3 

Torr. This is contrary to the hypothesis that ambient gasses mediate (or moderate) a cohesive 

interaction between grain particles that raises (or lowers) the effective repose angle of the heap 

over its vacuum value. 

While our data is consistent with the scaling hypothesis θ ∝ √ Fr, the data is clearly incomplete 

with respect to definitively addressing the validity of the scaling hypothesis. The scaling hypothesis 

remains an intriguing possible interpretation for repose behavior under variable gravity. If such 

a relation was statistically tenable, it would provide useful estimates of repose angles under arbitrary 

gravitational acceleration. Such data will inform the design of a variety of lunar exploration 

technologies including hoppers, excavators, and structures. Further, reliable estimates of repose 

angles for Martian soils under appropriate gravitational and pressure conditions would be of great 

utility in understanding geological surface features on Mars. 

Given the potential for such measurements, it is important to obtain statistically significant data 

for surface angles under variable gravity. Further flight and ground data is needed to reduce the 

uncertainty in angle measurements, and allow the scaling hypothesis to be fully tested. 

Acknowledgments 

The authors gratefully acknowledge the Wisconsin Space Grant Consortium for financial support, 

and the Reduced Gravity Office at NASA Johnson Space Center for support of the Systems Engineering 

Educational Discovery (SEED) program. 

References 

[Brucks et al., 2007] Brucks, A., Arndt, T., Ottino, J., and Lueptow, R., Behavior of flowing granular materials 

under variable g., Physical Review 75. 

[Castellanos et al., 2001] Castellanos, A., Sanchez, M.A., and Valverde, J.M., The onset of fluidization of 

fine powders in rotating drums Mater. Phys. Mech., 3. 

[Forsythe et al., 2001] Forsythe, A.J., Hutton, S.R., Rhodes, M.J., and Osborne, C.F., Effect of applied 

interparticle force on static and dynamic angles of repose of spherical granular material Phys. Rev. E, 

63. 

9

[ImageJ, 2008] Image Analysis Software in Java, http://rsbweb.nih.gov/ij/ 

[Klein et al., 1990] Klein, S. P., and White, B.R., Dynamic shear of granular material under variable gravity 

conditions AIAA Ann. 28. 

[Liu et al., 2005] Liu, X. Y., Specht, E., and Mellmann, J., Experimental study of the lower and upper 

angles of repose of granular materials in rotating drums Powder Technology, 154. 

[Orbitec, Inc.] JSC-1AF lunar regolith dust simulant manufactured by Orbitec, Inc. and provided by Juan 

Agui, NASA Glenn Research Center. 

[Orpe et al., 2001] Orpe, A. and Khakhar, D. V., Scaling relations for granular flow in quasi-twodimensional 

rotating cylinders Phys. Rev. E, 64. 

[Walton et al., 2007] Walton, O., De Moor, C. P., and Gill, K. S., Effects of gravity on cohesive behavior of 

fine powders: implications for processing Lunar regolith Granular Matter 9. 

10


Part Four 

Engineering

Dynamics Characterization of the Electron Beam Freeform Fabrication System 

Introduction 

Matthew Kallerud 


Abstract 

Electron beam freeform fabrication, a layer-additive manufacturing process, is a 

technology with promising applications in aerospace. The process melts metal 

wire with an electron beam and creates metallic parts. There are several 

parameters in this process that can affect build quality, including the system 

dynamics of a six axis robotic motion system. An understanding of the system 

dynamics was required to investigate its effect on build quality. High speed 

measurement equipment revealed that the robotic system consistently overshot its 

target velocity. This overshoot was reduced by one order of magnitude by 

utilizing existing commands within the control system. 

Electron b eam freeform fabrication (EBF 3 ) is a layer additive manufacturing process that uses 

wire feedstock to build metallic parts. The electron beam creates a m olten pool on a substrate, 

wire is fed into the molten pool, and the robotic system translates the molten pool to deposit the 

melted wire. This process is used to fabricate metallic parts and structures, of primarily titanium 

and a luminum, for a w ide v ariety o f aerospace applications. C ompared to c urrent me tallic 

fabrication methods, this process is more efficient in terms of energy, time, and material, which 

drives f urther r esearch of t he pr ocess. The p rocess, w hich o perates i n a v acuum, h as b een 

deployed in a portable system and proven to work in zero, Lunar, and Martian gravities. In the 

near future astronauts will be able to use it to construct structures in space or fabricate parts on 

the moon, Mars, or the International Space Station. It also has novel application in aeronautic 

structures; for ex ample, non-orthogonal structures ar e more eas ily built w ith E BF 3 than with 

conventional machining processes [1] [2] [3]. 

EBF 3 Dynamics 

The EBF 3 system includes a robotics system that moves on six axes. In addition to normal x, y, 

and z m ovements, t he s ystem i s cap able o f tiltin g th e electron b eam gun, tiltin g th e ta ble on 

which the part lies, and rotating the table. Six degrees of freedom allows for unusual shapes and 

formations when building. 

The r obot i s c ontrolled through p rogrammed commands c alled G -code. These a re s equential 

commands that can control robot movements via time or distant constraints. 

1

Research Goals 

At this point in the research of the EBF 3 process, little information is known about the system’s 

dynamics. A n investigation and characterization of the dynamics of the system was needed to 

provide a de tailed und erstanding of t he m ovements of t he r obotic axes a nd t o pr ovide a 

necessary foundation for further research on the effect dynamics has on build quality. 

Additionally, better control of the EBF 3 process would help to simplify the requirements for an 

eventual closed loop control system. E xamining and changing the current control s ystem will 

lead to the development of programming methods that more fully control the system dynamics 

and will likely improve build quality. 

Once the dynamics are more fully understood and controlled, the knowledge can be coupled with 

knowledge of the wire dynamics to synchronize the wire and robot, which would be beneficial to 

beginning and ending a deposition. 

Research Methods 

To capture data about the system’s velocity, two methods were considered. Originally, a laser 

positioning s ystem th at had b een in tegrated in to th e s ystem w as te sted. T he la ser f ailed to 

produce reliable data at a high enough rate to capture the period of interest. S ince the robotics 

moved quickly, a high sample rate was needed to capture data in a short period of time. Instead, 

a l inear v ariable d ifferential t ransformer ( LVDT) w as c hosen be cause i t c ould s upport d ata 

collection at a higher rate. An LVDT is a device that measures linear displacement by moving a 

magnetic rod past an electrically charged coil. A current is induced into two surrounding coils, 

which produce a di fferential voltage that is proportional to the linear displacement of the rod. 

The LV DT c an be s et up t o m easure a ny axis of l inear m otion. N ational Instruments da ta 

acquisition hardware and software reads the voltage data into a computer, where it was processed 

and analyzed by custom MATLAB scripts. 

In order to analyze velocity, the voltage data from the LVDT was converted into position data. 

A c alibration e quation w as pr oduced t o de scribe t he l inear r elationship be tween v oltage an d 

displacement, shown in Figure 1. 

2

Figure 1 - A calibration equation for the LVDT 

A calibration equation was produced by moving the robot in one-tenth of an inch increments and 

taking individual measurements from the LVDT. Figure 1 shows that the voltage reading of the 

LVDT and the linear displacement is linear, with R 2 = 1, affirming the quality of the linear fit. 

A MATLAB script imported the voltage data, converted it to position data using the calibration 

equation, di fferentiated t he pos ition da ta w ith r espect t o t ime, a nd ge nerated pl ots of velocity 

versus time. 

Results 

Distance (inches) 

LVDT Calibration 

1.5 

1 

0.5 

0 

-8 -6 -4 -2 0 2 4 

Voltage (volts) 

Linear (9-Jul) 

y = -0.1014x + 0.2466 

R² = 1 

Data w as collected f or a n ormal x -axis movement of one i nch. F igure 2 s hows t he up r amp, 

stabilization, and down ramp of a robotic move in the x direction. 

9-Jul 

Figure 2 - Ramp from 0 to 50 IPM and back to 0, X Axis, one inch move 

3

The system began movement at approximately 0.66 seconds. The target velocity of 50 i nches 

per minute was overshot by over 100%. It also overshot zero after beginning its down ramp at 

1.85 seconds, eventually stopping after 2.5 seconds. 

The velocity plot of a move in the x direction shows that the robotics system is under damped 

when ramping up to the target velocity and when ramping down to zero while stopping. 

Exploration of t he pr ogramming m anual r evealed a lternative c ommands f or c ontrolling t he 

movements. D ifferent combinations of ne w c ommands pr oduced uni que d ynamic responses. 

The overshoot of a normal move command can be eliminated using existing G-code commands 

that modify settings in the robotic control system. F igure 3 s hows velocity versus time of the 

same axis after new commands have been added to the programming. 

Figure 3 - Ramp from 0 to 50 IPM and back to 0, X Axis, one inch move, alternative commands 

Figure 3 shows that the percent overshoot has been reduced to under 10% from over 100%. The 

movement be gins a t 0.7 5 s econds and r eached target s peed i n unde r a qua rter of a s econd. 

However, this motion was much more controlled than the motion from the previous method. 

Testing t he y an d z ax es yielded similar r esults, an unde r damped motion that overshot target 

velocity. The less massive z axis had less overshoot because less mass is easier to control by the 

robot, while the more massive y axis (which consists of a table weighing nearly one thousand 

pounds) displayed a more erratic overshoot. Repeating the new programming method for the y 

and z axis yielded similar results, dampening the movement and reducing overshoot. 

4

Wire 

A secondary investigation into the dynamics of the wire feed was conducted, with the eventual 

goal o f s ynchronizing t he r obot m ovements w ith t he w ire w hen b eginning a nd ending a 

deposition. T he wire was fastened to the LVDT and a ramp up t o 100 inches per minute and 

ramp down to zero was recorded, shown in figure 4. 

Figure 4 - Wire feed, ramp to 100 IPM and back to 0 

Data collection was complicated by slippages between the LVDT-wire connection and between 

the wire and the drive wheels, which can occur during normal operation. F urther investigation 

into the wire dynamics is required. 

Combined 

To c apture a di fferential da ta s et o f t he s imultaneous m ovement of t he wire a nd t he r obot, a 

special set up was designed. First, the LVDT was mounted to the stationary table. The wire was 

fastened to the rod of the LVDT in the same way it was for the wire dynamics test. T he only 

difference between this test and the wire test was including the movement of the x axis. 

The wire fed in the negative x direction (by default), so by mounting the LVDT parallel to the x 

axis and moving the x axis in the positive direction, a d ifferential data s et was captured. T he 

wire was feeding, pushing into the LVDT at the same time the x axis was moving away from the 

5

LVDT (the w ire and r obot w ere m oving i n o pposite di rections). T his a llowed a w ay t o 

investigate the s ynchronization of the two simultaneous movements. The LVDT cap tured t he 

difference in position of the wire and the robot. 

Figure 5 s hows the ve locity di fference be tween t he w ire and t he robot a nd that th e tw o 

movements are not synchronized. 

Figure 5 - Differential Velocity between X Axis and Wire Feed, 50 IPM 

Figure 5 s hows m ovement be ginning at 0.75 s econds, w ith t he r obot s tarting b efore t he w ire 

(pulling away from the LVDT produces a positive velocity). From 1.25 to 2 seconds, the robot 

and the wire are traveling at the same speed (the LVDT is stationary). At 2 seconds the robot has 

stopped but the wire continues to feed (pushing into the LVDT produces a negative velocity). 

This plot shows that the movements are not synchronous but rather sequential. 

Conclusions/Further Research 

Dynamics characterization of the EBF3 system revealed the need for a better control method for 

the robotic movements. Better programming methods produced smoother dynamics, and more 

control ove r d ynamics s implifies t he r equirements f or a closed l oop c ontrol s ystem f or t his 

process. Further investigation into the synchronization of the wire feed and robot will likely lead 

to be tter bui ld qua lity; how ever, s ince t he G -code co mmands ar e s equential, m ore ad vanced 

programming techniques will be necessary. 

6

[1] Taminger, K. M. B.; and Hafley, R. A.: “Electron Beam Freeform Fabrication: A Rapid 

Metal Deposition Process.” Presented at the 3rd Annual Automotive Composites Conference; 

Troy, MI; September 9-10, 2003. In Proceedings. 

[2] Taminger, K. M. B.; Watson, J. K.; Hafley, R. A.; and Petersen, D. D.: “Low Voltage 

Electron Beam Solid Freeform Fabrication System.” US Patent number 7168935, issued January 

30, 2007. 

[3] Watson, J. K.; Taminger, K. M. B.; Hafley, R. A.; and Petersen, D. D.: “Development of a 

Prototype Electron Beam Freeform Fabrication S ystem.” P resented at t he 1 3th S olid Freeform 

Fabrication Symposium; Austin, TX; August 4-8, 2002. In Proceedings, 458-465. 

7

Abstract 

Analysis of Circulation Properties in Wake Vortices 

Vanessa L. Peterson 

University of Wisconsin- Madison 

Wake vortices are created as flying aircraft wing tips disrupt the surrounding air. These vortices 

can become quite dangerous to other aircraft that encounter them at low speeds. Lidar systems 

are us ed t o analyze t he w ind s peeds i n these vortices to c ollect d ata and f rom th is d ata, 

theoretical m odels h ave b een cr eated t o p redict t heir characteristics. H owever, t here ar e s till 

some d iscrepancies b etween t he an alytical m odels an d t he ex perimental d ata co llected w hich 

question th e r eliability of th e a nalytical mo dels. The p urpose o f t his p roject w as t o cr eate a 

program to analyze the experimental results of data collected by lidar on the circulation strength 

of wake vortices as they decay. This research allows for the possibility of comparison between 

analytical and experimental data collected on wake vortices and has the opportunity to increase 

confidence levels of wake vortex models and lidar simulator tools. 

9

Introduction 

The analysis of wake vortex data is a very important factor in airport efficiency. These vortices 

are normally not dangerous at typical flying altitudes, but become increasingly perilous to other 

aircraft when encountered close to the ground. If an aircraft were to fly through the wake vortex 

of a leading aircraft at a low speed and low altitude, there would be a high chance for the aircraft 

to become unstable and crash. 

To a void s ituations s uch a s t his, r egulations ha ve been i mposed t hat force d elays i n b etween 

landings. Since the wake vortices are invisible to the naked eye, the only information about them 

has been collected through lidar detection. There has been no reliable system developed thus far 

to predict their motion, so the regulations are based on the weight of the plane that created the 

wake vortices (2005). 

However, this system can be very inefficient and creates useless delays because in most cases, 

wake vortices decay much quicker than when predicted using the weight of the plane. If a model 

were to be developed that predicted wake vortex decay more efficiently, landing schedules could 

be pushed much closer together, leading to less delays and higher number of uses of each runway 

every day. 

Currently t here a re t hree accep ted m odels f or an alyzing w ake v ortices. T hese w ere d eveloped 

using information gathered from lidar detection. The three analytical models are the Burnham- 

Hallock Model¹, the Lamb-Oseen Model², and the TASS Initial Vortex Model³. The equations 

and variables for these models can be found in Appendix A. 

The purpose of this research has been to take steps toward the goal of developing a program to 

compare the analytical models of wake vortices to actual collected data. If the two data sets can 

be s hown t o co rrelate, t hese analytical m odels can be us ed t o m ake l anding pr ocedures m uch 

more e fficient. The s pecific f ocus o f m y p art i n t his r esearch w as t o create a p rogram th at 

compared t he experimental d ata co llected o n d ifferent as pects o f circulation pr operties in the 

wake vortices as they decay. 

Summary of Research 

To b egin m y research o n w ake v ortices, I created a joined program that co mpared t he t hree 

analytical models mentioned earlier with Ryan Coder and Tim Feyereisen. My part of the project 

was to create the velocity profile for the Burnham-Hallock Model. The program allowed for the 

input of variables affecting the velocity and for the placement of the lidar simulator. When run, 

the pr ogram out put a c ontour pl ot of bot h vor tices a nd t he ve locity pr ofile a s t he l idar s wept 

through the wake vortices. This program gave a baseline understanding of the project we were 

working on as we split into our respective parts. 

The purpose of m y final program was t o cr eate a G raphical U ser Interface ( GUI) t o i mport 

circulation d ata an d allow t he u ser t o m anipulate t hat d ata i n o rder t o create p lots f or l ater 

10

comparison. The GUI allows users to plot three different graphs at one time in order to directly 

compare ch aracteristics. The GUI also outputs the maximums and minimums of each data set. 

The incoming data can then be compared for both the positive and negative vortex include the 

age of the vortex, spatial coordinates, circulation strength, likeness, advection, and sink rate. The 

user is allowed to pick the dependent and independent variable for each data set and then has the 

option to save the view of the GUI under a title of their choosing. An image of the GUI before 

the data has been imported can be seen in Figure 1. 

Figure 1 : T he G UI Without D ata. This is a screenshot of the GUI before any data has been 

imported. As can be seen, the user is asked to first import a data set and is then able to choose 

between variables for plotting. The labels for the graph become visible only after the user has 

chosen the variables. 

After the data has been imported the user can plot up to three different data sets on the same plot. In the 

following example however, only two data sets have been used to compare the circulation strength decay 

of the positive and negative vortex. This is shown in Figure 2. 

11

Figure 2: Screen I mage o f G UI. This image shows the GUI created to allow the user to plot 

different characteristics of the circulation properties in wake vortices. In this instance, the 

circulation strength has been plotted against the age of the vortex for both positive and negative 

vortices. The maximum and minimum values are shown to the right side of the plot window and 

the image has been saved as Test1. 

Results & Conclusions 

The program MATLAB was used to create the GUI for this project. The final script has achieved 

everything i t w as m eant t o do a nd w orks a ccordingly. The da ta t hat has be en i mported i s 

experimental data collected through lidar detection of wake vortices. The next step in this project 

would be to incorporate the analytical models for wake vortex circulation and compare the two 

data s ets. Th e GUI w ould ha ve t o be c reated t o a llow t he us er t o i nput the different 

environmental variables of each test. 

Once this was done, the data could then be compared to determine the validity of the different 

analytical models. If it c ould be shown that the experimental data and a nalytical models were 

very close with very little percent error, it would be possible to use this information to impose 

new r egulations on t he timing of a ircraft l andings. T hese regulations would be m uch m ore 

precise t han t he ol d r egulations a nd w ould f luctuate de pending on t he weather a nd ot her airinfluencing 

circumstances at the time of landing. In the end, the increased knowledge on w ake 

12

vortex a ctivity w ould i mprove upon t he c urrent s ystem a nd i n t urn t he ove rall e fficiency o f 

runway use and airport landing schedules. 

Acknowledgements 

I would like to thank the following people who gave their time and guidance to me and helped in 

my r esearch t his s ummer. I w ould l ike t o t hank t he W isconsin S pace G rant C onsortium f or 

allowing me the opportunity to be a part of NASA, and Debbie Murray and Sarah Pauls for their 

work in finding accommodations and making everyone comfortable on c enter. I would like to 

thank m y mentor, D r. N arasimha Prasad for the opportunity to conduct t he research don e this 

summer and for giving the resources to be a part of his research team. I would like to thank Dr. 

Fred P roctor f or s haring w ith us hi s r esearch o n w ake vor tices a nd D r. J eff H inkley f or hi s 

guidance on the finishing touches of my research. Without any of these people and many more, 

my time here would have been much less productive and much less enjoyable. Thank you. 

13

References 

¹ Burnham, D.C., and Hallock, J.N., “Chicago Monostatic Acoustic Vortex Sensing System,” Report No. DOT-TSC- 

FAA-79-103. IV, July 1982, 206 pp. 

² Lamb, H., Hydrodynamics, 6 th Ed., Cambridge University Press, 1932, 738 pp. 

³ Proctor, F.H., “Interaction of Aircraft Wakes from Laterally Spaced Aircraft,” 47 th Aerospace Sciences Meeting & 

Exhibit, 5-8 January 2009, AIAA-2009-0343. 

(2005, Jan. 1). Monitoring Wake Vortices for More Efficient Airports. Originating Technology/ NASA Contribution 

- Transportation, 16-17. Retrieved July 29, 2009, from 

http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20060022051_2006145875.pdf 

14

Appendix A 

Variables 

radii 

large 

at 

n 

circulatio 

vortex 

velocity 

ngential 

vortex ta 

maximum 

velocity 

l 

tangentia 

of 

profile 

radial 

) 

( 

velocity 

gential 

vortex tan 

maximum) 

is 

V 

(Where 

radius 

core 

vortex 

center 

vortex 

from 

distance 

radial 

vortices 

rotating 

- 

co 

between 

distance 

separation 

initial 

ingspan 

aircraft w 

max 

0 

= 

Γ 

= 

= 

= 

= 

= 

= 

= 

∞ 

V 

r 

V 

V 

r 

r 

b 

B 

c 

Burnham-Hallock Model 

)} 

/( 

){ 

2 

/ 

( 

) 

( 

2 

2 

r 

r 

r 

r 

V c + 

Γ 

= ∞ π 

Lamb-Oseen Model 

]} 

) 

/ 

( 

2527 

. 

1 

exp[ 

1 

){ 

2 

/ 

( 

) 

( 

2 

c 

r 

r 

r 

V − 

− 

Γ 

= ∞ π 

TASS Initial Vortex Model 

c 

c 

c 

c 

r 

r 

r 

r 

B 

r 

r 

V 

r 

r 

B 

r 

r 

V 

4 

. 

1 

for 

]} 

) 

/ 

( 

2527 

. 

1 

exp[ 

1 

]}{ 

) 

/ 

4 

. 

1 

( 

10 

exp[ 

1 

{ 

0939 

. 

1 

* 

) 

2 

/ 

( 

) 

( 

4 

. 

1 

for 

]} 

) 

/ 

( 

10 

exp[ 

1 

){ 

2 

/ 

( 

) 

( 

2 

75 

. 

0 

75 

. 

0 

< 

− 

− 

− 

− 

Γ 

= 

≥ 

− 

− 

Γ 

= 

∞ 

∞ 

π 

π 

15

Validation of Novel Rigid Body Frictional Contact Algorithms using Tracked Vehicle 

Simulation: a Stepping Stone for Billion Body Dynamics 

Abstract 

Justin Madsen 1 

Simulation Based Engineering Laboratory 

Department of Mechanical Engineering 

University of Wisconsin – Madison 

Computer modeling and simulation of mechanical systems with many rigid body frictional contacts is currently 

limited due to inefficient formulation. A time-stepping method which describes frictional impacts and contacts as 

unilateral constraints and solves the resulting linear complementarity problem on the velocity-impulse level with a 

novel fixed-point iteration process has recently been introduced. Research has shown that this new method is 

computationally efficient and converges to a solution under most circumstances. 

This project applies the new methodology to a real engineering system, the tracked subsystem of a hydraulic 

excavator. A nearly identical computational model is formulated and simulated using both the new method and 

using industry-grade multibody dynamics software. The formulation of the models and simulation results are then 

compared. Differences and problems using the new methodology are addressed, and conclusions allude to the need 

for automatic decomposition of complex geometry. 

Introduction 

Over the last decade, simulation-based engineering in the form of virtual prototyping has been 

increasingly utilized by engineers in the design process of mechanical systems. This is due to 

economic factors such as cutting physical prototype costs and reduced time to market. In the case 

of space exploration, testing a prototype under actual conditions can be cost prohibitive or even 

impossible. As engineers apply virtual prototyping to increasingly complex systems, new 

numerical and computational methods must be leveraged to sustain or, preferably, increase the 

complexity of models that can be simulated. 

Systems with large numbers of rigid-body frictional contacts are a group of models which would 

greatly benefit from research into new mathematical and computational methods. Examples of 

such systems include automobiles and lunar landing craft driving on terrain such as gravel or 

sand. Pebble bed nuclear reactors or pharmaceutical drug packing processes could be accurately 

simulated as well. However, these types of models have yet to be effectively simulated because 

most multibody dynamics solvers handle rigid body frictional contacts in an inefficient manner, 

exploiting simplex or penalty-based methods which result in a quadratic worst-case time 

complexity (Tasora 2006). As the number of colliding bodies reaches well into the millions, an 

algorithm with exponential complexity makes the task practically impossible even on 

supercomputers. A time-stepping method which describes frictional impacts and contacts as 

unilateral constraints and solves the resulting linear complementarity problem (LCP) on the 

velocity-impulse level with a novel fixed-point iteration process has recently been proposed 

(Anitescu and Potra 1997; Anitescu, Potra et al. 1999; Anitescu 2006). The interest in this new 

method stems from the fact that it has been shown to have linear worst-case time complexity 

(Tasora 2006). The table below shows the CPU time taken to complete of two sets of simulations 

1 The author would like to thank the Wisconsin Space Grant Consortium for financial support 

of this work through a Graduate Fellowship Grant. 

17

which involved dropping an increasing number of spheres modeled with rigid body frictional 

contacts in a box using two different multibody dynamic engines. The ‘LCP’ column denotes the 

program which exploits the improved formulation, while the ‘Penalty’ column uses the 

inefficient penalty-based method (Madsen, Pechdimaljian et al. 2007). 

The motivation for this research project stems from the fact that a majority of the virtual 

prototyping software used by industry today relies on simple, brute force methods to solve rigidbody 

frictional contact problems. This creates a limit on the number of frictional contacts that 

can be simulated efficiently. High-fidelity models with many contacts, such as tracked vehicles 

or vibration feeders, cannot be used in the design process because even small simulations can 

take hours to days to complete on a state-of-the-art desktop computer. The primary goal of this 

project is to be able to model and simulate real-life mechanisms by leveraging this new rigidbody 

frictional contact formulation. Simulations are expected to take much less time, but 

solution accuracy must be maintained if the new method is to be useful. The mechanism of 

choice is a tracked sub-system of a hydraulic excavator similar to the model used in a previous 

project (Madsen 2007). There are many rigid-body frictional contacts present in the model, and 

short simulations using industry grade software took many hours to complete. 

Table 1: Number of rigid spheres vs. CPU time using LCP and Penalty contact approaches 

Number of Max Number of LCP CPU time (seconds) Penalty CPU time (seconds) 

Spheres Mutual Contacts 

1 1 0.7 0.41 

2 3 0.73 3.3 

4 14 0.73 7.75 

8 44 0.76 25.36 

16 152 0.82 102.78 

32 560 1.32 644.4 

The second goal of this research is to determine if the improved fixed-point iterative method will 

yield tracked vehicle simulation results that are accurate when compared to those from 

previously performed simulations. Concretely, the time evolution of the displacements, 

velocities, and accelerations of all bodies should be similar to the values previously obtained. 

This demonstrates that using the new fixed-point iterative method to solve the set of LCPs that 

arise from modeling frictional contacts as a set of unilateral constraints yields accurate results 

when compared to identical simulations carried out using proven methods. Because this new 

method enforces contact constraints on the velocity-level, there can be a drift in these constraint 

equations due to numerical errors (Studer and Glocker 2005). Also, the new formulation is not 

proven to have unique solution, although it has been shown to converge to a single solution 

under most circumstances (Anitescu 2006). 

Procedure 

The first step in the project was to select an appropriate model in which to apply the new 

multibody dynamic formulation. A hydraulic excavator model similar to that described in 

(Madsen 2007) was selected because the system contains many rigid-body frictional contacts. 

The computational model is described, and then the modeling procedure using two different 

multibody dynamic engines is discussed. The inefficient penalty based formulation for contact 

18

forces is implemented in the popular industry-grade software MSC/ADAMS, while the new LCP 

contact formulation and fixed-point iterative solver are implemented in the research-grade 

Software Development Kit (SDK) ChronoEngine. Difficulty was encountered while porting the 

original model, which was built using MSC/ADAMS, to ChronoEngine using the SDK. Details 

about the differences of the multibody models in each dynamics engine will be presented. 

Original Computational Model. The original model is a tracked subsystem of a hydraulic 

excavator which consists of the following rigid bodies: a center block, five bottom rollers, one 

main idler, three top rollers, a drive sprocket and 45 track shoes. Each of the five bottom rollers 

are modeled identically with a revolute joint that is at a fixed location from the center block, 

allowing the roller to spin around its own center axis with one Degree of Freedom (DOF). 

Friction is present in these joints. There are also contact forces between each roller and all the 

individual shoe elements. The three support rollers are modeled in a similar fashion as the 

bottom rollers. 

The drive sprocket is composed of three parts: two identical gears and a drive wheel. Both gears 

are rigidly attached to the drive wheel, and each gear has contact forces specified with all 45 

track shoes. The drive sprocket revolves around its central axis by a revolute joint which is a 

fixed distance from the center block. It is also where the driving torques and motions are applied. 

The front idler is modeled almost identically as the bottom rollers; it is constrained with a 

revolute joint and has contact forces specified with the track shoes. There is one main difference 

which is due to the presence of a simplified tensioning system. The revolute joint is not a fixed 

distance from the center block; it is constrained with a translational joint that represents the track 

tensioner system. This allows the idler to have an extra DOF so it can move forwards and 

backwards with respect to the center block. A single component horizontal force applied to the 

center of the idler keeps the shoes in tension. 

Each track shoe is connected to its neighboring shoes with revolute joints that rotate along the 

axis of the connecting pins. Contact forces are specified between each track shoe and all the 

running gear components, i.e. the rollers, idler and drive sprocket. There are also contact forces 

between the track shoes and the ground block, which is simply a large rectangle. Finally, the 

entire model is constrained to a single plane by applying a primitive joint to the center block 

which removes the ‘roll’ DOF of the vehicle. The assembled model as seen in MSC/ADAMS is 

shown in Figure 1. 

This type of tracked vehicle was originally selected because it has a simple suspension and low 

operating speeds (~2 km/h) which make its behavior easy to predict. There are also a total of 585 

rigid body contact forces specified in the model, which is the computational bottleneck. The 

original model was tested under multiple conditions using the ground block, the type of 

propulsion method and the rate of the propulsion method as independent variables. However for 

this comparison, a straight-line quasi-steady state dynamic simulation was used for simplicity. 

The next step was to create a model in ChronoEngine that is as similar to the original as possible. 

19

Figure 1. Assembled hydraulic excavator model as seen in MSC/ADAMS 

New Computational Model. Work was then done on replicating the MSC/ADAMS track 

model so it could be used by ChronoEngine. ChronoEngine is a research-grade dynamics engine 

which has no graphical user interface, unlike MSC/ADAMS which allows users to see and 

visually inspect models as they are created. There are classes implemented in the program which 

allow the user to write programs that model and simulate multibody systems. Functions were 

written using a C++ programming environment so that the rigid body information and kinematic 

constraints for the tracked vehicle model were identical in ChronoEngine and MSC/ADAMS. 

Rigid body frictional contacts using the new formulation were created between the track shoes 

and all the other bodies in the system. However, problems were encountered when attempting to 

create collision objects for the complex geometry in the model. More accurate collision objects 

were created by substituting a set of convex hulls for each rigid body collision object, also 

known as convex decomposition, which is discussed in the next section. 

Convex Decomposition. In the early stages of testing the track model using ChronoEngine, it 

became apparent that the program lacked the ability to represent complex geometry correctly as 

collision objects. There are only two types of bodies that have complex geometry in the model, 

the track shoes and the gears of the drive sprocket. Figure 2 shows a Computer-Aided Design 

(CAD) representation of the excavator track shoe. 

20

Figure 2. CAD representation of a track shoe 

The problem with complex geometry, e.g. the track shoe in Figure 2, arises from the fact that the 

collision detection algorithm can only detect contacts between objects that are convex. A body is 

considered convex if any straight line that passes through its volume has exactly one entry and 

one exit point. It is clear from Figure 2 that the track shoe is not convex; the program handles 

this by trying to break the body into a set of convex pieces, then “glues” them together so they 

behave as one body. However, this automated process is not well suited for detailed CAD 

models such as the track shoe, and the collision detection results were unexpected. The collision 

object was somewhat equivalent to covering Figure 2 tightly in plastic wrap that is infinitely 

rigid. It is this type of behavior that prevented the use of the native automatic convex 

decomposition algorithm on bodies with complex geometry such as the track shoe. In order to 

circumvent this shortcoming in the software, a manual convex decomposition was carried out on 

the affected geometry. 

Figure 3. Shaded surfaces collide with other bodies during simulation 

21

Figure 4. Convex decomposition collision geometry used for each track shoe 

The manual convex decomposition used as few convex hulls as possible that still accurately 

described the collision geometry. This was done by studying animations of the original excavator 

simulations to determine the surfaces on the track shoe that are actively in contact with other 

bodies during the simulation. Figure 3 shows both the top and bottom surfaces of the track shoe 

that are in active contact with other bodies as a darker shade. The white (or lightly shaded) 

surfaces saw no contact during the simulation and were omitted in the final collision shape. 

Figure 4 shows the final collision shape used for the track shoe body in the ChronoEngine 

simulations. 

A similar convex decomposition technique is applied to the two gears of the drive sprocket. Only 

the sections of the geometry that are actively colliding with other bodies during the simulation 

are included in the collision shape. In this case, each convex gear tooth is used as a collision 

object which is rigidly attached to the base body of the drive sprocket. Once the correct collision 

geometry was created for the track shoe and gear, it was applied to each appropriate body and the 

entire model was ready for simulation in ChronoEngine (Figure 5). 

22

Results 

Figure 5. Assembled hydraulic excavator model in ChronoEngine 

Simple straight-line simulations were run on the tracked vehicle models in both multibody 

dynamics programs as follows. At the onset of the simulation both models are in the same initial 

configuration, slightly above the ground as shown in Figure 1. In the first second of the 

simulation, the horizontal tensioner force is applied and the rotational speed of the drive sprocket 

is increased from zero to maximum speed. After the first second of simulation, the vehicle 

reaches an average speed and the simulation is allowed to continue in this quasi-steady-state 

motion for a few more seconds. 

Both the MSC/ADAMS and ChronoEngine simulations yield similar results as far as the general 

vehicle model is concerned. As expected, the ChronoEngine simulations took almost an order of 

magnitude less time to simulate. The forward velocity of both models is nearly the same, and 

both models are correctly driven by the track shoes engaging the gears attached to the drive 

sprocket. However, there were problems observed in the simulations, most likely due to do with 

the collision geometry and collision settings. As shown in the result animations, the track shoes 

have a tendency to ‘stick’ to the drive sprocket gears as they are disengaging. 

As described in the convex decomposition section, the actual CAD geometry could not be used 

and had to be broken into convex pieces manually. Each piece was created manually and this 

procedure is prone to error because it is difficult to measure the accuracy of a collision shape. 

Also, there is a relationship between some of the collision detection settings (e.g. outward 

envelope and inward safe margin) and the integrator step size. Modifying collision detection 

23

settings to create realistic contacts sometimes required the integrator step size to be modified as 

well. 

Another problem was encountered when using actual values for mass and inertia, which caused 

constraint drift as described in (Studer and Glocker 2005). Either reducing the mass or 

decreasing the step size below the value used for the results section alleviated this problem. 

Conclusions 

A new methodology for handling simulations with many rigid body frictional contacts has been 

shown to work well for systems with many bodies and simple geometry. A model of a tracked 

subsystem of a hydraulic excavator with complex geometry was created and simulated in two 

multibody dynamic engines and results were compared. In order to leverage the fast new method, 

only convex geometry was used, which led to a time-consuming manual convex decomposition 

of the complex geometry. This procedure is prone to error because the collision model must be 

created manually and there are limited metrics which can be used to determine how accurate a 

collision shape actually is. Also, large ratios of masses and inertias in the system can cause 

constraint drift, and the exact nature of the phenomenon is not well known and merits further 

study. Simulation results have shown that multibody dynamic models with many rigid body 

frictional contacts can be simulated in a computationally efficient manner with the new 

methodology; however, representing complex geometry easily and accurately is still one of the 

major hurdles. 

Currently, the Simulation Based Engineering Lab (SBEL) is investigating simulations in 

ChronoEngine in which bodies with complex geometry are represented as sphericaldecompositions. 

This is a process which takes the original CAD model, breaks it down into a 

mesh, then uses decomposition parameters to automatically create a collision model made 

entirely out of spheres. This allows for fast collision detection and collision forces to be 

computed in parallel on Graphics Processor Cards. Figure 6 shows a spherical decomposition of 

the track shoe CAD model from Figure 2. 

Figure 6. Spherical Decomposition of a track shoe collision model. 

24

References 

Anitescu, M. (2006). "Optimization-based simulation of nonsmooth dynamics." Mathematical 

Programming 105(1): 113-143. 

Anitescu, M. and F. A. Potra (1997). "Formulating dynamic multi-rigid-body contact problems 

with friction as solvable linear complementarity problems." Nonlinear Dynamics 14(3): 231-247. 

Anitescu, M., F. A. Potra, et al. (1999). "Time-stepping for three-dimensional rigid body 

dynamics." Computer Methods in Applied Mechanics and Engineering 177(3-4): 183-197. 

Madsen, J. (2007). High Fidelity Modeling and Simulation of Tracked Elements for Off-Road 

Applications Using MSC/ADAMS: Technical Report TR-2007-02. Madison, WI, Simulation- 

Based Engineering Laboratory, University of Wisconsin, Madison. 

Madsen, J., N. Pechdimaljian, et al. (2007). Penalty versus complementarity-based frictional 

contact of rigid bodies: a CPU time comparison: Technical Report TR-2007-05. Madison, WI, 

Simulation-Based Engineering Laboratory, University of Wisconsin, Madison. 

Studer, C. and C. Glocker (2005). Simulation of Non-smooth Mechanical Systems with many 

Unilateral Constraints. ENOC. Eindhoven, Netherlands. 

Tasora, A. (2006). An iterative fixed-point method for solving large complementarity problems 

in multibody systems. 16th Italian Congress on Computational Mechanics, Bologna, Italy. 

25

Modeling and Optimization of a Two Stage Mixed Gas Joule-Thomson Cryocooler 

Abstract 

H. M. Skye, G. F. Nellis, S. A. Klein 

University of Wisconsin, Madison, WI 53706 USA 

Cryocoolers are a common critical component for space flight missions that provide cooling for liquid propellant 

storage, infrared detector arrays, and a variety of other applications. Multiple stage Mixed Gas Joule-Thomson 

(MGJT) cryocoolers divide the large temperature range that must be spanned in most applications into two smaller 

temperature spans are each addressed more using a more compact system. The system offers high reliability and 

excellent vibration and electrical resistance because the compressors operate at warm temperatures and are 

decoupled from the load heat exchanger. The system operates in a continuous flow loop and is therefore the 

working fluid is readily transported to integrate to distant or spatially large loads. 

This paper presents a thermodynamic modeling tool that has been developed for a two stage, mixed gas Joule- 

Thomson (MGJT) cycle. A conventional vapor compression (VC) cycle using a pure refrigerant pre-cools the 

MGJT cycle that provides refrigeration at the load heat exchanger. The model is integrated with an optimization 

routine in order to investigate the optimal mixture composition for different load temperatures and identify the 

optimal precooling temperature. The system performance is reported in terms of heat exchanger conductance 

(related to size) and compressor displacement. The results for the two stage system are compared to the 

performance of a single stage MGJT cycle in order to demonstrate the benefits and limitations associated with the 

addition of the precooling cycle. 

Introduction 

Future instruments and platforms for NASA space applications will require increasingly 

sophisticated thermal control technology, and cryogenic applications will become increasingly 

more common. For example, the thermal management system used by the Single Aperture Far- 

IR (SAFIR) telescope and other cryogenic telescope missions must provide cooling and heat lift 

at multiple temperatures. Also, the management of cryogenic propellants requires cooling at 

integrated heat exchangers for zero boil off and densification as well as at structural members in 

order to intercept parasitics. One mechanism for providing this cooling is via a cascaded Mixed 

Gas Joule-Thomson (MGJT) cycle shown in FIGURE 1. A conventional vapor-compression 

cycle labeled “1 st stage” provides precooling for the 2 nd stage MGJT cycle. The working fluid 

for the 1 st stage is a single component synthetic refrigerant whereas the working fluid for the 2 nd 

stage is a mixture of components. The load heat exchanger provides cooling to the instrument; 

the refrigeration capacity of the system is Qload � and the nominal load temperature is T7. 


The authors would like to thank ASHRAE for funding this research project (project 1472-RP). Additional support 

provided by American Medical Systems as well as the Wisconsin Space Grant Consortium (WSGC) is gratefully 

acknowledged. 

27

Multi-stage Joule-Thomson cycles are used to divide the large temperature ranges that must be 

spanned into two smaller temperature spans that can each be addressed more using a more 

compact system. The system has the benefit of a continuous DC flow path and therefore the 

cooling fluid can be transported readily to accept distant loads (as opposed to oscillatory 

cryocoolers, where thermal integration is only possible at the cold head). Additionally, the load 

heat exchanger can be physically decoupled from compressors and therefore the system has 

excellent vibration and electrical isolation. A numerical model of this system has been 

developed for optimization of the gas mixture, as well as the other operating parameters 

Two Stage Mixed Gas JT Cycle 

FIGURE 1 provides a schematic of the primary components in the entire system, including 

numbered thermodynamic states. A conventional vapor-compression cycle labeled “1 st stage” 

provides precooling for the 2 nd stage mixed gas JT cycle. The working fluid for the 1 st stage 

analysis is R22 whereas the working fluid for the 2 nd stage is a mixture of nitrogen, ethane, 

methane, propane, isobutane, isopentane, and argon. The refrigeration capacity of the system is 

Q� and the nominal load temperature is T7. 

load 

The purpose of the thermodynamic model presented here is to investigate cycle design issues. 

For example, the model will allow the determination of the optimal mixture composition for the 

2 nd stage JT cycle as well as the optimum precooling temperature. This work is partially based 

on previously work at the University of Wisconsin at Madison that evaluated optimum gas 

mixtures for a single stage JT system [0]. This initial work has been used to optimize the design 

of a single-stage system [2]. This paper utilizes the same modeling methodology, but expands 

the approach to the two stage cycle shown in FIGURE 1. To our knowledge, the theoretical 

optimization of a cascade MGJT system has not previously been reported. Therefore, the model 

is used to identify the merits as well as the potential drawbacks associated with using a cascaded 

system as compared to a single stage, mixed gas JT cycle. 

The refrigeration capacity of a JT cycle is fundamentally limited by the Joule-Thomson effect 

associated with the working fluid. The Joule-Thomson effect is related to the isothermal 

enthalpy difference between the high and low pressure streams in the recuperator. Keppler et. al 

[0] and many others have demonstrated that the isothermal enthalpy difference exhibited by any 

pure fluid is large only over a small temperature span near the vapor dome. The vapor dome 

associated with a mixture of gases tends to extend over a larger temperature range corresponding 

to a temperature that is near the lowest boiling point of the components to one that is near the 

highest boiling point component. Therefore, the use of gas mixtures significantly extends the 

temperature range over which the isothermal enthalpy difference is large and therefore enhances 

the performance of the JT cycle. 

There is a tradeoff between the maximum cooling power that can be provided and the 

temperature range that must be spanned by the recuperator. For example, consider two 7component 

mixtures that could be used in the 2 nd stage of the cascaded system in FIGURE 1(a) 

where the load temperature is 140 K, and the high and low pressures are 1000 kPa and 100 kPa. 

The composition of mixtures A and B are summarized in TABLE 1 (by mole fraction); these 

mixtures have been optimized to produce the maximum JT effect over two different temperature 

28

spans, but both mixtures have the same constituents. Mixture A is carefully optimized for a 

temperature span of 285 K to 140 K, which would be typical of a single stage JT cycle (i.e., 

FIGURE 1 with the 1 st stage removed). Mixture B is optimized for a temperature span of 238 K 

to 140 K, which is typical of a JT cycle with some precooling that lowers the recuperator hot 

inlet temperature to 238 K. The maximum cooling effect that can be produced per unit of mass 

flow rate (i.e., the minimum value of the isothermal enthalpy change) over the temperature span 

for mixture A is 73 W/(g/s), whereas the maximum cooling effect for mixture B over its 

temperature span is 60% larger, 115 W/(g/s). In this example, by reducing the temperature range 

that must be spanned by the mixed gas JT system, it is possible to achieve a 60% increase in the 

amount of refrigeration provided by the JT cycle 

9 

10 

,1st W� 

comp 

compressor 

precooling 

evaporator 

condenser 

11 

8 

expansion valve 

load heat 

exchanger 

3 

5 

4 

6 

1 

7 

,2nd Wcomp � 

2 

compressor 

aftercooler 

recuperator 

Qload � 

FIGURE 1. Schematic of 

two stage refrigeration 

cycle showing the 

thermodynamic states 

associated with each stage. 

TABLE 1. Mixture operating 

temperatures and compositions 

Mixture A B 

Low Temp 140 K 140 K 

High Temp 285 K 238 K 

Nitrogen 0.11 % 0.0% 

Methane 43.3 % 50.1 % 

Ethane 40.3 % 39.3 % 

Propane 0.06 % 1.17 % 

Isobutane 6.67 % 9.38 % 

Isopentane 9.49 % 0.01 % 

Argon 0.07 % 0.0 % 

. 

The benefit of precooling must be evaluated based on whether the increase in cooling power that 

can be obtained is worth more than the increase in overall heat exchanger size (including the 

precooler and the recuperator) associated with the addition of the precooling heat exchanger. A 

figure of merit that describes the compactness of the heat exchangers relative to the refrigeration 

performance is the ratio of the refrigeration load to the total heat exchanger conductance 

( Q� load UAtotal 

), because the conductance is closely related to the size of the heat exchangers. 

The total conductance of the two stage system must include the recuperator and precooler, the 

conductance of the single stage system only includes the recuperator. 

The Optimization Results section demonstrates that the cascaded system does offer a more 

compact heat exchangers for a given refrigeration power over a reasonable range of precooling 

temperatures, provided the refrigerants used in both stages are correctly optimized. Other 

secondary parameters that must be considered when comparing the single- and two-stage 

systems include the overall compressor size and power consumption, contamination control, and 

the complexity and reliability of the system. The compressor requirements (power and 

displacement) can be precisely evaluated using the model discussed in this paper; the 

Optimization Results section shows that the two stage system can be implemented without a 

29

significant change in total compressor displacement although it is likely that the two compressors 

will take up substantially more space than a single compressor with the same displacement. 

An optimized mixture will provide a large JT cooling effect but require a relatively small system 

of heat exchangers. It is not possible to analytically or intuitively select: (1) a 2 nd stage mixture, 

(2) precooling temperature, and (3) recuperator and precooling evaporator pinch point 

temperatures for the two stage system that yield an optimum ratio of cooling power and heat 

exchanger size, as well as a system with a practical compressor sizes. The relationship between 

mixture composition and JT effect is affected by the complex mixture equation of state. The 

heat exchanger size that is required depends strongly on the specific heat capacity of the mixture, 

which varies substantially as a function of temperature, pressure, and mixture composition. 

Methods have been developed [3, 4] for selecting a mixture to yield the largest cooling effect. 

However these optimization methods don’t consider the heat exchanger size. Keppler et. al [0] 

demonstrated that a MGJT system which uses a mixture that is optimized for maximum cooling 

(or maximum efficiency) may be more than twice the size of a MGJT system using a mixture 

that is optimized for cooling per heat exchanger size. Therefore, a numerical optimization 

technique described here is used to design a MGJT system that is designed based on 

minimization of heat exchanger size. 

Optimization model 

This section presents the details of the thermodynamic model that provides the basis for the 

optimization of the 2 nd stage mixture and precooling temperature for the two stage system. The 

correlations used to evaluate the mixture properties are discussed. A freezing point model is 

described; the freezing point model is incorporated into the optimization routine in order to 

provide a constraint on the optimization so that a mixture that may freeze and clog the system is 

not considered. The numerical parameters that are required by the model and the optimization 

algorithm are investigated and appropriate parameters are selected. Finally, the optimization 

algorithm that is used to select an optimal mixture for a given load temperature is presented. 

Thermodynamic Model. The two stage refrigeration cycle shown in FIGURE 1 is evaluated 

using a numerical modeling tool discussed in this section. The Engineering Equation Solver 

(EES) software [5] is used to solve the governing system of equations that describe the 

performance of the system for a particular set of operating conditions and geometry. The 

modeling tool is used with an optimization algorithm in order to maximize the system 

performance in terms of the previously discussed figure of merit, the refrigeration load per total 

� ). 

heat exchanger conductance ( Qload UAtotal 

A variety of pure fluids or mixtures can be used in the 1 st stage; the working fluid analyzed here 

is R22. Property data for R22 are provided in EES. The 2 nd stage hydrocarbon mixture property 

data are obtained from the NIST4 (also called SUPERTRAPP) database [6]. A mixture of 

synthetic refrigerants was also considered, but is not discussed in this paper. The numerical EES 

model is interfaced with the FORTRAN routines provided in the SUPERTRAPP program from 

the National Institute of Standards and Technology (NIST) with a separate interface routine [5]. 

A comparison of the NIST4 and REFPROP [7] mixture property data over the range of 

30

temperatures/pressures considered in this paper is described in [0]. The model is designed with 

the flexibility to use a mixture or a pure fluid in the 1 st stage, however for this paper, only pure 

refrigerants are used in the 1 st stage. 

The inputs to the model are chosen based on the operating conditions that are achievable using 

conventional equipment. The 2 nd stage aftercooler and 1 st stage condenser are not explicitly 

modeled, rather they are assumed to be sufficiently large that the fluid exiting the compressors is 

cooled to ambient temperature( T amb ) . Additionally, the 1 st stage refrigerant leaving the 

precooling evaporator (at state 8) is assumed to be saturated vapor. The pressure drop in the heat 

exchangers is neglected; therefore the working fluids change pressure only across the 

compressors and expansion valves. Operating pressures representing the high and low pressures 

of each cycle are defined: ( st, st, nd, nd 

high,1 low,1 high,2 low,2 

) 

P P P P . Other inputs to the model include: 

2 nd stage fluid compositions ( y nd , a vector of molar concentrations of each component that will 

2 

be controlled and adjusted, eventually, by the optimization algorithm), the ambient temperature, 

the load temperature ( T load ) , and the precooling and recuperative heat exchangers pinch-point 

temperature differences ( ∆Tpp, pc and ∆ Tpp, 

rec ). 

1 st Stage Analysis. An iterative process is required to solve the governing equations to 

determine the performance of the cycle. The iteration procedure begins by considering the 

enthalpy of the fluid entering the 1st stage expansion valve (h10), which is computed according 

to: 

( amb, st, st 

high ) 

h = enthalpy T P y 

(1) 

10 ,1 1 

where enthalpy represents the correlation in NIST4 or EES that evaluates the specific enthalpy at 

the given state. The enthalpy (h11) and temperature (T11) for the 1 st stage fluid entering the 

precooling heat exchanger can be computed assuming isenthalpic expansion across the valve: 

h = h 

(2) 

11 10 

T11 temperature h11 , P 

low,1 

, y 

1 

( st st ) 

= (3) 

where temperature represent the correlations in NIST4 or EES that evaluates the temperature at 

the given state. An assumed cold-end temperature difference for the precooling evaporator 

(∆Tcold,pc) is systematically varied until the specified precooling evaporator pinch-point 

temperature ( ∆ Tpp, 

rec ) difference is achieved. The pinch point temperature difference is defined 

below. The temperature (T4) and enthalpy (h4) of the 2 nd stage fluid leaving the precooling 

evaporator are calculated: 

T = T +∆ T 

(4) 

4 11 cold , pc 

( , nd , nd 

high ) 


(5) 

4 4 ,2 2 

The 1 st stage working fluid (which is assumed here to be a pure refrigerant) is assumed to exit the 

precooling evaporator as a saturated vapor, so the enthalpy (h8) is computed as: 

h = enthalpy x = 1, P , y 

(6) 

( st st 

low ) 

8 8 ,1 1 

31

The enthalpy (h3) of the 2 nd stage fluid entering the precooling evaporator is calculated using: 

( amb, nd , nd 

high ) 


(7) 

3 ,2 2 

The ratio of the mass flow rate in the 1 st to the mass flow rate in the 2 nd stage (MR) is defined as: 

MR = m� m� 

(8) 

st nd 

1 2 

and is computed using an energy balance on the precooling evaporator: 

MR = h −h h − h 

(9) 

( ) ( ) 

3 4 8 11 

The rate of precooling heat transfer as well as all subsequent energy transfer rates are computed 

on a per unit of 2 nd stage mass flow rate basis. 

Q� m� = MR h −h 

(10) 

pc 

2 nd 

( ) 

8 11 

The precooling heat exchanger is divided into a number ( N pc ) of small heat exchangers, as 

shown in FIGURE 2(a), where each section transfers an equal fraction (1/Npc) of the total 

precooling load. Dividing the heat exchanger into equal heat transfer segments rather than equal 

physical sizes facilitates direct computation of the enthalpy distribution in the heat exchangers 

and significantly improves computation speed and convergence. The first heat exchanger section 

is located at the hot end of the precooling evaporator and is shown in FIGURE 2(b). The 

enthalpy of the 1 st stage working fluid leaving the precooling evaporator is equal to the enthalpy 

of the 1 st stage fluid at the first node of the heat exchanger. 

h = h 

(11) 

st 

1 , pc,0 

The enthalpy of the mixture entering the precooling evaporator is equal to the enthalpy for the 

mixture at the first node of the heat exchanger. 

h2 nd, pc,0 

= h3 

(12) 

The enthalpies of the hot and cold exit streams at the interface of each segment are computed 

using an energy balance. 

heat exchanger 

section index - i 

1 

2 

3 

N pc -1 

N pc 

h h 

st 

1 , pc,0 

h h 

= 8 nd 

2 , pc,0 

3 

. 

. 

h = h 

heat exchanger 

node index - i 

0 

1 

2 

3 

h = h 

N pc -2 

N pc -1 

N pc 

st = 

1 , pc, N 11 nd 

4 

pc 2 , pc, N pc 

8 

1st of N differential segments of HX 

h st = h 

1 , pc,0 

8 

Q� 

pc 

m�nd 2 

1 

Npc 

h nd = h 

2 , pc,0 

3 

Q� 

1 

h st = h st − 

1 , pc,1 1 , pc,0 

m N MR 

Q� 

pc 1 

= − 

m� N 

pc 

h nd h nd 

� 2 , pc,1 2 , pc,0 

nd 

2 pc 

(a) (b) 

FIGURE 2. a) Precooling heat exchanger divided into Npc sections and (Npc + 1) nodes. b) First differential heat 

exchanger element. 

( ) 

h = h −Q m N MR 

� � i = 1…Npc (13) 

st st nd 

1 , pc, i 1 , pc, i−1 

pc 2 pc 

32 

nd 

2 

pc

( ) 

h = h −Q m N � � i = 1…Npc (14) 

nd nd 

2 , , 2 , , 1 

nd 

pc i pc i− pc 2 pc 

The temperatures at the inlet and exit of each side of each section (i.e., heat exchanger node 

index in FIGURE 2(a)) are computed based on the enthalpy and pressure: 

( , , ) 

( , , ) 

T = temperature h P y 

i = 0…Npc (15) 

st st st st 

1 , pc, i 1 , pc, i low,1 

1 

T = temperature h P y i = 0…Npc (16) 

nd nd nd nd 

2 , pc, i 2 , pc, i high,2 

2 

The pinch-point temperature difference is defined as the minimum temperature difference 

between the 1 st and 2 nd stage streams anywhere within the precooling heat exchanger. 

( ) 

∆ T = T − T 

i = 0…Npc (17) 

min nd st 

pp, pc 2 , pc, i 1 , pc, i 

The size of the heat exchangers is a function of the pinch-point temperature difference; a smaller 

pinch point temperature corresponds to a larger value of overall conductance (UA, the overall 

heat transfer coefficient-area product). The refrigeration load also depends on the pinch point 

temperatures; as the pinch point temperature differences in either the recuperator or precooling 

evaporator decreases, the load increases. Therefore a compact system (where Q� load UAtotal 

is 

maximum) results by optimizing the pinch point temperature difference in order to balance the 

heat exchanger size against the refrigeration load. The model uses 2 K as the pinch point 

temperatures for both the precooling evaporator and the recuperator based on previous 

observation that the optimal pinch point temperature for MGJT systems is about 2-6 K. 

The conductance of the precooler (UApc) can be calculated using an effectiveness-NTU 

relationship for a counterflow heat exchanger [8] if the specific heat capacities of the fluids are 

constant throughout the heat exchanger. However, the specific heat of the mixture is very 

sensitive to the temperature and therefore it varies significantly within the heat exchanger. If a 

sufficient number of heat exchanger sections are used (i.e., if Npc is large) then the specific heat 

capacity within each section is very nearly constant and so the effectiveness-NTU solution can be 

used to compute the conductance of each section. A numerical study ensured that Npc was 

sufficiently large that the results are insensitive to this parameter. The total heat exchanger 

conductance is calculated by summing the conductance of each section. The fluid specific heat 

within a section is represented by an average specific heat defined for each stage as: 

( − ) ( − ) 

( − ) ( − ) 

c = h −h T − T i = 1…Npc (18) 

st st st st st 

1 , pc, i 1 , pc, i 1 1 , pc, i 1 , pc, i 1 1 , pc, i 

c = h −h T − T i = 1…Npc (19) 

nd nd nd nd nd 

2 , pc, i 2 , pc, i 1 2 , pc, i 2 , pc, i 1 2 , pc, i 

The effectiveness of each segment ( ε pc, i ) is defined as the ratio of the actual heat transfer rate to 

the maximum possible heat transfer rate that could occur in that section. The maximum heat 

transfer rate in each section occurs when the outlet temperature of the minimum capacity rate 

stream reaches the inlet temperature of the maximum capacity rate stream. 

⎡Q� pc 1 ⎤ 

ε pc, i = ⎢ ⎥ 

⎡min ( c nd , c st MR 

2 , , 1 , , ) ( T st −T 

⎤ 

nd 

pc i pc i 1 , pc, i 1 2 , pc, i ) i = 1…Npc (20) 

m nd N ⎢ − ⎥ 

⎢ 2 pc ⎥ ⎣ ⎦ 

⎣ 

� 

⎦ 

Note that the capacity of the 1 st stage fluid stream must be scaled by MR in order to compare the 

capacity rates of the two streams. The conductance of each section is calculated: 

33

UA 

m� 

⎛ ε −1 

⎞ 

( c nd c st MR 

2 , pc, i 1 , pc, i ) ⎜ ⎟ ( Crpci 

, , ) 

pc, i pc, i 

2 

nd 

= min , ln −1 

⎜εC−1⎟ ⎝ pc, i r, pc, i ⎠ 

where Cr,pc,i is the capacity ratio characterizing the section: 

( nd st ) ( nd st ) 

rpci , , 2 , pc, i 1 , pc, i 2 , pc, i 1 , pc, i 

i = 1… Npc (21) 

C = min c , c MR max c , c MR i = 1… Npc (22) 

The overall conductance of the precooler per unit of 2 nd stage mass flow rate is computed by 

summing the conductances of each of the segments. 

UA N 

pc UApc, 

i 

= ∑ 

i = 1... Npc (23) 

m� m� 

2 i= 

1 2 

nd nd 

2 nd Stage Analysis. The thermodynamic states of the 2 nd stage, the refrigeration load, and the 

temperature distribution in the recuperator are solved using a process similar to that described in 

the 1 st stage analysis. The solution process is nearly the same as that of the single stage MGJT 

cycle presented in [0]. The only difference is that here, the high pressure gas mixture is 

precooled before entering the recuperator, whereas in the single stage system the high pressure 

gas mixture enters the recuperator near ambient temperature. 

− is iteratively adjusted to achieve a 

The recuperator hot end temperature difference ( T T ) 

4 7 

specified recuperator pinch point temperature difference (∆Tpp,rec). The load temperature (T7) is 

specified and isenthalpic expansion is assumed across the expansion valve. A numerical model 

of the recuperator was created by dividing it into sections of equal heat transfer to calculate an 

enthalpy distribution. The enthalpies at states 1, 4, 5, and 7 are used as boundary conditions for 

the numerical recuperator model. The enthalpy distribution and recuperator pressures (Ph,2nd and 

Pl,2nd) are used to calculate a temperature distribution that facilitates the calculation of mixture 

heat capacities and recuperator conductance (UArec). 

Overall Thermodynamic Analysis – figures of Merit. The overall system performance can be 

quantified using several figures of merit of importance to a MGJT system. From a heat 

exchanger size standpoint, an optimal MGJT system is small and generates a large amount of 

cooling power. Therefore an appropriate figure of merit [0] that is used to optimize the system is 

the total cooling load provided per total heat exchanger conductance, which is indicative of the 

heat exchanger size. 

Q� load UAtotal = ( Q� load m� nd 

2 ) ( UArec m� nd + UA nd 

2 pc m� 

2 ) 

(24) 

It is also of interest to reduce the size of the other hardware required; particularly the 

compressors. The compressors can be connected to the heat exchangers via flexible tubing and 

physically decoupled from the precooling and recuperative heat exchangers. Therefore, the size 

of the compressors is less important than the size of these heat exchangers. However, the size of 

the compressors largely dictates the size and weight of the enclosure that houses the 

compressors, as well as the 2 nd stage aftercooler, and 1 st stage condenser. Smaller compressors 

will therefore lead to a small system that can more readily be integrated with flight payloads. 

The compressor suction side flow rate determines the required displaced volume and therefore, 

to first order, the size of the compressor. The figure of merit that captures the combined 

compressor size is the refrigeration load per unit of total compressor displacement: 

34

load total load 

( nd ) ( 1 nd 8 nd ) 

Q� v� = Q� m� v� m� + v� m� 

(25) 

2 2 2 

Optimization Algorithm. The optimization process for the results presented in this paper 

identify the optimal mixture composition in the 2 nd stage ( y nd , the vector of compositions of 

2 

each mixture component) for a particular set of operating conditions. The total heat exchanger 

compactness figure of merit ( QUA � 

total ) calculated by the thermodynamic model varies 

significantly as the mixture mole fraction vector ( y nd ) changes, so it is necessary to evaluate a 

2 

wide range of mole fraction combinations. It is not computationally efficient to parametrically 

evaluate the mole fractions for all possible y nd combinations. Therefore, an optimization 

2 

algorithm that is able to select the optimal mixture using significantly less computations than a 

parametric study is utilized here. 

The optimization routine selected is the PIKAIA 1.2 [9] genetic algorithm that finds the 

maximum of the objective function using an algorithm that mimics biological evolution. A 

detailed description and demonstration of the routine as it is used to select a best mixture is 

found in [0]. As shown by [0], other optimization techniques such as the direct search and 

variable metric strategies do not reliably converge because of the sharp discontinuities in mixture 

properties near phase boundaries as well as other constraints that are placed on the mixture. 

Note that the reliability of the genetic optimization routine comes at the expense of computation 

speed; the genetic algorithm should only be used when other, faster, routines have failed. 

The optimization routine excludes mixtures based on two practical considerations: 1) A mixture 

is excluded from further consideration if the temperature at the exit of the 2 nd stage expansion 

valve is below the freezing point temperature. 2) Mixtures in a saturated or liquid state leaving 

the recuperator (state 1) are excluded to avoid the introduction of liquid into the 2 nd stage 

compressor. The freezing point is calculated as the weighted average of the triple point of the 

mixture constituents, as described in [0] 

Optimization Results 

The results of the optimization model using a pure refrigerant in the 1 st stage and a hydrocarbon 

based mixture in the 2 nd stage are presented in this section. The refrigerant used for analysis in 

the 1 st stage is R22. The Q� load UAtotal 

figure of merit is optimized in order to yield the MGJT 

system that provides the most cooling for a given geometric size for load temperatures spanning 

100 K to 180 K. The effect of the precooling temperature (T4) on Q� load UAtotal 

is studied in 

order to show the optimal balance between the precooler and the recuperator. The other figure of 

merit, the load specific compressor volumetric flow rate is not explicitly optimized but is 

reported as it is important in the design of a practical system. The performance of the two stage 

system is normalized against the performance of a single stage system in order to show the 

relative benefit and penalty associated with the addition of precooling. 

The Q� load UAtotal 

for a two stage system in which an optimal mixture has been selected for each 

precooling and load temperature is shown in FIGURE 3(a). As the precooling temperature is 

35

educed, the temperature range that must be spanned by the recuperator decreases and the 

performance of the 2 nd stage cycle increases. The overall heat exchanger size for an optimized 

system remains relatively constant over the range of precooling temperatures studied here. 

Therefore, as the precooling temperature is reduced, Q� load UAtotal 

increases due to the 

improvement in the efficiency of the 2 nd stage cycle. 

FIGURE 3(b) shows the results in FIGURE 3(a) normalized by the refrigeration load per heat 

Q� UA 

as reported in [0]. 

exchanger size for an optimized single stage system, ( ) 

load total singlestage 

FIGURE 3(b) shows clearly that the two stage system offers a more compact heat exchangers 

compared to the single stage system and this advantage increases as the precooling temperature 

is reduced. 

(a) (b) 

FIGURE 3. (a) Q� load UAtotal 

for the two stage system over a range of precooling and load temperatures. 

(b) Q� load UAtotal 

for the two stage system normalized by the Q� load UAtotal 

of a single stage system. 

FIGURE 3 suggests that the precooling temperature should be made as low as possible in order 

to achieve an optimized MGJT system. However, other considerations related to the compressor 

power and size limit the range of practical precooling temperatures. FIGURE 4(a) shows the 

ratio of the refrigeration to the total volumetric flow rate at the suction to the compressors, 

Q� load v� 

total (which provides an indication of the size of the compressors required). The required 

compressor sizes increase as the precooling temperature is reduced. The suction pressure 

( P ) for the R22 must be reduced in order to achieve the desired precooling temperature; the 

,1 st 

low 

specific volume of the R22 at the compressor suction side increases and the compressor power 

and size therefore increases. The selection of precooling temperature must balance the reduction 

in heat exchanger size that can be achieved against the increased compressor size that is required. 

FIGURE 4(b) shows the Q� load v� 

total for the two stage system normalized by these quantities for 

a single stage system. Note that the increased volumetric flow rate requires additional power and 

Q� v� 

trend. 

the overall system COP decreases; the trend very nearly follows load total 

36

FIGURE 4. (a) Q� load v�total 

for the two stage system over a range of precooling and load temperatures (b) 

Q� load v�total 

for the two stage system normalized by the Q� load v�total 

of a single stage system. 

Conclusion 

A thermodynamic model was developed to evaluate the theoretical performance of the two-stage 

mixed gas JT cycle. The model was interfaced with the genetic optimization algorithm in order 

to select the 2 nd stage mixture composition that maximizes the refrigeration load per heat 

exchanger conductance. The primary advantage associated with precooling is a reduction in the 

temperature span of the recuperator which allows a mixture with a higher JT effect (more 

cooling capability) to be selected for the second stage JT cycle. The increased cooling capability 

results in a more compact heat exchangers, as the overall heat exchanger size remains about the 

same with the addition of precooling. The two stage results compared with the single stage 

system performance show that the heat exchangers. can be made significantly smaller with the 

addition of precooling. The model also demonstrates how the compressor size limits the range of 

practical precooling temperatures. The model is therefore a useful tool for designing a two stage 

system that balances the heat exchanger and compressor hardware parameters. 

References 

1. Keppler, F.; Nellis, G.; Klein, S. A. Optimization of the Composition of a Gas Mixture in a Joule-Thomson 

cycle. HVAC&R Research 2004, vol. 10, 213-230. 

2. Fredrikson, K.; Nellis, G.; Klein, S. A. A Design Method for Cryosurgical Probes. International Journal of 

Refrigeration 2006, vol. 29, 700-715. 

3. Gong, M. Q.; Luo, E. C.; Zhou, Y.; Liang, J. T.; Zhang, L. Optimum composition calculation for 

multicomponent cryogenic mixture used in Joule-Thomson refrigerators. Advances in Cryogenic Engineering 

2000, 45, 283. 

4. Alexeev, A.; Haberstroh, C.; Quack, H. Further Development of a Mixed Gas Joule Thomson Refrigerator. in 

Advances in Cryogenic Engineering: Proceedings of the 1997 Cryogenic Engineering Conference 1997, vol 

43, 1667-1674. 

5. Klein, S. A. EES - Engineering Equation Solver. 2007, 7.982. 

6. Ely, J. F.; Huber, M. L. NIST Thermophysical Properties of Hydrocarbon Mixtures Database 

(SUPERTRAPP). 1992, 3.2. 

37

7. Lemmon, E. W.; Huber, M. L.; McLinden, M. O. NIST Reference Fluid Thermodynamic and Transport 

Properties - REFPROP. 2007, 8.0. 

8. Incropera, F. P.; DeWitt, D. P. Fundamentals of Heat and Mass Transfer, Fourth Edition; John Wiley & Sons: 

New York, 2002 

9. Charbonneau, P. Version 1.2 2002, PIKAIA Homepage. 

http://www.hao.ucar.edu/Public/models/pikaia/pikaia.html (accessed 11/3, 2007). 

38

Marshall Space Flight Center NASA Academy 2009 

ECLSS Vapor Compression Distillation System and Brine Processing 

PI: Keith Parrish 

RA: Cheryl Perich 

ABSTRACT: 

In order to conserve liquid hydration resources during extended space missions, Vapor Compression Distillation 

(VCD) is used to process condensate, hygiene byproducts, and urine waste. In conjunction with proper pre and 

post-treatments used to remove any extraneous organic or volatile matter, the VCD utilizes a simulated gravity 

environment in a thermally treated rotating drum in order to remove approximately 85% of water components 

from waste solutions, producing drinkable water to a purity of approximately 97% 1 . While this technology has 

been customized for use in micro-gravitational fields, such as those experienced on the International Space 

Station, the basic water filtration and extraction is to be used in conjunction with a brine processor in attempt to 

completely dehydrate waste brine in lunar gravity for future use on the lunar outpost. 

ANALYSIS: 

I. VCD 

The basic principle behind the VCD processor (See Figure 1) involves simple phase changes in order to 

separate water from the solution waste. The system boils wastewater to produce and collect the water vapor 

before it uses a de-mister and compressor to condense the separated solution. In order to produce maximum 

efficiency the pressure inside the VCD is lowered to 0.7 psi to lower the boiling point. To begin, the 

wastewater is fed into a rotating drum (220 rpm) to form a thin film coating the walls of the drum. The liquid 

then boils off between 90-105 degrees Fahrenheit. The water vapor is then collected by a de-mister in the 

center of the drum and fed into a compressor. The collected sample is then injected onto a thin wall around the 

outside of the drum where it condenses. The water is then collected for testing, and the remaining brine is 

removed for processing and testing. Samples were collected at the beginning and end of the processing cycle 

and sent for processing. Tests included: pH/conductivity, TOC, TIC, alcohol, anion/cation, metals, acids, semivolatiles, 

volatiles, glycol, non-volatiles, aldehydes, total bacteria, and surface tension. 

Figure 1: Vapor Compression Distillation System for ECLSS 

Water Recycling. 

39

Throughout each of the run cycles, hourly processing parameters were observed and recorded in order to ensure 

proper processing conditions. These data include values for product and solution masses, as well as 

temperature, pressure, and current at specific access points in the assembly. See Figures 2 and 3 for Solution 1 

and Solution 2 for data respectively. Note the only main difference between the two Solution systems, besides 

that of normal fluctuations, is the TM01 and Prod. This directly correlates to the larger input mass of 

wastewater due to hygiene waste incorporation of Solution 2. 

(a) (b) 

(c) (d) 

Figure 2: VCD Data trends taken daily for Solution 1 taken 6/15/2009. 12.7 lb urine, 17.2 lb condensate. (a) I1= 

Current (Amps); P16=Centrifuge pressure absolute (psi); P1=Delta pressure (psi); (b) P5=Housing pressure of Fluid 

Control Pump Assembly (psi); K1=Conductivity of product fluid (µohms); (c) TM01=Mass in feed tank (lbs); 

Prod=Mass of product (lbs); (d) T1=Temp inside centrifuge (°F); SE=Specific Energy (Watt-lbs/hr). 

(a) (b) 

40

(c) (d) 

Figure 3: VCD Data trends taken daily for Solution 2 taken 7/16/09. 13.2 lb urine, 65.5 lb hygiene waste, 17.2 lb 

condensate. (a) I1= Current (Amps); P16=Centrifuge pressure absolute (psi); P1=Delta pressure (psi); (b) 

P5=Housing pressure of Fluid Control Pump Assembly (psi); K1=Conductivity of product fluid (µohms); (c) 

TM01=Mass in feed tank (lbs); Prod=Mass of product (lbs); (d) T1=Temp inside centrifuge (°F); SE=Specific 

Energy (Watt-lbs/hr). 

II. Brine Processing: 

The secondary brine processor is in design stages in order to completely remove the remaining water from the 

VCD processing via an Air Evaporation System (AES). The AES will incorporate similar thermal conditions 

to that of the VCD, yet will fully utilize the experienced lunar gravity in conjunction with vacuum technology to 

extract the remaining water without the scaling or fouling experienced by the primary VCD system. The output 

product is to be tested per standards previously set by the VCD system. In order to properly choose an 

experimental brine processing assembly, a number of potential water recycling technologies were considered 

(See Figure 4 for comparison). The listed technologies include processes considered while designing the VCD. 

These technologies with proper alterations were considered as possible brine processing facilitators. 

Figure 4: Brine Processing options considered for initial experiment 2 . 

The AES chosen was modified in order to remove the consumable wicks. The system consists of two large 

tanks connected in series with a vacuum pump, chiller, heat exchanger, and a number of data measuring devices 

including scales, thermocouples, and ammeters (See Figure 5). The initial brine filled tank is heated using a 

41

and heater adjusted to accommodate the varying brine volume and time constraints. In order to obtain the 

most efficient processing ratio, the following equations were used. 

Equations: 

Figure 5: Brine Processing Modified Air Evaporation System 

(Variables: R---cylinder radius (m); T---temperature (K); q---heat rate (W); L---length (m)) 

Equation of State 3 

Boundary Conditions for Constant Surface Temperature (Tedge) 3 

Boundary Conditions for Constant Heat Flux (q”) 3 

In this system of equations, linearity and homogeneous parameters were assumed. Using separation of 

variables, it is possible to solve the equations, given the assumption of either constant surface temperature or 

constant heat flux. Using this system, an equation analogous to the Biot number in the dimensionless solutions 

(θ* and x*) was found. The final volume and time ratio was selected based off of available equipment and total 

VCD brine output. 

FUTURE WORK: 

After the Brine Processor assembly has been completed, daily test runs will begin in accordance with VCD 

product output. Flow rate and tank size may be altered to accommodate daily VCD output and for time/product 

volume efficiency maximization. Other technology options may be used to compare efficiency and purity of 

42

output water product. Ultimately, a complete integration of the two processors (VCD and brine processors) is 

desired in order to maximize efficiency and minimize space and monetary requirements for use on the 

International Space Station and Lunar Outpost. Since the simulated gravity centrifuge of the VCD is not 

necessary in lunar gravity, a less energy consuming and vibration inducing tank is to be considered to replace 

the VCD rotating drum for use at the Lunar Outpost. The ISS brine processor would integrate the rotating drum 

technology. 

REFERENCES: 

1. Carter, L., Personal Communication: Discussion on brine dewatering technologies. E. Thomas, Editor. 

2008: Houston. 

2. Thomas, E., NASA-JSC ELS Brine Dewatering Paper. 2008: Houston. 

3. Incropera, Frank. Fundamentals of Heat and Mass Transfer. 6 th ed. New York: Wiley, 2006. 

ACKNOWLEDGEMENTS: 

Special Thanks to Dr. Frank Six, Dr. Gerald Karr, Keith J. Parrish, David A. Long, D. Layne Carter, Matthew 

W. Pruitt, NASA Academy, Teresa Shurtz, Sharon Brandt, Wisconsin Space Grant Consortium 

43

Phaeton Mast Dynamics Mechanical Systems 

Adam Harden 



Introduction 

This document is a summary report of the mechanical 

engineering internship undertaken by the author during the 

summer of 2009 at the NASA Jet Propulsion Laboratory with 

the Phaeton Mast Dynamics (PMD) Project. The primary 

goals of the summer, their method of completion, and the 

results of the approaches taken are presented herein. 

Additionally, comments on the results are presenting with a 

focus on future work to be accomplished. 

Project Overview 

Phaeton Mast Dynamics is an early career hire program at the 

Jet Propulsion Laboratory designed to establish recent 

graduates in engineering and science with solid technical 

experience at the start of their careers. This summer, I am working with Phaeton’s Mast 

Dynamics (PMD) project. The objective of PMD is to fly an array of tri-axial accelerometers on 

the upcoming NuSTAR X-Ray Observatory satellite in order to characterize the dynamics of an 

extended mast on orbit. In particular, the behavior of the mast due to motion of the spacecraft 

and thermally-induced material expansions and contractions will be studied. PMD is currently in 

the test and integration phase. 

Project Objectives 

The Phaeton Mast Dynamics project by its very nature is a quickly developing flight project with 

a fast-paced schedule and rapidly changing needs that were not always known weeks or months 

in advance. Under these circumstances it is better to describe the objectives of my internship in 

the context of my role of supporting mechanical subsystem development and testing, not by a 

single task or group of tasks designed to extend ten or so weeks. My success can be measured by 

my ability to agilely and successfully complete tasks assigned to me. This is not to say that some 

of my tasks this summer could not be known in advance, namely:. 

• Learning the UGS NX 5 CAD software; 

• Creating mechanical drawings for various purposes; 

• Planning out hardware assembly and integration; 

• Development of thermal-vacuum/vibration test procedures; 

• Design and fabrication of test fixtures; and 

• Design and/or fabrication of supporting mechanical hardware. 

In retrospect, however, much of this work was realigned or simplified in order to accommodate 

the following new objectives: 

45

• Fabrication of stereolithography models for use in subsequent mechanical fit checks; 

• Development of NuSTAR kinematics model and PMD acceleration data modeling; 

and 

• Research into electromagnetic interference mitigation techniques and adhesives 

usage. 

Project Approach 

A unique aspect of the Phaeton Mast Dynamics project, perhaps more so than other flight 

projects, is the significant amount of knowledge that must come from outside the project team in 

order to reach project milestones. The small size of the project and the accelerated timeline 

dictate this aspect. Therefore, the primary method of accomplishing any major task was as 

follows: 

• Define successful outcome of task and any relevant constraints; 

• Discuss task with JPL subject matter expert to refine constraints and gain insight into 

process; 

• Implement recommendations of expert and refine implementation with mentor; 

followed by 

• Task completion. 

In addition, tools (software, hardware, or otherwise) occasionally had to be acquired in order to 

complete the task. This would typically run concurrently with the task definition. 

Because of the wide variety of work undertaken this summer, the general approach outlined 

above cannot be refined further without introducing an example. Also note that in many cases 

my mentor or other project staff knew the relevant information, thus eliminating the need for an 

external engineer or technician. 

Project Results 

Integration and Test 

Vibration 

Preparation for the dynamics testing proceeded 

smoothly. A test fixture to interface the PMD hardware 

with the Environmental Test Laboratory dynamics 

equipment was successfully designed and fabricated. In 

addition, work on the vibration test AIDS document was 

started. No work was started on the vibration test plan, 

as this document will be produced by the 

NuSTAR/PMD dynamist. No testing was performed as 

this test is scheduled for November or December of 

2009. 

Thermal-Vacuum 

Thermal-vacuum test preparation yielded fewer results. A draft of the thermal test procedure was 

produced. In addition, test thermocouple placement was decided and, in collaboration with the 

PMD team, removable thermocouples were decided upon. No analysis was performed due to 

project schedule and pending decisions regarding the test temperature profile, though some 

46

exposure to the FEMAP with NASTRAN software was gained. In addition, I had the opportunity 

to contribute the development of thermal-vacuum test planning via numerous discussions with 

the PMD mechanical team and Bob Krylo, a JPL thermal engineer. 

PMD Kinematics Model 

Work on the PMD kinematics model was largely successful. A model was successfully 

implemented in MATLAB that is capable of generating point accelerations on the NuSTAR 

optics bench at the locations of the PMD instrument. Additionally, a second script was developed 

that is capable of using nine discrete accelerometer measurements (three locations, three 

orthogonal axes per location) to calculate body accelerations both linear and angular. Difficulties 

that still remain with the second code will be discussed in further detail below. However, after 

these issues are resolved, this code would theoretically be capable of generating data that could 

be used in other dynamics models to help simulate the behavior of the mast. 

Both codes implement standard rigid body kinematics, 4th order Runge- Kutta integration 

methods, and (where needed) cubic interpolation schemes. The input of the “body to point” 

script is capable of handing any combination of time-dependent elementary functions to describe 

X-, Y-, and Z-axis linear and angular accelerations and allows for variable step size and 

simulation duration to the extent allowed by onboard computer processing power and RAM. The 

output of the same is three plots that show point accelerations at all three PMD accelerometer 

units and a rendering showing the motion of the PMD instrument as it experiences these 

accelerations. Note that this script was written primarily to gain experience with the rigid body 

kinematics relationships and to generate test output for the script that generates body 

accelerations from point accelerations. 

47

The input of the “point to body” script is nine sets acceleration measurements (three axis at three 

locations) as well as estimated initial velocity vectors. The outout is two plots that show the 

change in body angular velocity and acceleration around the three principle axis. Simple 

modifications will also allow for the linear acceleration of the centroid to be output as a function 

of time, as well. The same variability of time step and simulation duration exists with this 

implementation, but as these parameters will be dictated by the input measurements, they are 

only relevant in the calculations themselves. In the figures below, a constant angular acceleration 

of 1 rad/sec/sec was applied about the Z-axis in the corresponding simulation. 

48

The irregularity in the plots above will be explained in further detail below, as noted above. 

Additional Tasks 

The following tasks were completed in addition to those presented above: 

• Test cables for the thermal vacuum testing were successfully designed (fabrication 

pending); 

• Stereo lithography models implementing the post-mechanical CDR changes were 

quoted, fabricated, delivered, and assembled (drawings required by the vendor were 

completed as well); 

49

• CAD models of thermal hardware, including Vishay-Dale wire-wound resistor 

survival heaters and Honeywell 706S thermal switches, were created; and 

• Material for various critical design review presentations was created. 

Remarks on Project Results 

Integration and Test 

Two major challenges were encountered when assisting with the preparations for the thermalvacuum 

and vibration testing. These challenges are outlined below. 

Vibration 

A fixture is required in order to fix the PMD instrument to the vibration test apparatus. This 

fixture took the shape of a thick plate with numerous hole patterns through which test hardware 

and PMD fasteners will be applied. As outlined in the previous progress report, the first iteration 

of this plate was returned from the vendor with numerous errors. Primarily, the hole pattern that 

allowed for a bolted interface between the test hardware and the test fixture was found to be 

partially misaligned in such a way that properly bolting down the plate was impossible. More 

minor issues with the plate involved a missing counterbore on one hole and an irregular 

counterbore on another. In addition, an issue with the design arose in that the counterbores were 

not large enough in general to fit the washers that were required. 

In order to rectify this, the manufacturing engineer that was acting as liaison between the vendor 

and PMD was contacted. The problems with the current plate were explained and the existing 

work was compared to the provided mechanical drawing. A new drawing was created that 

included an updated, larger counterbore diameter with tolerances assigned to prevent a smaller 

hole from being drilled. The updated drawing was then forwarded on to the vendor, who had 

previously agreed to replace the part at minimal cost to PMD. The following Monday, the new 

fixture arrived and was fit checked against the vibration test equipment. No issues were found 

and the fixture will be stored until it is needed to support dynamics testing. 

Thermal-Vacuum Testing 

Preparation for thermal-vacuum testing proved to be more challenging than previously thought. 

The difficultly with both running thermal analysis and writing thermal test plans was 

significantly underestimated in the initial work project for the summer. In addition, undefined 

thermal test requirements in the NuSTAR environmental requirements document (to which PMD 

must adhere) delayed developing the test plan and invalidated a significant amount of the draft 

test procedure. Coupled with pressure from pending critical design reviews, this resulted in a 

significant delay that preventing some analyses from being completed. 

PMD Kinematics Model 

The PMD kinematics model, while mostly a successful endeavor, encountered problems with 

numerical stability in the point accelerations to body accelerations calculations. The model used 

a technique developed by Dr. Jorge Angeles in a paper entitled “Computation of Rigid-Body 

Angular Acceleration From Point-Acceleration Measurements” [1]. A detailed analysis was 

conducted of the instability and compared to the technique developed in the paper. It was found 

that the technique was properly implemented in MATLAB and the source of the stability was the 

50

inversion of a particular matrix, the elements of which would grow to large values when a 

particular condition in the model developed. This is illustrated below. Notice that when the 

intermediate value �� 

is 0, the calculated angular acceleration goes to essentially 

infinity. The value of has no physical meaning in and of itself, but is a necessary 

computation in the solution. 

While it has yet to be seen to be seen if the condition exists when the model is used with 

measured data (versus pre-calculated data), based on testing it would appear that at the sampling 

rate the PMD instrument will operate at, the technique used in the model will not yield valid 

results across the entire time domain of interest. 

51

Resolutions to this particular problem involving solving the fundamental equation of motion in a 

different manner, hopefully yielding results that do not exhibit any instability; using higher order 

integration schemes to generate more accurate velocity and position data (4th order Runge-Kutta 

is currently implemented), and monitoring for the condition that causes this issue and attempting 

to correct for it directly. Additionally, the author of aforementioned paper has been contacted in 

an attempt to gain insight. 

Having researched various solutions to the computation of body accelerations “problem”, it does 

appear that the method implemented described above is similar to others that have been 

documented by the ASME, AIAA, etc. This being the case, and as PMD implements the 

minimum nine accelerometers necessary for unambiguous results, caution should be taken when 

selecting an method of data reduction in order to ensure accurate results and avoid the numerical 

issue described above. 

Where ai is the local acceleration vector at point i (where i =1,2,3), ac is the linear acceleration 

vector of the centroid of the i points, α is the angular acceleration vector, ω is the angular 

velocity vector, and ri is the position vector from the centroid to point i. Progress has been made 

on a solution of this approach and while it is not ready at the time of writing this document, it is 

possible that it will be implemented in MATLAB before the end of the internship session. 

Additional Tasks 

All other additional tasks assigned were completed without noteworthy problems arising. Given 

the typically small, short-duration nature of these tasks, this is not unexpected. 

Conclusion 

A variety of tasks were undertaken this summer in support of the Phaeton Mast Dynamics 

project. The main project goals at the start of the summer were the writing of the thermal vacuum 

test plan, developing test fixtures as necessary, assisting with acceleration data modeling, and 

assisting with other tasks as required. Ten weeks later a vibration test fixture has been fabricated, 

a kinematics model of the NuSTAR instrument has been implemented in MATLAB, stereo 

lithography models of the PMD instrument have been updated and fabricated, thermal-vacuum 

test cables have been quoted and will soon be fabricated, and information regarding 

electromagnetic interference (as it pertains to mechanical design) and flight adhesives has been 

gathered and implemented. 

The thermal-vacuum test plan was not written as previously thought due to undefined test 

requirements and underestimated complexity not appropriate for an engineering student 

managing other tasks. The thermal-vacuum test procedure, a separate document, is currently in 

draft form but will require modification after numerous elements of the test have been defined or 

52

clarified. Additionally, the PMD kinematics model still suffers from numerical instability at 

various points. An alternative solution is currently being studied and implemented. 


I have to first thank Case Bradford and Deb Sigel, my mentors, for providing a great experience 

this summer, allowing me to experience a broad variety of topics in mechanical engineering, and 

letting me participate in the Phaeton Mast Dynamics project in such a dynamic role. Thank you! 

In addition, I would like to thank the Jet Propulsion Laboratory, the California Institute of 

Technology, and the Wisconsin Space Grant Consortium for running a program that allows 

students such as me to have their ‘dream job’ for the summer. 

Finally, a thank you to Dr. William Farrow at Milwaukee School of Engineering for providing 

gracious help in preparation of this summer. 

This research was carried out at the Jet Propulsion Laboratory, California Institute of 

Technology, and was sponsored by the Wisconsin Space Grant Consortium and the National 

Aeronautics and Space Administration. 

References 

[1] Angeles, Jorge. “Computation of Rigid-Body Angular Acceleration From Point-Acceleration 

Measurements.” J. Dyn. Sys., Meas., Control. 109.2 (1987): 124-127. 

53

Aaron D. Olson 


Risk Mitigation Study for ISIM Systems Engineering, Code 443 

Background 

The James Webb Space Telescope (JWST), when it is launched in 2014, will be the 

world’s premier space telescope. The JWST will study each phase in the history of our 

Universe, ranging from the first luminous glows after the Big Bang, to the formation of 

solar systems capable of supporting life on planets like Earth, to the evolution of our own 

Solar System. The JWST will be able to do all this because of the state of the art 

technology used on the observatory. The JWST will have the largest primary mirror ever 

sent into space at 6.5 meters in diameter. It is a folding, segmented mirror, adjusted to 

shape after launch that will utilize lightweight beryllium optics to maintain its shape in its 

cold environment. To detect the very faint light sources, the telescope and its instruments 

must be kept at a very cold operating temperature under 50 K. To achieve this 

temperature passively, JWST has a tennis court sized sunshield that will keep the optical 

element and science instruments in a stable and cold environment, away from the Sun, 

Earth and Moon. Despite the innovative mirror and sunshield, the Integrated Science 

Instrument Module (ISIM) is the main payload of the observatory. The ISIM’s four 

instruments, aligned along the chief ray of the telescope, receive light from the mirrors 

and process it to detect distant cosmic sources. The Mid-Infrared Instrument (MIRI), 

Near-Infrared Camera (NIRCam), Near-Infrared Spectrograph (NIRSpec) and Fine 

Guidance Sensor (FGS) of the ISIM could not do their jobs without the state of the art 

Infrared detectors that convert light photons into an electrical signal that can be 

processed. 

55

A. Abstract 

For hardware to be flight certified; it’s verification must either be by test, a 

precise analysis, significant heritage (history of similar hardware) or a combination of 

these to infer that the hardware will perform in its intended environment. In general, of 

the three methods of certification; testing of the hardware is the preferred over analysis 

and heritage. The infrared detectors of the JWST are part of a larger Sensor Chip 

Assembly (SCA) which in turn is a part of a larger Focal Plane Assembly (FPA). The 

FPA structure is the hardware that needs to be shock qualified to be certified for space 

missions. Shock certification for the FPA involves more risk than for other hardware. 

The qualification methods do not provide complete assurance because the shock 

environment is difficult to quantify. Furthermore these efforts do not always simulate an 

accurate shock response environment. This results in the margin between the acceptance 

and qualification level to be large. In many cases the qualification level is twice the 

acceptance. These levels have also been known to change in the course of a mission. 

Shock testing the FPAs is costly because it involves replicating high frequency 

environments. This testing involves tailored setups that require significant effort to tune. 

Avoiding the risk of testing is preferred because of the testing setup and the uncertainty 

in shock levels. Precise structural analysis of the FPA involves risk as well because it 

relies on the same uncertainty in shock. As far as heritage, prior to this document, 

detailed FPA heritage has not been compiled. 

56

B. Microshutter History of Risk in Testing 

The Microshutters are a brand new technology that was created for the NIRSpec 

instrument on the JWST. This technology provides the most precise in space 

spectroscopy to date. The Microshutters are an array of individual micron scale cells 

with lids that open and close with a magnetic field in order to survey the spectra of 100 

objects at a time as well as block out the other portions in space. 

The Microshutter Assembly (MSA) is the enclosure for the array that must be 

flight certified for shock. The Microshutters are new technology with no flight heritage 

and because of this it was determined that the MSA be tested for shock certification. 

Similar to the FPAs, the MSA is on the Science Instrument (SI) level of the ISIM. The 

MSA was shocked to the SI specification. The Microshutters failed their initial testing as 

the array was severely fractured. This failure had several costly repercussions. Broken 

hardware, the need for more testing and a delay in the project schedule were the results. 

It was confirmed that the testing did not simulate the shock that the MSA will really 

encounter. The predicted levels of the shock that were given for the MSA were much 

higher than they should have been. Since the initial failure of the MSA, more precise and 

expensive tests were done at facilities in Europe and the U.S. until the MSA could finally 

be qualified for shock. The entire ordeal of qualifying the MSA demonstrates the risks of 

direct testing of shock for structures at the SI level. 

The Microshutter Assembly (MSA) Individual microscopic shutters 

C. Testing deferral to Observatory level 

From previous examples, there is a definite risk for shock testing at the SI level. 

To avoid this, testing can be deferred to the next larger level. The observatory level is the 

next level for shock testing the FPAs and the decision to defer shock testing for FPAs is 

based on an assessment of the strength and robustness of the assembly. Dominant 

assembly resonance is one area of consideration for the sturdiness of a structure. A Jet 

Propulsion Laboratory (JPL), analysis of the predicted shock on the FPMs for MIRI 

(Focal Plane Module and Focal Plane Assembly are synonyms), from the MIRI 

Requirements Document (JWST-IRD-000782 Revision K), showed that the response of 

the FPM and the internal detector assembly, were not sensitive to the shock levels given 

from the Observatory Base requirements Document (JWST OBA RD). As a result, JPL 

concluded that a SI level shock test of the FPM was not necessary and that shock testing 

57

should be performed at the JWST observatory level. This approach has validity; however 

it moves the risk of failure to a future test. This deferral incurs the chance of failure, at a 

point in the project that would be very costly. Consequently, the JWST OBA RD still has 

the requirement that the FPAs be tested and qualified for flight. Strong evidence for 

deferment is needed to warrant the waiver of these tests. 

D. Risk Mitigation from Shock Heritage 

The NIRSpec instrument has had its fair share of issues with shock certification. 

The MSA has demonstrated that. The assessments of JPL and outside sources like the 

Space Systems Division of ITT suggest that the FPAs of JWST are not shock susceptible. 

The past uncertainty of shock levels and the risk of delaying the project with FPA failure 

at the observatory level strongly suggests that NIRSpec and the entire ISIM need more 

than analysis to defer testing. FPA heritage for shock can assure that shock testing of the 

JWST detectors can be deferred based on previous structures’ similarity and the 

comparison of their shock requirements. JWST is the first to use photonic detectors of 

this precision for infrared astronomy in outer space. Future space telescopes and 

missions with similar detectors could use this heritage as well for deferment assurance. 

[18] 

Nuclear Spectroscopic Telescope Array (NuSTAR) Space Interferometry Mission (SIM) Lite Astrometric Observatory 

Terrestrial Planet Finder (TPF) Darwin Mission Telescope (1 of 4) 

Discussion 

58

Shock 

Mechanical shock, a physical shock, is the form of shock that will affect detector 

structures. The definition of mechanical shock is the response of a structure to high 

frequency, high magnitude stress waves that propagate throughout the structure as a 

result of an explosive event. The nature of shock is transient, or very brief. The response 

of the structure from the stress waves or shock pulse is a sudden acceleration. This 

acceleration is measured with an accelerometer, which measures a shock pulse as a plot 

of acceleration, in terms of gs, versus time. 

A. Quantification of Shock 

A shock response spectrum (SRS) is the method used for representing a 

mechanical shock event. A SRS is a graphical representation of shock in terms of how a 

single degree of freedom (SDOF) system responds to the shock pulse at a defined 

amplification. It is a graph of the peak acceleration response of SDOF systems at each of 

their own natural frequency. 

An SRS is generated from a shock waveform using the following process: 

1. Pick a damping ratio for your SRS to be based on and assume a hypothetical Single 

Degree of Freedom System (SDOF), with a damped natural frequency of x Hz 

2. Calculate the maximum instantaneous absolute acceleration experienced by the mass 

element during (or after) exposure to the shock in question. 

3. Plot this in g's (g's are standard, but pick any unit of acceleration you want) against the 

frequency (x) of the hypothetical system. 

4. Repeat steps 2 and 3 for other values of x, say logarithmically up to 1000x. 

59

B. Attenuation and Uncertainty 

Shock attenuates when it propagates through a material. Attenuation, the 

reduction of shock, is critical in calculating the peak levels of shock that a particular 

piece of a larger structure will be exposed to. Attenuation analysis of shock is based on 

the length of the dispersion path through the structure and the number and types of 

structural joints in that structure. According to a NASA Technical Standards for Pyro- 

Shock Requirements Document (RD 23) there is a reduction in shock magnitudes in the 

realm of 20 to 75 percent from propagation through structural joints. This percentage 

range depends on the material of the joint and structure, and the path of transmission 

through the joint. This shows the uncertainty in quantifying shock, creating large 

margins between acceptance and qualification SRS. 

C. Methods to Assess Shock 

There are general methods of assessing the risk of damage from shock. Coverage 

of shock using random vibration loads and the two methods for structural and electrical 

components from the ESA Shock Handbook are widely accepted. The 0.8f rule for 

structural components is a simple comparison of a 0.8 sloped limit line, with respect to 

frequency, where structures are not at risk if their SRS lie below the limit. The 

6db/octave rule for electronics shares a similar concept, except it has a smaller slope of 

6db/octave and tops out at 500g for 2 kHz. The other method of assessing shock allows 

one to compare the shock specification to random vibration. This is the most accepted 

method among the space community. This approach converts random vibration loads 

into a random response spectrum or RRS that can be compared to a SRS. This is done 

with the Miles formula, which is a simplified approximation of an RRS. The equation 

uses Q; an amplification factor which is typically 10 or 25, fo; the frequency at which the 

acceleration is calculated and W (fo); the value in g 2 /Hz of the Power Spectral Density 

(PSD or random specification) at the frequency fo in the calculations. Shock will not be 

an issue if the RRS covers the SRS levels. By using the formula below, the frequency 

values for random vibration loads and the corresponding PSD, RRS can be compared to 

SRS to asses shock and determine if direct testing is needed. 

RRSMiles (fo) = 3 x SQRT [π/2 * Q * fo * W (fo)] 

The James Webb Space Telescope Detectors 

JWST uses new detectors that are top of the line with very low noise and darkcurrent, 

and high quantum efficiency. These qualities measure the precision, accuracy 

and reliability of detectors, respectively. There are a total of 18 infrared detectors 

between the four instruments. Of these, 15 are HAWAII 2RG detector arrays from 

Teledyne Imaging Sensors and the other three are SB-375 Si:As detector arrays from 

Raytheon. HAWAII is an acronym for HgCdTe Astronomical Wide Area Infrared 

Imager and 2RG denotes the size (2048x2028 pixel), and that the array has reference and 

guidance pixels included. H2RG detectors are used in the Near-Infrared Camera 

60

(NIRCam), Near-Infrared Spectrograph (NIRSpec) and the Fine Guidance Sensor (FGS). 

The Silicon Arsenic detectors are 1024x1024 pixel arrays with reference and guidance as 

well, that are only used in the Mid-Infrared Instrument (MIRI). The detectors on 

NIRCam, NIRSpec and FGS are used for imaging, spectroscopy, object tracking and 

narrow band imaging respectively. The detectors on MIRI are used for mid-infrared 

imaging and spectroscopy. The 18 detector arrays are the most impressive, however not 

the only part of the Focal Plane Assembly (FPA). The FPAs are positioned to capture 

light while being mounted on the base of their instruments, which also sit on top of the 

ISIM base by way of kinematic mounts. [2][6] 

A. FPA Structure 

The FPAs for each instrument consist of the Sensor Chip Assembly (SCA) along 

with its support structure and housing enclosure. The SCA is made up of the infrared 

detector material, a Silicon readout integrated circuit (ROIC), and a SCA mount 

connecting to a flex cable link. Microscopic Indium bumps connect the infrared material 

to the ROIC. The Silicon ROIC is responsible for charge to voltage conversion, signal 

transfer through the circuit, and digitization of the signal. The flex cable connects to the 

ROIC and sends out the digitized signal to be processed. The SCA mount supports the 

ROIC, infrared array and flex cable. Above the SCA is a stray light mask which protects 

the detector array from unwanted sources of light. The SCA is epoxied to an insert or 

interface plate below it, this provides electrical isolation as well as precision alignment 

for the SCA. The interface plate rests on top of the mosaic plate to provide a sturdy 

connection of the SCA to lower supporting structure. The mosaic plate by way of flexure 

struts connects to the baseplate. The baseplate is the support of the assembly housing 

and interface to the instrument. The FPAs in the ISIM have different configurations, 

some with 1 SCA and others with 2 or 4. The FPAs with 4 SCAs are set as 2x2 mosaics. 

[2][6] 

61

Detector Array and ROIC SCA Mount 

Levels of FPA Structure 

SCA with Flex Cable 

62

B. FPA Material Comparison 

The FPAs of JWST all share the same core components. Each FPA, however, is 

its own version of the shared design. These versions utilize different materials for some 

components. The ROIC for all the FPAs is made of silicon because it is an integrated 

circuit similar to one used for a computer chip. The SCA mount is made of a TMZ 

Molybdenum hybrid for all the instruments, except MIRI which uses a copper alloy. 

These were chosen because they provide good coefficient of thermal expansion (CTE) 

match with the ROIC. The interface plate between the SCA mount and mosaic plate is 

also made of Molybdenum for similar purposes. The three flexure struts, which also 

serve as standoffs, that connect the mosaic plate to the baseplate are made of Titanium. 

The struts accommodate thermal mismatches and provide active temperature control. 

The baseplate is made of Titanium, Molybdenum for MIRI, and provides a strong base 

for the assembly. The assembly housing for radiation shielding and shadow mask stray 

light mitigation are unique to each FPA. [3][11] 

NIRSpec FPA NIRCam FPA 

MIRI FPA FGS FPA [16][17][8][4] 

63

Detector Shock Environment 

The levels of shock on the observatory are specified as the shock environment. 

This environment comes from the different sources of shock that affect the JWST. These 

levels of shock are from the launch and the internal events that affect the observatory 

while it’s in space. The shock from launch is from the effects of launch vehicle and 

fairing separation. The internal events that self-induce shock on to the observatory are 

from the launch restraint mechanisms (LRMs) that deploy the JWST. There are specified 

acceptance and qualification SRS for the Spacecraft, Optical Telescope Element (OTE), 

ISIM and their interfaces. The ISIM/OTE interface, referred to as the Region 1 interface 

is the environment that is of interest to the detectors. The shock environment will occur 

at approximately 200K when the observatory is in orbit. ISIM/OTE Interface shock 

specifications have recently been updated by Northrop Grumman in ECR 155. [14] 

A. Shock from Launch 

The shock from launch comes from the Ariane 5 interface; this includes PAF and 

fairing separation. Arianespace has developed a new fairing and clampband separation 

device that have made the shock input to the spacecraft negligible. These new 

developments were demonstrated by several on-ground tests. The shock levels from 

launch are now enveloped by the shock from the observatory internal events. [11][5] 

Freq (Hz) 

Ariane 5 Interface Shock 

Acceptance Qualification Acceptance Qualification 

Freq (Hz) 

SRS (g) SRS (g) SRS (g) SRS (g) 

100 20 40 100 10 20 

1000 111 222 3000 700 1400 

10000 80 160 10000 700 1400 

B. Shock from Internal Events 

Updated ISIM Shock Specifications (ECR 155) 

Launch restraint mechanisms (LRMs) used to release the observatory deployables 

utilize non-explosive actuators to release the appendages, allowing them to deploy to 

their on-orbit positions. The LRMs chosen produce a very low shock impulse into the 

attached hardware compared to heritage pyrotechnic devices. However, significant shock 

impulses occur when the stored strain energy of the restrained hardware is released. The 

Primary mirror wing LRMs are the closest source to ISIM. The detectors are on the 

Science Instrument (SI) level of the observatory, thus the shock from the LRMs is 

64 

[14] 

Observatory LRM Shock

attenuated from its source. The attenuated shock levels for the detectors are in the table 

below along with a graph comparing current acceptance to the previous. [11][13] 

Updated SI Acceptance Shock Specification 

Freq (Hz) 

A. Previous Projects 

Acceptance Qualification 

SRS (g) SRS (g) 

100 10 20 

1700 350 700 

10000 350 700 

SRS (g) 

1000 

100 

SI Acceptance Shock Levels 

[1] 

Detector Heritage 

Stored Strain Release LRMs 

10 

100 1000 

Freq (Hz) 

10000 

Updated Specification Previous Specification 

Teledyne Sensors and Raytheon are producing the deliverable FPAs for JWST. 

These companies are the main detector producers and have significant experience in 

building the assemblies. They have built FPA packages for the Spitzer Space Telescope, 

Deep Impact Comet mission, Hubble Space Telescope (HST), Mars Reconnaissance 

65

Orbiter (MRO), Moon Mineralogy Mapper and several classified missions. These 

missions have all shared the same level by level assembly design as JWST. [2][6] 

Wide Field Camera 3 FPA on HST CRISM FPA on MRO 

B. Establish Similarity 

The FPAs used in these previous projects span a large range of the kinds of 

missions being done. They are all fairly recent projects and have the same basic design 

for the detector assembly. This and talk with detector specialists like Murzy Jhabvala 

lead to the conclusion that FPAs, manufactured by the main producers for NASA, are not 

only similar, but structurally the same. The shortwave NIRCam FPA is a good example 

of the structure that is shared among today’s detector assembly. A typical FPA mount 

refers to the buildup connecting the baseplate to the struts, mosaic plate, insert and finally 

to the SCA and detector array. [16] 

SW NIRCam FPA Mount Assembly: Typical FPA Mount 

66

C. Structural Fortitude 

Detector assemblies have been known to be sturdy structures. Previous designs 

from Teledyne have had extensive thermal, structural, electrical, and optical modeling 

analyses by Ball Aerospace. The SCA and flex cable have been shown to be resistant of 

high cycle fatigue for 1000 thermal cycles. The thermal requirements for the SCA were 

thought to be the most difficult for the component, yet it passed with a considerable 

margin. The SCA is thought to be a robust component for spaceflight. The struts or 

standoffs are also a stout part of the FPA. Their focus is to provide thermal isolation, but 

they are very good in mechanical compliance between the molybdenum mosaic plate and 

titanium baseplate. The struts are sized for optimum dynamic stiffness as well. [6][7] 

D. Previous Shock Levels 

A database of previous acceptance and qualification shock levels for FPAs would 

have great use. This database would be a graph of collected SRS from past missions. 

This could be used by projects like JWST to compare qualification levels of their project 

to successful ones of the past. This comparison could lead to complete deferment of FPA 

structures to shock testing or just a waiver to be used by one project. Either way, this 

database will be a risk and cost saving tool in shock certification. 

Results 

Analysis of the detectors from several sources states that testing should be 

deferred to the observatory level. The FPA structure is believed to be a strong design 

with several components that support the idea that the design isn’t susceptible to shock. 

Heritage of the structure shows that the design is not unique to JWST, even though the 

ISIM’s infrared arrays have never been flown. Additionally, there have not been 

documented shock failures of FPA structures. With the new Ariane fairing and LRM 

devices lowering shock levels, previous projects infer that shock testing can be deferred. 

A compilation of shock levels from previous missions will augment this mitigation study, 

References 

1. NASA GSFC. ISIM SI Shock Update 05/22. s.l. : JWST, 2009. 

2. Hall, Donald N. B. HgCdTe Optical & Infrared Focal Plane Array Development in the Next Decade. s.l. : 

Institute of Astronomy, University of Hawaii. 

3. JPL MIRI FPS Team. Focal Plane System (FPS) Critical Design Review (CDR). JPL D-31387 DRD PM-20. 

Pasadena, CA : NASA JPL, 2006. 

67

4. NASA JPL. MIRI FOCAL PLANE SYSTEM (FPS) QUALIFICATION. JWST-PLAN-011762. Pasadena, CA : 

JPL, 2008. 

5. NASA Detector Requirements. Marshall, Cheryl. Los Angeles : NASA GSFC, 2002. 

6. Teledyne Imaging Sensors: Infrared imaging technologies. James W. Beletic, Richard Blank. Camarillo, 

CA : Society of Photo-Optical Instrumentation Engineers, 2008. 

7. Kaufman, Dan. MSS Shock Prediction Status Report. JWST-RPT-011728. s.l. : NASA GSFC, 2009. 

8. Garnett, Dr. James. JWST NIRSpec Focal Plane Arrays. Capabilities Paper for the JWST NIRSpec Focal 

Plane Arrays. 2003. 

9. NASA JPL. Focal Plane System Critical Design review. JPL D-31387 DRD PM-20. Pasedena, CA : s.n., 

2006. 

10. Young, Eric. NIRCam FPA Critical Design Review. s.l. : University of Arizona, 2006. JWST-RVW- 

006193. 

11. Salvignol, J-C. SRE-PJ/12793/JCS Shock NRI. SRE-PJ/12793/JCS. s.l. : ESA. 

12. JWST ISIM. Fine Guidance Sensor Instrument Requirements Document. JWST-IRD-000783. s.l. : NASA 

GSFC, 2009. 

13. Gomez, Hector. Observatory Shock Environments at Cryogenic Temperatures. 08-JWST-0275D. s.l. : 

NASA GSFC, 2008. 

14. Northrop Grumman. James Webb Space Telescope (JWST) Engineering Change Request 155 

Revision C. JWST-ECR-0155 Rev. C. s.l. : NASA GSFC, 2009. 

15. JWST ISIM. MIRI IRD. JWST-IRD-000782 Rev. K. s.l. : NASA GSFC, 2008. 

16. —. Near-Infrared Camera Interface Requirements Document. JWST-IRD-003272 Rev. J. s.l. : NASA 

GSFC, 2008. 

17. —. Near-Infrared Spectrograph Interface Requirements Document. JWST-IRD-000781. s.l. : NASA 

GSFC, 2009. 

18. ITT. JWST NIRSpec FPA Math Model & Dynamic Analysis. FPA SHOCK T1600-0063. s.l. : NASA GSFC, 

2009. 

68


Part Five 

Atmospheric Science

Is Lake Superior a significant source of atmospheric carbon dioxide? 

Val Bennington 1 , Galen A. McKinley, Nazan Atilla 

Department of Atmospheric and Oceanic Sciences, University of Wisconsin, Madison 

Abstract 

Lake Superior is the largest freshwater lake in the world by surface area, yet very little is known 

about its carbon cycle. The air-lake flux of carbon dioxide (CO2) influences nearby observations 

of the terrestrial carbon cycle, and constraining the air-lake flux would improve understanding of 

terrestrial, lake, and regional carbon cycling. Previously, large lakes were believed to be a large 

source of carbon to the atmosphere, but direct measurements and indirect estimates of open-lake 

partial pressure of CO2 (pCO2) in Lake Superior do not support a large lake-wide efflux. We 

used a hydrodynamic model (MITgcm) coupled to an ecosystem module to understand the 

seasonal cycle of lake pCO2 and estimate the open lake air-lake CO2 flux. The model is able to 

reasonably capture lake temperature, circulation, productivity, and pCO2. We conclude the open 

lake is not a significant source of CO2, but terrestrial inputs of carbon likely support a near-shore 

efflux. River and runoff inputs of carbon must be added to estimate the lake-wide flux. 

Introduction 

Lakes are an important component to a regional carbon budget, because they process and respire 

terrestrial carbon the runs off from the surrounding watershed. Large boreal lakes of the world 

were previously considered to be significant sources of atmospheric carbon [Alin et al., 2007; 

Cole et al., 2004], and Lake Superior is the largest lake in the world by surface area, with a 

volume greater than all other Great Lakes combined. Field studies done on Lake Superior 

indicate the lake is heterotrophic except during a brief period during summer [Urban et al., 2004; 

Russ et al., 2004] and a net source of CO2 to the atmosphere. However, the most comprehensive 

carbon budget put together for Lake Superior [Urban et al., 2005, 2006a] is out of balance, likely 

due to extrapolating from a small region by the Keweenaw Peninsula over the whole lake and 

over the entire year. 

In situ measurements are subject to both spatial and temporal bias in Lake Superior. There are 

almost no wintertime observations of temperature or chemistry, due to the extreme conditions. 

The EPA measures pH, alkalinity, and nutrients twice yearly (April, August) at 19 open lake 

stations. Some stations are visited at night when the ecosystem would be respiring, and others 

during the day. All stations are visited twice a year, and this sampling misses blooms. Significant 

changes in nutrients and pCO2 from one year to the next could be caused by small changes in 

bloom timing or the time of day the station was visited. To compound the problems of 

understanding a yearly air-lake flux from observations taken twice yearly, estimating pCO2 from 

measurements of pH and alkalinity in freshwater is highly uncertain. pH electrodes have an 

average bias of -0.13 pH units (an uncertainty of +100-2000µatm) in freshwater [French et al., 

2001], but the bias is unknown for any of the EPA measurements (or other field measurements). 

1 The author would like to acknowledge the Wisconsin Space Grant Consortium and the National 

Science Foundation (OCE-0628560 and OCE-0628545) for financial support of this project. 

1

EPA protocol also requires the water sample be heated to 25 C before measuring the pH and 

alkalinity. Heating does not change the alkalinity of water but can drastically change the pH, so 

pH must first be corrected to in situ lake temperature to estimate lake pCO2. Previous studies 

[Alin et al., 2007; Urban et al., 2005] have been unaware of this protocol, leading to large 

overestimates of springtime lake pCO2 (hundreds of µatm) and moderate overestimates in 

August (~50 µatm). Such extreme changes in estimates of lake pCO2 can completely change the 

understanding of whether the lake is a significant source of carbon dioxide or a sink during 

certain times of the year. 

There are two time series of direct measurements of Lake Superior pCO2. These observations are 

limited in space, because they measure pCO2 at one point in the lake, but you can see daily and 

hourly changes in pCO2 at the point location. Direct observations of lake pCO2 suggest that the 

open lake is a source of carbon dioxide during spring and a sink during the bloom. During the 

cold months, water and respired carbon dioxide are able to mix to the surface and efflux. When 

the lake warms, biological productivity removes carbon dioxide from the surface waters, leading 

to reduced pCO2 [Atilla et al., 2009]. 

Primary productivity and carbon fluxes are orders of magnitude different in the ocean coastal 

region than in the open ocean, and significant spatial variability in Lake Superior in satellite 

chlorophyll are observed as well. However, the majority of physical and chemical observations 

in Lake Superior have been taken near shore along the Keweenaw Peninsula or near Duluth in 

the western arm of the lake. Scientists must extrapolate over the largest lake in the world in both 

space and time and cannot be expected to balance the carbon budget or capture the expected 

spatial variability. In an attempt to balance the lake carbon budget and understand the 

mechanisms of the air-lake CO2 flux, we used a coupled model to close the open-lake carbon 

budget, estimate lake efflux, and understand mechanisms controlling the efflux. We simulated 

lake conditions for 2001, because both indirect estimates and direct observations of lake pCO2 

exist for 2001. 

Methods 

An ecosystem model of Lake Superior was coupled to a three dimensional hydrodynamic model 

of the lake to estimate the air-lake flux of carbon dioxide in the absence of any terrestrial carbon 

inputs. The couple model simulated lake circulation, productivity, and air-lake gas exchange for 

2001. 

Physical Model. We used the MIT general ocean circulation model [Marshall et al., 1997a, 

1997b] configured to the bathymetry of Lake Superior [Schwab and Seller, 1996] with a uniform 

horizontal resolution of 10 x 10 kilometers. The model uses a z-coordinate system of 29 vertical 

layers. The top 50 meters have finest vertical resolution, with layer thicknesses of 5 meters. 

Vertical resolution gradually becomes coarser with depth to a thickness of 33.8 meters at 322 

meters depth. The Smagorinsky [1963] horizontal diffusivity scheme and the KPP vertical 

mixing scheme [Large et al., 1994] simulate the effects of sub-grid scale processes. The time 

step of numerical integration is 200 seconds. 

A bulk formula is used to calculate momentum exchange and net heat fluxes between the 

atmosphere and lake. Ice cover data from NOAA [Assel, 2003] is applied as a fractional mask to 

2

each grid cell at daily resolution. The ice mask alters the exchange of heat, gas and momentum 

between the lake and the atmosphere. Temporal increases/decreases in fractional ice coverage 

create a heat flux to/from the lake. For this purpose, lake ice is assumed to have a constant 

thickness of 0.25 meters when present. 

Ecosystem Model. The ecosystem model is that of Dutkiewicz et al. [2005] and Bennington et 

al. [2009] updated so phosphorous is the only limiting nutrient. It includes the cycling of carbon, 

alkalinity, and oxygen. The model tracks the states of phosphorous and carbon as they pass from 

dissolved inorganic forms to phytoplankton, to zooplankton, and to detritus in both dissolved and 

sinking particle forms. A schematic of the ecosystem is presented in Figure 1. 

Figure 1. Schematic of the model ecosystem adapted from Dutkiewicz et al., [2005]. 

Two size classes of phytoplankton are included, and phytoplankton growth is limited by 

phosphorous availability, lake temperature, and light. Phytoplankton are assumed to assimilate 

carbon into biomass at a fixed C:P ratio of 200 [Urban et al., 2006b]. Particulate organic carbon 

and phosphorous remineralize at a maximum rate of 5 d -1 to include the effects of the microbial 

loop and sink 0.5 m/day [Chai and Urban, 2004]. Growth and remineralization rates are 

modified by temperature according to Eppley [1974]. One class of zooplankton preys upon both 

phytoplankton classes, but assimilates only a portion of what it grazes into its biomass. To 

include the effects of the microbial loop without an explicit tracer for bacteria, a significant 

portion of the lake’s ecosystem [Biddinda et al., 2001], 95% of deceased material goes directly 

into dissolved form and rapidly remineralizes. Chlorophyll is parameterized according to the 

mass of phytoplankton present and light availability during the last 24 hours. Photosynthetically 

active radiation (PAR) is assumed to be 45% of shortwave radiation [Frouin and Pinker, 1995]. 

The exchange of CO2 between the lake and atmosphere is dependent on the difference in partial 

pressures across the air-lake interface and a piston velocity. We used the parameterization of 

Wanninkhof [1992] to include the effects of wind speed on the air-lake CO2 flux. For coupled 

model runs, the model was spun up for three years so that the lake may outgas excess carbon 

from the initial conditions. 

Model Forcing. The physical model was forced with hourly winds, downward short and long 

wave radiation at the surface, air temperature, atmospheric pressure, and specific humidity 

created by interpolating meteorological observations at three open lake buoys and from nearby 

3

land stations over the lake [Schwab and Bedford, 1994; Hsu, 1986]. The concentration of 

atmospheric CO2 above the lake was assumed equal to concentrations at the nearby WLEF tall 

tower and was provided at monthly resolution by Ankur Desai [Desai et al., 2008]. 

Results 

Lake Circulation. Few observations of large-scale circulation of Lake Superior exits. Most 

current observations are restricted to the Keweenaw Current, which is not adequately captured at 

such a coarse resolution. The mean circulation of Lake Superior is unknown, but the the gyres in 

the modeled circulation agree with the previous modeling study of Lam [1978]. We expect year- 

to-year variability in flow driven by changes in wind direction and speed, so it is interesting that 

most of the modeled flows also agree with observed winter and summertime integrated flow 

from one year during the late 1960s [Beletsky et al., 1999] (Figure 2). 

Figure 2. Current observations from Beletsky et al., [1999] (left) during summer of 1967 (top) and winter of 1966- 

1967 (bottom). Model column-integrated currents (right) during summer (top) and winter (bottom). Bold arrows on 

model depict major flow directions in common between the model (2001) and observations (1967). 

Lake Temperature. The physical model captures the seasonal cycle of lake surface temperature 

reasonably well (Figure 3a). The seasonal cycle of temperature in the deepest portion of the lake 

is 2 C too small in the model, and the seasonal cycle along the southern coast is 2 C too large. 

Model surface temperatures are lower than satellite temperatures in the open lake during 

summer, but the satellite detects skin temperature and the model predicts temperature for the top 

5 meters. Phytoplankton need light and nutrients for growth. Stratification keep phytoplankton 

from mixing out of the sunlight, and lake surface temperatures reach 4 C and begin stratifying at 

the three NOAA buoys at the observed time (Figure 3b). Model mixed layer depths also agree 

with EPA temperature profiles in August (not shown). 

4

Figure 3. (a) Seasonal cycle of lake surface temperature (JAS-FMA) from AVHRR (top) and in the model (bottom). 

(b) Daily surface temperatures (April-Dec) at the three NOAA buoys. 

Lake Productivity and Oxygen Consumption. Model primary production of 78 gC/m 2 /yr is 

on the low end of estimates of Great Lakes productivity (60-300 gC/m 2 /yr) [Vollenweider et al., 

1974], but within the range of productivity estimated for Lake Superior [Urban et al., 2005]. 

Similar to the ocean, productivity is highest along the coasts (Figure 4). The data also show 

hotspots of productivity at the mouths of the St. Louis and Ontonagon Rivers, but nutrient supply 

from rivers is neglected in the model. The model also captures the onshore-offshore progression 

of the bloom seen in satellite images of chlorophyll and noted by observers. The algorithm for 

retrieving Lake Superior chlorophyll is being developed by Colleen Mouw at the SSEC in 

Madison, Wisconsin. Satellite images are preliminary and provided by Colleen Mouw. 

During the KITES project in the late 1990s, consumption of oxygen in the hypolimnion was 

measured between July and October at three locations off the Keweenaw Peninsula. Model 

hypolimnetic oxygen consumption at the three ADCP locations was 0.13-0.15 µgO/L/hr, on the 

lower end of observed values (0.1- 1.3 ugO/L/hr) [Urban et al., 2004]. This difference is 

reasonable considering the model does not include respiration of river runoff or the background 

terrestrial carbon in the lake. 

5

Figure 4. Chlorophyll estimated from SeaWiFS imagery (personal communication with Colleen Mouw) and model 

chlorophyll (mg Chl / m 3 ). 

Lake pCO2 and CO2 Flux. The difference in the partial pressures of carbon dioxide between 

the atmosphere and the lake surface drive the air-lake CO2 gas exchange. The partial pressure of 

CO2 (pCO2) has been directly measured only twice in Lake Superior. Direct measurements of 

pCO2 indicate the lake is under-saturated and a sink of carbon dioxide during the summer bloom. 

pCO2 is indirectly estimated by measuring lake pH and alkalinity. Measuring pH with an 

electrode has a significant bias (-0.13) in freshwater [French et al., 2001] that introduces an 

uncertainty in pCO2 up to hundreds of µatm (if bias is present, actual pCO2 would be 

significantly lower than shown) during summer. This can completely change the scientific 

understanding of whether the lake is a sink or source of atmospheric CO2. To further complicate 

scientific understanding of the lake’s pCO2, the EPA measurements of pH are taken after the lake 

water is heated to 25 C. When estimating pCO2 from EPA measurements of pH and alkalinity, 

the pH must be corrected to the in situ temperature. The model pCO2 compares extremely well to 

directly measured pCO2 provided by M. Baehr (Figure 5). The model would not be expected to 

capture the extreme magnitude of the CO2 drawdown seen in the data, as the model is a 10km by 

10km region. The model pCO2 is also reasonable compared to indirect estimates of pCO2 from 

the nineteen stations visited by the EPA each year in April and August (Figure 5). Uncertainties 

of indirect pCO2 estimates are unknown, as the pH bias is unknown. 

6

Figure 5. Model pCO2 at the surface (thick black) and at 12m (SAMI, thin grey) compared to direct measurements 

of pCO2 (thick grey) during 2001. Indirect estimates of pCO2 from EPA pH and alkalinity at the nearest EPA station 

are shown in shapes. Note that the EPA station was visited overnight. 

Model pCO2 agrees with the open-lake direct observations and indirect estimates of pCO2 and 

estimates a lake-wide efflux of only 0.03 TgC/yr. For comparison, Urban et al. [2005] estimated 

the lake efflux to be 3 TgC/yr from indirect estimates of pCO2 taken near the coastline along the 

Keweenaw Peninsula. Significant spatial variability in productivity, pCO2, and carbon fluxes 

exist in the model, suggesting that extrapolating spatially limited observations can be 

problematic. The model is able to provide an estimate of the daily air-lake flux for all of the lake 

to fill in observational gaps. 

Discussion and Conclusions 

The model adequately simulates lake temperature, productivity and pCO2. The open lake is 

supersaturated with carbon during winter and spring and under-saturated or near equilibrium 

during summer. Lake Superior mixes to the bottom twice yearly, and no water mass within the 

lake is able to store carbon dioxide away from the atmosphere for long periods. The open lake is 

not a significant source of carbon dioxide (0.03 TgC/yr), and the near shore waters would be 

unable to maintain super-saturation without continued inputs of carbon. 

Nearshore observations along the Keweenaw Current [Urban et al., 2005] suggest the lake is 

heterotrophic and a source of carbon to the atmosphere in the near shore zone. The magnitude of 

this source is extremely uncertain, because all pCO2 estimates were indirectly calculated from 

electrode-measured pH, temperature, and alkalinity. Direct measurements of near-shore pCO2 

are needed for a true understanding of the lake’s carbon cycle. When the model can agree with 

both near shore and open lake pCO2 observations and estimates, the whole lake carbon budget 

may be understood. The model does not include terrestrial inputs of carbon. Due to its long 

residence time (176 years [Quinn, 1992]), 90% of the allochthonous carbon is respired within the 

lake. The terrestrial carbon must be respired mainly in the near shore zone; otherwise, model 

pCO2 in the offshore regions would be low compared to observations. The next step is to add 

river and runoff inputs of terrestrial carbon to the model. 

7

The understanding of Lake Superior is seriously impacted by the scarcity of in situ observations. 

There are no observations of winter lake temperature below the surface or nutrients to evaluate 

the model results. Open lake stations are only visited twice yearly by the EPA and could easily 

miss a significant bloom or mixing event. The data from these twice-yearly cruises must be 

interpreted with care, since inter-annual variability in lake stratification [Austin and Colman, 

2007] and bloom onset exist. Satellite observations provide the only wintertime temperature 

observations. Algorithms for retrieving lake chlorophyll from satellite observations are currently 

being developed [Schuchman et al., 2006; Colleen Mouw, personal communication], and as part 

of this project, we are working together with the algorithm developers (Colleen Mouw, SSEC, 

Univeristy of Wisconsin-Madison) to improve and evaluate the chlorophyll algorithm for Lake 

Superior. When finalized, satellites will be able to provide scientists with productivity 

observations at significantly higher spatial and temporal frequency than is currently possible 

from direct measurements and help to improve our understanding of Lake Superior’s carbon 

cycle. 

References 

Alin, S. R., and T. C. Johnson (2007), Carbon cycling in large lakes of the world: A \synthesis of production, burial, 

and lake-atmosphere exchange estimates, Global Biogeochem. Cycles, 21, GB3002, 

doi:10.1029/2006GB002881. 

Assel, R.A. (2003), Great Lakes Ice Cover, First Ice, Last Ice, and Ice Duration: Winters 1973-2002. NOAA TM 

GLERL-125. Great Lakes Environmental Research Laboratory, Ann Arbor, MI. 

ftp://ftp.glerl.noaa.gov/publications/tech_reports/glerl-125/tm-125.pdf 

Atilla, N., G.A. McKinley, V. Bennington, M. Baehr, N. Urban, A. Desai, and C Wu (2009), Observed variability of 

Lake Superior pCO2, submitted to Global Biogeochemical Cycles. 

Austin, J.A. and S.M. Colman (2007), Lake Superior summer water temperatures are increasing more rapidly than 

regional air temperatures: A positive ice-albedo feedback, Geophys Res. Lett., 34 ,L06604, 

doi:10.1029/2006GL029021 

Beletsky, D., J.H. Saylor, and D.J. Schwab (1999), Mean Circulation in the Great Lakes, J. Great Lakes Res., 25(1), 

78-93. 

Bennington, V., G. A. McKinley, S. Dutkiewicz, and D. Ullman (2009), What does chlorophyll variability tell us 

about export and air-sea CO2 flux variability in the North Atlantic?, Global Biogeochem. Cycles, 23, 

GB3002, doi:10.1029/2008GB003241. 

Chai, Y. and N.R. Urban (2004) 120Po and 210Pb distributions and residence times in the near-shore region of Lake 

Superior. J. Geophys. Res. 109(C10S07), doi:10.1029/2003JC002081. 

Cole, J.T., N.F. Caraco, G.W. Kling, and T.K. Kratz (1994), Carbon dioxide supersaturation in the surface waters of 

lakes, Science, 265, 5178, 1568-1570. 

Desai, A.R., A. Noormets, P.V. Bolstad, J. Chen, B.D. Cook, K.J. Davis, E.S. Euskirchen, C.M. Gough, J.M. 

Martin, D.M. Ricciuto, H.P. Schmid, J. Tang, and W. Wang (2008), Influence of vegetation and seasonal 

forcing on carbon dioxide fluxes across the Upper Midwest, USA: Implications for regional scaling, 

Agricultural and Forest Meteorology, 148(2), 288-308. 

Dutkiewicz, S., M.J. Follows, and P. Parekh (2005), Interactions of the iron and phosphorous cycles: A threedimensional 

model study, Global Biogeochem. Cycles, 19, GB1021, doi:10.1029/2004GB002342. 

Eppley, R. W. (1972), Temperature and phytoplankton growth in the sea, Fish. Bull., 70, 1063 – 1085. 

French, C.R., J.J. Carr, E.M. Dougherty, L.A.K. Eidson, J.C. Reynolds, and M.D. DeGrandpre (2001), 

Spectrophotometric pH measurements of freshwater, Analytica Chimica Acta, 453, 13-20. 

Frouin, R., and R.T. Pinker (1995), Estimating photosynthetically active radiation (PAR) at the Earth’s surface from 

satellite observations, Remote Sensing of Environment, 51, 98-107. 

Hsu, S. A. (1986), Correction of land-based wind data for offshore application: A further evaluation, J. Phys. 

Oceanogr., 16, 390–394. 

Lam, D.C.L. (1978), Simulation of water circulations and chloride transports in Lake Superior for summer 1973, J. 

Great Lakes Res., 4(3-4), 343-349. 

Large, W. G., J. C. McWilliams, and S. C. Doney, (1994), Oceanic vertical mixing: a review and a model with a 

nonlocal boundary layer parameterization. Rev. Geophys., 32, 363-403. 

8

Marshall, J., A. Adcroft, C. Hill, L. Perelman, and C. Heisey (1997a), A finite volume, incompressible Navier- 

Stokes model for studies of the ocean on parallel computers, J. Geophys. Res., 102, 5753-5766. 

Marshall, J., C. Hill, L. Perelman, and A. Adcroft (1997b), Hydrostatic, quasi-hydrostatic, and nonhydrostatic ocean 

modeling, J. Geophys. Res., 102, 5733-5752. 

Quinn, F.H. (1992), Hydraulic residence times for the Laurentian Great Lakes, J. Great Lakes Res., 18(1), 22-28. 

Russ M.E., Ostrom N.E., Gandhi H., Ostrom P.H., and Urban N.R. (2004) Temporal and spatial variations in R:P 

ratios in Lake Superior, an oligotrophic freshwater environment. J. Geophys. Res. 109(C10S12), 

doi:10.1029/2003JC001890. 

Shuchman, R., A. Korosov, et al. (2006), Verification and application of bio-optical algorithm for Lake Michigan 

using SeaWiFS: a 7-year inter-annual analysis. J. Great Lakes Res., 32: 258-279. 

Schwab, D.J., and K.W. Bedford (1994), Initial implementation of the Great Lakes Forecasting System: a real-time s 

ystem for predicting lake circulation and thermal structure. Water Poll. Res. J. Canada, 29:203-220. 

Schwab, D.J. and D.L. Seller (1996), Computerized Bathymetry and Shorelines of the Great Lakes, NOAA Data 

Report ERL GLERL-16. 

Smagorinsky, J. (1963), General circulation experiments with the primitive equations. I. The basic experiments, 

Mon. Weather Rev., 91, 99 –164. 

Urban N.R., Apul D.S., and Auer M.T. (2004) Planktonic respiration rates in Lake Superior. J Great Lakes Res. 30 

(Suppl. 1), 230-244. 

Urban N.R., Green S.A., Auer M.T., Lu X., Apul D.S., Powell K., and Bub L. (2005) Carbon cycling in Lake 

Superior. J. Geophys. Res. 110(C6), doi:10.1029/2003JC002230. 

Urban N.R. (2006a) Carbon Cycling in Lake Superior: A Regional and Ecosystem Perspective. In State of Lake 

Superior in thet 21 st Century: Health, Integrity, Sustainability, Backhuys Publ. 

Urban N.R. (2006b) Nutrient Concentrations and cycles in Lake Superior: a retrospective. In State of Lake Superior 

in the 21 st Century: Health, Integrity, Sustainability, Backhuys Publ. 

Vollenweider, R.A., M. Munawar, and P. Stadelmann (1974), A comparative review of phytoplankton and 

primary production in thet Laurentian Great Lakes, J. Fish Res. Bd. Can., 31, 739-762. 

Wanninkhof, R. (1992), Relationship between wind speed and gas exchange over the ocean, J. Geophys Research, 

97, C5, 7373-7382. 

9

Realizing a Better Hydrostatic Response in NWP with MODIS Products 

Jordan J. Gerth 

Department of Atmospheric and Oceanic Sciences 

University of Wisconsin – Madison 

Madison, Wisconsin 

ABSTRACT 

Hydrostatic processes initiated along shorelines (herein, lake breezes) have been a neglected area 

of study over the recent decade as the power and functionality of numerical weather prediction 

(NWP) models has significantly grown. The Great Lakes have a substantial impact on life and 

commerce in the large industrial cities on its coast. One particularly unique manifestation of the 

lake breeze is the pneumonia front, which is a boundary containing a sharp temperature contrast 

that occasions eastern Wisconsin during the spring and summer months. This paper provides a 

review of important Great Lakes studies to date and examines how a coupling of spatially highresolution 

data from weather satellites and the Weather Research and Forecast (WRF) model can 

produce better forecasts through the enhancement of mesoscale and synoptic scale features in 

marine-modified climates across the Great Lakes. 

Introduction 

Sea breezes, complex hydrostatic processes resulting from differential skin surface heating 

between land and water, are sub-synoptic meteorological phenomena that significantly influence 

the weather of coastal communities. The study of sea breezes in tropical climates, such as those 

along the Florida peninsula, is ongoing (LaCasse 2008). While the behavior of sea breezes 

occurring within mid-latitude maritime climates compared to those rooted in tropical maritime 

climates is similar, and the dynamical and physical factors behind the genesis of each is 

identical, there are some particular local aspects and manifestations of the former which are 

worthy of study. The lesser studied category of sea breezes, lake breezes, are a relatively 

common occurrence on the Great Lakes and other large, non-oceanic bodies of water in the midlatitudes 

during the spring, summer, and fall months, but recent research on them is lacking. 

The Great Lakes collectively form one of the largest reservoirs for fresh water in the world. 

There are numerous large metropolitan areas with significant industry harboring on the Great 

Lakes, including: Green Bay, Wisconsin; Milwaukee, Wisconsin; Chicago, Illinois; Gary, 

Indiana; Detroit, Michigan; Buffalo, New York; and Cleveland, Ohio. In the absence of strong 

synoptic flow, the weather in these major cities, like surrounding coastal communities, is 

conditioned as a function of the Great Lakes water temperature. To what extent a lake breeze 

influences or is expected to influence the shoreline communities has a significant impact on 

energy production via anticipated usage. Using a simple temperature-pressure proportionality, 

established atmospheric science arguments show lake breezes modify shoreline wind conditions 

in response to a differential horizontal heat flux between land and water as the land heats warmer 

than water. The atmosphere seeks to assuage the inland heat with lake-cooled air through the 

lake breeze response. Like all mesoscale meteorological features, the resulting temperature and 

wind can differ between coastal and near-coastal locales less than five miles apart. This variety 

in conditions poses numerous problems in the forecast process and is consistently liable to deter 

recreation and commerce near the Great Lakes through conditions which are rarely hazardous, 

but often uncomfortable. This problem can be remedied through accessible Great Lakes-focused 

11

numerical weather prediction (NWP) models and high-resolution satellite observation, along 

with a greater understanding of the dynamical and physical processes that influence maritime 

climates. Lake breezes in the vicinity of Lake Michigan and Lake Superior are particularly 

worthy of study due to the sink of cold air they access during the spring and early summer. 

Pneumonia Fronts 

There is a unique manifestation of the lake breeze from Lake Superior that occasionally impacts 

the western shore of Lake Michigan in the presence of a typically weak synoptic cold front. This 

feature, known as a pneumonia front, has been loosely and variably defined by National Weather 

Service forecasters, herein referred to as the working definition, to be a ―rapid wind [shift] to the 

northeast‖, an increase in surface wind speed, and a ―rapid [decrease] in surface temperature‖ 

(Behnke 2005). For the purpose of his study, Behnke offered a less subjective definition of a 

pneumonia front, which has since been adopted by the National Weather Service. The quantified 

Behnke definition of a pneumonia front has two parts in order to narrow his research. First, 

hourly temperature falls at least 16 degrees Fahrenheit are required at a lakeshore observing 

station, and no more than 10 degrees Fahrenheit at a station well inland. Second, this 

temperature fall must be during the months of May through August. 

There are some stipulations that both the working definition and Behnke definition make which 

can be problematic in application. First, there are three types of boundaries which can produce a 

fairly large temperature drop over a short period of time other than a lake breeze: an outflow 

boundary separating rain-cooled air from a warm, moist, conditionally-unstable environment; a 

strong, sharp cold front, particularly during the spring and fall months; an early spring lake 

breeze where lake-modified air is substantially colder than the air over land heated diurnally. 

Second, Behnke quantifies a complex dynamical and physical process via a point change in 

temperature at two locations. This approach can lead to difficulties in producing climatology, as 

Behnke partially relents, because high-resolution temperature data at one-minute intervals is 

generally not available. Therefore, a pneumonia front crossing the observing station at reporting 

time may fail to meet the criteria if the adjacent reports are one hour prior and later. In addition, 

it is possible that a temperature decrease of less than the quantified amount could be experienced 

with a replica pneumonia front under a different synoptic situation or after sunset as the ground 

cools, even discounting the lack of high-resolution data. While Behnke’s attempt to reconcile an 

overbroad working definition of the pneumonia front is a good forward step to respect the 

responsible coordinate dynamical and physical processes, more investigation is needed. In 

addition, his attempts to objectively narrow his set of potential case studies proved to be 

somewhat fruitless, as he eventually examined 94 potential days surface sea-level pressure and 

wind speed from the National Centers’ (NCEP/NCAR) reanalysis data subjectively. 

Behnke’s method in finding potential pneumonia front cases focuses solely on the temperature 

drops at General Mitchell International Airport in Milwaukee, Wisconsin, (KMKE) without 

similar declines at Dane County Regional Airport outside Madison, Wisconsin, (KMSN). In his 

study, he features one case of a particularly viable pneumonia front on 17 July 2003. On this 

day, the synoptic cold front was synoptically quite weak though discernable. Earlier, surface 

observations suggested there was little in the form of a cross-frontal temperature gradient. 

However, later in the day, there was a strong temperature drop and increase in wind at KMKE 

during the evening hours, when a one-hour temperature decrease of 85 degrees Fahrenheit to 65 

degrees was realized associated with the pneumonia front enhanced by Lake Michigan. 

12

More generally, Behnke focuses on two synoptic situations which are favorable for pneumonia 

front occurrences, though they are similar. He emphasizes the presence of a seasonally strong 

surface low east of Hudson Bay, with a cold front transitioning into a stationary boundary 

between two areas of high pressure, one over western Ontario, and the other in central Indiana. 

Still, as Behnke shows through a statistical validation, the synoptic situations are not reliable 

indictors of pneumonia front occurrences. Nor are the synoptic situations closely identical to 

those which promote Lake Michigan lake breezes (Laird 2001). 

Developments in Numerical Air and Sea Analyses and Predictions 

There has been a notable development in computing power and NWP since the work of Behnke. 

Most particularly, there has been a marked shift away from case studies applying the fifth 

generation of the Pennsylvania State University—National Center for Atmospheric Research 

Mesoscale Model (MM5) toward the Weather Research and Forecast Model (WRF). 

Furthermore, following development, the WRF infrastructure was subsequently augmented into 

user-friendly software and web platforms such as the Linked Environments for Atmospheric 

Discovery (LEAD) and WRF Environmental Modeling System (EMS). This effectively allowed 

easy access to running individualized simulations with limited knowledge of the Linux operating 

system and the model cores. Furthermore, the process of running a model became much more 

streamlined and efficient. 

In a study by Song et al., the MM5 produced output on a three-kilometer spatial grid resolution 

to study the influence of water surface temperatures (WSTs; also known as sea surface 

temperatures, SSTs) on the marine boundary layer (MBL). There was a decent correspondence 

between the simulated MBL and the introduced WSTs, signifying that high-resolution WSTs can 

be used to accurately simulate the corresponding MBL. Chelton (2005) used the European 

Centre for Medium-Range Weather Forecast (ECMWF) model for investigating the impact of 

the spatial resolution of WSTs on NWP accuracy. In his experiment, he compared models run 

with the WST analyses of one degree and half degree spatial resolution. The finding was that the 

forecast surface wind in the model run using the higher-resolution WST product had a stronger 

correlation to the near-surface wind approximations from the Quick Scatterometer instrument 

onboard the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI). While 

the ECMWF is not a high-resolution model, as computing power continues to grow, 

investigations of mesoscale features will commence and expand rapidly in NWP efforts. 

The MODerate-Resolution Imaging Spectroradiometer (MODIS) instrument is onboard two 

National Aeronautics and Space Administration (NASA) polar-orbiting satellites: Aqua and 

Terra. The instrument uses one-kilometer infrared channels to approximate SSTs in clear pixels. 

This approximation relies heavily on the brightness temperature from the 11 and 12 micron 

channels (Davies) with respect for the climatologically normal SST value. A study from Minnett 

et al. (2007) has found that the MODIS SST products have an error of around a half of a degree 

Celsius when compared to buoys. The accuracy of daytime MODIS SST products is higher than 

nighttime due to the addition of visible channels into the SST computation (Strabala). 

The Real-Time Global (RTG) SST analysis is a product of the NCEP Marine Modeling and 

Analysis Branch. In producing an analysis at the resolution of 0.5-degree and 0.083-degree 

spatial resolution, it uses in situ data from buoys and ships as well as satellite-derived SST from 

the polar-orbiting National Oceanic and Atmospheric Administration (NOAA) Advanced Very 

13

High Resolution Radiometer (AVHRR) satellites. All data is no more than 36 hours old. The 

background field is provided from a climatology-influenced previous analysis and built upon 

through a two-dimensional assimilation scheme (Thiebaux et al. 2005). In addition, ―SST 

observations are given a smaller (larger) radius of influence in regions of strong (weak) SST 

gradients.‖ (LaCasse 2008) The RTG SST analysis at 0.5-degree is the background SST field for 

operational and experimental weather prediction models, including the WRF. The WRF is 

becoming increasingly useful in mesoscale studies of Great Lakes meteorological phenomena. 

However, there are a few reasons why the RTG SST analysis, even at 0.083-degree resolution, is 

not suitable for WRF simulations with domains over the Great Lakes. First, there are a limited 

number of in situ data points. Relatively few ships report water temperature. Over the winter 

time particularly, many of the open water buoys are collected and stored or moved out of the 

Great Lakes during the ice season. Moving these buoys takes a reasonable amount of time and it 

can be well into the spring before they are replaced. In addition, many of the thermal eddies 

within the Great Lakes occur at much smaller scales than the thermal eddies which comprise the 

Gulf Stream. Finally, there is significantly strong upwelling within about five miles of shore in 

certain synoptic situations. This upwelling typically leads to a strong WST gradient and fog in 

the upwelling area. With no buoys in the near-shore areas of the Great Lakes, and fog to cover 

the upwelling from satellite detection, this very important bathymetric response is precluded 

from the RTG SST analysis. Furthermore, the RTG SST has developed with a global emphasis 

and, despite the use of a spatially varying Gaussian error decorrelation length scale (Thiebaux 

2003), the enclosed nature of the lakes prevents effective incorporation of all available data. 

In contrast, the Great Lakes Environmental Research Laboratory runs a Great Lakes Coastal 

Forecasting System (GLCFS) on a five-kilometer spatial resolution grid which produces an 

initial period SST product, which heavily incorporates data from the AVHRR. While there are 

some slightly notable gradients present in initial GLCFS conditions, spatial variability tests 

normalize the product to remove important SST gradients which are useful in NWP. Spatial 

variability testing is important to remove anomalous cold points which could often be clouds 

obstructing the clear view of the satellite, but the Great Lakes bathymetry allows for 

exceptionally tight SST contrasts in near-shore and open waters. The GLCFS also has to contain 

low error, since the initial SST product is used to run a Great Lakes Wave Model for producing 

National Weather Service marine wind forecasts distributed operationally. 

High-Resolution Meteorological Models 

As Mass et al. (2002) emphasizes, high-resolution models, those with a spatial resolution below 

ten kilometers, may better represent mesoscale features, but they are not captured through 

current methods of verification. The result is that the accuracy of the forecast cannot be 

improved through higher-resolution model runs, assuming that the approximations utilized from 

parameterizations still hold applicability over smaller grid scales when checked against explicit 

diagnostics. Mass’ underlying, important claim is a detailed, refined statement and applicable 

case study that NWP is an initial-value problem. There can be no more gained from a model 

than what is ingested. A model running with a spatial resolution of 20 kilometers cannot benefit 

from WST data at 10-kilometer spatial resolution. Likewise, running at 20-kilometer model with 

an initial WST dataset with a 40-kilometer spatial resolution can gain no better verification 

statistics than a model run at 40 kilometers, the resolution of the datasets that initialized the 

model. While the horizontal resolution of NWP models is on the increase, it is critical to 

14

modeling efforts that the highest possible initial-value datasets are provided. This challenge is 

already underway in terms of land-surface modeling, where there are dozens of land types which 

influence the surface heat and moisture flux, sometimes significantly depending on the soil 

moisture and land use. As LaCasse (2008) showed in her investigation of WRF simulations over 

the Florida peninsula, coastal mesoscale gradients in SST can lead to distinct changes in cloud, 

precipitation, and temperature forecasts for the immediate onshore and offshore areas. The 

weather and water conditions near the shore have a significant impact on the near-term forecast 

for the shoreline communities. More precision is needed for accurate forecasts in assessing the 

movement of lake breezes. 

Furthermore, per Minnett (2003), diurnal differences in SST are typically much less than those 

over land, but under calm conditions, can be notable. Increases in SST have been realized over 

the Great Lakes under full sun, high pressure, and calm winds. For example, on 23 July 2007, 

MODIS SST data collaborated with midlake automated buoy observations over Lake Michigan, 

found skin water temperature increases can exceed 1 degree Fahrenheit per hour. Such large 

changes can have significant impacts on NWP accuracy which models are not currently 

parameterized to correct. Water temperatures generally remain fixed in models applying landsurface 

schemes. This alters the differential flux which could be important for the (de)generation 

of a lake breeze. Tijm et al. (1999) found that the pressure field responds as the surface flux 

changes at the land-sea interface upon the inception of the sea breeze. 

Changes in WSTs lead to a difference in land-sea surface flux comparison and thus signal a 

response within the MBL. The sea-breeze circulation model is similar to the pressure forcing 

model, in which changes in the pressure gradient influence the surface wind field. The WRF 

boundary layer height is an approximation based on a bulk Richardson formulation (Troen and 

Mahrt 1986), which depends on the conditions in the near-surface layer. As a consequence, 

changes in the WRF planetary boundary layer (PBL) height are a byproduct of changes in the 

sensible heat flux (LaCasse 2008). This allows for the PBL height to be used as a surrogate for 

locating marine-modified air masses in the WRF during the day, regardless of the mechanism, 

mesoscale or synoptic scale, for advecting marine air onshore. 

Case Study: 26 May 2008 

During the afternoon and evening hours of 26 May 2008, a strong convergence boundary 

descended south out of Lake Superior and accelerated down the western coast of Lake Michigan. 

There was no precipitation associated with this boundary, but the temperature and wind shift was 

significant, with coastal temperature declines of 20 degrees Fahrenheit and larger within an hour 

period. Reporting stations well inland did not experience significant temperature falls, though 

winds statewide become increasingly northeasterly during the late afternoon and evening. The 

National Weather Service in Milwaukee, Wisconsin, described this boundary as a pneumonia 

front consistent with both the working definition and Behnke’s definition. As shown in figures 

later, the boundary was first recognized on base reflectivity data from the Marquette, Michigan, 

(KMQT) Weather Surveillance Radar (WSR) as a hybrid of convective outflow and a lake 

breeze. The WSRs at Green Bay, Wisconsin, (KGRB) and Sullivan, Wisconsin, (KMKX) also 

captured the convergence boundary as it progressed into their scan area (124 nautical miles from 

the radar site). Since the boundary was relatively shallow, it was well captured using the halfdegree 

base reflectivity slice near radar sites, but difficult to detect beyond 80 nautical miles 

from the site, especially when it was not well-established over northern Wisconsin. 

15

On the morning of 26 May 2008, there was an area of low pressure over western Lake Superior. 

There was a weak cold front extending from Marquette, Michigan, to Madison, Wisconsin, to 

Des Moines, Iowa, and large area of high pressure building southward out of central Canada to 

Montana and the Dakotas. By the morning of 27 May 2008, the area of low pressure had 

reached the Newfoundland province of Canada and high pressure was firmly established across 

much of the northern United States. A weak cold front extended from upstate New York through 

central Indiana and into central Oklahoma. The period of study in Wisconsin is the 24 hours 

between 1200 UTC on 26 May 2008 and 1200 UTC on 27 May 2008. 

Green Bay, Wisconsin, had a morning with clear skies and temperatures warming to 81 degrees 

Fahrenheit. During the late afternoon, the surface wind direction changed from westerly to 

northeasterly, skies became overcast, and the surface temperature at Green Bay (KGRB) fell 

sharply from 78 degrees to 61 degrees between 1653 LST and 1743 LST (2253 UTC and 2343 

UTC), according to quality-controlled data from the National Climatic Data Center (NCDC). 

Milwaukee, Wisconsin, had partly cloudy skies during most of the day with temperatures similar 

to Green Bay in the upper 70s and lower 80s. Winds prevailed out of the southwest. At 1852 

LST (0052 UTC), the temperature at KMKE was 78 degrees. At 1926 LST (0126 UTC), clouds 

had increased and the temperature had decreased to 55 degrees per NCDC reports. Unlike at 

Green Bay, the Milwaukee observer reported a squall as winds changed from Southwest at 8 

mph prior to the temperature drop to North at 29 mph gusting to 43 mph after. Winds remained 

strong with gusts in excess of 30 mph for the balance of the evening. 

The WRF was employed in an attempt to accurately model the 26 May 2008 case. The model 

was run using the Advanced Research WRF (ARW) version 2.2.0 core to 24 hours from 

initialization on an 800 km by 800 km localized domain centered on Oconto, Wisconsin, with 

grid spacing of four kilometers. The domain covered most of the states of Wisconsin and 

Michigan, as well as Lake Michigan, Lake Huron, and southern Lake Superior. The WRF model 

with the ARW core is user-configurable and optimized for parallel computing environments. A 

similar WRF model was established in LaCasse (2008) to study the impact of MODIS SST data 

on the nocturnal Floridian marine boundary layer. LaCasse’s WRF configuration was modified 

only slightly in the localized model runs configured over the western Great Lakes. The model 

used the Lin microphysics scheme, Dudhia shortware radiation scheme, Rapid Radiative 

Transfer Model (RRTM) longwave radiation scheme, Monin-Obukhov surface-layer scheme, 

Noah land-surface model, and the Yonsei University boundary-layer scheme. Cumulus and gridscale 

convection was parameterized with the Kain-Fritsch scheme under vertical velocity 

damping and mixing terms evaluated in physical space. There were no vertical or horizontal 

diffusion constants. The model initialized with data from the 40-kilometer North American 

Mesoscale (NAM) model. The model accurately resolved a morning lake breeze boundary south 

of Lake Superior over the Upper Peninsula of Michigan and evolved a convergence boundary 

sweeping down the western shore of Lake Michigan at a speed observed with radar. 

The WRF model simulation shows strong surface heating across most of the land domain at 1500 

UTC (1000 AM CDT) on 26 May 2008. However, by 1800 UTC, there is a notable contrast in 

surface temperature across far northern Wisconsin and the Upper Peninsula of Michigan. Strong 

northerly winds fan out from Lake Superior and the boundary accelerates southward, visible 

through the height of the boundary layer. The intensity of the surface boundary, measured 

16

through the cross-front temperature drop, is less late in the day. There is a broad decrease in 

temperature across the boundary as it sags southward over inland Wisconsin during the early to 

middle part of the afternoon (1800 UTC to 2100 UTC). At 2100 UTC, and especially by 0000 

UTC on 27 May 2008 (0700 PM CDT 26 May 2008), this boundary is clearly visible operating 

on Lake Michigan’s MBL as a density current, where it is discernibly vigorous with strong 

northerly winds advecting air from Lake Superior. This boundary is at a 45-degree angle to 

shore at 0300 UTC and reaches the bottom of Lake Michigan to dissipate through mixing 

starting at 0600 UTC (0100 AM CDT). Figure 1 shows the evolution of the two-meter surface 

temperature at three-hour intervals as the pneumonia front slides down the lake. Figure 2 shows 

the height of the boundary layer and 10-meter wind direction during the same three-hourly times. 

The figures’ time series illustrates a successful attempt of NWP to accurately render a sharp, 

mesoscale convergence boundary evolving from a lake breeze. It shows the importance of highresolution 

simulations on forecast precision. 

The results of the WRF model output over the local domain were compared to the surface 

observations and radar data. The WRF model did a sufficient job of evolving the structure of the 

pneumonia front. However, the genesis of the boundary is approximately two to three hours 

later than observed after verification with WSR base reflectivity (Figure 3), though the 

temperature difference across the boundary is generally similar to the observations. Since the 

boundary is later than expected transgressing the southern coastline of Lake Michigan, the inland 

boundary layer had begun to cool with the sun setting and loss of incoming radiation. The WRF 

model data for the grid point nearest Milwaukee, Wisconsin, had a pre-frontal temperature at two 

meters of 64 degrees Fahrenheit, and a post-frontal temperature of 48 degrees, for a net drop of 

18 degrees. This compares to 23 degrees in the observation. The simulation’s representation of 

the prevailing wind direction both before and after the passage of the front is fairly accurate after 

a cursory comparison with observational data. 

A WRF simulation of this case without Lake Superior will determine if Lake Superior plays a 

role in the formation of this feature. It is anticipated based on aforementioned findings that it 

does. Thus, applying the theories first formulated from Behnke, it is hypothesized that the 

pneumonia front will be nonexistent or significantly weakened. Other future simulations will 

include better sea-surface temperature data in the initial conditions, with input from MODIS. 

Conclusions 

Recent advancements in NWP systems, such as WRF, along with computer power, have enabled 

high-resolution simulations of physical and dynamical processes on marine boundary layers. 

The only way to improve these forecasts is to better the initial conditions. Weather satellites 

which deliver data of one-kilometer spatial resolution have the potential to significantly improve 

the near future of meteorological modeling. When these systems are interlinked with studies of 

synoptic and mesoscale processes, such as the study of lake breezes and pneumonia fronts, they 

can be a powerful tool for understanding the atmosphere and lead to advancements in forecasting 

for coastal locations. At the current time, Behnke’s lone, decade-old study of pneumonia fronts 

using the MM5 leaves many questions unanswered about the origins and interactions of marinebased 

convergence boundaries. However, the WRF has provided valuable output on which 

better scientific conclusions can be reached. The accurate representation of the 26 May 2008 

case is a testament to this fact. 

17

References 

Behnke, C., 2005: Synoptic and Local Controls of the Lake Michigan Pneumonia Front. 

Chelton, D. B., 2005: The impact of SST specification on ECMWF surface wind stress fields in the eastern tropical 

Pacific. J. Climate, 18, 530–550. 

Davies, J., 2004: The IMAPP MODIS Sea Surface Temperature Algorithm. 

Garratt, J. R., 1986: Boundary-layer effects on cold fronts at a coastline. Bound.-Layer Meteor., 36, 101–105. 

LaCasse, K.M., M.E. Splitt, S.M. Lazarus, and W.M. Lapenta, 2008: The Impact of High-Resolution Sea Surface 

Temperatures on the Simulated Nocturnal Florida Marine Boundary Layer. Mon. Wea. Rev., 136, 1349–1372. 

Laird, N.F., D.A.R. Kristovich, X.Z. Liang, R.W. Arritt, and K. Labas, 2001: Lake Michigan Lake Breezes: Climatology, 

Local Forcing, and Synoptic Environment. J. Appl. Meteor., 40, 409–424. 

Mass, C.F., D. Ovens, K. Westrick, and B.A. Colle, 2002: Does Increasing Horizontal Resolution Produce More Skillful 

Forecasts? Bull. Amer. Meteor. Soc., 83, 407–430. 

Song, Q., T. Hara, P. Cornillon, and C. A. Friehe, 2004: A comparison between observations and MM5 simulations of the 

marine atmospheric boundary layer across a temperature front. J. Atmos. Oceanic Technol., 21, 170–178. 

Strabala, K., 2008: Personal interview. 

Thiébaux, J., E. Rogers, W. Wang, and B. Katz, 2003: A new high-resolution blended real-time global sea surface 

temperature analysis. Bull. Amer. Meteor. Soc., 84, 645–656. 

Tijm, A. B. C., A. A. M. Holtslag, and A. J. van Delden, 1999: Observations and modeling of the sea breeze with the 

return current. Mon. Wea. Rev., 127, 625–640. 


I would like to thank the Wisconsin Space Grant Consortium, the National Aeronautics and Space Administration, and the 

Space Science and Engineering Center at the University of Wisconsin for the ability to conduct this undergraduate 

research. This important opportunity has made it possible for me to explore the environment in terms of my own, 

unconfined interests, and provides a conduit for eventual graduate-level research. 

Figures 

Figure 1 

Figure 1 shows four-kilometer resolution, two-meter temperature data plotted at a 3-degree Celsius contour interval 

from the WRF-ARW at 3-hourly intervals, starting at 1500 UTC on 26 May 2008 and ending at 0600 UTC on 27 

May 2008. Notice the evolution of a strong temperature gradient over eastern Wisconsin. 

18

Figure 2 

Figure 2 shows the height of the planetary boundary layer (150m shading) and ten-meter surface wind data from the 

four-kilometer WRF-ARW at 3-hourly intervals, starting at 1500 UTC on 26 May 2008 and ending at 0600 UTC on 

27 May 2008. Notice the evolution of a lake-modified air mass over northern Wisconsin early in the period. 

19

Figure 3 

Figure 3 contains base reflectivity (0.5-degree) radar data showing a boundary, from upper-left to lower-right: 1501 

UTC and 1759 UTC on 26 May 2008 from Marquette, Michigan, KMQT, 2058 UTC and 0003 UTC 27 May 2008 

from Green Bay, Wisconsin, KGRB, and 0300 UTC and 0604 UTC from Sullivan, Wisconsin, KMKX. 

20


Part Six 

Astronomy

A Comparative Study of Type IIb Supernovae 

Bradley T. Rentz 1 , Cyrus M. Vandrevala 1,2 , and Christopher J. Stockdale 2 

Department of Physics, Marquette University, Milwaukee, WI 

Kurt W. Weiler 3 

Naval Research Laboratory, Washington, DC 

ABSTRACT 

We analyzed radio and X-ray data from supernova SN 2008ax and SN 2008bo 

and compared the results with SN 1993J, SN 2001ig and SN 2001gd. The radio 

data was obtained from the Very Large Array and the X-ray from the Chandra 

X-ray observatory and the Swift satellite X-ray telescope. The radio and X-ray 

data were taken at several frequencies. 

Introduction 

During the summer period I analyzed archival data of type IIb Supernovae (SNe) from 

the Very Large Array (VLA) 4 radio telescope in order to determine the radio properties of 

the SNe. Concurrently, my colleague, Cyrus Vandrevala, analyzed archival X-ray data of 

the same SNe taken from the Chandra and Swift X-ray space observatories to determine 

the X-ray properties of the SNe. The results of both projects were compared for a better 

understanding of the type IIb SNe. 

Researching SNe is important, because understanding how such stars decay and then die 

can answer general questions about how the Universe and solar systems were formed and 

how they will evolve. Knowledge of how SNe develop and evolve can also lead to answers 

about how life was formed, since many of the elements necessary for life are formed in SN 

explosions. 

Type IIb SNe are especially important to study because they do not conform to the standard 

model for SNe. Type IIb SNe are unique among type II SNe in that they have a very depleted 

hydrogen envelope. The outer layer of hydrogen of the type IIb SNe was shed in a stellar 

wind, a very gentle slow process, as a single evolving star or through interaction with a 

binary companion star (Filippenko, et al., 1988). As a result of the lack of hydrogen, the 

radio emission from these SNe evolve at a quicker rate and the radiation from the SNe can be 

1 Work on this project was funded by the Wisconsin Space Grant Consortium. 

2 Research sponsored by NASA Award NNX09AC90G. 

3 KWW thanks the Office of Naval Research for the 6.1 funding, which supports his research. 

4 The VLA telescope of the National Radio Astronomy Observatory is operated by Associated Universities, 

Inc. under a cooperative agreement with the National Science Foundation. 

1

detected earlier than other types of SNe (Stockdale, et al., 2009). This phenomenon can be 

observed when comparing a type IIb SN (Figure 1) with other type II SNe (Figure 2). 

Fig. 1.— Light Curve for SN 2008ax. 

Fig. 2.— Light Curve for SN 1995N (Chandra, 

et al., 2009). 

The radio analysis method involves measuring the intensities of radio emission at certain 

wavelengths from the blastwave of the SN. The intensity of the wavelength directly corresponds 

to the material shed by the SN, the circumstellar material (CSM). The CSM will 

absorb certain wavelengths of radio emission as continuum absorption in contrast to an optical 

absorption line spectrum, where each wavelength represents certain energy and materials 

(such as hydrogen or oxygen). Thus, the emission observed by the radio telescope does not 

indicate which materials are present in the CSM, but can be used to measure the density 

of the CSM. Then comparing when the radio emission is observed relative to the initial 

blast, and assuming the speed of the blastwave and the CSM ejected the approximate time 

a material ejected before the SN can be calculated. This knowledge is called the mass-loss 

history. Knowing how the CSM was formed in the SN leads to knowledge of the life and 

death processes of the SN. This knowledge can then be used to determine characteristics and 

properties of type IIb SNe. The radio emission created is formed when the blastwave, due 

to the massive amount of energy it contains, accelerates and the electrons in the CSM to 

relativistic speeds and heating them from ∼10,000 K to ∼10,000,000 K. The blastwave itself 

also creates a strong magnetic field that traps the excited relativistic particles and causes 

them to emit radiation called synchrotron radiation (Figure 3). This radiation is detected 

by radio observations, which in turn depict the density make up of the CSM. 

The CSM is often considered to be “fog-like” in nature, because the “fog” particles (free 

electrons in a cloud of ionized hydrogen) obscure radio radiation. This “fog of particles” is 

2

Fig. 3.— Cartoon Image of SN blastwave interacting with CSM (Stockdale, et al., 2007). 

generally regarded as being evenly distributed, with a homogenously decreasing density as 

the distance from the SN increases but this is not always the case, and can lead to variances 

in the data (Weiler, et al., 2002). The clumps in the CSM depicted in Figure 3 are areas 

of an uneven density distribution and can be the result of the stars natural loss of CSM or 

because of the interaction with a binary companion star. 

The CSM “fog” is different from normal fog in that long wavelength radiation is obscured 

more easily, whereas the short wavelength optical radiation that is normally obscured by a 

classical “fog”. Thus, the results of the radio emission generally show that over time, as the 

energy of the blastwave disperses and the column density of the CSM decreases, the intensity 

of the longer wavelength frequencies increase; whereas the short wavelengths then decrease 

over time. The larger intensity of the longer wavelengths over time is an indication that 

the radiation from the SN is not thermal, but rather synchrotron (radiation coming from 

relativistic particles trapped in magnetic fields). The radiation emissions from the blastwave 

are also obscured by the distant ionized hydrogen (HII) regions as seen in Figure 3. This 

absorption, unlike that of the CSM, is constant with time and can easily be accounted for 

in order to find the effects by the CSM alone. 

Procedure 

The radio data for the SNe were analyzed and processed using the computer program called 

Astronomical Imaging Processing System (AIPS). The first step of the program after loading 

the data onto the computer was to flag any erroneous data points that could skew the outcome 

of the analysis. The erroneous data stemmed from several possible sources, such as weather 

3

or galactic effects. Once the data were flagged, they were calibrated since the data were 

collected by several radio dishes of the telescope. The calibration process started first with 

a task called SETJY that set the flux (brightness per unit area) of the primary calibrator 

(a very bright, nearly constant radio source). The secondary calibrator was then adjusted 

relative the the primary calibrator using the task SETJY. Next the phases were calibrated 

by the task CALIB. Finally, the task CLCAL applied the calibrations to the data set. 

After calibrating the data, the data were imaged and the images were “cleaned” to reduce 

noise and other effects. Several calculations could then be made on the final images, such 

as calculating the spectral luminosity. The results of the calculations were then placed in a 

spreadsheet. 

After processing all the data sets for a SN, the log of the spectral luminosity vs. the log of 

the time since explosion for each data period were graphed, thus producing what is called a 

light curve. The light curves were then used to make determinations about the SN. 

Results 

SN 2008ax in NGC 4490 was found to have reached its peak at around 10 days after the 

initial blast and faded on the 22 GHz band after about 20 days (Figure 1). Most SN do 

not reach their peak until at least 500 days after the initial blast and do not fade until after 

several thousand days (cf. Figure 2). Further SN 2008ax exhibits irregular development 

after its peak, especially on the 8.46 GHz and 4.86 GHz bands, where the data points show 

irregular dips and rises that do not correspond to the normal parameterized curve. 

SN 2008bo in NGC 6643 is very faint in the radio spectrum, far away and early in its 

progression as a SN. It appears that SN 2008bo may be similar to SN 2008ax, since it is 

progressing quickly (Figure 4). Further, SN 2008bo has brighter X-ray emissions than SN 

2008ax, since SN 2008bo is fairly isolated and SN 2008ax is near a bright X-ray source. 

Conclusions 

The deviance from the parameterized curve and the short lifespan of SN 2008ax are similar 

to those of SN 2001ig (Figure 5). The similarities may indicate that SN 2008ax is not 

unique, rather part of a new subclass of type IIb SNe. The fluctuations in the light curve by 

SN 2008ax and SN 2001ig have been attributed to the presence of a companion star. The 

gravitational field of the companion star cause ripples in the CSM of the SN progenitor star. 

When the SN progenitor star explodes, the resulting radio wave emissions deviate from the 

standard model at places where the companion star’s gravitational field caused ripples in 

the CSM (Figure 7). 

4

Fig. 4.— Light curve of SN 2008bo. 

Fig. 5.— Light curve of SN 2001ig (Ryder, et al., 2004). 

5

Fig. 6.— Pinwheel effect caused by binary interaction 

in WR 104 (Tuthill, et al., 1999). 

Fig. 7.— Cartoon of binary wind model of 

WR 104 (Tuthill, et al., 1999). 

The presence of a companion star with the type IIb SN 2001ig has been observed previously. 

The companion star of SN 2001ig itself was not visible, but the resulting pinwheel effect was, 

as seen in Figure 6 (Ryder, et al., 2004). 

Future Plans 

SN 2008bo is early in its progression and it has not been extensively observed, so further 

study of SN 2008bo is warranted on both the radio and X-ray spectra. SN 2008ax is also 

not well studied on all bands, so further observations are need. 

During the fall semester, I will continue to work on studying SN 2008ax, SN 2008bo and 

other type IIb SN. I will focus mostly on the X-ray spectrum instead of the radio. 

The data also indicate there might be two subclasses of type IIb SNe. One subclass progresses 

rapidly and has short-term periodic oscillations (SN 2008ax and SN 2001ig), as a result of a 

binary interaction. The other subclass exhibits a sudden drop after a few thousand days (SN 

1993J and SN 2001gd), which may also be a link between type II and type Ib/c SNe. 

REFERENCES 

Chandra, P., Stockdale, C. J., Chevalier, R. A., Van Dyk, S. D., Ray, A., Kelley, M. T., 

Weiler, K. W., Panagia, N., Sramek, R. A., “Eleven Years of Radio Monitoring of the 

6

type IIn Supernova SN 1995N,” 2009. Astrophysical Journal, 690, p. 1839-1846. 

Filippenko, A., “Supernova 1987K - Type II in youth, type Ib in old age”, 1988, Astronomical 

Journal, p. 1914-1948. 

Kelley, M., Weiler, K., Panagia, N., Sramek, R., Marcaide, J., Williams, C., Van Dyk, S., 

“Light Curves of Radio Supernovae,” 2007, Supernova 1987A: 20 Years After, p. 

269-271. 

Ryder, S. D., Sadler, E. M., Subrahmanyan, R., Weiler, K. W., Panagia, N., Stockdale, C. 

“Modulations in the radio light curve of the Type IIb supernova 2001ig: evidence for 

a Wolf-Rayet binary progenitor?,” 2004, Monthly Notices of the Royal Astronomical 

Society, 349, p. 1093. 

Stockdale, C., Kelley, M., Weiler, K, Panagia, N., Sramek, R., Marcaide, J., Williams, C., 

Van Dyk, S, “Recent Type II Radio Supernovae”, 2007, Supernova 1987A: 20 Years 

After, p. 264-268. 

Stockdale, C., Vandrevala, C., Weiler, K., Sramek, R. A., Marcaide, J., Immler, S., Van 

Dyk, S., Pooley, D., Pooley, D., “VLA Radio Observations of the Type II Supernova 

2008ax”, 2009, American Astronomical Society. 

Tuthill, P., Monnier, J., Danchi, W., “A dusty pinwheel nebula around the massive star 

WR104”, 1999, Nature. 

Weiler, K., Panagia, N., Montes, M., Sramek, R., “Radio Emission from Supernovae and 

Gamma-Ray Bursters”, 2002, Annual Review Astrophysics, p. 393. 

This preprint was prepared with the AAS L ATEX macros v5.2. 

7

Gravitational Heating and Evolution of Galaxy Groups 

Melania Riabokin 

University of Wisconsin Madison, Department of Astronomy, 475 N. Charter Street, Madison WI 

53706 USA 

Eric M. Wilcots 

University of Wisconsin Madison, Department of Astronomy, 475 N. Charter Street, Madison WI 

53706 USA 

ABSTRACT 

It has been established that there is a subgroup of galaxy groups and clusters with 

properties that differ from the standard predictions of cold dark matter models of structure 

formation. This subgroup tends to have excess emission in the x-ray for a given 

luminosity as compared to similar groups, suggesting additional heating and/or energy 

in the over-luminous groups. A wealth of information abounds describing various 

proposed methods for providing heating in groups and clusters. We have studied the 

evolution of a galaxy group from inception to the present epoch, based on a simple evolutionary 

model. We investigated the amount of heating generated from gravitational 

collapse for groups of various masses and find that groups of mass> 8x 10 13 M 0 have 

sufficient gravitational heating to bring them to currently observed temperatures. We 

also find that for groups of lower mass, a simple AGN energy injection model yields 

realistic results as compared with recent surveys of temperature and x-ray luminosity. 

Subject headings: galaxies: groups, radio galaxy; AGN 

1. Introduction 

Extensive studies have been conducted which investigate the formation, evolution and nature 

of galaxy groups. Nearly 70 % of all observed galaxies live in galaxy groups (Thlly et al 1986). As 

gravitating bodies, galaxies interact and eventually fall into the same potential well and become 

bound gravitationally to form small congregations of 2-10 galaxies. These small groups of galaxies 

can interact, one group with another to form larger galaxy aggregates to which are known as clusters 

and can have anywhere from 20 to thousands of galaxies. Groups tend to live along the filaments 

that make up the large scale structure of the universe, and clusters live at the intersections of these 

filaments. 

In essence, galaxy groups are the basic building blocks of the large scale structure of the 

universe, therefore it is important to understand the formation and evolution of galaxy groups in 

the context of an evolving universe. 

9

and then can spiral in to the black hole. This accretion disk feeds the black hole and increases its 

mass and once the black hole has accreted enough mass the AGN is triggered 'on'. The turning 

on of the AGN sends out highly collimated beams of particles, accelerating to relativistic velocities 

along magnetic field lines. These structures are called jets. The exact physics of how the AGN is 

triggered, and the black hole transitions from dormant to active, are not quite clear and this is an 

active area of study in astrophysics. For example, it remains unclear as to whether or not an AGN 

turns on only once in it's lifetime, or whether it has a duty cyele where it turns on and off several 

times. Also, although every galaxy has a black hole at the center, very few of them are actually 

active. This may be simply because not all galaxies can have an AGN, or because the AGN in a 

particular galaxy may be in a dormant phase of it's duty cycle. 

It is through the interaction of the jet with the IGM that the AGN is believed to heat the IGM. 

The exact method of energy transfer between the relativistic particles of the jet and the cool IGM 

is not known exactly, but it is probable that the turbulent mixing resulting from the jet running in 

to the surrounding medium is what heats the IGM gas. 

What we have done is to consider the lowest mass group in our study at 10 13 Mev. Assuming 

that the group virialized and began to cool llGyr ago, we allow it to cool for one cooling time 

which for such a group is 4Gyr. During this time, the potential AGN host galaxy has enough time 

to accrete a significant amount of mass on it's central black hole. We assume the AGN turns on 

after one cooling time and remains active for 2 x 10 6 years, imparting lkeV of energy per particle 

in the group. With this additional energy, the temperature - and therefore the cooling time - of 

the group is increased. Figure 4 outlines the evolution of such a group, and it is evident from the 

cooling times and temperatures that if the group evolves in this way then the temperature after 

cooling to the present day is consistent with those of observed hot groups. 

2.4. Emissivity & Luminosity 

We performed another consistency check by calculating the peak luminosity of a group, given 

the present day temperature. We used equations 6 & 7 to calculate the emissivity and luminosity 

as a function of frequency. The frequency we used to calculate these was the peak frequency found 

using Wien's Law, Amax 0.0029jT m·K. 

Pavg = Eff 

6.8e 38Z 2 ngff -1 -3 -1 

TO.5 e hvjkT ergss em Hz (6) 

L (4j3)1T R3P avg (7) 

The peak frequency came out in the soft x-rays, which is expected. The actual luminosities we 

calculated were on the order of Lx = 2 X 10 16 W H z-l which is significantly lower than the observed 

10 40 W H z-l of Croston et al 2005. But this is also consistent with our expectations in that we 

did not include metallicity, and the most effective way of cooling and most prominent source of 

14

R is the core radius, which is arbitrary, so we replace it with r again 

GMf.,imp -a 

) a+lk ,Pgas = r (A9) 

(1 + a r BPgas 

GMf.,imp 

Tgas(r) = kB(l + a)r 

AlO is the working temperature profile seen as equation 2 above. 

B. Appendix B 

C. Appendix C 

Our expression for luminosity in equation 7 above came from a very simple subsitution 

(AlO) 

(Bl) 

(Cl) 

4 

P = T 3lT -+ T = [RPjl/4 (C2) 

R 3lT 

L = (4/3)1rR3Pavg (C3) 

17

Understanding the Evolution of Supernova Progenitors 

Cyrus M. Vandrevala 1, 2 , Bradley T. Rentz 1 , Christopher J. Stockdale 2 

Milwaukee, WI 


Stefan Immler 

Greenbelt, MD 

NASA Goddard Space Flight Center/Universities Space Research Association 

Kurt W. Weiler 3 

Washington D.C. 

Naval Research Laboratory 

Abstract 

We present the results of radio and Xray observations of the Type IIb supernovae SN 2008ax,  

SN 1993J, and SN 2001ig.  Though they are all Type IIb supernovae, SN 2008ax and SN 2001ig  

have radio decay patterns which show bumps on the decline of the curve while SN 1993J has a  

smooth radio light curve which suddenly drops off.   Additionally, SN 2008ax and SN2001ig  

have fast radio and Xray evolutions on the order of hundreds of days, while SN 1993J shows a  

much longer evolution on the order of thousands of days.  

Introduction 

Type IIb supernovae (SNe), first discovered by Woosley et al. (1987), were thought to be the 

result of the collapse of a Type Ib SN progenitor that has a small hydrogen envelope. They are 

very rare since only ~40 have been discovered in the last couple of decades and only a few are 

well observed. Our research concentrates on these supernovae for two primary reasons: 

1) These supernovae may be the evolutionary link between Type II supernovae (which have 

observed hydrogen spectra) and Type Ib/c supernovae (which have no observed hydrogen 

lines) as suggested by Woosley et al. (1987). 

2) However, not all Type IIb supernovae evolve in the same manner; we are trying to 

determine if the current classification of Type IIb supernovae needs to be re-evaluated. 

Background  

Supernovae   

A  supernova  is  the  catastrophic  end  of  the  life  of  a  very  massive  star  that  can  create  

                                                         

1 

Funding provided by the Wisconsin Space Grant Consortium (WSGC) and the National Space Grant College and 

Fellowship Program (NSGC) 

2 

Sponsored by NASA award NNX 09AC90G 

3 

KWW thanks the Office of Naval Research for the 6.1 funding that supports his work 

19

enormous shock waves in a short amount of time.  They can release more than 10 51  ergs of  

energy  in  just  a  few  seconds  with  shock  speeds  greater  than  10,000  km/s  and  are  

responsible for many of the phenomena in the universe like the formation of black holes  

and neutron stars, the evolution of a host galaxy, cosmological probes, and the creation and  

distribution of heavy elements and energy in the interstellar medium (ISM).  We often see  

two different kinds of supernovae – those that die as a thermonuclear explosion of a white  

dwarf star (Type Ia) and core collapse supernovae (Type Ib/c and Type II). It is thought that 

Type IIb supernovae may be an evolutionary link between Type Ib/c and Type II supernovae due 

to the fact that they start with some observed hydrogen in their spectra, like a Type II, but lose it 

later in their lives and resemble a Type Ib/c. Our research specifically targets the evolution  

core collapse supernovae. 

  

During the final 10,000 years of the life of a star larger than eight times the mass of the Sun,  

it releases large amounts of its hydrogen envelope; it will swell into a red giant star as it  

exhausts the hydrogen fuel in its core (Chevalier, 1982).  Based on the evolution of the star,  

this ejected circumstellar material (CSM) can be smooth or clumpy with varying density as  

it moves away from the star at a speed of about 10 km/s.  Meanwhile, the core of the star  

will yield to gravity and begin shrinking inward, growing hotter and denser.  A series of  

nuclear reactions will temporarily halt the inward collapse of the star, but when the core is  

essentially just iron, it will not permit its atoms to fuse into heavier elements, and fusion  

will cease.  In less than a second, the core temperature rises to over one billion degrees as  

the  iron  atoms  are  forced  together.    In  a  process  that  is  still  not  fully  understood,  the  

repulsive  force  between  the  iron  nuclei  overcomes  the  force  of  gravity,  so  the  core  

compresses,  and  then  recoils  with  an  outward  shockwave  with  a  speed  of  more  than  

10,000 km/s.  This is called a core collapse supernova.  

  

When this shock encounters the star's outer layers, the material is heated and fuses to form  

new  elements  and  radioactive  isotopes.    The  shock  wave  then  ejects  this  newly  created  

matter  out  into  space.    The  material  that  is  ejected  from  the  star  is  now  known  as  a  

supernova remnant.  All that will remain of the original star is a small, super‐dense core  

composed almost entirely of neutrons called a neutron star.  If the original star was 25 or  

more times the mass of our Sun, even the neutrons cannot withstand the gravitational pull  

inwards, and the core collapses, forming a black hole.  

  

The hot material given off by the supernova, the radioactive isotopes, and the free electrons  

moving in the strong magnetic field that is produced by the neutron star produce X‐rays  

and  gamma  rays.    They  are  caught  in  the  strong  magnetic  field  causing  them  to  spiral  

quickly in a relativistic manner.  This acceleration is offset by emissions of photons called  

synchrotron emission.  However, sometimes the acceleration is so great that these particles  

re‐absorb their own photons in a process called synchrotron self‐absorption.  An electron  

might  also  get  caught  in  the  electromagnetic  field  of  an  ion  in  a  process  called  free‐free  

absorption.    These  high‐energy  photons  can  be  observed  in  order  to  describe  the  

properties of the preceding supernovae.    

  

Radio Analysis  

20

Radio  measurements  of  supernovae  can  create  a  detailed  image  of  the  front  of  the  blast  

wave where synchrotron self‐absorption and free‐free absorption take place (Weiler et al.,  

2002).    By  detecting  the  intensity  and  wavelength  of  the  radio  waves  given  off  by  a  

supernova, scientists are able to draw conclusions about the characteristics of the CSM of  

the star such as if it was uniform or clumpy and how far from the star it was before the star  

exploded (see Figure 1).    

  

Radio antennae are a good way to detect photons from these processes.  However, one of  

the  main  disadvantages  of  radio  antennae  is  that  they  have  low  resolving  power.    Radio  

wavelengths are about 10 5  times larger than visible light.  Therefore, if an optical and radio  

telescope were built with the same diameter, the radio telescope would have 10 5  times less  

resolving  power.    This  means  that  some  radio  telescopes  would  have  to  be  built  on  the  

order  of  10‐100  km  to  get  the  same  resolving  power  as  an  optical  telescope.    Radio  

astronomers use a technique called interferometry to rectify this problem.  If two “normal”  

sized radio antennae were placed a few kilometers apart and their received signals were  

synchronized, the separate dishes could act like a single dish and give an image of a thin  

strip  of  the  sky.    If  many  antennae  were  placed  near  each  other,  and  each  of  them  

synchronized their image with each of the other ones, then one could create a clear image  

of the sky.  

  

The Very Large Array (VLA) 4  is a collection of 27 radio antennae, each about 25 meters in  

diameter,  which  monitors  various  celestial  phenomena  such  as  quasars,  supernovae  and  

gamma  ray  bursts.  Normally,  in  order  to  study  celestial  bodies  at  a  great  distance  with  

acceptable image clarity, a huge dish must be used.  However, the VLA fixes this problem by  

having  groups  of  satellites  take  images  together  in  different  configurations.    In  the  A  

configuration, all of the antenna are spread out with each arm at 21km.  This simulates a  

single radio dish that is 36 km in diameter.  The size of the array decreases slowly with the  

B and C configurations until it is finally in the D configuration where all of the antennae are  

placed within 0.6km of the center.  When the antenna are in the A configuration, the radio  

array has the most magnification and can pick up the greatest amount of detail.  When the  

size of the array shrinks, scientists can study the overall structure of the celestial object.  

  

The Weiler et al. (2002) parameterized model is based on six parameters that describe the  

properties of the observed radio emissions.  The basis for these parameters comes from the  

work  of  Chevalier  (1990)  on  the  interactions  of  the  CSM  with  a  supernova  blast  wave.   

These parameters can provide many details about the CSM.  The features the parameters  

represent are shown in Figure 1.  

                                                         

4 The VLA of the National Radio Astronomy Observatory is Operated by Associated Universities, Inc. under a 

cooperative agreement with the National Science Foundation 

21

  

  

  

  

Figure 1: The mechanism of a core‐collapse SN  

  

22 

  

  

  

  

  

  

  

  

 

In the equations on the previous page, K1 refers to the flux of the object.  K2 is a property of  

the uniform CSM represented as τuniform.  This parameter provides information about the  

material  that  has  been  constantly  drifting  off  the  progenitor  over  the  final  thousands  of  

years of the life of the SNe.  K3 refers to the clumpy material represented by τclumpy above.   

This  material  has  been  leaving  the  progenitor  in  bursts  and  is  therefore  not  uniformly  

dense.  K4 represents the material that didn’t come from the progenitor but is still between  

the SN and Earth.  This term is generally associated with a supernova that occurs in an area  

of a dense star population and is represented as τdistant above.  The K5 term is a parameter  

that  is  affected  by  synchrotron  self‐absorption  and  free‐free  absorption.    All  of  these  

parameters are inserted into the equations on the previous page to create a fit of the radio  

data.  

  

X‐ray Analysis  

X‐ray  analysis  is  a  second  method  by  which  scientists  can  probe  the  blast  wave  of  a  

supernova.  Unlike radio analysis, X‐ray analysis probes the reverse shock and the back of  

the blast wave because there is a higher concentration of high‐energy photons produced in  

those regions (see Figure 1).  Additionally, X‐ray analysis deals with very low numbers of  

photons  –  on  the  order  of  10 ‐3   to  10 1   counts  per  second.    This  means  that  unlike  radio  

analysis (where a data set may contain thousands of data points), X‐ray analysis must use  

low photon statistics.  When a supernova is analyzed using radio and X‐ray data, a good  

model of the evolution of the supernova and its progenitor can be created.  

  

In our research we mainly receive X‐ray data from two sources – the Swift X‐ray telescope  

(XRT) and Chandra satellites.  The Swift‐XRT satellite has the advantage of being able to  

focus in on a source in a very short amount of time.  Within hours of detecting an X‐ray  

source  in  the  sky,  the  Swift‐XRT  satellite  can  focus  in  on  it  and  make  observations.   

However, its main disadvantage is that it has relatively low spatial resolution.  On the other  

hand,  Chandra  has  excellent  spatial  resolution;  its  main  disadvantage  is  due  to  the  

complicated  algorithm  that  determines  where  the  satellite  points  in  the  sky,  it  can  take  

days before Chandra focuses on an object of interest.  

23

Results 

Only five confirmed Type IIb supernovae have been measured using radio and X-ray techniques: 

SN 1993J, SN 2001ig, SN 2001gd, SN 2008ax, and SN 2008bo. Below, we present the results 

of the analysis of SN 1993J, SN 2008ax, and SN 2001ig. 

SN 1993J in NGC 3031 

SN1993J is one of the most well observed supernovae with over 680 observations taken over the 

course of thousands of days. Figure 2 shows the radio flux density of SN1993J versus time since 

the explosion; the x-axis represents the log time since the explosion occurred (measured in days), 

and the y-axis represents the electromagnetic flux density of the source (measured in milli- 

Janskys). Each colored line represents a different frequency band of data as seen in Figure 1. 

The data seems to fit the parameterized model very well up until around day 3000. After that, 

the data points sharply tend downward representing a sudden and dramatic decrease in radio flux 

intensity. This might have been caused by a sudden change in the mass loss rate of the star. The 

data fitting shows that there has to be a significant synchrotron self-absorption component to the 

evolution of SN 1993J in addition to free-free absorption. 

Figure 2: Radio light curve of SN 1993J in NGC 3031 

As a side note, the work done by Stockdale et al. (2007) shows that SN 2001gd follows a similar 

decay pattern to SN 1993J early in its evolution. It was determined from the fitting of the curve 

that a significant synchrotron self-absorption component had to be present in SN 2001gd. 

24

SN 2008ax in NGC 4490 

SN2008ax is a unique Type IIb supernova because it does not seem to follow the same pattern as 

SN1993. Firstly, its characteristic decay pattern occurs much faster than normal. Supernovae 

often take thousands of days to progress through the radio light curve cycle that we can detect on 

Earth. However, SN2008ax has already begun the downward slope of the curve for all 

wavelengths after less than 100 days. This suggests that it may actually be a yellow giant star 

rather than a red giant star due to its higher temperature, apparently smaller radius, and thinner 

solar wind. Secondly, we see that there are many fluctuations in its decline, indicating that the 

mass is not uniformly distributed throughout the star. This may be caused by a non-uniform 

CSM or by an interaction with a nearby star. Figure 3 shows the radio flux density of SN2008ax 

versus time since explosion; the x-axis represents the log time since the explosion occurred (in 

days), and the y-axis represents the electromagnetic flux density of the source (measured in 

milli-Janskys). 

Figure 3: Radio light curve of SN 2008ax in NGC 4490 

25

SN 2001ig in NGC 7424 

Like SN2008ax, SN2001ig does not seem to follow the normal parameterized model in the radio 

spectrum. As the radio flux decays we see many fluctuations in its evolution that indicate that 

the mass is not uniformly distributed throughout the star. We predict that the bumps in the radio 

light curve are caused by a nearby star that is pulling in some of the supernova remnant as it is 

ejected outward. Figure 4 below shows the radio flux density of SN2001ig versus time since 

explosion; the x-axis represents the log time since the explosion occurred (in days), and the yaxis 

represents the electromagnetic flux density of the source (measured in milli-Janskys). 

We also note that the decay pattern of SN2001ig occurs much faster than normal. Though it did 

not evolve as fast as SN2008ax, SN2001ig has already begun the downward slope of the curve 

for all wavelengths after about 300 days, indicating a thinner solar wind (see Figure 4). 

Figure 4: Radio light curve of SN 2001ig in NGC 7424 

Conclusions 

It is thought that Type IIb supernovae may be the link between Type Ib/c and Type II 

supernovae. In our analysis, we are seeing two different variations of Type IIb supernovae; 

SN1993J shows a sudden drop-off in radio flux density after the initial steady decline while 

SN2008ax and SN2001ig show bumps in the radio light curve that are not explained by the 

standard parameterized model. Additionally, more testing is needed on a bigger sample of Type 

IIb supernovae to determine if Type IIb supernovae are an evolutionary link between Type Ib/c 

and Type II supernovae and if there are different types of Type IIb supernovae. 

26

References 

Chevalier, R. A., “Interaction of supernovae with circumstellar matter,” 1990, Supernovae, p. 91 

Chevalier, R. A., “The radio and X-ray emission from type II supernovae,” 1982, Astrophysical 

Journal, Vol. 259, pg. 302-310 

Stockdale, C. J.; Kaster, B. C.; Kelley, M. T.; Panagia, N.; Sramek, R. A.; Van Dyk, S. D.; Weiler,  

K.  W.;  Williams,  C.  L.  M.,  “Recent  Type  II  Radio  Supernovae,”  2006c,  Kavli  Institute  for  

Theoretical Physics (KITP) Conference: Supernova and Gamma‐Ray Burst Remnants (Feb  

6‐10, 2006) Coordinators: Roger Chevalier, Una Hwang, Martin Laming  

  

Stockdale, C. J.; Williams, C. L.; Weiler, K. W.; Panagia, N.; Sramek, R. A.; Van Dyk, S. D.; 

Kelley, M. T., “The Radio Evolution of SN 2001gd,” 2007, Astrophysical Journal, Volume 671, 

pp. 689-694 

  

Weiler,  K.,  Panagia,  N.,  Montes,  M.,  Sramek,  R.,  “Radio  Emission  from  Supernovae  and  

Gamma‐Ray Bursters”, 2002, Annual Review Astrophysics, p. 393  

  

Woosley, S. E., Pluto, P. A., Martin, P. G., & Weaver, T. A., “Supernova 1987A in the Large  

Magellanic  Cloud  ‐  The  explosion  of  an  approximately  20  solar  mass  star  which  has  

experienced mass loss?” 1987, Astrophysical Journal, Vol. 318, pg. 664 

27


Part Seven 

Physics

Computational Fluid Dynamical Model of a Cyclone 

Separator in Microgravity 

Kevin M Crosby and Brad Frye 

Department of Physics, Carthage College, Kenosha, WI 

Abstract 

Collection efficiencies and operational characteristics of a small air cyclone with a low-density 

lunar dust load are calculated under different gravitational conditions using computational fluid 

dynamics (CFD) methods. In agreement with our experimental results obtained on a NASA 

microgravity research aircraft, the collection efficiency is largely independent of the strength of 

the external gravitational field. We propose a simple analytical model that explains these results. 

Introduction 

An air cyclone is a device that separates particles from a carrier air stream by means of a centrifugal 

force acting on the particles. The essential geometry of an air cyclone is depicted in Fig. 1. Dust 

particles, initially entrained in the air flow, enter the tangential inlet near the top of the cyclone, 

and follow the downward spiral of the air vortex. Centrifugal force and inertial effects act on the 

particles to move them outward toward the inner wall of the cyclone where they are trapped in the 

boundary flow. Trapped particles eventually move down the inner wall and are collected in a dust 

cup at the base of the cyclone while the air flow reverses direction near the base of the cyclone, 

and exits the through the vortex finder at the top of the cyclone. 

Air cyclones are a promising technology for first stage air filtration in future lunar habitats where 

lunar dust mitigation is a mission critical concern. Primary among the advantages of the air cyclone 

as a potential technology for lunar habitats are design simplicity, the reduced operating costs associated 

with the lack of consumables (filtration media), and the general robustness against failure of 

air cyclones due to the lack of moving parts. While much research has been directed at microgravity 

studies of liquid phase separation in multiphase fluids in liquid cyclones, comparatively little 

is known about the operation of air cyclones in microgravity[Ahn et al., 2000]. Our experimental 

work with cyclones in microgravity as part of NASAs Systems Engineering Educational Discovery 

(SEED) program demonstrated that the operational characteristics of an air cyclone are not 

significantly different in lunar gravity [Pennington et al., 2008]. This result suggests that further 

engineering studies are warranted in order to establish the viability of cyclone filtration in lunar 

habitats. In this paper, we report the results of a computational fluid dynamics (CFD) study of the 

cyclone used in our experimental work in reduced gravity. 

1

Figure 1: The geometry of the air cyclone used in this study. The cyclone consists of a straight 

cylinder with diameter 5.08 cm and length L cylinder = 10.5 cm, a cone of length Lcone = 14.0 

cm, and a cylindrical vortex finder of diameter 2.54 cm. Dust-laden air is introduced through the 

inlet of diameter 2.54 cm. A dust cup is attached to the bottom of the cyclone. 

CFD Workflow and Boundary Conditions 

The intent of this study is to model the specific operating characteristics of the cyclone used in our 

experimental work aboard the NASA C-9 microgravity aircraft. For this reason, we created a geometrically 

accurate, 3-dimensional CAD model of our cyclone using SolidWorks Design software 

[SolidWorks, Inc.]. Our cyclone is illustrated in Fig. 1. The overall length of the cyclone body is 

L = Lcone + L cylinder = 24.5 cm. 

The CAD model was imported into CFDesign which was used to create the volume mesh for our 

calculations, and to perform the CFD calculations [Blue Ridge Numerics, Inc.]. The volume mesh 

consists of 44,500 nodes. CFDesign implements the standard k − ɛ model for evolving solutions of 

the Navier Stokes equations for turbulent fluid flow [Launder et al., 1974]. Fluid dynamics in the 

2

context of the k − ɛ model involves the solution of transport equations for turbulent kinetic energy 

within the isotropic turbulent viscosity approximation. This model is well-validated for small-scale 

turbulence, but is known to fail to reproduce accurate inner vortex dynamics in larger cyclones 

(L > 1 m). The failure of the isotropic turbulent viscosity approximation to accurately model flow 

with high vorticity and turbulent flow that occurs in large systems is not a concern in the present 

study; our intent is not to accurately reproduce experimental results, but rather to investigate the 

qualitative features of particle motion in the presence of different gravitational fields. 

Boundary conditions for the CFD calculations were derived from the experiments performed on 

the C-9 aircraft. In particular, an outflow of vout = 10.0 m/s is imposed on the top of the vortex 

finder, and a zero gauge pressure condition is imposed on the inlet to correspond to the open 

atmosphere condition in the experiment. The inlet velocity is sufficiently low that the flow can be 

considered incompressible, greatly simplifying the CFD calculation. The bottom of the cyclone is 

sealed against losses, while the entire inner surface of the cyclone is presumed to have a coefficient 

of restitution, e = 0.5. A coefficient of restitution, e < 1 ensures that particles lose energy on 

contact with the walls, and so are eventually captured by the walls, as happens in the operation of 

a real cyclone. 

Collection Efficiency 

The collection efficiency for a cyclone separator is a measure of the relative number of particles 

trapped in the cyclone at a given particle diameter. Given a discrete spectrum of particle diameters 

di, the collection efficiency for the i − th particle type is: 

ɛi = Ni − N ′ i 

Ni 

where Ni is the number of particles of diameter di present at the inlet, and N ′ i 

(1) 

is the number of 

particles of diameter di that escape collection and exit the cyclone through the vortex finder. In our 

CFD calculations, we introduce a monodisperse spectrum of particles with diameters between 0.1 

and 15 µm to the inlet of the cyclone. The particles have mass densities ρp = 2900 kg/m 3 , a value 

chosen to match the mass density of the lunar dust simulant, JSC-1AF used in our experiment 

[Orbitec, Inc.]. In each calculation run, 100 particles at each diameter are injected into the cyclone 

with a common initial speed that match the inlet gas flow rate of Vi = 10.0 m/s. 

Two sets of calculations are performed for each set of particles. In the first calculation, the particles 

are subject only to the centrifugal and inertial effects that result from their mass. Gravity does not 

act on the particles in this case. In the second set of calculations, the same spectrum of particles is 

introduced to the cyclone with the same initial conditions, but with a 1-g gravitational acceleration 

acting in the axial direction. Representative particle traces for particles of diameter dp = 0.1µm 

and dp = 10µm are shown in Fig. 2. The qualitative features of particle motion in a cyclone are 

reproduced in our CFD calculations. In particular, the smaller mass associated with the dp = 0.1µm 

3

particles results in reduced centrifugal and inertial forces on the particles, and the particles stay 

largely entrained in the air flow as it travels through the cyclone. Most of these smaller particles 

escape with the air through the axial vortex finder. The heavier (dp = 10µm) particles experience 

larger inertial and centrifugal forces, and travel to the walls of the cyclone where they are eventually 

trapped. 

Figure 2: Representative particle traces for particles of diameter (a) 0.1 µm, and (b) 10.0 µm. The 

larger particles are more efficiently captured in the cyclone, while most of the smaller particles 

escape. 

Collection efficiencies for each set of particles are calculated according to Eqn. 1. The efficiency 

results for both zero and 1-g calculations are displayed in Fig. 3. There is no statistically significant 

difference between the collection efficiencies obtained under the different gravitational fields. This 

result agrees well with our experimental data which are also displayed in Fig. 3. For particles 

with 1.0µm < dp < 5µm, the CFD calculations underestimate the efficiency of particle capture 

relative to experimental data. This is due to the inability of the standard k − ɛ model to reproduce 

the strong inner vortex present in real cyclones. In a real cyclone, the inner vortex core extends 

nearly the entire length of the cyclone and serves to provide particle trajectories that can also span 

the length of the cyclone. In contrast, the CFD calculations produce an axially compressed inner 

vortex that results in a “short-circuit” of the flow for particles entrained in the inner vortex. The 

4

Figure 3: Experimental results (• and ⋄), CFD results(� and ×), and Lapple Model predictions 

(�) for the performance of the model cyclone used in this study. The width of the error bars 

on the experimental data represents the uncertainty in particle size measurements in the particle 

detector used. The heights of the error bars on the experimental data are the standard deviations of 

the collection efficiency measurements. The numerical data has sampling errors smaller than the 

symbol size. The Lapple model data is a fit of the cyclone performance predictions derived from 

the work in Ref. [Shepherd et al., 1940]. 

smaller particles do not travel far down the cyclone before the truncated vortex flow carries them 

out the vortex finder. 

Analytic Model of Particle Collection 

We can understand the somewhat unexpected result that gravity does not play a significant role in 

particle capture in our cyclone through a heuristic model that incorporates the important physics of 

particle motion in a vortex flow. We make the reasonable assumptions that (a) the particles do not 

5

interact with each other (low density dust load), and (b) the particles are sufficiently small that they 

do not affect the flow characteristics. The latter condition leads to a requirement that the air flow 

is laminar in the presence of the particles, and the drag force acting on the particle is governed by 

Stokes’ Law, FD = −3πηdpv. Here, η = 1.75 × 10 −5 Pa-sec. is the kinematic viscosity of dry air, 

dp is the particle’s aerodynamic diameter, and v is the particle’s velocity. 

The motion of a particle initially entrained in the air flow is subject to the forces identified in Fig. 

4. Let the density of the carrier air stream be ρg, and the particle density be ρp. The axial forces 

are gravity F weight = −(πd 3 pρpg/6)ˆz, the gravitational buoyancy force, FB = (πρgd 3 pg/6)ˆz, and a 

Stokes drag force FD = 3πηdp˙zˆz. In the rotating frame of the particle, the radial forces include 

the centrifugal force exerted by the air stream, FC = (πρpd 3 pr˙θ 2 /6)ˆr, the opposing radial buoyancy 

force F ′ B = (πd3 pρgr˙θ 2 /6)ˆr, and the Stokes drag F ′ D = −3πηdp˙rˆr. For the small particles of interest 

here, the tangential velocity of the particle is the same as that of the carrier air stream. Finally, we 

make the simplifying assumption that the tangential motion of the particle is rigid body-like, so 

that we can define the constant angular speed of the particle as ω ≡ ˙θ. 

Figure 4: Free body diagram of a particle subject to buoyancy and drag forces in the radial and 

axial directions. Weight F weight , Stokes drag FD, and buoyancy force FB govern axial motion. 

In the (non-inertial) frame of the particle, an outward centrifugal force Fc acts in opposition to a 

radial buoyancy force F ′ B , and a radial drag force F′ D . 

Axial Motion 

For the small particles considered here, axial accelerations act only briefly, and the axial motion of 

a particle is largely governed by the terminal velocity condition ¨z = 0. Force balance in the axial 

6

direction results in the terminal speed 

which is on the order of 10 −4 m/s for 1 µm particles. 

Radial Motion 

vzT ≡ ˙z| terminal = − (ρp − ρg)d2 p 

g (2) 

18η 

The radial and tangential motions in the cyclone are coupled, but the analysis simplifies considerably 

under the rigid body assumption of constant angular speed. In this case, we find the radial 

motion to satisfy the equation of motion, 

¨r = − 

� 

1 − ρg 

� 

rω 

ρp 

2 − 18η 

ρpd2 ˙r. (3) 

p 

Again, for the small particles considered here, radial accelerations are transient, and we can safely 

consider the case ¨r ≈ 0. In this case, Eqn. 3 simplifies to 

Residence Time 

˙r = (ρp − ρg) 

18η ω2 d 2 pr. (4) 

The residence time of a particle in a a cyclone is defined to be the time spent by the particle from 

entry at the inlet to entrainment at the boundary flow near the wall or dust cup. A particle is 

considered to have been captured if the particle residence time is less than the residence time of the 

air as it flows through the cyclone. This condition is a rough guide to expected collection efficiency 

and results in an empirical estimate of the collection efficiencies for a given particle diameter and a 

given cyclone geometry. The cyclone consists of two segments, a cylinder with fixed outer radius 

R, and a cone with radius R(z). We can estimate the residence time of particles in our cyclone by 

integrating Eqn. 4 from an initial radial coordinate R0 to the outer radius of the cyclone R. For 

simplicity, we initially consider motion constrained to the fixed-radius cylinder. Typically, R0 is 

considered to be the radius of the vortex finder. We find the particle residence time 

18η 

τresidence = 

(ρp − ρg)ω2d2 � � 

R 

log . (5) 

p R0 

Axial and radial motions are decoupled, so that τ residence is independent of gravitational acceleration 

g. It is important to note that, although we’ve made a gross simplification of the tangential 

motion by assuming rigid body motion (ω ≡ ˙θ = constant), the essential result that particle residence 

times are independent of gravity does not rely on the specific model of tangential motion. In 

general, the residence time of a particle in a cyclone scales as (ωdp) −2 . 

7

We’ve assumed that particle motions are confined to the cylinder segment of the cyclone. Relaxing 

this constraint implies that we consider the wall of the cyclone to be described by the function 

� 

R Lcone ≤ z ≤ Lcone + L 

R(z) = 

cylinder . (6) 

R0 + ((R − R0)/Lcone)z z ≤ Lcone 

In this case, residence times will be a weak function of z through the log(R(z)/R0) dependence, 

and are therefore weakly coupled to the gravity-driven axial motion (Eqn. 2). We expect in this 

case that reduced gravity may have a “second order” effect on residence times, slightly reducing 

capture efficiencies. Our experiments with lunar dust simulant did not have the resolution to discern 

a difference in collection efficiencies between lunar and earth gravity. We are currently designing 

more sensitive cyclone experiments that will have the resolution to identify gravitationally induced 

differences in cyclone performance if they are actually present. 

Summary and Future Directions 

We have performed CFD calculations on a model of the small cyclone separator used in our microgravity 

experiments on cyclone performance in lunar gravity. Our CFD calculations agree well 

with the central finding of the experiment: Gravity does not play a significant role in determining 

collection efficiencies over the range of particles diameters typically present in lunar dust. By 

means of a simple, heuristic model of particle motion in a cyclone, we can understand these results 

in terms of the magnitudes of drag and buoyancy forces acting on small particles in the Stokes 

regime appropriate to our test particles. 

The results obtained in this paper suggest that experiments with higher resolution in particle size 

and collection efficiency may establish useful bounds on the effect of gravity on collection efficiency 

in full-scale cyclone separators currently being considered for deployment in future lunar 

habitats. 


The authors are grateful to the Wisconsin Space Grant Foundation and to Carthage College for 

the financial support of this work. The authors would also like to acknowledge the contributions 

of Juan Agui at NASA Glenn Research Center for conceiving of and initiating the experimental 

program of cyclone performance in reduced gravity. 

References 

[Ahn et al., 2000] Ahn, H., Tanaka, K, Tsuge, H., Terasaka, K., and Tsukada, K., Centrifugal gas 

- liquid separation under low gravity conditions, Separation and Purification Technology 19, 

8

121 (2000). 

[Blue Ridge Numerics, Inc.] Blue Ridge Numerics, Inc., 650 Peter Jefferson Place, Suite 250, 

Charlottesville, VA 22911. 

[Launder et al., 1974] Launder, B. E. and Sharma, B. I., Application of the energy-dissipation 

model of turbulence to the calculation of flow near a spinning disc, Letters in Heat and Mass 

Transfer 1, 131 (1974). 

[Orbitec, Inc.] JSC-1AF lunar regolith dust simulant manufactured by Orbitec, Inc. and provided 

by Juan Agui, NASA Glenn Research Center. 

[Pennington et al., 2008] Pennington, C., Martin, E., Sorensen, E., Fritz, I, Frye, B., 

and Crosby, K. M., Inertial Filtration in Lunar Gravity: SEED Project Report: 

http://www.carthage.edu/dept/physics/flight/Final Report.pdf (2008). 

[Shepherd et al., 1940] Shepherd, C. B., and Lapple, C. E., Flow Pattern and Pressure Drop in 

Cyclone Dust Collectors, Industrial and Engineering Chemistry 32, 1246 (1940). 

[SolidWorks, Inc.] SolidWorks Corporation, 300 Baker Avenue, Concord, MA 01742. 

9

Simulation of Fast Magnetic Reconnection using a Two-Fluid Model of Collisionless Pair 

Plasma without Anomalous Resistivity 1 

Overview 

E. Alec Johnson 2 

James A. Rossmanith 3 

Department of Mathematics 

UW-Madison 

Abstract 

For the first time to our knowledge, we demonstrate fast magnetic reconnection near a 

magnetic null point in a fluid model of collisionless pair plasma without resorting to the contrivance 

of anomalous resistivity. In particular, we demonstrate that fast reconnection occurs 

in an anisotropic adiabatic two-fluid model of collisionless pair plasma with relaxation toward 

isotropy for a broad range of isotropization rates. For very rapid isotropization we see fast 

reconnection, but instabilities eventually arise that cause numerical error and cast doubt on the 

simulated behavior. 

Motivating Problem. We have been working to develop algorithms that efficiently model 

fast magnetic reconnection in collisionless space plasmas. A plasma is a gas of charged particles 

and the accompanying magnetic field that it carries. The ability to simulate plasmas efficiently over 

a wide range of phenomena and scales is essential to understanding and predicting the behavior 

both of space plasmas and of industrial fusion plasmas. 

The two basic types of plasma models are kinetic models and fluid models. Kinetic models represent 

particles or evolve the space-velocity distribution of particles. Fluid models evolve moment 

averages (e.g. density, momentum, or energy) of assumed (e.g. Maxwellian) velocity distributions. 

Fluid models give good accuracy for highly collisional plasmas. For rarefied space plasmas, however, 

particularly near magnetic null points, the regular velocity distributions assumed by fluid 

models often fail to hold, since the particles move essentially unconstrained. 

The phenomenon for which fluid models of plasma have been most apt to fail is collisionless fast 

magnetic reconnection. It is precisely this phenomenon which has proved most critical to understanding 

and predicting the volatile dynamics of astrophysical plasmas, including solar storms and 

geomagnetic substorms in Earth’s magnetosphere. The critical physical role of fast reconnection 

and the failure of fluid models to capture it has prompted extensive studies using particle-based 

simulations of collisionless magnetic reconnection. Our objective is to study the ability of fluid 

models to match particle-based simulations of collisionless fast magnetic reconnection. We have 

1 This research was supported by a Wisconsin Space Grant Consortium Graduate Fellowship for 2008-2009. 

2 ejohnson@math.wisc.edu 

3 rossmani@math.wisc.edu 

11

concentrated our effort specifically on fast magnetic reconnection in collisionless pair plasmas. A 

pair plasma is a plasma whose positively and negatively charged particles have the same chargeto-mass 

ratio. The physical example of a pair plasma is an electron-positron plasma, of interest to 

astrophysicists. 

Historical development. The GEM magnetic reconnection challenge problem [3] identified 

Hall effects as critical to fast magnetic reconnection in electron-ion plasmas. Since Hall effects 

are absent for electron-positron (pair) plasmas, this prompted Bessho and Bhattacharjee [1, 2] to 

demonstrate via particle simulations that fast magnetic reconnection occurs even in collisionless 

pair plasma, which they attributed to pressure anisotropy. 

The next challenge was to demonstrate fast reconnection in a fluid model of pair plasma. Assuming 

the ubiquitous presence of a strong background magnetic guide field (which constrains charged 

particles to move in tight spirals) allowed Chacón et al. [4] to develop an analytical fluid theory of 

fast reconnection in magnetized pair plasma. 

For the case where there is a magnetic null point, Zenitani et al. [10] demonstrated fast reconnection 

in a two-fluid model of relativistic isotropic pair plasma. Their model assumes a spatially 

dependent anomalous resistivity, as has been used with resistive single-fluid MHD to simulate fast 

reconnection. By selecting a resistivity with an anomalously high value near the X-point, one can 

essentially prescribe the desired rate of reconnection (as determined from PIC simulations); the 

elusive goal is to find a simple and generic expression for anomalous resistivity that works for a 

broad range of problem conditions. We remark that resistive single-fluid MHD does not assume 

any particular ratio of mass-to-charge ratios of the two species and thus (with an appropriate choice 

of anomalous resistivity) could be used as a fluid model of pair plasma. 

Our work. Rather than resort to an anomalous resistivity, we seek generic two-fluid 

moment closures that give reconnection behavior in agreement with particle-based simulations. 

We have studied reconnection in five-moment (isotropic) and ten-moment (anisotropic) adiabatic 

fluid models of collisionless pair plasma with varying rates of relaxation toward isotropy. We implemented 

conservative shock-capturing Discontinuous Galerkin two-fluid five-moment and tenmoment 

plasma models, following Hakim et al. [5, 6]. 

We initially adopted the modified GEM settings of [1, 2], but found that for pair plasmas the 

large aspect ratio of the reconnection region gives rise to a secondary instability, namely, the unpredictable 

formation of magnetic islands, making it difficult to obtain demonstrably converged 

results, as seen in our paper, [8]. Essentially, for the pair plasma case of the GEM problem, in contrast 

to the electron-proton case, the tearing instability wants to produce smaller magnetic islands; 

since the GEM problem is the formation of one big magnetic island, we can avoid the instability 

by reducing the size of the domain. Pair plasma involves no need to resolve scale separation between 

species, so arguably a smaller domain is acceptable. Therefore, to eliminate the secondary 

instability and to reduce computational expense we multiplied the dimensions of the domain by 

one half. 

We also chose to focus on the case where both species have the same temperature. In this case there 

is complete symmetry between the two species, and the number of equations needed is halved. In 

12

the case of zero guide field the GEM problem is symmetric about both the horizontal and vertical 

axes. We enforced all these symmetries, reducing computational expense by a factor of eight. 

We simulated the GEM problem using the collisionless adiabatic ten-moment pair plasma model 

supplemented with a globally prescribed rate of pressure tensor isotropization. We have neglected 

all diffusivities and relaxation terms except isotropization. We varied the rate of pressure 

isotropization and studied the resulting variation in the rate of reconnection and the contributions 

of the terms in Ohm’s law. 

Model 

Generic physical equations for the ten-moment two-fluid model are: 

• conservation of mass for each species: 

∂tρs + ∇ · (ρsus) = 0, 

• conservation of momentum for each species: 

∂t(ρsus) + ∇ · (ρsus ⊗ us + Ps) = qs 

ρs(E + us × B) + Rs, 

• evolution of the pressure tensor for each species: 4 

∂tPs + 3∇ · (us ∨ Ps) + ∇ · Qs = 2Sym( qs 

Ps × B) + Rs, 

• Maxwell’s equations for evolution of electromagnetic field: 

∂tB + ∇ × E = 0, 

∂tE − c 2 ∇ × B = −J/ε, 

• and Maxwell’s divergence constraints: 

∇ · B = 0, 

∇ · E = σ/ε. 

ms 

In these equations, Sym denotes the symmetric part of the argument tensor (i.e. the average over 

all permutations of subscripts), ∨ denotes symmetric outer product (i.e. the symmetric part of 

the tensor product), and i and e are positive (“ion”) and negative (“electron”) species indices; for 

species s ∈ {i,e}, qs = ±e is particle charge, ms is particle mass, ns is particle number density, 

ρs = msns is mass density, σs = qsns is charge density, usρs is momentum, Js = usσs is current 

4 For conservation and shock-capturing purposes we actually evolve the energy tensor Es := Ps +ρsusus rather than 

directly evolving the pressure tensor. 

13 

ms

density, and Ps is a pressure tensor; B is magnetic field, E is electric field, c is the speed of light, 

ε is vacuum permittivity, J = Ji + Je is net current density, and σ = σi + σe is net charge density. 

To close the system, constitutive relations must be supplied for the generalized heat fluxes Qs, 

the interspecies drag force on the ions Ri = −Re, and Rs, the production of generalized thermal 

energy due to collisions. We nondimensionalize these equations, choosing the timescale to be the 

gyroperiod of a typical particle of mass 1, and choosing the typical velocity to be a typical Alfvén 

speed. The nondimensionalized equations retain the form of the dimensional equations above, with 

the simplifications that e = 1, mi + me = 1, and 1 ε = c2 . 

In our closure we assume that Rs = 0, and to provide for isotropization we let 

Rs = 1 

� 

1 

3 (trPs)I 

� 

− Ps , 

τs 

where τs is the isotropization period of species s, tr denotes tensor trace, and I is the identity tensor. 

In our present work we assume that Qs = 0. This assumption might not be satisfactory, though: the 

particle simulations of Hesse et al. [7] showed, at least for the case of guide-field electron-proton 

reconnection, that generalized heat flux contributions to the evolution of the pressure tensor are 

necessary to obtain an appropriate approximation for the pressure nongyrotropy near the X-point. 

We therefore plan to investigate C. David Levermore’s closure [9], 

Qs = 9 

5 (ν0 − ν1)I ∨ tr � ∇ ∨ Θ −1� 

� 

s + 3ν1 ∇ ∨ Θ −1 

� 

s , 

where Θs := Ps/ρs and ν0 � ν1 is proportional to collision frequency. We expect to determine 

whether we can get fast reconnection without isotropization by using such a non-vanishing generalized 

heat flux. 

Ohm’s law. Combining the momentum equations gives net current balance. Assuming 

quasineutrality (zero net charge) and solving for electric field gives Ohm’s law for the electric 

field: 

E = ˜mi + ˜me 

(−Ri) 

ρ 

(resistive term) 

+ B × u (ideal term) 

+ ˜mi − ˜me 

J × B 

ρ 

(Hall term) 

+ 1 

ρ ∇ · ( ˜mePi − ˜miPe) (pressure term) 

+ ˜mi ˜me 

ρ 

� 

∂tJ + ∇ · � uJ + Ju + ˜me − ˜mi 

ρ 

JJ �� 

(inertial term), 

where ˜mi := mi 

e and ˜me := me 

e and the resistive term is usually assumed to be of the form η · J, i.e., 

a linear function of current. 

14

GEM magnetic reconnection challenge problem 

The GEM magnetic reconnection challenge problem studies the evolution of 2-dimensional plasma 

in a rectangular box aligned with the coordinate axes and centered at the origin. The top and 

bottom of the box are conducting walls and periodic symmetry in the x-axis defines the width of 

the box. The plasma is initially in near-equilibrium. The upper half of the box is occupied by 

strong magnetic field lines pointing to the right and the lower half is occupied by strong magnetic 

field lines pointing to the left, separated by a thin, potentially volatile transition layer along the 

x-axis. Some studies (e.g. [4]) add a constant out-of-plane component to the magnetic field, called 

a “guide field”. We do not have a guide field (nor did the original GEM problem or the studies we 

are trying to replicate). 

Domain. The computational domain is the rectangular domain [−Lx/2,Lx/2]×[−Ly/2,Ly/2]. 

The problem is symmetric under 180 degree rotation around the origin, and in the case of zero guide 

field is also symmetric under reflection across either the horizontal or vertical axis. In the original 

GEM problem, Lx = 8π and Ly = 4π. We halved the dimensions, so that Lx = 4π and Ly = 2π. 

Boundary conditions. The domain is periodic along the x-axis. The boundaries parallel 

to the x-axis are thermally insulating conducting wall boundaries. A conducting wall boundary 

is a solid wall boundary (with slip boundary conditions in the case of ideal plasma) for the fluid 

variables, and the electric field at the boundary has no component parallel to the boundary. We 

also assume that magnetic field does not penetrate the boundary. 

Model Parameters. We set the speed of light to 10 (rather than 20 as in [1]) and set the 

mass of each species to 0.5 (rather than the GEM values of 25/26 for ions and 1/26 for electrons). 

Initial conditions. The initial conditions are a perturbed Harris sheet equilibrium. The 

unperturbed equilibrium is given by 

B(y) = B0 tanh(y/λ)ex, p(y) = B20 n(y), 

2n0 

ni(y) = ne(y) = n0(1/5 + sech 2 (y/λ)), pe(y) = Te 

Ti + Te 

E = 0, pi(y) = Ti 

On top of this the magnetic field is perturbed by 

δB = −ez × ∇(ψ), where ψ(x,y) = ψ0 cos(2πx/Lx)cos(πy/Ly). 

In the GEM problem the initial condition constants are 

Ti + Te 

p(y), 

p(y). 

Ti/Te = 5, λ = 0.5, B0 = 1, n0 = 1, ψ0 = B0/10; 

we reset the initial temperature ratio to 1 get symmetry between the species, and we set ψ0 to 

r 2 s B0/10, where rs = 0.5 is our domain rescaling factor, so that in the vicinity of the X-point our 

initial conditions agree (up to first-order Taylor expansion) with the initial conditions of the GEM 

problem. 

15

Properties of the GEM problem 

Reconnected flux. We define magnetic reconnection to be the loss of magnetic flux 

through the vertical axis into the first quadrant. Using Faraday’s law, ∂tB + ∇ × E = 0, one can 

show that the rate of reconnection is minus the value of the out-of-plane component of the electric 

field at the origin (i.e. the X-point) [8]. 

Ohm’s law at the origin. Since the electric field at the origin is the rate of reconnection, 

we are lead to study the terms of Ohm’s law for the electric field at the origin. Since the problem 

is symmetric under 180 degree rotational symmetry about the origin, only the out-of-plane component 

of vectors is nonzero at the origin. In Ohm’s law only the out-of-plane components of the 

resistive, pressure, and inertial terms survive. 

As a proxy for Ohm’s law5 , we select a species and write the momentum equation solved for the 

electric field: 

� 

� −Ri 

dt(reconnected flux) = E3 

� = + origin eni 

∇ · Pi 

+ 

eni 

mi 

e ∂tui 

� � 

� 

� 

� . 

3� 

origin 

In a perfectly collisionless, gyrotropic plasma, the resistive term and pressure divergence vanish, 

and reconnected flux should exactly track with species velocity (a proxy for the current) at the 

origin. 

Results 

We simulated the GEM magnetic reconnection challenge problem for pair plasmas, varying the 

mesh resolution and varying the rate of isotropization from zero to instantaneous. We plotted 

the contribution of proxy Ohm’s law terms to the reconnected flux. We find that for a broad 

intermediate range of isotropization rates reconnection is fast and that the pressure term makes the 

dominant contribution to reconnected flux. 

When isotropization is very slow or absent, there is oscillatory exchange between the inertial term 

and the pressure term at roughly a typical gyrofrequency, and reconnection proceeds at a slow 

to moderate rate (our simulations for the case of no isotropization are not sufficiently resolved). 

For a broad intermediate range of isotropization, there is little oscillation and, in agreement with 

PIC simulations (see [1, 2]), the pressure divergence dominates and provides for faster reconnection. 

As the rate of isotropization becomes very fast, however, the pressure divergence is forced to 

vanish and the inertial term is the only remaining term in the equation that can provide for reconnected 

flux. This forces the current at the origin to ramp up in track with reconnected flux. The 

system seems unable to sustain this ramp-up in current, however, and numerical instability kicks 

in, as evidenced by the sudden appearance of strong, very rapid oscillatory exchange (less evident 

in the accumulation integrals shown) between the inertial term and the residual. The numerical 

5 Ohm’s law involves the approximating assumption of quasineutrality, whereas the momentum equation holds 

exactly and the inertial term reduces to a simpler form at the origin. 

16

esidual displaces the inertial term, allowing the the current to peak and then decay while the reconnected 

flux maintains the same smooth, rapid ascent seen for intermediate isotropization rates, 

as if unconcerned whether pressure, inertia, numerical resistivity – or even anomalous resistivity, 

as our cursory investigations suggest – provides for its determined course. We can conclude that 

fast reconnection at least commences in an isotropic pair plasma model, and we conjecture that 

the numerical instability we see corresponds to some physical (perhaps streaming?) instability 

that provides for an effective anomalous resistivity (whose functional form we have not analyzed). 

As expected, the five-moment simulations show agreement with instantaneous relaxation of the 

ten-moment system to isotropy. 


I thank the Wisconsin Space Grant Consortium for their support. In addition to my advisor, James 

Rossmanith, I also thank Nick Murphy, Ellen Zweibel, and Ping Zhu for helpful conversations. 

References 

[1] N. Bessho and A. Bhattacharjee. Collisionless reconnection in an electron-positron plasma. 

Phys. Rev. Letters, 95:245001, December 2005. 

[2] N. Bessho and A. Bhattacharjee. Fast collisionless reconnection in electron-positron plasmas. 

Physics of Plasmas, 14:056503, 2007. 

[3] J. Birn, J.F. Drake, M.A. Shay, B.N. Rogers, R.E. Denton, M. Hesse, M. Kuznetsova, 

Z.W. Ma, A. Bhattacharjee, A. Otto, and P.L. Pritchett. Geospace environmental modeling 

(GEM) magnetic reconnection challenge. Journal of Geophysical Research – Space Physics, 

106:3715–3719, 2001. 

[4] L. Chacon, Andrei N. Simakov, V.S. Lukin, and A. Zocco. Fast reconnection in nonrelativistic 

2D electron-positron plasmas. Phys. Rev. Letters, 101:025003, July 2008. 

[5] A. Hakim, J. Loverich, and U. Shumlak. A high-resolution wave propagation scheme for 

ideal two-fluid plasma equations. J. Comp. Phys., 219:418–442, 2006. 

[6] A.H. Hakim. Extended MHD modelling with the ten-moment equations. J. Fusion Energy, 

27(1–2):36–43, June 2007. 

[7] M. Hesse, M. Kuznetsova, and J. Birn. The role of electron heat flux in guide-field magnetic 

reconnection. Physics of Plasmas, 11(12):5387–5397, 2004. 

[8] E.A. Johnson and J.A. Rossmanith. Collisionless magnetic reconnection in a five-moment 

two-fluid electron-positron plasma. submitted, 2008. 

[9] C. David Levermore. Kinetic theory, Gaussian moment closures, and fluid approximations. 

Presented at IPAM KT2009 Culminating Retreat, Lake Arrowhead, California, June 2009. 

[10] S. Zenitani, M. Hesse, and A. Klimas. Two-fluid magnetohydrodynamic simulations of relativistic 

magnetic reconnection. The Astrophysical Journal, 696:1385–1401, May 2009. 

17

Figure 1: Increase in reconnection rate as isotropization increases. For very slow or no isotropization 

the simulations we report not sufficiently resolved. For instantaneous relaxation the positron bulk velocity 

ceases to track with reconnected flux and the residual becomes unacceptably large after t = 15, indicating 

that we are no longer solving the momentum equation faithfully. 

18

Figure 2: Convergence comparisons for coarse (left) and fine (right) meshes. For intermediate isotropization 

rates a coarser mesh is satisfactory, but for no isotropization and for fast isotropization a finer mesh 

is required. For fast and instantaneous isotropization, even for a very fine mesh, the residual in “Ohm’s 

law” (the momentum equation) becomes unacceptably large, overwhelming and displacing the inertial and 

pressure terms. The five-moment simulations and the instantaneously relaxed ten-moment simulations agree 

well where convergence is demonstrated. 

19

Figure 3: Examples of agreement of reconnected flux with minus the accumulation integral of the outof-plane 

electric field at the origin for fine and coarse mesh. (For the fine mesh the two plots are exactly 

superimposed and so are indistinguishable.) Flux across the vertical axis represents the portion of field lines 

that have not reconnected. We obtained similar excellent agreement in all our simulations. 

Figure 4: Snapshots of magnetic field lines. The reconnected flux is proportional to (the change in) the 

number of field lines through the horizontal axis of symmetry, and the unreconnected flux is proportional to 

the number of field lines through the vertical axis of symmetry. 

20


Part Eight 

Biological Science

Reflection and Refraction of Vortex Rings ∗ 

Kerry Kuehn, Matthew Moeller, Michael Schulz and Daniel Sanfelippo 

Department of Physical Sciences 

Wisconsin Lutheran College, Milwaukee, WI 

Abstract 

We have experimentally studied the impact of a planar axisymmetric vortex ring, 

incident at an oblique angle, upon a sharp gravity-induced interface separating two 

fluids of differing densities. After impact, the vortex ring was found to exhibit a variety 

of subsequent trajectories, which we have organized according to both the incidence 

angle, and the ratio of the Atwood and Froude numbers, A/F . For relatively small 

angles of incidence, the vortices tended to penetrate the interface. In such cases, the 

more slowly moving vortices, having values of A/F � 0.004, tended to subsequently 

curve back up toward the interface. Quickly moving vortices, on the other hand, tended 

to refract downward, similar to a light ray entering a medium having a higher refractive 

index. A simplistic application of Snell’s law of refraction cannot, however, account 

for the observed trajectories. For grazing angles of incidence, fast moving vortices 

tended to penetrate the interface, whereas slower vortices tended to reflect from the 

interface. In some cases, the reflected vortices executed damped oscillations before 

finally disintegrating. 

Introduction 

One of the early successes of the wave theory of light was its ability to explain the law of 

refraction by appealing to the variation of the speed of light when crossing a boundary separating 

media of different densities [Huygens, 1952]. Early models of light which emphasized 

its corpuscular nature proved less convincing in explaining refraction, largely because these 

models required that light travel more rapidly when entering a denser media, such as water 

or glass, an ostensibly counterintuitive postulate which was later rejected on experimental 

grounds [Smith, 1987]. 

With the advent of Hamilton-Jacobi theory [Wolf and Born, 1965], and later wave mechanics 

[Schrödinger, 1964], the refraction of light was eventually reconciled with a corpuscular 

nature, albeit only in a statistical sense. Quantum theory, of course, makes no predictions 

as to the trajectory of an individual photon. Nonetheless, recent experiments in high energy 

physics and quantum optics have revived a significant amount of speculation as to the 

nature, and even the structure, of the photon [Nisius, 2000, Tiwari, 2002]. 

tium. 

∗ This work was supported by a Research Infrastructure Grant from the Wisconsin Space Grant Consor- 

1

Regardless of the nature of the photon, it is interesting to inquire whether, apart from waves, 

collective excitations can exhibit refraction at the boundary separating two media. To this 

end, we have measured the trajectory of a vortex ring launched at an oblique angle toward a 

sharp gravity-induced interface separating two fluids. Based on our observations, the answer 

at which we have arrived is: yes, under certain conditions. 

Vortex ring studies. Due to their ubiquity in natural phenomena, vortex rings have been 

studied in the laboratory for decades. A detailed summary of the structure and stability of 

vortex rings may be found in review articles by Maxworthy [1972] or Shariff and Leonard 

[1992]. Many studies have focused on the interaction of a vortex ring with either a solid 

surface [Lim, 1989] or a free surface [Bernal and Kwon, 1989, Sarpkaya, 1996]. Others have 

focused on the trajectory of a vortex ring when propagating through fluid with a gradual 

vertical density gradient [Honji and Tatsuno, 1976, Scase and Dalziel, 2006]. Comparatively 

few studies have been performed on the interaction of a vortex ring with a sharp density 

interface between two fluids. We describe these in the following paragraph. It should be 

emphasized at this point that here has been no systematic study of the oblique incidence of 

a vortex ring upon a density interface separating two fluids. 

Linden [1973] has studied the normal incidence of a vortex ring on a sharp density interface. 

He found that the penetration depth of the ring into the lower, more dense fluid, depended 

upon the Froude number of the vortex ring. Dahm et al. [1989] have performed a comprehensive 

study of the normal incidence of a vortex ring on a sharp density gradient. They found 

that for thin interfaces, the interaction of the vortex ring with the interface was governed 

by an effective inverse Froude number, which they denote as R, and the Atwood number, 

A. They found that in the Boussinesq limit (A � 0) the interaction could be characterized 

by the product AR, which represents a dimensionless interface strength: for small values of 

AR, the vortex ring could penetrate the interface; as AR was increased, the interface began 

to act more like a solid interface, preventing penetration. 

Dimensionless quantities. The Atwood number, A = (ρ2 − ρ1)/(ρ2 + ρ1), is a dimensionless 

quantity which measures the density difference across an interface between two fluids. 

Here, ρ1 and ρ2 are the densities of the fluids on the top and bottom sides of the interface, 

respectively. The length scale, w/l, is a dimensionless quantity formed from the interface 

thickness, w, and the vortex ring diameter, l. The Froude number, F , is a dimensionless 

quantity which measures the relative importance of the inertial and gravitational effects on 

the flow of a mass of fluid. It has several formulations. We will use the one provided in 

Landau and Lifshitz [1987], F = v/ √ lg. Here, l and v are the diameter and speed of the 

vortex ring relative to the surrounding fluid, respectively, and g is the acceleration of gravity. 

We now combine the Froude and Atwood numbers into a single dimensionless quantity, 

A/F = ∆ρ √ lg/2ρav, which plays a similar role as the interface strength, AR, in Dahm 

et al. [1989]. Here, ρa = (ρ2 + ρ1) /2 is the average fluid density, and ∆ρ = (ρ2 − ρ1) is the 

difference between the top and bottom fluid densities. 

2

Experimental Procedure 

The left hand side of Fig. 1 shows a schematic diagram of the experimental configuration; the 

right hand side is a detailed photograph of the vortex ring launcher and some of its nearby 

components. The design of our vortex ring launcher was largely inspired by that of Lim 

[1989]. Briefly, the vortex ring launcher (a) was suspended inside of a 30 ′′ L x 12 ′′ W x 22 ′′ D 

glass aquarium (b) by an adjustable platform whose height could be controlled with a screw 

jack driven by a servo motor (c). The declination of the launcher could be adjusted manually 

using a notched protractor machined from marine brass (d). A stepper motor sealed in a 

water tight chamber in the rear of the vortex launcher was used to actuate a piston so as 

to force a precisely controlled volume of fluid through the barrel of the launcher (e). This 

formed a vortex ring which propagated at the desired angle and speed toward the density 

interface. 

h 

i 

g 

a 

b 

c 

d 

f 

e 

Figure 1: Left: A schematic diagram of the experimental configuration. Right: A photograph of 

the vortex ring launcher. See the text for identification of the labels. 

Just prior to actuation of the piston, a small quantity of neutrally buoyant red ink was injected 

into the barrel using a motor actuated syringe (f). The vortex ring launcher and ink 

injection system were controlled using a multifunction data acquisition system (g). Experimental 

control software ran on a desktop computer (h). Concurrently running video capture 

software was used to obtain movies of the vortex ring trajectory using a high-definition 

digital video camcorder (i). The image capture rate was 30 frames per second. 

3

Density Interface. To establish a sharp fluid density interface, we used a procedure similar 

to the one described in Dahm et al. [1989]. The lower fluid layer consisted of a mixture of 

deionized water and salt, the upper layer of pure deionized water. To measure the interface 

thickness, we incrementally lowered the tip of a commercially available salinity sensor across 

the interface. 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

Figure 2: The salinity profile of an air/saltwater interface (circles), a freshwater/saltwater interface 

(squares), a freshwater/saltwater interface immediately after disruption by vortex launches 

(triangles), and a freshwater/saltwater interface after allowing diffusion for six days (diamonds). 

A normalized salinity profile of the interface thus obtained is shown in Fig. 2. The origin of 

the abscissa has been shifted so as to correspond to the position of the initially established 

interface. To illustrate the resolution of the probe, we show the salinity profile of a saltwater/air 

interface, before adding a layer of freshwater (circles). The interface thickness was 4 

mm. Next, we added a layer of fresh water atop the saltwater. Immediately, we measured 

the profile (squares), and found a thickness of 12 mm. After launching several vortices, 

we again measured the profile (triangles), finding that it had widened to 25 mm. Finally, 

we allowed the saltwater to diffuse into the freshwater over six days, without any further 

disruption of the interface (diamonds). In all cases we define the interface thickness as the 

height of the region over which the salinity differed by more than 5% from its maximum, and 

its minimum concentrations. To avoid excessive interface broadening during the course of 

our experiments, we routinely drained the tank and reestablished an interface after three to 

five launches. Given the instrumental broadening of 4 mm, the average interface thickness, 

w, was assumed to be approximately 1.5 cm. 

Image processing. Movies of the vortex ring trajectories were analyzed using commercially 

available image processing software. For each movie frame, a background image was 

subtracted, and an ellipse was fit to the region of the image comprising the vortex ring. In 

this way, various vortex ring parameters, such as its major axis, orientation and center were 

calculated. A montage depicting the trajectory of a vortex ring is shown in Fig. 3. The 

circles indicate the center of the vortex ring at each instant. The line segments transecting 

them indicate the length and orientation of the major axis of the ellipse. The first three 

4 

�� 

��

frames in the upper left corner, which lack transecting lines, were identified by hand so as 

to seed the automated vortex ring tracking code. The horizontal dotted line indicates the 

position of the fluid density interface. Notable in Fig. 3 is a slight change in the trajectory 

and orientation of the vortex ring just after penetrating the fluid density interface. These 

changes are accompanied by a slight change in its velocity and diameter. 

�� 

� 

� 

�� 

�� 

�� 

� 

� 

�� 

�� 

Figure 3: A montage depicting the position, and orientation, of a vortex ring launched from the 

upper left corner of the image at an angle of 45 degrees with respect to a normal to the horizontal 

density interface (dotted line). 

Results and Discussion 

Vortex ring refraction. In the right column of Fig. 4 are shown nine representative 

trajectories of vortex rings when launched at incidence angles θi = 35 (row a), 45 (row b), 

and 60 degrees (row c). In each row, three separate launches are shown, represented from 

right to left by triangles, squares and circles. The precise value of θi for each launch is 

determined by fitting a line to the data which lies above the interface (the horizontal axis). 

It is then measured with respect to a normal to the fluid density interface. 

In the left column of Fig. 4 is shown the depth dependence of A/F for each launch. A 

logarithmic scale has been used on the abscissa to facilitate depiction of a wide range of 

data. Notice that in each row, the rightmost vortex strikes the interface with the smallest 

velocity, and hence the largest value of A/F . In such cases, once below the interface, the 

vortex begins to curve upward. This suggests that for small Froude numbers, the buoyancy, 

rather than inertia, determines its trajectory. Proceeding leftward in each row, the vortices 

strike the interface with diminishing values of A/F . In particular, notice that the leftmost 

trajectory (circles) exhibits a distinct downward deflection, as emphasized by a broken line 

fit to the data beneath the interface. Such refraction occurs at relatively large values of the 

Froude number, when A/F � 0.004. 

5 

�� 

�� 

��

�� 

�� 

�� 

� 

� 

�� 

�� 

�� 

�� 

�� 

� 

� 

�� 

�� 

�� 

� 

� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

Figure 4: Right column: The trajectories of vortex rings when launched at incidence angles 35 (row 

a), 45 (row b), and 60 degrees (row c). In each row, three separate launches are shown, represented 

from right to left by triangles, squares and circles. Left column: The depth dependence of log(A/F ) 

for each vortex launch. 

6 

�� 

�� 

��

Vortex ring reflection. Similarly, the right column of Fig. 5 depicts representative trajectories 

for vortex rings launched at larger incidence angles: θi = 72 (row a) and 82 degrees 

(row b). And the left column depicts the corresponding depth dependence of A/F . Notice 

that at each incidence angle, the rapidly moving vortices (squares and circles) penetrate the 

interface. But the slowly moving vortex (triangles) is reflected from the interface. Also, in 

row (a), the reflected vortex exhibits one cycle of damped oscillations prior to disintegrating. 

This is likely due to entrainment of the surrounding fluid, which leads to alternating upward 

and downward forces on the ring. 

�� 

�� 

� 

�� 

�� 

�� 

�� 

�� 

�� 

�� 

Figure 5: Right column: Trajectories for vortex rings when launched at incidence angles 72 (row a) 

and 82 degrees (row b). Left column: The depth dependence of log(A/F ) for each vortex launch. 

A summary of the dimensionless quantities calculated from the trajectory data represented 

by various symbols in Figs. 4 and 5 is provided in Tab. 1. There, the listed values of the 

dimensionless length scale, w/l, Froude number, F and interface strength, A/F , indicate 

their values an instant before the vortex ring strikes the interface. The refraction angle is 

only provided in cases where refraction is clearly discernible. As revealed in the final two 

columns of Tab. 1, a simplistic application of the law of refraction cannot account for the 

data. In particular, the ratio of the sines of the angles of incidence and refraction differs 

significantly from the ratio of the average values of the velocities measured above, and below 

the interface. 

On the one hand, it may seem intuitively clear that a vortex ring would experience refraction 

when striking an interface. After all, it is an extended structure, the different components of 

which can travel at different velocities, similar to a wavefront. On the other hand, a theory 

of vortex ring refraction would likely need to account for vortex ring tension, which offers 

resistance to deformation of its structure. A generalized form of the law of refraction may 

also need to explicitly incorporate the Froude and Atwood numbers, or a ratio of the two. 

7 

�� 

��

Table 1: Dimensionless quantities for the data represented by various symbols in Figs. 4 and 5. 

Fig. Row Sym. w/l F A/F θ ◦ i θ ◦ r 

sin θi 

sin θr 

4 a • 0.8 1.29 0.00199 35.4 33.9 1.04 1.42 

� 0.9 0.60 0.0042 35.0 − − − 

� 0.8 0.27 0.0096 35.3 − − − 

b • 0.9 0.72 0.0033 45.1 41.7 1.06 1.33 

� 0.8 0.51 0.0047 44.7 − − − 

� 1.0 0.27 0.0087 44.7 − − − 

c • 0.7 1.90 0.00225 59.2 58.4 1.01 1.44 

� 0.8 0.98 0.0044 59.9 − − − 

� 0.7 0.10 0.044 60.7 − − − 

5 a • 2.5 0.84 0.0076 72.2 − − 

� 2.5 0.44 0.015 72.4 − − − 

� 3.0 0.02 0.3 72.3 − − − 

b • 1.5 0.56 0.0026 81.6 − − 

� 1.5 0.18 0.0081 83.0 − − − 

� 1.5 0.02 0.07 83.2 − − − 

8 

vi 

vr

References 

L Bernal and J Kwon. Vortex ring dynamics at a free surface. Physics of Fluids A: Fluid 

Dynamics, 1:449, Jan 1989. 

W Dahm, C Scheil, and G Tryggvason. Dynamics of vortex interaction with a density 

interface. Journal of Fluid Mechanics, 205:1–43, Jan 1989. 

H Honji and M Tatsuno. Vortex rings in a stratified fluid. Journal of the Physical Society 

of Japan, 41:2121–25, Jan 1976. 

Christiaan Huygens. Treatise on light. In Robert Maynard Hutchins, editor, Newton, Huygens, 

volume 34 of Great Books of the Western World, page 551. Encyclopaedia Britannica, 

Inc., 1952. 

L. D. Landau and E. M. Lifshitz. Fluid Mechanics, Second Edition: Volume 6 (Course of 

Theoretical Physics), page 58. Butterworth-Heinemann, 2 edition, January 1987. 

T Lim. An experimental study of a vortex ring interacting with an inclined wall. Experiments 

in Fluids, 7:453–463, Jan 1989. 

P Linden. The interaction of a vortex ring with a sharp density interface: a model for 

turbulent entrainment. Journal of Fluid Mechanics, 60(3):467–480, Feb 1973. 

T Maxworthy. The structure and stability of vortex rings. Journal of Fluid Mechanics, 51: 

15, Jan 1972. 

R Nisius. The photon structure from deep inelastic electron-photon scattering. Physics 

Reports, 332:165, Jul 2000. 

Turgut Sarpkaya. Vorticity, free surface, and surfactants. Annual Review of Fluid Mechanics, 

28:83, Jan 1996. 

M Scase and S Dalziel. An experimental study of the bulk properties of vortex rings translating 

through a stratified fluid. European Journal of Mechanics/B Fluids, 25:302–320, 

Jan 2006. 

Erwin Schrödinger. The fundamental idea of wave mechanics. In Nobel Lectures: Physics 

1922-1941, pages 305–316. Elsevier Publishing Company, Amsterdam, 1964. 

K Shariff and A Leonard. Vortex rings. Annual Review of Fluid Mechanics, 24:235–279, Jan 

1992. 

A Smith. Descartes’s theory of light and refraction: A discourse on method. Transactions 

of the American Philosophical Society, 77(3):i–92, Jan 1987. 

S. C Tiwari. Relativity, entanglement and the physical reality of the photon. Journal of 

Optics B: Quantum and Semiclassical Optics, 4:39, Apr 2002. 

E Wolf and M Born. Principles of optics, page 734. Pergamon Press, Jan 1965. 

9

Understanding risk determinants of Chagas disease in peri-urban Peru 

Megan Christenson 

Nelson Institute of Environmental Studies 


Madison, Wisconsin 

Abstract 

Infestation o f ho mes with Triatoma infestans, an important vector of Chagas disease in southern Peru, is 

common in the peri-urban shantytown communities of Arequipa, Peru. P revalence rates of Chagas disease are not 

well-known b ecause o f t he disease’s i nsidious nature. This u nknown el ement o f a p otentially fatal d isease 

necessitates a proactive and preventative approach. Household vector sprayings have been conducted to eliminate T. 

infestans infestation i n c ommunities within a nd s urrounding Arequipa. A t t he t ime o f t he s praying, ho usehold 

surveys were c onducted i n Nueva A lborada, one o f th ese c ommunities, to d etermine th e d istribution o f c ertain 

animal an d p eridomestic a nd d omestic h ousing t ypes. S patial an d s tatistical a nalyses o f t his d ata i ndicate t hat 

guinea pigs, dogs, sheep, chickens, and cracks on household walls are associated with T. infestans infestation. The 

distance of a home from another containing 20 or more bugs is also very significant in explaining infestation. This 

finding suggests that highly infested homes contribute to vector dispersal. U nderstanding how the spatial, animal, 

and household predictors relate to household vector infestation in Nueva Alborada can potentially provide insight 

for other Peruvian peri-urban communities with T. infestans in order to improve Chagas disease prevention efforts. 

Introduction 

Chagas d isease i s a z oonotic d isease considered t he s ixth-most “neglected t ropical 

disease” in the world as ranked by mortality (Hotez et al., 2006). “Neglected tropical diseases” 

are chronic infectious diseases that affect rural and urban poor people in developing countries. 

Although a century has passed since Carlos Chagas discovered the triatomine vector and disease 

agent of Chagas disease, there are still many facets of the disease that remain mysteries because 

of its complex epidemiology. 

Chagas disease is caused by the parasite Trypanosoma cruzi and is principally transmitted 

by a triatomine v ector, or ‘kissing bu g,’ as it is commonly known. V ectorial t ransmission t o 

humans oc curs w hen t he t riatomine f eeds on t he bl ood of a n i nfected host a nd t hen bi tes a 

human, passing the parasite through defecation into the wound; the vector-borne route accounts 

for about 80% of transmission in humans (Beard, 2005). Approximately 50,000 people die from 

Chagas d isease e ach year with d eath t ypically resulting f rom congestive h eart f ailure 

(Anonymous, 2006). 

Many species of triatomines spread Chagas disease throughout Latin America, but only 

one t riatomine s pecies, Triatoma infestans, tr ansmits th e p arasite in th e a rea o f in terest in 

Arequipa, P eru. A lthough C hagas di sease i s hi storically associated with t he r ural poo r, t he 

domestic T. infestans vector has successfully invaded peri-urban and urban areas (Vallvé et al., 

1996 and Levy et al., 2006). Peri-urban areas are defined as regions on the periphery of urban 

areas, which are often characterized by unplanned growth. 

Since Chagas disease is potentially fatal and lacks a vaccine, it is crucial to find ways to 

prevent it. Drugs are available to treat the acute and early chronic phases of the infection, but the 

medications have undesirable side effects and are not always effective. Vector control has been 

a common prevention technique, and the “Southern Cone Initiative,” an intergovernmental effort 

initiated in 1991 to eliminate T. infestans by insecticide sprayings, was successful in Uruguay, 

Chile, and Brazil but only regionally successful in Bolivia, Argentina, and Paraguay (Schofield 

et a l., 2 006). P eru w as p art o f a s imilar affiliation c alled th e Andean Countries’ Initiative, a 

11

group started in 1997 with the goal of eliminating vectorial Chagas disease transmission from its 

region by 2010 (Guhl, 2007). One of the biggest challenges is coordinating governmental efforts 

to s ystematically c ontrol t he ve ctors; w ithout t horough control efforts, r einfestation is lik ely, 

especially when homes are close in proximity. P roximity is considered an important factor not 

only in bugs spreading from one home to another, but also in determining which host species a 

triatomine chooses to bite (Minter, 1976a). 

Because of Chagas disease’s complex epidemiology, it is important to carefully examine 

all potential risk factors; I have focused on the environmental factors that affect risk. T he host 

species t hat pr ovide bl ood m eals f or t riatomines a re a n i mportant c omponent of t he C hagas 

disease transmission. T. infestans feeds on a variety of mammals and birds, but birds cannot be 

infected b y t he T. cruzi parasite, s o t hey pl ay a uni que role i n C hagas disease ep idemiology. 

Research i n t he C haco region o f northern Argentina ha s s uggested t hat during hot w eather T. 

infestans prefer feeding on c hickens and dogs over humans when all are present (Gürtler et al., 

1997). Dogs and cats also are significantly more infectious to bugs compared to humans making 

them important reservoir sources of T. cruzi infection (Gürtler et al., 2007). B ecause of guinea 

pigs’ abundance a nd hi gh i nfectivity o f T. cruzi, they occupy a c rucial r ole i n t he C hagas 

epidemiology i n B olivia a nd P eru ( Minter, 1976b). Sheep h ave a T. cruzi seroprevalence o f 

approximately 19% ( Cetron, 1995) . C attle, pi gs, a nd ot her large livestock ha ve m uch l ower 

infectivities to T. cruzi, and l ess research ha s focused on t heir r ole i n transmission (Minter, 

1976a). However, the importance of pigs in Chagas disease transmission has been suggested by 

Pizarro and Stevens’ study of T. infestans in Bolivia (2008). Studying the factors of host animal 

competence t o T. cruzi, T. infestans animal pr eference, a nd hos t be havior i s i mportant f or 

understanding the role of animals in Chagas disease transmission to humans. 

Survey d ata f rom a p eri-urban c ommunity of Arequipa, P eru a llows one to ex amine 

explanatory v ariables o f hous ehold i nfestation o f T. infestans, a likely measure o f t he r isk o f 

contracting C hagas d isease. W hile s ome Latin A merican c ountries ha ve ha d s uccessful 

insecticide c ampaigns to e liminate th e T. infestans vector, s outhern P eru s till ha s T. infestans 

vector presence, which is conducive for analyzing spatial patterns of infestation (Fraser, 2008). 

Arequipa may be representative for other dry areas such as the arid Chaco region of Argentina 

where the T. infestans vector is also present. Examining type and number of household animals 

in conjunction with socio-economic indicators such as housing type in a spatial analysis allows 

us to explain why certain houses are more susceptible to infestation than others. 

Methods and Materials 

Study Area 

The s tudy was c arried out i n N ueva A lborada (Figure 1 ), a pe ri-urban c ommunity of 

Arequipa, Peru, located at 16.433˚S, -71.492˚W (Levy et al., 2008). 

12

Figure 1. High-resolution Google Earth aerial photograph of the study site, Nueva Alborada. 

This particular region of southern Peru is arid with rainfall typically only occurring in the months 

of January through March. From 1997 to 2001 the maximum temperatures ranged between 19.1 

and 29.4˚C while minimum temperatures ranged from 6.3 to 13.7˚C (Polk et al., 2005). 

Nueva Alborada is one of hundreds of pueblos jóvenes, or shantytowns, on the periphery 

of Arequipa; the establishment of these squatter communities is intricately related to political and 

social conditions of Peru’s history. In 1969 radical agricultural reform in Peru triggered a wave 

of migration of rural residents to peri-urban areas, establishing pueblos jóvenes. A lthough the 

reform o ptimistically aimed to r edistribute w ealth b y e liminating h aciendas, it w as n ot v ery 

successful i n a lleviating pove rty i n t he poo rest pa rts o f t he c ountry, t he s outhern hi ghlands 

(Klarén, 2000). Continued poverty conditions coupled with low potato prices between 1969 and 

1974, a s taple hi ghland c rop, t riggered r ural m igration of pe asants t o pueblos jóvenes to s eek 

employment (Klarén, 2000). Violence associated with the Shining Path movement also played a 

role in continued rural to urban migration patterns in the 1980s (Pease, 1995). 

Study Design 

This study was carried out in collaboration with the Arequipa Ministry of Health Chagas 

Disease Control Program (Levy et al., 2008). In July 2006 a preliminary search of consenting 

households was c onducted b y entomologic c ollectors t o de termine pr esence or a bsence of 

triatomines i n pe ridomestic a reas, r egions w ithin t he hom eowner’s l ot t hat a re not pa rt of t he 

daily living space. The collectors used a tetramethrin 0.15% aerosol insecticide to spray for the 

triatomines. S everal m onths la ter th e M inistry o f H ealth s ystematically sprayed hous ehold 

rooms, peridomestic areas, and animal enclosures of consenting households with an insecticide 

treatment o f d eltamethrin co ncentrate, w hich was d iluted a t a r ate o f 2 5 mg /m 2 (Levy et al ., 

2008). The insecticide application took place from November 27, 2006 to January 18, 2007 and 

two triatomine collectors spent a total of one person-hour per household searching for vectors 

immediately after the spraying. The Chagas research team assisted the Ministry of Health with 

the bug c ollection i n c ertain pa rts of N ueva A lborada. The num ber of ha bitants a nd t ype o f 

animal (guinea pig, dog, cat, bird, rabbit, sheep, other) were recorded at the time of the spraying. 

The s urveyors not ed t he t ype o f d omestic an d p eridomestic h ousing material ( sillar (white 

volcanic rock), stucco, unmortared brick, adobe, other). They also noted the presence of cracks 

in the material of walls. Triatomines were tested for T. cruzi following the protocol described in 

the study by Gürtler et al (1998). 

13

I conducted five semi-structured interviews with some of the founders of Nueva Alborada 

in July and August of 2008. T he purpose of these interviews was to learn about the history of 

Nueva Alborada, especially in the context of the establishment of triatomines in the area. 

Data Analysis 

Latitude and longitude coordinates of each house were determined using high-resolution 

aerial phot ographs f rom G oogleEarth (GoogleEarth, 2007) . I converted t he l atitude a nd 

longitude coordinates into the Universal Transverse Mercator (UTM) coordinate system (WGS 

84, Zone 19 South) to facilitate analysis in meters. I created shapefiles in ArcGIS 9.3 to spatially 

display the housing and animal types by household. 

Because there were many houses without the vector, the outcome variable of number of 

bugs was not normally distributed. Although there were many small integers containing zeros in 

the out come, t he P oisson m odel w as n ot ap propriate b ecause t he m ean an d v ariance w ere n ot 

equal. S tatistical an alyses w ere co nducted i n t he s tatistical p ackage R ( R D evelopment C ore 

Team, 2008) . I used a two s tep a pproach i n a nalyzing t he da ta ba sed on t he a ssumption t hat 

there may be two different mechanisms at work: one mechanism that determined whether any 

triatomines were present or not, and one mechanism that determined how many triatomines were 

present a mong hous es w ith a non -zero num ber of t riatomines. F irst I performed uni variate 

logistic regressions using presence or absence of triatomines as the response, and the animal and 

housing type (again as presence/absence of that particular animal or housing type) as predictors. 

Then I ran a backward stepwise selection procedure starting with all predictors to select the best 

fitting model. T he f inal m odel w as c hosen us ing t he lowest Akaike’s Information Criterion 

(AIC). The second step was an analysis of just the houses with bugs using linear regression and 

backward stepwise selection on the logarithmically transformed non-zero outcome data. I also 

tested the correlations b etween al l predictors an d the in teractions b etween th e s ignificant 

predictors from the univariate regressions. 

I created semi-variograms of residuals to check for spatial autocorrelation in the residuals 

for all models. Some models showed evidence of spatial patterns in the data; to account for this I 

added a predictor variable measuring the distance from each house to the nearest house with 20 

or m ore i nsects. T his metric w as c alculated u sing th e D istance B etween P oints f unction in 

Hawth’s A nalysis T ools f or A rcGIS ( Beyer, 2004 ). I then r edid t he previously m entioned 

statistical tests including this predictor. In ArcGIS, I also tested Moran’s I Index and Ripley’s K 

Function t o t est f or s patial a utocorrelation a nd clustering, r espectively. Moran’s I Index i s a 

spatial statistic used to evaluate whether homes with bugs are clustered, dispersed, or random. 

Ripley’s K Function i s a nother w ay t o t est w hether t he hous es w ith bugs a re c lustered or 

dispersed but provides a simpler method for finding the distance to which a feature is clustered. 

It uses a multi-distance analysis to assess the observed clustering among infested households as 

compared to what would be expected in randomly distributed infested households. 

Results 

499 l ots w ere i n t he c ommunity of Nueva A lborada a t t he t ime of t he i nsecticide 

spraying. 452 households participated in the household sprayings; the remaining ones were not 

sprayed b ecause t he lots w ere e ither uni nhabited, t he ow ners w ere una vailable, or t he ow ners 

refused. O wner refusal accounted for only seven of the non-participating households and these 

homes were randomly distributed. In the community 4,111 total bugs were counted, which were 

distributed among 165 h ouses (37% of total). No bugs were infected with T. cruzi. Figure 3 

displays a map of the surveyed houses and triatomine presence in the community. 

14

Figure 3. Map of 452 surveyed homes in the peri-urban community of Nueva Alborada, Peru. Red dots indicate 

homes with one or more triatomines. Spatial patterns of infestation are suggested by the clusters of homes with 

triatomines. The southern region had no bugs, which may be because this section of Nueva Alborada is newer. 

The map shows clusters of homes that had bug presence. 

Before conducting statistical regressions, the animal category of other, and housing types 

of pe ridomestic a nd dom estic a dobe, pe ridomestic a nd dom estic ot her, unm ortared dom estic 

brick, and peridomestic cracks on walls were removed from the data because they either had less 

than 5 h omes i n t he cat egory o r w ere co llinear w ith an other cat egory. T he u nivariate an d 

multivariate regressions were then analyzed in a two step approach: first, a logistic regression of 

all surveyed houses with a bug presence/absence outcome and second, a linear regression of just 

the homes with bugs to observe the effect on the outcome of number of bugs. 

Presence of guinea pigs, dogs, and distance to a house with 20+ bugs were statistically 

significantly related to presence of bugs (P

In the example of the guinea pig predictor, the odds ratio of 2.35 means that a home with guinea 

pig pr esence i s 2.35 t imes a s l ikely t o ha ve t riatomine pr esence. In t he uni variate l inear 

regression analyses of animal types, presence of guinea pig, sheep, and bird were positively and 

statistically s ignificantly r elated to n umber o f bugs when considering ju st th e h omes with 

triatomines (Table 2). 

Animal Intercept Slope P-value 

Guinea Pig 17.88 18.13 0.0028 

Sheep 2.17 1.44 0.019 

Bird 2.036 0.51 0.033 

Distance to 20+ bugs 

(meters) 

2.95 -0.044

einfestation of T. infestans also found distance to infested areas to be significant (McGwire et 

al., 2006). 

The statistical analyses indicate that guinea pigs could also be a p otential predictor of T. 

infestans infestation in N ueva A lborada. G uinea pi gs are kno wn t o be an i mportant factor i n 

Chagas disease epidemiology (Herrer, 1955 and Levy et al., 2006), especially in Peru where they 

are raised as a f ood source. Because guinea pigs feed on leftover kitchen scraps, it is easier for 

poor families to raise them (Rath, B., pers. comm.). S everal factors affirm the epidemiological 

importance of guinea pi gs: hi gh abundance, hu sbandry c onditions, hi gh i nfection r ates of T. 

cruzi, and easy transportability (Herrer, 1955 a nd Minter, 1976b). Families often have a much 

greater number of guinea pigs compared to other domestic animals. T he guinea pigs are often 

raised in cages, which provide favorable living conditions for T. infestans; the cages offer places 

to hide from the sun, to lay eggs, and to easily access numerous sources of blood meals (Herrer, 

1955). R ates of g uinea pi g i nfection c an r each 60% ( as c ited i n M inter, 1976b a nd C etron, 

1995). The ease of transporting guinea pigs may also favor the spread of the T. cruzi parasite to 

new areas through migration (Herrer, 1955). 

Chickens, turkeys or du cks, represented b y the bird predictor in the analysis, may also 

help t o e xplain t he pr esence of t riatomines. A ccording t o an e xperimental s tudy, T. infestans 

preferred feeding on chickens over guinea pigs when both were equally available (Vázquez et al., 

1999). Because chickens are blood-meal sources for T. infestans and are not susceptible to T. 

cruzi infection, th eir e pidemiological r ole is u nique. W hile s ome r esearchers o ptimistically 

hoped t hat c hickens m ight s erve i n a z ooprophylactic r ole, o ne t hat w ould d ecrease t he 

probability o f hum ans be ing f ed upon, t his e ffect ha s not be en s cientifically pr oven. S ome 

studies h ave i ndicated a d ecrease i n b ug i nfection r ates i f T. infestans feed ex clusively o n 

chickens; however, since T. infestans shift between host species that are susceptible to T. cruzi 

infection, the effectiveness of zooprophylaxis is limited (Vázquez et al., 1999 and Gürtler et al., 

1998). The presence of domestic chickens may significantly increase the density of domestic T. 

infestans (Cécere et al., 1997), which is confirmed by my results from Nueva Alborada. 

Although t he s heep va riable w as a s ignificant pr edictor i n t he uni variate a nd s tepwise 

linear regressions, this is not a consistent finding with previous studies. Minter states that the 

main blood sources for T. infestans are humans, chickens, dogs, and cats (1976a). M inter also 

states that host proximity may be a more important factor in blood source selection than animal 

preference (1976a). T. infestans rarely feed on the larger livestock; in fact, no T. cruzi infections 

of sheep had even been detected at the time of Minter’s study (1976b). One reason for this lack 

of T. infestans blood m eals f rom s heep m ay be t he di fficulty of s uccessfully bi ting a s heep 

because of the thick wool. F urther research in other sites would need to be conducted to see if 

sheep are associated with house infestation. 

Presence of cracks in the wall material of domestic areas was positively and significantly 

related to the number of bugs in the linear stepwise regression. Because triatomines hide in wall 

cracks during the day, it makes sense that they would prefer this type of housing. Poor housing 

conditions a re kno wn t o be a ssociated w ith C hagas di sease ( Zeledón a nd R abinovich, 198 1). 

Although presence of wall cracks was the only significant housing predictor in the regressions, it 

is possible that the dog predictor substituted for some housing t ypes since it was significantly 

correlated to four of the housing categories. However, these correlation coefficients were small 

(

explain t his t rend. F rom t alking t o r esidents of N ueva A lborada a nd o thers f amiliar w ith t he 

area, I know that the southern section is newer, so perhaps the bugs simply had not spread to this 

region yet. Vector dispersion is common among homes that are close to one another (Levy et al., 

2008), so it is likely the bugs will infest the southern region. 

Because m any elements o f C hagas d isease t ransmission ar e s till u nknown, a m ain 

strength of t his s tudy i s f urthering t he und erstanding t he i nfluence of a nimals, hous ehold 

materials, a nd s patial f actors on ve ctor i nfestation of T. infestans in pe ri-urban communities. 

Although no bugs tested positive for T. cruzi in Nueva Alborada at the time of the study, other 

pueblos jóvenes in A requipa ha ve c onfirmed t he pr esence of t he pa rasite, s o i mplementing 

preventative tactics is important. H owever, some limitations of this study restrict the extent to 

which t he r esults c an b e us ed t o pr edict i nfestation i n ot her a reas. Because t he s tudy w as 

conducted in a small site with specific climatic conditions and domestic animal behaviors, the 

findings may not accurately be generalized to other sites. Since the researchers assisted with the 

bug collection in certain parts of Nueva Alborada, some sampling bias exists in the data. Simple 

data on presence or absence of animals was recorded in the surveys, so the analysis neglects the 

effect of a nimal a bundance on bug i nfestation. A lso, t he bug popul ations m ay not a ccurately 

correspond to the surveyed animals but may be an effect of animals no longer present (Levy et 

al., 2006). 

Precise prevalence rates of Chagas disease are not known for Arequipa since large-scale 

surveys have not been conducted. However, a survey in one peri-urban community of Arequipa 

indicated a 5.3% rate of infection of 2-18 year old children (Levy et al., 2007). P resence of T. 

cruzi in t he hum an pop ulation c oupled w ith do mestic a nimals a nd hous ehold i nfestation of 

triatomines in the pueblos jóvenes of Arequipa facilitates Chagas disease transmission; thus, it is 

important to have an effective strategy to minimize the disease risk. I nsecticide treatments are 

quite e ffective i n ridding hous eholds of T. infestans, but t hey are a s hort-term s olution s ince 

reinfestation from n eighboring c ommunities is likely. S ome mo re s ustainable pr eventative 

approaches i nclude pr eventing dom estic a nimals f rom s leeping i n t he s ame r oom a s pe ople, 

stuccoing walls, and educating communities about Chagas disease (Cohen and Gürtler, 2001). 

To advance the understanding of Chagas disease prevention, future studies might address 

the host competence of sheep and other large livestock to T. cruzi and their association with T. 

infestans infestation. T he pr esence of c hickens s hould c ontinue t o be studied f or i ts pot ential 

zooprophylactic e ffect w hen other a nimals a re e xcluded f rom dom estic hous ehold a reas. A 

comparative s tudy between N ueva A lborada and a n eighboring T. cruzi-positive community 

would be a us eful a nalysis be cause i t c ould pr ovide i nsight r egarding w hy only s ome 

communities have the parasite. 

Conclusion 

This t hesis ha s s hown how e xamining hous ehold s urvey i nformation combined w ith 

interview data can explain the risk factors involved with Chagas disease in pueblos jóvenes of 

Peru. C hapter 1 e xamined t he s ocial a nd pol itical ba ckground of periurban a reas i n P eru, 

especially A requipa, i n order t o co ntextualize C hagas d isease em ergence i n pueblos jóvenes. 

Interviews with community founders of several pueblos jóvenes of Arequipa revealed historical 

information about how these pueblos jóvenes formed and what the residents understand about the 

Chagas disease vector. Focusing on household survey data from one such periurban community, 

Nueva A lborada, i ndicated how hous ehold c haracteristics s uch a s c racks i n w alls a nd t he 

18

presence of guinea pigs and dogs are statistically significant determinants of the presence of the 

vector of Chagas disease in southern Peru, Triatoma infestans. 

Many challenges face eradication of Chagas disease: i t has no vaccine, treatment of the 

disease is complicated with undesirable side effects, and reinfestation of homes with T. infestans 

following i nsecticide t reatment i s co mmon. Given t hese ch allenges, t here i s a n eed f or 

affordable, sustainable prevention techniques. As the results from Chapter 2 show, prevalence of 

the T. infestans vector is correlated with the presence of dogs and guinea pigs; thus, restricting 

dogs and guinea pigs from bedrooms might help reduce the number of triatomines likely to bite 

humans. S ealing wall cracks i s another effective p revention strategy. N ueva Alborada i s 

representative o f o ther pueblos jóvenes and t his a nalysis pr ovides a n example of how t he 

distribution of the T. infestans Chagas d isease v ector can analyzed with t he goal o f i nforming 

prevention efforts. 

References 

[Anonymous], 2006. Chagas’ disease - an epidemic that can no longer be ignored. Lancet 368: 619. 

Beard CB, 2005. Kissing bugs and bedbugs, the Heteroptera. Marquardt WC, ed. Biology of Disease Vectors. San 

Francisco, CA: Elsevier Academic Press, 57-65. 

Beyer HL, 2004. Hawth’s Analysis Tools for ArcGIS. Available at: http://www.spatialecology.com/htools. 

Accessed March 12, 2009. 

Cécere MC, Gürtler RE, Chuit R, Cohen JE, 1997. Effects of chickens on the prevalence of infestation and 

population density of Triatoma infestans in rural houses of north-west Argentina. Med Vet Entomol 11: 383-388. 

Cetron MS, 1995. Chagas’ disease. Jong, EC, McMullen R, eds. The Travel and Tropical Medicine Manual. 

Philadelphia, PA: W.B. Saunders Company, 270-281. 

Cohen JE, Gürtler RE, 2001. Modeling household transmission of American Trypanosomiasis. Science 293: 694- 

698. 

Fraser B, 2008. Controlling Chagas’ disease in urban Peru. Lancet 372: 16-17. 

GoogleEarth, 2007. Satellite Photo Google. Available at: http://earth.google.com. Accessed June 19, 2007. 

Guhl F, 2007. Chagas disease in the Andean countries. Mem Inst Oswaldo Cruz 102 Suppl 1: 29-37. 

Gürtler RE, Cohen JE, Cécere MC, Chuit R, 1997. Shifting host choices of the vector of Chagas disease, Triatoma 

infestans, in relation to the availability of hosts in houses in north-west Argentina. J Appl Ecol 34: 699-715. 

Gürtler RE, Cohen JE, Cécere MC, Lauricella MA, Chuit R, Segura EL, 1998. Influence of humans and domestic 

animals on the household prevalence of Trypanosoma cruzi in Triatoma infestans populations in northwest 

Argentina. Am J Med Hyg 58: 748-758. 

Gürtler RE, Cécere MC, Lauricella MA, Cardinal MV, Kitron U, Cohen JE, 2007. Domestic dogs and cats as 

sources of Trypanosoma cruzi infection in rural northwestern Argentina. Parasitology 134: 69-82. 

Herrer A, 1955. Importancia del cobayo como reservorio de la enfermedad de Chagas en la región sudoccidental. 

Rev Med Exp 9: 45-55. 

Hotez PJ, Molyneux DH, Fenwick A, Ottesen E, Ehrlich Sachs S, Sachs JD, 2006. Incorporating a rapid-impact 

package for neglected tropical diseases with programs for HIV/AIDS, tuberculosis, and malaria. PLoS Med 3: 

e102. 

Klarén PF, 2000. Peru: Society and Nationhood in the Andes. New York, NY: Oxford University Press. 

Levy MZ, Bowman NM, Kawai V, Waller LA, Cornejo del Carpio JG, Cordova Benzaquen E, Gilman RH, Bern C, 

2006. Periurban Trypanosoma cruzi-infected Triatoma infestans, Arequipa, Peru. Emerg Infect Dis 12: 1345- 

1352. 

Levy MZ, Kawai V, Bowman NM, Waller LA, Cabrera L, Pinedo-Cancino VV, Seitz AE, Steurer FJ, Cornejo del 

Carpio JG, Cordova-Benzaquen E, Maguire JH, Gilman RH, Bern C, 2007. Targeted screening strategies to 

detect Trypanosoma cruzi infection in children. PLoS Negl Trop Dis 1: e103. 

Levy MZ, Quíspe-Machaca VR, Ylla-Velasquez JL, Waller LA, Richards JM, Rath B, Borrini-Mayori K, Cornejo 

del Carpio JG, Cordova-Benzaquen E, McKenzie FE, Wirtz RA, Maguire JH, Gilman RH, Bern C, 2008. 

Impregnated netting slows infestation by Triatoma infestans. Am J Trop Med Hyg 79: 528-534. 

McGwire K, Segura EL, Scavuzzo M, Gómez AA, Lamfri M, 2006. Spatial pattern of reinfestation by Triatoma 

infestans in Chancaní, Argentina. Journal of Vector Ecology 31: 17-28. 

Minter DM, 1976a. Feeding patterns of some triatominae vectors. New Approaches in American Trypanosomiasis 

19

Research. Washington, DC: Pan American Health Organization. Scientific Publication No. 318, 33-47. 

Minter DM, 1976b. Effects on transmission to man of the presence of domestic animals in infested households. 

New Approaches in American Trypanosomiasis Research. Washington, DC: Pan American Health 

Organization. Scientific Publication No. 318, 330-337. 

Pease GYF, 1995. Breve Historia Contemporánea del Perú. 1st ed. Distrito Federal, México: Fondo de Cultura 

Económica. 

Pizarro JC, Stevens L, 2008. A new method for forensic DNA analysis of the blood meal in Chagas disease vectors 

demonstrated using Triatoma infestans from Chuquisaca, Bolivia. PLoS ONE 3: e3585. 

Polk MH, Young KR, Crews-Meyer KA, 2005. Biodiversity conservation implications of landscape change in an 

urbanizing desert of southwestern Peru. Urban Ecosystems 8: 313-334. 

R Development Core Team, 2008. R: A language and environment for statistical computing. R Foundation for 

Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. Available at: www.r-project.org. 

Rath, B. E-mail interview. 5 June 2009. 

Schofield CJ, Jannin J, Salvatella R, 2006. The future of Chagas disease control. Trends Parasitol 22: 583-588. 

Vallvé SL, Rojo H, Wisnivesky-Colli C, 1996. Urban ecology of Triatoma infestans in San Juan, Argentina. Mem 

Inst Oswaldo Cruz 91: 405-408. 

Vázquez DP, Canale D, Gürtler RE, 1999. Effects of non-susceptible hosts on the infection with Trypanosoma cruzi 

of the vector Triatoma infestans: an experimental model. Mem Inst Oswaldo Cruz 94: 413-419. 

Vazquez-Prokopec, GM, Ceballos LA, Kitron U, Gürtler RE, 2004. Active dispersal of natural populations of 

Triatoma infestans (Hemiptera: Reduviidae) in rural northwestern Argentina. J Med Entomol 41: 614-621. 

Zeledón R, Rabinovich JE, 1981. Chagas’ disease: an ecological appraisal with special emphasis on its insect 

vectors. Ann Rev Entomol 26: 101-133. 

20

Astronaut Advanced Life Support: 

Engineering Extra Terrestrial Extremophile Plants 

Submitted By: Dan Hawk, PI 1 ; Dr Cindi Schmitt, PI 2 ; and Dr Gertrud Konings-Dudin, PI 3 

. 

Abstract Phase 2: This paper directly supports NASAs Exploration Systems Mission 

Directorate (ESMD) of returning to the moon 4 

. Going to the moon is not a redundant 

undertaking; we are here to support bigger lunar missions and longer stays. In this phase we 

continue to amend JSC-1A lunar regolith simulant with pyrogenic carbon 

(Cpyr); the main ingredient of sustainable anthropogenic Amazon Black Earth (ABE) soil. 

In addition to growing food in a lunar habitat demonstration, we expand research to include 

engineering the lunar habitat by creating an extremophile plant that can survive on the lunar 

surface known as “Star”. Herein we report that our research directly supports an unmanned lunar 

Lander whereby Star is to be placed inside a growth module on the surface of the moon, and 

having a >90% chance of surviving a minimum of one lunar cycle (~28 earth days). Star will be 

of great scientific potential yet on earth, extremophile research allows us to explore reintroducing 

ABE for high-altitude sustainability agriculture, in response to global climate change. 

Summary of Contents: 

1. Star. 

a. Introduction: Star Mission to the Moon 

b. KAG, Divide, Colorado Experiments 

c. EPCC, El Paso, Texas Experiments 

d. CMN, Green Bay, Wisconsin Experiments 

2. MAX-X. 

3. Future Experiments. 

4. Plant Module (Concept Drawings). 

Star Mission to the Moon. The addition of ABE and liquid fertilizer amended soils can produce 

a yield of 880% 5 6 more than the baseline. ABE increases plant stress hormones which allow the 

plant to survive under very harsh conditions. In addition ABE is independent of climate and soil 

type 

7 8 

allowing us to research ABE technology for use on an extra-terrestrial body. Extra- 

terrestrial research simulations guide us in developing extremophile plants for advanced 

astronaut life support (food production), engineering extra-terrestrial atmospheres, and highaltitude 

plant growth experiments. 

1 College of Menominee Nation. 

2 Kalagesi AniNoquisi Gatusi. 

3 El Paso Community College. 

4 http://www.nasa.gov/exploration/home/index.html. 

5 http://www.youtube.com/watch?v=T1eYn76bO4E. 

6 Steiner, C., (Dissertation) Slash & Char: As Alternative To Slash & Burn. Cuvillier Verlag Götingen. 

7 Lehmann, J., (2003). Amazonian Dark Earths: Origin, Properties, Management. Netherlands:Kluwer Academic 

Publishers. 

8 Lunar regolith is not considered soil and by definition cannot become Amazon Black Earth (ABE). 

21

Our team has two goals: 

1. To put Star on the moon by 2016, ensuring that Star will have a >90% chance of survival 

for a minimum of one lunar cycle (~28 earth days). 

2. To reintroduce a food crop using ABE for high-altitude sustainable agriculture, for which 

under wild-type conditions it will not produce a yield yet, will produce a substantial yield 

using ABE amended soil. 

Before returning humans to the moon many scientists believe continued unmanned lunar 

exploration is needed. NASA believes that precursor robotic lunar Lander missions will provide 

valuable information that will ultimately minimize astronaut risk, increase their efficiency, and 

lower lunar architectural costs. 9 

Star would be an important technology demonstration for; local; 

gamma, X, UV, and IR spectroscopy, local magnetic field, thermal gradients, biological 

mechanism @ 1/6g performance, ECLSS loop @ 1/6g performance, E(semi-C)LSS, (open- 

E)LSS, biological lunar dust impact, biological dust characterization, biological dust mitigation, 

material science dust compatibility, radiator dust exposure, dust filter, thermal surface influence, 

surface processes for an airless planetary body, gravimetry; g-field as a function of lunar 

rotation, lighting perspective; permanent low incidence at lunar poles, in situ thermal wadi 

(oasis) effects on biologics, and Star experiment surface communication. 

KAG, Divide, CO. To reach the goal of putting an extremophile plant on the surface of the 

moon we directed our research at the objectives however, like most experiments many paths and 

opportunities presented themselves. To maintain focus we needed to create a name for our 

extremophile plant, the same plant that will ultimately be planted on the surface of the moon; 

Optunia noquisi. Dr Cindi Schmitt decided on the name Noquisi which means “Star” in the 

Cherokee language. Star then is an engineered extremophile cactus designed to be both lunar day 

and night hardy 10 

. 

Dr Cindi Schmitt has her research base at ~9,800 feet (70.2kPa) with research extending to 

10,100 feet or 69.3kPa (Pressure Loss 31.03% from Sea Level). Numerous papers indicate using 

10psi or 69kPa for proposed inflatable lunar structures. In fact Dr Schmitt’s research work in 

lunar partial pressure is remarkable. Furthermore Dr Schmitt’s high-altitude research includes; 

two kinds of tomato plants, mesclun, mondara, bell peppers, Brussels sprouts, herbs, 

strawberries, beans, chia, spinach, beets, cucumber, cilantro, lichen, red and white clover, mint, 

oregano, savory, cress, nasturtium, two varieties of O. ficus-indica, three varieties of O. fragilis, 

O. phaeacantha, O. polyacantha, grafting, and 10 types of soil amendments; JSC Lunar-1A 

regolith simulant, JSC Mars-1A simulant, Pikes Peak Regolith (PPR), Cpyr, N liquid, Vitamin B, 

perlite, wild-type scat, wild-type soil, and greenhouse prepared mulch which contains; egg shells, 

tea leaves, coffee grounds, and aged horse manure. 

The high elevations found in Teller County and Ute Pass greatly influences plant growth. The 

length of the growing season varies from 150 days at 6,000 feet to 70 days at 10,000 feet. At 

6,000 feet the frost-free season extends from May 15th to October 10th. Given great latitude, in 

9 NASA Lunar Precursor Robotic Program (LPRP). http://moon.msfc.nasa.gov/. 

10 LUNAX, Nightspan Dark Hardiness Experiment. Controls lighting schedules to follow the twentyeight 

day cycle of sunlight and darkness on the Moon’s surface. www.lunarreclamation.org/lunax/index.htm. 

22

general for an increase of 100 feet from sea level, the growing season is shortened by two days; 

one day in the spring and another in fall. 11 

Along with this research, Dr. Schmitt proposes a 

specialized, high altitude greenhouse design which is applicable to both sustainable agriculture 

and to extra-terrestrial habitats and colonization. 

Soils in the Teller County portion of the Rocky Mountains are generally poor in nutrients, very 

low in organic materials, and have an unusually high pH and calcium content. The higher 

calcium content makes it difficult to lower the pH using gypsum, lime, sulfur, and to augment 

iron deficiencies. The mountains were mined for heavy and precious metals; many of the 

groundwater caches contain these substances. Semi-precious gemstones and jewelry-grade 

minerals are found due to past volcanic history. 

PPR is a widespread geologic formation found in the Front Range of Colorado and comprises a 

much larger deposit of Pikes Peak batholith formed about 1.08 billion years ago. Molten magma 

from volcanic activity cooled creating PPR. About 60 million years ago the Rocky Mountains 

were formed raising Pikes Peak to its current height. 12 PPR ranges in color from pink to almost 

red due to large amounts of microcline feldspar and iron minerals. In many places PPR is very 

coarse grained made up of large crystals of feldspar, typically about 1 cm across. PPR is easily 

weathered making it crumbly. 13 

The average analysis of PPR indicates that it is similar to 

Martian regolith characteristics. 

Growing Constraints at 10,000 feet. Teller County, Colorado. The ground cover is shallow, 

fragile tundra. Permafrost found at 18-24” in August, with ambient temperatures of ~80°F. 

Daytime temperature swings with rolling storms average 40°F with evening rarely above 65°F. 

Storm rainfall is often 2” with concomitant pea-sized hail, snow, high winds, and cloud-toground 

lightening. Low atmospheric pressure (69.3kPa @ 10,100 feet), high UV exposure, 

11 Mountain Gardening Advisories by Teller County Master Gardeners, Colorado. 

12 www.1911encyclopedia.org/Granite. 

13 http://en.wikipedia.org/wiki/Pikes_Peak_granite. 

23

fluctuating humidity, drying winds, periods of drought, cold soils, lack of gentle rains, and 

hungry wildlife makes high-altitude agriculture in this region very difficult. It is at this stage we 

are investigating not only sustainability techniques for transfer to high-altitude populations, but 

its applicability to extra-terrestrial human colonies and long-duration spaceflight. 

Hypothesis0. Pyrogenic carbon enables greater nitrogen and micronutrient 

utilization at the root tips thereby allowing higher concentrations of stress hormones 

to be produced which increase biomass production. 

Teller County as a region is indicative of human-populated, high elevation places needing 

sustainable agricultural solutions; transferable to deep space and planetary exploration 

horticultural needs. 

Results of 2008 Cpyr Augmentation. Mesclun leaves attained 10” in length, were not stringy and 

tasted sweet. All JSC-1A died. Red Cherry (hybrid) tomato plant had tremendous biomass gain. 

Final speculation is that tomato plant soil media needs to be warmed. 

2009 Data as of this report. Tomato varieties galina and yellow cherry are thriving. All 

galina were started in lunar+Cpyr+mulch and companion planted with clover (to increase N) 

plus nasturtium (an edible green and flower, for a symbiotic relationship). Seedlings are 10-h and 

healthy. Lunar grown galina were transplanted August 4, 2009. Brussels sprouts started with 

PPR+ Cpyr+mulch, no results to report. Heirloom spinach started with lunar+ Cpyr was 

harvested and tasted fine. A later spinach plant is now getting ready to flower. Chia did not grow 

to flower but tasted fine. Final speculation is that chia soil needs to be warmed to make it flower. 

2010 plan will use warming tray for chia. Mondara (bee balm, bergamot) showed the greatest 

biomass gain, at 20-h, which is significant at 10,000 feet. Cilantro, oregano, mints, and savory 

were started in PPR+ Cpyr+mulch with control plantings in PPR unaugmented. Hanging 

strawberries are also successful with flowering and berry set. Strawberries planted outside 

survived the winter with temperatures as low as -28°F and are growing. 

Nitrogen Stimulation. galina and nasturtium grows with companion plantings of clover and rock 

cress for testing N content; post growing season. Both clover and rock cress are found at 10,000 

feet and higher so it is not N limiting factor. Plan to test spotted dog lichen N-fixer in 2010. 

Cpyr Optimum Amendment. 

Hypothesis1. That there is an optimum amount of Cpyr amendment necessary to 

achieve maximum plant growth and, that amount varies by plant type. 

There is a plant requirement for balancing the soil nutrition for optimal growth. The above 

varieties tolerate a higher amount of C in the soil than do Opuntia varieties. 

Opuntia. O. varieties tend to tolerate less C in the soil. The usability of O. as food, fodder, and 

medicine over the centuries is well documented and piqued our research interest as an 

extremophile plant. O. is extremely adaptable to hostile environments and soils. O. is a prime 

candidate for extraterrestrial utilization. We are developing a very hardy variety that could be 

24

planted on the surface of the moon or Mars; Optunia noquisi or Star. There is abundant research 

for O. horticulture references and medicinal applications. 

O. Varieties @ High-Altitude. During the course of experimentation, discussions, and seminars 

through the months of 2008, the possibility of using cacti as another food source for sustainable 

high-altitude and extra-terrestrial use came to the forefront for viable consideration. Through a 

collaborative effort, several experiments were developed and run simultaneously at elevations of; 

10,100 feet in Divide, CO, 4,090’ in El Paso, TX, and 695’ in Green Bay, WI using Cpyr 

amended JSC Lunar-1A Simulant. 

Hypothesis2. Cpyr amended soils, PPR, and JSC Lunar-1A Simulant will make a 

difference in O. growth at cold, high-altitude temperatures. 

Hypothesis3. O. will not grow as rapidly in unaugmented media at cold, highaltitude 

regions. 

Dr Konings-Dudin guided our team with our study of O., providing us each six pads (6 

nopals) which were subsequently planted in late September. Two pads were planted in 

PPR+Cpyr+sand+mulch, two pads in PPR+Cpyr+mulch, and two in Lunar+Cpyr+mulch. 

Through 2008 to today, the lower elevation pads remain at ambient air of

western border of the Arkansas River Valley and the eastern border is the Mosquito Range. The 

objective was to find O. in their natural distribution, photograph them, and obtain samples. 

Samples were collected and sent to the team on July 27, 2009 for collaborated experiments. The 

highest O. fragilis colony was found at 10,100 feet in a Montane biome, with a sub-alpine 

distribution. This coincides with a distinct habitat change; at that elevation the Ponderosa Pine 

tree habitat gives way to the Lodgepole Pine habitat. This colony had flower buds whereas the 

lower elevation colonies had finished blooming one week prior. The healthiest plants in this 

colony were sprinkled with elk or deer scat. 

Another colony was found within 400 feet of the Avalanche Trailhead. The regolith is granite 

scree with some decomposed pine matter making tundra. Subsurface permafrost and mini 

glaciers were present. This was the only colony thus samples were not collected however, 

photographs were taken. 

The Montane biome samples were sent with their soil and scat to team members for their use. Dr. 

Schmitt planted in PPR+Cpyr+mulch+wild-type scat+wild-type soil, and is allowing them to rest 

and root before moving forward with experimentation. 

EPCC, El Paso, Texas. We need an extremopohile plant that can withstand severe thermal 

cycling; from very hot temperatures to deep freeze. Our initial test plant was O. ficus-indica 

which is well known in Mexico but, also grows in El Paso, Texas. Summer days in El Paso can 

go above 100°F and winter nights can below 32°F. O. ficus-indica is reported to survive 

temperatures up to 140°F 15 

. O. ficus-indica is known for food production and as fodder in arid to 

semi-arid lands all over the world. 

Results. O. ficus-indica cladodes were removed from a wild-type in an arroyo which were an El 

Paso (EP) variety, and from a domestic Mexican (Mex) variety growing in El 

Paso used as a supply plant for team research. Seeds were collected from the wild-type 

EP and from domestic Mex tunas. 

1. O. ficus-indica develops a good root system and grows well on plain JSC-1A simulant as 

well as Cpyr amended JSC-1A up to 10% by volume. 6% Cpyr amended JSC-1A 

provides the best results. Plants were given distilled or rainwater and supplemented with 

Miracle-Gro®. Resultscomparable at different altitudes and to mulch amendment. 

2. Seedlings of O. ficus-indica germinated and developed in JSC-1A simulant and 6% Cpyr 

amended JSC-1A. Several seedlings are thriving. Seedlings were provided rainwater and 

supplemented with Miracle-Gro®. Results comparable at different altitudes and with 

seedlings grown in mulch amendments. Germination rates were low in all cases. 

3. O. ficus-indica rooted cladodes in 6 and 10% Cpyr amended JSC-1A simulant were 

exposed to night/day cycles in a 14/14 and 5/23 night/day lunar cycle rhythm. The plants 

were exposed to temperatures up to 104°F. Since the Start of lunar cycling two plants 

have died. Speculate that these plants died to carbon toxicity from the previous 

experiment, one plant was the control. Cladodes develop new growth. Experiments are 

done in two locations. 

15 Nobel P.S., Recent Ecophysiological Findings for O. ficus-indica. J PACD, 1997, 89-96. 

26

4. New growth points at different stages on cycling O. ficus-indica cladodes in 6 and 10% 

Cpyr amended JSC-1A simulant, were exposed to similar lunar night/ day rhythms. 

Beginning stages showed developmental problems from re-absorption to stunted growth 

of the new pad. Once the newly developing pad shows green growth from the mother 

cladode it develops normally and demonstrates a surprising developmental spurt when 

entering the day cycle. Experiments are done in two locations. 

5. O. ficus-indica seedlings transferred to 6 and 10% Cpyr amended JSC-1A simulant were 

exposed to one 5/23 night/day rhythm with a temperature as low as 37.4°F and high as 

104°F. Seedlings appear to be developing normally. Experiment is ongoing. 

Conclusion. 6 and 10% Cpyr amended JSC-1A simulant supports normal growth stages for 

O. ficus-indica to include seed germination and seedling development. All stages of O. 

ficus-indica survive lunar cycle rhythms of 5/23 and 14/14 with temperatures between 

37.4 and 104°F. New growth depends on the progress made before entering the night 

cycle which stimulates growth in well developed cladodes. Experiments should continue 

for a minimum of 3 years. 

CMN, Green Bay, Wisconsin. Our goal is to put Star on the moon by 2016! What does 

our research team need to do, to make that happen? 

Mitigating Star Concerns. Star will need to be a lunar engineered extremophile plant designed 

16 17 

for an extremely hostile environment. In addition the “Plant Module” will have a selfcontained 

biological environment controlled life support system (ECLSS). As an example, Dr 

Larry Taylor suggests using a small 2.45 or 5.8GHZ dc pulsed magnetron to heat the JSC-1A 

simulant creating a thermal sink; maintaining Star above -100°F while night cycling. Dr Taylor 

however, is concerned with radiation, micro meteorites, and the nearly nonexistent lunar 

atmosphere; more specifically the intensity of the vacuum. 

In addition Dr Schmitt is concerned with space IR, increased plant stress, and reversing callous 

cell formation by glycol re absorption. 

Considering the 2016 lunar Lander deadline, how can our team poke holes in these significant 

concerns in order to protect Star and complete our mission? 

1. Star Shield. The Plant Module is designed with a protective shield. The shield must 

protect Star from micrometeorites weighing one gram or less. The shield must allow 

sufficient sunlight to cause natural photosynthesis. The shield should provide radiation 

shielding sufficient to allow Star to survive for no less than one lunar cycle (~28 earth 

days). The shield must be an open design, allowing Star to interact with the lunar 

elements for lunar dust mitigation experiments. Currently welding lens technology can 

filter out up to 100% UV/IR wavelengths. In addition auto-darkening welding lens can be 

16 

Dr Larry Taylor, Dept. of Geological Sciences, Planetary Geosciences Institute, University of 

Tennessee, Knoxville. http://web.utk.edu/~pgi, lataylor@utk.edu. 

17 

Dr Taylor identifies Star containment as a “Plant Module” by email, "Taylor, Lawrence A" 08/03/09 8:28 

PM. 

27

adjusted to less than .4milliseconds. A custom auto-darkening Star shield would be clear 

while in low-light and night cycling modes. Hydrosight Corporation 18 

has been contacted 

to supply the Star shield. 

2. Growing Star in a Vacuum. At

Cpyr is amended to JSC-1A because it has been fertile for hundreds of years with no crop 

rotation and still produces sustainable fertile soil. Three questions and concomitant experiments 

were performed to answer these questions. 

1. Can Cpyr amended soils heal, mend, or bring back to life seemingly dead plants? 

Answer: In Indian Fig Experiment 12, on September 27, 2008, seemingly dead cacti 

plants were planted in the St. Norbert College greenhouse. The experiment showed that 

Cpyr cannot bring seemingly dead plants back to life. 

1. In Star Cacti Experiment 18-C (Cold Tolerance Testing), during 4-LNC, Star went into 

freezing temperatures for the first time. Starting on May 13th cold temperatures reached 

27°F for two days consequently, Star appeared to have freezer burn. 

Answer: The damaged pad produced a bud showing the potential to overcome the frost 

damage. However, the bud soon regressed and the pad stopped growing altogether 

whereby a younger bud soon surpassed the growth of the damaged pad. 

2. Can Cpyr amended JSC-1A increase plant growth while night cycling? 

Answer: In Star Cacti Experiment 18-B, on July 5, 2009, 6-hour time lapse photographs 

were taken through a modified plate-glass porthole (cut into the refrigerator) while Star 

was night cycling for 14 days. The beginning photograph was compared to the ending 

photograph taken on July 19, 2009. On July 20, 2009 the following email was sent to the 

team, “I have attached a combined photo of two Star photos one at the beginning of 

night cycling and the other at the end of night cycling. I used a photoshop program 

to overlap them to see if the last one was bigger than the first one... indicating... 

possible growth while Star is in night cycling mode for 14 days of cooler 

temperatures and darkness. The results from the photo; very little if any growth 

while in night cycling mode.” 

Reproductive or Vegetative Organs? Nobel 22 

describes some comparisons between temperatures 

and the likelihood of what kind a new organ (Opuntia growth) might be. Nobel writes, “How 

low day/night temperatures favor initiation of reproductive organs and high temperatures 

favor vegetative organs is not clear, but future investigation using unrooted cladodes 

maintained under various temperatures, thermoperiods, and photoperiods may help 

elucidate the mechanisms underlying such environmental responses.” 

Nobel’s description of day/night temperatures, thermoperiods, and photoperiods are exactly the 

kinds of long-term experiments our team is now investigating. Currently Star 

22 Nobel P.S., Recent Ecophysiological Findings for O. ficus-indica. J. PACD – 1997. 

29

Cacti Experiment 18 has produced 100% daughter pads while in a 14/14 night/day lunar cycle 

with temperatures between 27 and 104°F. In addition, Dr Konings-Dudin at EPCC 23 

is preparing 

a 2009 fall biology class to further elucidate the mechanisms Nobel has observed. 

High Soil Salinity. Nobel and Bobich 24 

explain that high soil salinity has a negative effect on net 

CO2 uptake. Nobel writes, “Net CO2 uptake for O. ficus"indica decreases by about 50% 

after exposure to a 150mM NaCl solution and 83% after exposure to a 200mM solution for 

10 weeks. Longer term exposure to high concentrations of NaCl has an even more 

profound effect, with exposure to a solution of 100mM for six months causing a net CO2 

efflux for O. ficus-indica.” 

What would be the value of Cpyr if it could “filter” the salt before it gets to plant roots? In 

Soil Salinity Experiment 19, we attempt to answer this question. 500ml of 200mM 

(5.845gm) NaCl solution was poured into 40gms Cpyr using two single strand layers of gauze to 

minimize Cpyr loss. ~350ml of the solution was absorbed by Cpyr leaving 150ml of solute to be 

evaporated. Tiny bits of carbon were observed in the solution. Using a hot plate the temperature 

was slowly increased from 212 to 662°F (100 to 350°C). A glass rod was used to stir 

occasionally. 1.38gms of carbon-salt precipitate remained. The results of this experiment showed 

a >80% 25 reduction in NaCl concentration; from 200mM to just 40mM. What does this 

experiment mean? By amending plant soils with Cpyr the concentration of NaCl at the root tips 

would be reduced and therefore according to Nobel increase the net CO2 uptake of O. ficusindica. 

The following email 26 

was sent to the team, “In conclusion, pyrogenic carbon could be 

used as a barrier layer below and around high-salt soils to prevent the negative effects of 

NaCl exposure to O. ficus-indica. In addition, the positive attributes of Cpyr could induce 

higher growth potential for O. ficus-indica.” This translates to sustainable agriculture in 

highly saline soils; reclaiming habitat destroyed by salinity and pollutants. 

Pyro-DE Insecticide. When we think of crops we also think about pesticides and sometimes 

normality is put aside. In an email to Chrissy Paape 27 

we mentioned invasive species and Cpyr 

capabilities. The following was written, “I was thinking about Cpyr adsorption properties. If 

you apply an EAB insecticide (adsorbate) on Cpyr it will stay put and protect the Ash tree 

for an extended period of time. Currently insecticides are applied annually. And, you can 

see the problem in that Furthermore, trees succumb to EAB when they are under stress, as 

you know pyrogenic carbon reduces plant stress up to 4 fold. Also as you know when Cpyr 

is sequestered (Carbon Sequestration) in and around ash trees then it becomes a global 

warming solution, enhancing tree growth and increasing CO2 uptake. Any study should 

look at the life expectancy of the different types of EAB insecticides when adsorbed to 

23 

El Paso Community College, Introductory Biology, Classroom Projects; 1) Germination of 

Opuntia ficus-indica seedlings, 2) O. ficus-indica pad growth, 3) Vegetative propagation, 4) 

Symbiotic bacteria on roots (SEM). 

24 

Nobel and Bobich, Nutrients and Salinity. Additional information regarding this reference is 

unknown at time of writing. 

25 

Some of the precipitate was carbon; therefore the reduction in NaCl was greater than 80%. 

26 

Thursday - July 30, 2009, 11:08 AM. 

27 

chrissy@space-explorers.com, Tuesday – November 4, 2008, 6:20 PM. 

30

pyrogenic carbon. Perhaps instead of annual insecticide applications it could be stretched 

to 3 years, perhaps 5 years or more when applied to Cpyr.” 

In Cpyr-DE Slurry Experiment, EAB Mitigation, we attempted to make a long lasting non-toxic 

mechanical insecticide by combining Diatomaceous Earth (DE) and Cpyr. 10ml of 

Cpyr and 10ml of DE were mixed together in 200ml of water. DE and Cpyr in water did not bind 

at STP when stirred. Also the slurry stuck to the glass rod when pulled from the slurry. Then, the 

slurry was occasionally stirred with a glass rod while heating to 464°F (240°C). At this point the 

glass rod was coming out of the slurry relatively clean. A drop of the heated slurry was placed on 

a microscope slide. The following was noted in an email 28 

to the team, “I put a sample on a 

slide and the DE in some places was absorbing the carbon into the DE cells and in some 

places tiny specs of carbon were adhering to the DE. It seems that if I "cooked" it long 

enough that the DE would assimilate the carbon. In any case I think this is a significant 

find... that I can get Cpyr and DE to bind. It might turn out to be a valuable insecticide.” 

Applications: Pyro-DE injected under the bark where EAB and Mountain Pine Beetle 

infestations are suspected could mitigate that infestation. Furthermore, the value of Cpyr cannot 

be underestimated for increasing plant fertility and reducing plant stress, especially in infested 

trees. Infested, dying and decaying trees are now considered a massive source of CO2 emission 

rather than them being a natural carbon sink. We propose that research along these lines is 

applicable to forestry habitat and agricultural sustainability. 

MAX-X. Below is a list of future experiments but, to place Star on the moon by 2016, we will 

need to conduct all-encompassing or maximum experiments (MAX-X) over the next couple of 

years. So, what would a MAX-X experiment look like? During the lunar day Star would be 

exposed to high-heat and radiation while growing in a vacuum. During the lunar night Star 

would be exposed to cryogenic temperatures and radiation while growing in a vacuum. MAX-X 

acclimation experiments would give Star the greatest chance to survive on the moon. 

Future Experiments. 

1. To increase the germination rate of O. seeds. 

2. Study effects of lunar night/day cycling on germination rate of O. seeds. 

3. Do flowers develop during lunar cycling? 

4. After artificial pollination, how do fruits develop during lunar cycling and how much 

time do they need? 

5. Test O. survival in extreme thermal cycling under specified conditions such as humidity. 

6. Metrics of O. adaptability. 

7. Does lunar cycling increase O. adaptability to extreme conditions? 

8. Can O. fragilis, O. phaeacantha, and O. polyacantha tolerate the heat as well as the cold? 

9. Can O. fragilis, O. phaeacantha, and O. polyacantha pass the “cold tolerance factor” 

gene to an O. ficus-indica seedling via grafting? 

10. How will the O. ficus-indica seedling develop on a cold tolerant plant? 

11. How will and O. fragilis, O. phaeacantha or O. polyacantha seedlings grow on a O. 

ficus-indica graft and would they be edible? 

12. Can organic wastes supplement N in lunar regolith? 

28 Wednesday – March 11, 2009, 5:13 PM. 

31

13. What N decomposer's are needed and will they survive on the moon? 

14. Can Cpyr replace CO2 for plant growth on the moon? 

15. Will a plant grow in a vacuum using Cpyr? 

16. Is plant life, life in general, possible in a vacuum? 

17. How long can Star live in a vacuum? 

18. What are the protective measures for growing a plant in a vacuum? 

19. What are the advantages and disadvantages of using mineral supplements in JSC-1A 

simulant? 

20. Would Amaranth be a viable Star alternative? 

21. Would Double Claw be a viable Star alternative? 

22. Would corn be a viable Star alternative? 

32


Part Nine 

Chemistry

Abstract 

Toxic Offgassing Analysis at Marshall Space Flight Center 

Kenion Blakeman 

Department of Chemistry 


The toxicity laboratory tests flight hardware and materials to ensure offgassing, primarily 

volatile organic compounds, in space vehicles and the International Space Station does not 

compromise astronaut safety. This is accomplished using analytical chemistry methods 

including gas chromatography and gas chromatography-mass spectrometry to analyze known 

volumes of gas samples taken from the headspace above the test object in a sealed test chambers. 

Successful experiments this summer focused on better understanding the offgassing process and 

common products, and included response factors, limits of quantitation, and peak area versus 

temperature. This new information will further improve test methods that help ensure astronaut 

safety. 

Introduction 

Offgassing causes gases to be released into the atmosphere at normal temperatures humans live 

in every day. Earth’s atmosphere contains enough air to prevent this posing a problem for 

human health. However, during missions into space the air in a spacecraft must be recycled, and 

offgassing becomes important because it can affect astronaut health. The toxicity laboratory at 

Marshall Space Flight Center (MSFC) analyzes offgassing from flight hardware materials and 

hardware to ensure it does not compromise astronaut safety. 

In toxicity testing a sample of material or piece of flight hardware undergoes test preparation and 

documentation before being placed in a test chamber chosen at a density of 5.00±0.25g of the 

test item per liter of chamber volume. The sample is baked at 120±5ºF for 72±1 hours before 

analysis, which simulates the harshest conditions astronauts would face in space during an 

emergency. Gas chromatography-mass spectroscopy (GM-MS) served as the primary method 

used to identify offgassing products. Gas chromatography (GC) with a flame ionization detector 

(FID) quantifies offgassing components, and requires peaks identification from GC-MS library 

identification. The instruments used are an Agilent 6890 GC/FID and Autosystem GC/FID, 

Finnigan Voyager GC-MS, and Incos GC-MS. Fourier transformed infrared spectroscopy (FT- 

IR) identifies ammonia offgas concentration. 

A set of formulas determine offgas concentrations for a material or flight hardware. The 

spacecraft maximum allowable concentration (SMAC) value is considered to be the greatest 

concentration of a specific compound to which astronauts can safely be exposed to. SMAC 

______________________________________________________________________________ 

This research was generously supported by the Wisconsin Space Grant Consortium. I would also like to thank my 

summer research mentors, Eddie Davis and Robin Moore from NASA-MSFC. 

1

values are determined through toxicological evaluation performed by Johnson Space Center 

(JSC) Scientists. Most offgass concentrations are found safely below the SMAC value, and a 

summation formula for toxicity determines how much material or how many units can be flown 

safely. Results are entered into a database, the Materials and Process Technical Information 

System (MAPTIS), for future reference. 

A number of experiments were run to better understand the offgas process, and continue 

improving toxicity testing. The main procedure variation was instead of baking the chamber for 

72 hours the experimental chambers were baked anywhere between 4 and 24 hours to save time. 

The first experiment investigated how combining a volatile organic mix solution and flight 

material would affect offgassing. The Mix, material, and a combination of both were placed in 

separate chambers and analyzed by identical methods to look for differences in peak areas. 

Limits of quantitation (LOQ) tell the minimum amount of analyte that can be quantified by a 

particular instrument or method. Offgassing must be quantifiable below its SMAC values to 

ensure offgassing can be identified before it poses health hazards. Two Mixes were analyzed for 

limits of quantitation by injecting the smallest volume possible that gave integrable peaks. 

A response factor, represented by the letter K, is the ratio of compound concentration to the 

corresponding peak area. In the past, response factors were calculated with results from flight 

hardware, because this was a fast and easy method that gave quality results. In this research 

standard volatile organics Mixes were analyzed to compare response factors to standard solutions 

as opposed to test data, as well as the difference between the GCs, an Agilent 6890 GC/FID and 

Autosystem GC/FID. 

Under normal toxicity test procedures, samples cool to ambient room temperature before gas 

chromatography analysis. Another experiment looks at the effects of chamber cooling on peak 

area. Peak area should decrease over time as the chamber reaches room temperature. Slight 

leaks in the chamber could cause the peak areas to decrease slightly over a long period of time 

due to offgas loss to the air. 

Procedure 

Two 5.0g samples of material log number 108053 were weighed in weighing boats. Since the 

material was a large piece of cloth, pieces of it were cut until 5.0g was reached. The material 

was transferred to 1.00L sealed chambers in the weighing dishes. The chambers had previously 

been purged for at least five minutes with high quality air and were examined for seal strength. 

Gloves were worn while handling the sample material. A syringe was rinsed once with Mix 4 

and used to transfer Mix 4 to two chambers, 30uL to the chamber that already contained 108053, 

and 50uL to the chamber with only a weighing bowl. 

Only one rinse with volatile organics Mix 4 was used to clean the syringe to conserve the 

chemical mix as quantities were limited. Isopropanol was used to rinse the syringe after the Mix 

4 had been transferred to the chambers. When not in use Mix 4 was kept in a refrigerator. For 

transferring Mix 4, a syringe that had been baked for at least 24 hours was selected. 

2

The chambers were sealed and baked at 120±5ºF for 24 hours, after which analysis could be 

peformed. An alkane standard was run on both gas chromatographs (GCs) before running the 

actual samples to ensure proper instrument performance. Samples were also allowed to cool to 

room temperature before analysis, around 75ºF. For the Autosystem, 2.0cm 3 of headspace 

sample was injected, while 1.0cm 3 was injected for the 6890, both in duplicate. Standard toxic 

lab procedures were used for all runs. When samples were taken from the chambers to be 

analyzed, the syringe was rinsed at least three times with sample before injecting into the GC. 

Two 2.0cm 3 syringes were used for all the transfers as they were graduated and could be used to 

accommodate both injection volumes. For identification purposes each of the three samples was 

run on GC-MS as well. A similar procedure was also used to look at the interactions between 

108193 and Mix 1. The difference was 50μL of Mix 1 was used in each of the two chambers 

opposed to unequal Mix volumes. 

To determine response factors 10μL of Mix 1 was added to a purged 1.00L test chamber and 

baked overnight. After cooling the chamber to room temperature four 1.0cm 3 samples were run 

on the Agilent 6890 GC/FID and three 2.0cm 3 samples were run on the Autosystem GC/FID. 

Following these runs 10μL of Mix 2 was added to the chamber and after baking for four hours 

and cooling once again, three 1.0cc samples were run on the Agilent 6890 GC/FID and three 

2.0cm 3 samples were run on the Autosystem GC/FID. The chamber was purged with high 

quality air and 10uL of Mix 2 was added to the chamber. After allowing it to bake overnight 

four 1.0cc samples were run on the Agilent 6890/FID and three 2.0cm 3 samples were run on the 

Autosystem GC/FID. To identify peaks 40mL samples of Mix 1, Mix 2, and Mixes 1 and 2 

combined were run on the Incos GC-MS. 

In a 1.00L test chamber 10μL of Mix 1 was added and baked for four hours. After cooling to 

ambient room temperature 100μL of the sample was injected in the Autosystem GC/FID. The 

resulting peaks were too large, and a 25μL sample was run, which resulted in too small of peaks. 

When a 50uL sample was run the peaks were ideal. Seven trials were run using one syringe for 

both injections because one of the two 50μL syringes had poor airflow. The same procedure was 

used in analyzing Mix 2 and six sets of data were taken for it. Incos GC-MS spectra from 

previous experiments were used for identification of peaks. To quantify the detection limits an 

Excel template was used. 

The first sample tested was log number 108148, a green foam. A 5.0g sample was placed in a 

1.00L chamber and baked for 24 hours after which it was removed from the oven and a 1.0μL 

sample was injected into the Autosystem GC/FID three minutes later. Four more samples were 

injected at approximately 48 minute intervals as soon as the GC cooled. The next day after 

removing the chamber from the oven a run was done 19.5 hours later. Times for each run and 

the temperature of the chamber were recorded at the time samples were injected into the GC. A 

60mL sample was also run on the Incos GC-MS after it had cooled for about an hour for 

identification purposes. 

The next sample analyzed was composite material, log number 107999, three 5.0g white foam 

cubes. The same basic procedure was used as before with several exceptions. Room 

temperature was also recorded as the temperature fluctuated before, and air conditioning setting 

3

changes caused the room to be about 3ºF cooler than before. Since the GC-MS peaks for 108148 

were small a 100μL sample was injected into the Incos GC-MS. In total seven trials were run on 

the 107999 sample. 

Results 

Table 1 shows the response factors that have been calculate to date. 

Table 1. Mix 1 Response Factors 

Gas Code Component Previous Response Factor New Response Factor 

061500 Chlorobenzene 2.0 0.95 

035500 Xylene 1.05 1.48 

030800 Isopropylbenzene 0.75 1.55 

068600 Chlorotoluene 0.50 0.85 

034000 Propylbenzene 0.50 (default) 2.45 

012401 Tert-butylbenzene 0.50 (default) 1.50 

068801 Dichlorobenzene 0.50 (default) 0.51 

0300915 Sec-butylbenzene 0.50 (default) 1.60 

Tables 2 and 3 show the limit of quantization for Mixes 1 and 2, and the method detection limits 

(MDL) intermediate results needed to reach it. The SMAC value are provided for each 

component. 

Component MDL (ppm) MDL 

(µg/g) 

Table 2. Mix 1 Limits of Quantitation Summary 

MDL 

(mg/ 

LOQ 

(mg/ 

SMAC 

(mg/ 

Chlorobenzene 0.057 0.26 0.18 0.46 45.85 

Xylene 0.011 0.05 0.03 0.09 73.00 

Xylene 0.019 0.08 0.06 0.14 73.00 

Isopropylbenzene 0.024 0.12 0.08 0.21 73.27 

Chlorotoluene 0.090 0.47 0.33 0.82 0.10 

N-propylbenzene 0.022 0.11 0.08 0.19 48.92 

Tert-butylbenzene 0.038 0.21 0.15 0.37 0.10 

Sec-butylbenzene 0.060 0.36 0.25 0.64 0.10 

Dichlorobenzene 0.021 0.12 0.08 0.20 30.00 

Component MDL (ppm) MDL 

(µg/g) 

Table 3. Mix 2 Limits of Quantitation Summary 

MDL 

(mg/ 

LOQ 

(mg/ 

SMAC 

(mg/ 

Styrene 0.011 0.05 0.03 0.08 42.50 

Trimethylbenzene 0.010 0.05 0.04 0.09 60.30 

Trimethylbenzene 0.106 0.52 0.36 0.91 60.30 

Isopropyltoluene 0.017 0.09 0.07 0.17 13.70 

Butylbenzene 0.021 0.09 0.07 0.16 5.50 

Trichlorobenzene 0.105 0.78 0.55 1.36 0.10 

Naphthalene 0.031 0.16 0.11 0.29 N/A 

4

Tables 4 and 5 show offgassing components and peak areas for material 108148. Blank spaces 

correspond to peaks that were not integrated by the computer software, usually because of their 

small area. 

Table 4. Peak Intensity as a Function of Temperature 

Time Removed 

Methanol 

Peak Areas 

Methanol Freon 113 Freon 113 

From Oven (min) Temperature (F) (Channel A) (Channel B) (Channel A) (Channel B) 

3 106.7±2.0 2748292 2950781 20450 21150 

50 80.7±2.0 1472803 1899586 

97 77.0±2.0 1746735 556734 

145 76.4±2.0 1020035 1469113 

193 76.1±2.0 1197274 1469495 

1171 76.1±2.0 1451378 1137074 

Table 5. More Peak Intensities as a Function of Temperature 

Peak Areas 

Time Removed 

Phenol Phenol Ethanol-2-(2Ethanol-2-(2- From Oven Temperature (Channel (Channel ethoxyethoxy)ethoxyethoxy) (min) 

(F) A) B) acetate (Channel A) acetate (Channel B) 

3 106.7±2.0 126980 81014 698483 509213 

50 80.7±2.0 24631 27797 118966 196505 

97 77.0±2.0 25476 147102 72896 

145 76.4±2.0 16595 19631 87619 168685 

193 76.1±2.0 20605 19152 98511 139476 

1171 76.1±2.0 16296 129293 155580 

Tables 6 and 7 show peak areas for the offgassing of sample 107999. 

Table 6. Peak Areas as a Function of Temperature 

Peak Areas 

Time Removed 

Methanol Methanol Xylene (2) Xylene (2) 

From Oven (min) Temperature (F) (Channel A) (Channel B) (Channel A) (Channel B) 

1 105.8±2.0 221369 103281 63066 50465 

48 78.8±2.0 218732 93573 61905 48217 

96 74.1±2.0 120907 137732 33725 56761 

144 73.2±2.0 189578 58842 52246 29195 

224 72.6±2.0 181453 52436 53133 27178 

273 72.6±2.0 182577 41552 48558 23042 

1437 71.9±2.0 142189 21832 38150 

Table 7. More Peak Areas for 107999 Offgassing 

Peak Areas 

D-Limonene D-Limonene Ethanol-2-(2-ethoxyethoxy)Ethanol-2-(2-ethoxyethoxy)- (Channel A) (Channel B) acetate (e.e.a) (Channel A) acetate (e.e.a) (Channel B) 

47320 42150 70331 

40319 48217 58133 21151 

24521 42979 21433 46025 

34341 20051 36491 

37022 19196 29037 31496 

35290 28413 

26487 22507 

5

Figure 1 shows peak areas for offgassing from 107999 as a function of time when a significant 

number of integrated peaks are present. 

Discussion 

Figure 1: Peak Areas Over Time 

The response factors calculated using the new data were different from the previous response 

factors. There was no trend as to whether they were higher or lower as they varied between 

components. Since these values were calculated based on standard solutions this would suggest 

these values would have high accuracy. For a lot of compounds from offgassing there were only 

default response factors which have little accuracy, so new calculated response factors give a 

better idea of what they should be. 

When the Mixes for response factors were placed in separate chambers and baked, the peak areas 

were inconsistent due to different chambers having different seals. Unequal seals were expected 

but the extent to which chambers were different was unexpected. To get more consistent results 

a single chamber had to be used for each different test. While this did allow more accurate 

results to be found the drawback was the time involved with running tests one at a time. Since 

the seals were unequal, there was no way around this problem, which was important to discover 

for future experimentation. 

When samples of 108193 and Mix 1 were run, it was clear that 108193 did not offgas well. The 

largest peak detected on the Autosystem GC/FID was the injection peak, followed by the column 

peak. Injection and column peaks were the only peaks seen on the Agilent 6890 GC/FID. When 

the sample was run on the Incos GC-MS, there was no flat baseline and only a few small peaks 

were observed, consistent with the GC results. Spectra of Mix 1 showed strong peaks 

corresponding to the components Mix 1, and the spectra of the Mix and 108193 only showed 

peaks corresponding to Mix 1. Any interactions between Mix 1 and 108193 were too small to 

identify. 

Both 108053 and Mix 4 offgasssing produced dichlorobenzene, and for 108053, it was the only 

quantifiable offgas. However, when they were baked together in a chamber and run on GCMS 

6

there was no dichlorobenzene peak. GC spectra confirmed that no dichlorobenzene could be 

found in the sample. Considering dichlorobenzene was present in both samples, a large peak 

was expected but not detected. 

One of the challenges of this data was comparing effects of the sample and mix on peaks areas 

when the amount of Mix 4 used was different between chambers. Peak areas varied a lot simply 

due to the volume difference of Mix 4, making it difficult to analyze. Unfortunately by the time 

this was realized the solution of Mix 4 had been used up and there was no way to further analyze 

the interactions of the Mix and sample. 

Peak areas for offgassing of 108148 were expected to decrease with decreasing temperature. 

This happened for the first two runs, and after that there was a general decreasing trend. 

However, different peaks increased and decreased at different points, suggesting the relationship 

was not that simple between peak area and temperature. 

Data from 107999 showed more identifiable offgassing which was helped by using 15.0g of 

sample as opposed to 5.0g. For the most part these results showed a more consistent decrease in 

peak area over time. For both 107999 and 108148 the peaks at 97 and 96 minutes had 

inconsistent areas for each column. The 107999 front peaks had small areas while the back 

column peaks had large areas. The opposite was true for 108148 which had large front column 

peak areas and small back column peak areas. While there is no clear explanation for why these 

peaks at about 96 minutes were inconsistent, the fact that it was observed in both samples 

suggested it was significant. 

Conclusion 

These experiments involved studying a variety of aspects of toxicity testing. Standard Mixes 

were used to look at response factors and detection limits of a variety of molecules. Mixes were 

also used to look at their interactions with materials, with limited success. Peak area dependence 

on temperature was investigated using material samples. Knowledge learned from these 

experiments served as part of continuous efforts to improve toxicity testing and the 

understanding of it. 

References 

EM10-OWI-CHM-039, Revision B (Marshall Space Flight Center) “Determination of Offgassed 

Products Testing (Toxicity),” November 28, 2007. 

NASA-STD-6001 “Flammability, Odor, Offgassing, and Compatibility Requirements and Test 

Procedures for Materials in Environments that Support Combustion,” February 9, 1998. 

7

Preparation and Characterization of Platinized Electrodes for Use in Oxygen 

Extraction from Lunar Regolith 

Abstract 

RA: Nathan Wong 

PI: Dr. Peter A Curreri EM30 

Lunar oxygen is an important resource that can be utilized by humans returning to 

the moon, for both life support on the lunar surface, and propulsion to return back to 

Earth or to continue exploration of the solar system. Oxygen is an abundant resource on 

the moon, but it is chemically bound to other elements in the form of oxides in the lunar 

regolith. One process being studied for the extraction of oxygen from lunar regolith 

involves dissolving the regolith in an ionic liquid. This reaction produces a solution of 

water and metal ions. The water is then separated by distillation and electrolysis is 

performed to extract the oxygen as well as the hydrogen to regenerate the ionic liquid. A 

prime candidate material for the electrodes is platinum because it is highly conductive 

and resistant to corrosion. The oxygen and hydrogen production rate are proportional to 

the electrode surface area. The process of platinization of a platinum electrode involves 

electrochemically adhering a coating of platinum black on the platinum electrode. 

Platinum black has a significantly higher surface area than sheet platinum. This study 

used cyclic voltammetry to compare the amount of current produced between two 

electrodes for a given voltage. It was found that the current between blackened 

electrodes can be as much as a factor of 100 greater than the current between 

unblackened electrodes. 

Introduction 

Oxygen has many uses, from supporting life to providing a fuel for propulsion. 

On Earth oxygen is an abundant resource that is part of our atmosphere. On our moon, 

oxygen is also an abundant resource but it is in the form of oxides in the lunar regolith. 

Lunar regolith contains about forty-two percent oxygen by weight, but methods are 

needed to release the oxygen from the chemical bonds. 

One method is molten oxide electrolysis. According to Curreri et al. (2006) 

Molten oxide electrolysis involves heating lunar regolith until it becomes molten. Then 

electrolysis is performed on the molten oxide to extract the oxygen. The benefit of 

molten oxide electrolysis is that it requires no consumable reagents that would need to be 

launched to the moon on a regular basis. The drawback of molten oxide electrolysis is 

that in order to melt and process the lunar regolith temperatures of 1,500-1,600°C are 

needed. This requires a large amount of initial energy to be expended, and can provide 

safety and maintenance problems. 

Another method involves the use of ionic liquids in chemical beneficiation of the 

lunar regolith. Ionic liquids are comprised solely of oppositely charged ions, and 

according to Paley et al. (2009) ionic liquids have characteristics that make them ideal for 

space exploration. Ionic liquids have low vapor pressure and low flammability. This 

makes them easy to store, and safe to handle compared to other possible chemical 

9

eneficiation reagents. Ionic liquids are stable in extreme temperatures and in a hard 

vacuum. These characteristics make them ideal for space travel where the ionic liquid 

will be subject to these conditions. Ionic liquids’ molecular structure can be altered to be 

task specific. This provides versatility to the work that ionic liquids can do. 

The use of ionic liquids in the production of oxygen from lunar regolith involves 

three steps. The first is to alter the ionic liquid so it has the characteristics of an acid. 

The ionic liquid is then used to break down the metal oxides into water and metal ions, 

followed by distillation of the water. This is typically done at temperatures of about 

150°C. Secondly, electrolysis is performed on the water to separate the hydrogen and 

oxygen, and lastly, the ionic liquid is regenerated using the hydrogen generated from the 

electrolysis. 

In the electrolysis a platinum electrode is used due to its high conductivity and 

resistance to corrosion. The hydrogen and oxygen production rates are directly 

proportional to the electrode surface area. One method of increasing the surface area of 

the electrode is by platinizing the electrode. This involves coating the electrode with a 

layer of platinum black. Platinum black has a much higher true surface area than the 

geometric surface area. In this study the effects of using a blackened platinum electrode 

compared to an unblackened electrode are studied using cyclic voltammetry. 

Experimental methods 

Platinum foil was heated to 1000°C in a Fisher Isotemp Muffle Furnace to clean 

the surface of the platinum of debris and organic material. The cleaned platinum was 

fastened to a wire to create an electrode that would be used in testing. 

Figure 1: Platinum foil is cleaned by heating to 1000°C. 

The electrodes were placed in an aqueous solution of .072 mol/kg chloroplatinic 

acid and .00013 mol/kg lead acetate. Two volts were run through the system for ten 

10

minutes and the negative platinum electrode became coated with a layer of platinum 

black. To increase the adherence of the platinum black the process was performed in a 

sonicator. In studies done by Marrese (1987) this removed loose particles and kept the 

strongly adhered particles. Alternation of the polarity on the electrodes was also 

performed in some tests to platinize two electrodes at the same time similar to the work 

done in MacNevin and Levitysky (1987. 

Figure 2: (Left) On the bottom is the power supply used to supply the system with two 

volts, and on top is the waveform generator used to alternate the polarity of the 

electrodes. (Right) Platinum electrodes in the platinization solution. 

A current profile of each of the electrodes was found by using cyclic voltammetry 

on the voltage range of -.2 volts to 1.2 volts. This range was chosen so that the current 

profiles could be compared to previous studies performed on platinized electrodes in 

Marrese (1987). The measurements obtained were current and voltage compared to a 

silver/silver chloride reference electrode. 

Results 

Cyclic voltammetry data was recorded and plotted using a spreadsheet program. 

This data was used to generate a current versus voltage plot. This was done for both 

blackened and unblackened electrodes. The results for the unblackened electrode can be 

seen in figure 3 and the results for the blackened electrode can be seen in figure 4. The 

current for the unblackened electrode is in micro amps and the current for the blackened 

electrode is in milliamps. This shows that the current produced by the blackened 

electrode is up to one hundred times that greater than that of the unblackened electrode. 

11

Figure 3: Current versus voltage data for the unblackened electrode. The voltage range 

is from -.2 to 1.2 volts and the current is in micro amps. 

Figure 4: Current versus voltage data for the blackened electrode. The voltage goes 

from -.2 to 1.2 volts and the current is in milliamps. The current produced by the 

blackened electrode is up to one hundred times greater than that produced by the 

unblackened electrode. 

12

Conclusion 

The use of blackened electrodes in the production of oxygen from lunar regolith 

will increase the rate of oxygen and hydrogen production. This is due to the increased 

surface area of the blackened electrode. The use of a sonicator during the platinization 

process provides for a more durable, longer lasting electrode. With the procedures used 

in this study, platinized electrodes can be easily produced in the lab that provide a good 

candidate for the electrode to be used in the electrolysis of water produced from lunar 

regolith and the ionic liquid. 


I would like to thank Dr. Peter A. Curreri and Dr. Laurel Karr for their mentorship 

and guidance in this project. I would also like to thank Dr. Steve Paley and Dr. Matt 

Marone for their help in the lab, and for answering my endless questions. This work is 

supported by the NASA Academy program at the Marshall Space Flight Center and the 

Wisconsin Space Grant Consortium. Without everyone and all the programs involved in 

the project, none of this would be possible, and I am very grateful for the opportunity of 

this experience. 

References 

Curreri, Peter A et al. Process Demonstration for Lunar In Situ Resource Utilization – 

Molten Oxide Electrolysis. MSFC Independent Research and Development 

Project. August 2006 

MacNevin, William M, Levitsky, M. “Reproducible Platinized Platinum Electrode for 

Anodic Polarography”.Analytical Chemistry 24.6 (1952) 973-975 

Marrese, Carl A. “Preparation of Strongly Adherent Platinum Black Coating”. Anal. 

Chem. 59.1 (1987) 217-219 

Paley, Mark Steven et al. Oxygen Production from Lunar Regolith using Ionic Liquids. 

Space, Propulsion, and Energy Sciences International Forum. February 2009. 

13

Abstract 

New Initiatives in the Project on Fossilization via Silicification 

Vera M. Kolb 1 and Patrick J. Liesch 2 

1 Department of Chemistry 

University of Wisconsin-Parkside 

2 Department of Entomology 

University of Wisconsin Madison 

In t his pa per w e r eport o n o ur n ew in itiative in th e p roject o n f ossilization v ia 

silicification. We focus on fossilization of insects and the role that chitin might have in 

facilitating th e s ilicification. In th e i ntroduction w e of fer r ationalization f or t he 

experiments. In the experimental section we report the initial results of the silicification 

of insects, which were successful. 

Introduction 

In th e p ast w e h ave s tudied th e in teraction o f s odium s ilicate w ith various s mall 

molecules, s uch as s ugars, a mino a cids, a lcohols, a nd ot hers. O ur recent review 

summarizes our work, which was continuously sponsored by the Wisconsin Space Grant 

Consortium (Kolb and Liesch, 2008). We became interested in the possible silicification 

of insects via chitin, which is a polysaccharide they utilize extensively. This would be the 

natural extension of our work on the silicates of sugars (Kolb and Zhu, 2004; Lambert et 

al. 2004). Chitin is a long-chain polymer of N-acetylglucosamine in which the units are 

linked via beta-1,4-linkage (Voet et al., 2006). Chitin is a constituent of the exoskeletons 

of i nvertebrates s uch a s crustaceans a nd i nsects, and i s a lso f ound i n t he c ell w alls of 

most f ungi and ma ny algae. C hemically, c hitin is s imilar to c ellulose in w hich o ne 

hydroxyl group on each m onomer i s r eplaced by a n aminoacetyl f unction. C hitin is 

seemingly ubiquitous in nature, and is almost as abundant as cellulose (Voet et al., 2006). 

Chitin is a la bile mo lecule, b ut b ecomes r esistant to d ecay w hen c omplexed w ith 

proteins. Such complexation occurs in arthropod cuticles, in which chitin is cross-linked 

with pr oteins. S tankiewicz a nd B riggs (1997) ha ve s hown b y t he p yrolysis-gas 

chromatography-mass s pectrometry (py-GC-MS) a nalysis t hat c hitin s urvived i n 25 - 

million-year old fossilized insects. They have identified pyrolytic remnants of the chitin 

and the associated protein. Flannery et al. (2001) have also shown that chitin can persist 

under f avorable c onditions i n f ossils, a nd ha ve a dditionally i dentified t he g lucosamine 

moiety b y G C-MS s elected i on m onitoring ( GC-MS-SIM). I nfra-red ( IR) s pectroscopy 

was used by various investigators to observe and assign the chitin bands (Biniaś et al., 

2007; Briggs et al., 1998; Brugnerotto et al., 2001; Gow et al., 1987). This technique is 

available t o us , a nd w e have s tudied t he IR s pectra o f s ilica gels th at were formed b y 

silicification of va rious organic s amples ( Kolb a nd Liesch, 2008) . Thus, i nformation 

exists for the IRs of both chitin and silicates. 

15

Various modes of fossilizations of insects are reviewed (Grimaldi and Engel, 2005). We 

are interested the most i n the fossils that are pr eserved in chert, such as in the Rhynie 

chert (from t he O ld R ed S andstone i n S cotland; P ragian; a bout 396 -407 millio n years 

ago). The R hynie ch ert contains t he f ossil o f what i s b elieved t o b e t he o ldest i nsect 

(Engel and Grimaldi, 2004). Chert is microcrystalline silica, SiO2. It is translucent and 

provides good qua lity fossils, c omparable t o t hose i n a mber, but f rom a m uch ol der 

period. The insects were trapped in ancient shallow pools created by hot springs which 

then rapidly silicified. Similar to chert is the Tertiary onyx, which is also a form of silica. 

In the initial stages of our project we attempted to silicify insects in the laboratory with 

sodium silicate at room temperature. We believe that the chitin is a good candidate to 

undergo silicification due to its hydroxyl groups that can both hydrogen-bond to the silica 

and can be entombed inside the silica gel. D ue to the fact that proteins are also present 

together with the chitin, the probability for silicification is increased, since proteins are 

known t o i nduce pol ymerization of s ilicates to s ilica ge l. Below are s hown ou r 

experiments. 

Experimental 

For experiments JB1-JB4B, beetles were collected in Shawano, WI on May 29, 2008, and 

had originally been preserved in 70% EtOH. Before the experiments, the beetles were 

removed from the alcohol and allowed to air dry for one week. At the time that 

experiments JB1-JB4B were started (after the week of drying) there was still some 

moisture remaining inside of the beetles--either due to water or residual alcohol. 

JB1: 24. March. 2009 

Phyllophaga anxia (adult); 0.41g. The beetle was placed in scintillation vial and 4 ml of 

Na-Silicate w ere a dded. This v ial w as k ept ti ghtly sealed to prevent desiccation. 

Initially the beetle floated in the milky, viscous fluid. A rubbery gel appeared after 48 

hours. The gel w as milky and translucent. The beetle w as embedded i n gel, with the 

dorsal s urface of t he be etle e xposed a bove t he surface of t he gel. A s mall a mount o f 

clear, yellowish liquid was present above the gel. This liquid has pH of ~11. More liquid 

was noticeable above the gel after 5 more days (7 days total). The gel became discolored 

near th e to p ( slightly yellowish). S lightly mo re liq uid w as n oticeable o ver t he n ext 7 

days (14 days total). Observations were carried out for another week (21 days total) but 

no more changes were noted. 

JB2: 24. March. 2009 

Phyllophaga anxia (adult); 0.25g. The beetle was placed in a scintillation vial and 4 ml of 

Na-Silicate w ere ad ded. The cap of t his vi al was ke pt l oosely s ecured t o allow for 

desiccation. Initially the beetle floated in the milky, viscous fluid. After 22 hours, a fine 

ring of whitish gel was observed on the side of the vial at the surface of the Na-Silicate. 

This ring increased in thickness over the next 20 hours (42 hours total). The main mass 

of the Na-Silicate was also noticeably thicker at this time, but still flowed. This increased 

viscosity resembled that of our alcohol-gel experiments (Liesch and Kolb, 2007). By the 

third day of observations, a gel had formed and the beetle was embedded in the gel, with 

part of the beetle exposed. The gel was milky and translucent. A small amount of liquid 

16

was noticed above the gel. A discoloration (yellowing) of the gel was observed by the 7 th 

day of observations. Observations were carried out for 21 days total, although no further 

changes were noted. 

JB3: 24. March. 2009 

Phyllophaga anxia ( adult; w ith le gs removed to a llow f or easier p ositioning o f th e 

specimen); 0.24g. The beetle was placed in a scintillation vial and coated with 0.5 m l 

Na-Silicate. The Na-Silicate was added to the ventral surface of the thorax to coat the 

layer of fine hairs p resent. The cap o f this vial was kept loosely secured to allow for 

desiccation. Initially, t here w as a cl ear l ayer o f N a-Silicate c overing th e in sect b ody. 

After 40 hours, some gel was noticed at the bottom of the vial, where the vial came into 

contact with the dorsal surface of the insect (beetle was placed into the vial upside down). 

No fluid was observed with the gel. By the 6 th day of observations the gel had solidified 

on the hair-covered ventral thorax. A thin, second coat of Na-Silicate was applied on the 

7 th day. This second application appeared to have dried and solidified within a day (8 th 

day overall), and no more liquid could be observed. Observations were carried out for 13 

more days (21 days total), with no further changes noted. 

JB4A: 24. March. 2009 

The hard and heavily sclerotized wing covers (elytra) were removed from a Phyllophaga 

anxia adult beetle. The elytra were first viewed through an Omano® stereo microscope 

at 30x power, and were observed to be sparsely fringed with hairs along the medial edges 

and heavily fringed on t he ventroanterior surface. The elytra were covered with 0.5ml 

Na-Silicate, which formed a thin, uniform layer on the bottom of the vial. The cap of this 

vial was kept loosely secured to allow for desiccation. Initially the elytra were sitting in 

the clear Na-Silicate. After 40 hours, a thin layer of firm gel was noticed at the bottom of 

the vial. No liquid was observed above the gel. Some white/opaque gel was observed on 

the sides of the vial. Observation under the microscope revealed that the fine hairs were 

preserved, and had not been dissolved or destroyed by the basic nature of the Na-silicate 

solution. O bservations were carried out for 21 days total, and no further changes were 

noted. 

JB4B: 24. March. 2009 

The s ofter an d lightly s clerotized wings were r emoved from a Phyllophaga anxia adult 

beetle. The wings w ere f irst vi ewed t hrough an O mano® s tereo m icroscope at 30x 

power. The wings were observed to have reduced venation, and a few hairs were found 

at the base of the wing. The wings were placed into a scintillation vial and were covered 

with 0.5m l N a-Silicate, which f ormed a t hin, un iform l ayer on t he bot tom of t he vi al. 

The cap o f this vial w as k ept l oosely s ecured t o allow for desiccation. I nitially th e 

wings were sitting in the clear Na-Silicate. S imilar to experiment JB4A, a thin layer of 

firm gel was noticed at the bottom of the vial. No liquid was observed above the gel. 

Some w hite/opaque g el w as obs erved on t he s ides of t he vi al. O bservation unde r t he 

microscope r evealed t hat t he f ine ha irs w ere pr eserved, a nd ha d not be en di ssolved or 

destroyed by the basic nature of the Na-silicate solution. O bservations were carried out 

for 21 days total, and no further changes were noted. 

17

------------------------------------------------------------------------------------------------------------ 

For experiments JB5-JB10, beetles had been collected in Shawano, WI on June 6, 2009, 

and were not placed into alcohol. Instead, these beetles were allowed to “air dry” for 14 

days before using them in experiments. By the time the experiments began, the beetles 

had desiccated to the point moisture was not noticeable inside of the body cavity. 

JB5: 22. June. 2009 

Phyllophaga anxia (adult); 0.15g. The beetle was placed in a scintillation vial and 4 ml 

Na-Silicate was added. This vial was kept tightly sealed to prevent desiccation. Initially 

the beetle floated in the milky, viscous fluid. No changes were noted in the first 14 days. 

On day 14, specimen was sealed with Parafilm and shipped to UW-Parkside. No gel was 

present at that time, and the liquid was still milky, and had a viscosity similar to that of 

the original Na-Silicate. (Note: this sample is currently under observation at UW-Parkside 

for possible gel formation). 

JB6: 22. June. 2009 

Phyllophaga anxia (adult); 0.11g. The beetle was placed in a scintillation vial and 4 ml 

Na-Silicate w as ad ded. The cap of t his vi al was ke pt l oosely s ecured t o allow for 

desiccation. Initially the beetle floated in the milky, viscous fluid. After 7 days, the Na- 

Silicate was noticeably thicker and more viscous. The gel was slightly thicker by day 14. 

This gel seems to resemble those from our alcohol experiments (Liesch and Kolb, 2007). 

The v ial was sealed a nd s hipped t o U W-Parkside on da y 14. ( Note: t his s ample i s 

currently under observation at Parkside for possible gel formation). 

JB7: 22. June. 2009 

The hard and heavily sclerotized wing covers (elytra) were removed from a Phyllophaga 

anxia adult beetle. The elytra were first viewed through an Omano® stereo microscope 

at 30x power, and were observed to be sparsely fringed with hairs along the medial edges 

and heavily fringed on t he ventroanterior surface. The elytra were covered with 0.5ml 

Na-Silicate, which formed a thin, uniform layer on the bottom of the vial. The cap of this 

vial was kept loosely secured to allow for desiccation. After 36 hou rs a clear, rubbery 

gel was observed on the bottom of the vial. Observation under the microscope revealed 

the preservation of delicate structures (hairs). No further changes were observed through 

day 14. The vial was sealed and shipped to UW-Parkside on day 14. 

JB8: 22. June. 2009 

The softer and lightly s clerotized wings were r emoved from a Phyllophaga anxia adult 

beetle. The wings w ere f irst vi ewed t hrough an O mano® s tereo m icroscope at 30x 

power. The wings were observed to have reduced venation, and a few hairs were found 

at the base of the wing. The wings were placed into a scintillation vial and were covered 

with 0.5m l N a-Silicate, which f ormed a t hin, un iform l ayer on t he bot tom of t he vi al. 

This cap of vial was kept loosely secured to allow for desiccation. Initially the wings 

were sitting in the clear Na-Silicate. After 36 hours a clear, rubbery gel was observed on 

the bottom of the vial. Observation under the microscope revealed the preservation of 

18

delicate structures (hairs). N o further changes were observed through day 14. The vial 

was sealed and shipped to UW-Parkside on day 14. 

JB9: 22. June. 2009 

Phyllophaga a nxia ( adult); 0.16g. The b eetle w as pul verized t o a f ine p owder us ing a 

mortar an d p estle. A few l arger “flakes” o f exoskeleton remained in the powder. The 

crushed beetle was placed to a scintillation vial and 3 ml Na-Silicate was added. The cap 

of this vial was sealed tightly to prevent desiccation. Initially, most of the particulate 

matter s ettled to th e b ottom o f the v ial, a nd w as c overed w ith th e mil ky N a-silicate 

solution. S ome pieces of exoskeleton floated on the surface. The vial was tilted gently 

by hand to mix contents. The viscosity of the Na-silicate was noticeably higher after 24 

hours, but a true gel had not formed at that time (see experiment JB10). By 48 hours, a 

gel ha d formed. G el w as da rk ( brownish i n s pots) a nd contained vi sible pa rts o f t he 

crushed b eetle. This gel w as not e ntirely s olid. If vi al was t ilted on s ide, gel s lowly 

flowed ove r t ime--similar t o our alcohol e xperiments ( Liesch a nd K olb, 2007) . N o 

further changes were observed through day 14. The vial was sealed and shipped to UW- 

Parkside on day 14. (Note: this sample is currently under observation at UW-Parkside for 

further changes in gel consistency). 

JB10: 22. June. 2009 

Phyllophaga a nxia ( adult); 0.21g. The b eetle w as pul verized t o a f ine p owder us ing a 

mortar and pestle. A few larger “flakes” of exoskeleton remained in the powder. The 

crushed beetle was placed to a scintillation vial along with 3 ml Na-Silicate. The cap of 

this v ial w as k ept l oosely secured t o allow for desiccation. Initially, mo st o f th e 

particulate matter settled to the bottom of the vial, and was covered with the milky Nasilicate 

solution. Some pieces of exoskeleton floated on the surface. The vial was tilted 

gently by hand to mix contents. A gel had formed within 24 hours (1 day faster that when 

vial was sealed). The gel was dark (brownish in spots) and parts of the crushed beetle 

were visible throughout. Interestingly, the gel was not entirely solid: if vial was tilted on 

side, gel slowly flowed over time--similar to our alcohol experiments (Liesch and Kolb, 

2007). No further changes observed through day 14. The vial was sealed and shipped to 

UW-Parkside on da y 1 4. ( Note: t his s ample is c urrently unde r obs ervation a t UW- 

Parkside for further changes in gel consistency). 

Work done at UW-Parkside 

As t he s amples were r eceived, t hey were phot ographed b y the U W-Parkside’s 

professional phot ographer, D on Lintner, from t he M edia S ervices. Thus, a p ermanent 

record was made o f the appearance of each sample in terms of color, transparency and 

other details. M onitoring of the samples in which gel has not yet solidified continues. 

The IR studies of the insect parts prior to and after the silicification will begin shortly. 

The latter studies will disturb the gel structure and appearance and this is the reason why 

the photographs of the samples had to be taken first. We shall be looking for the IR bands 

specifically associated with chitin, and will try to assign as many other bands in the IR 

spectra as we can. We shall then repeat the IR studies after 2-3 months and afterwards, at 

19

the a pproximately s ame pe riods of t ime, t o obs erve a ny pos sible de terioration of t he 

chitin within the silica gel matrix. 

Conclusions 

Our in itial a ttempts in s ilicification o f in sects w ere successful. Our experiments w ill 

elucidate the insect fossilization process in the silicate –rich environments. The IR study 

of the chitin within the silica matrix will enable us to monitor any possible deterioration 

of chitin over a period of time. 


Thanks are expressed to the WSGC for a steady support of our projects on silicification 

in t he a strobiological context. We a re grateful t o D on Lintner f or hi gh qu ality 

photographs of our samples. 

References 

Biniaś, D., M. Wyszomirski, W. Biniaś, and S. Boryniec, 

“Supermolecular S tructure of C hitin a nd i ts D erivatives i n F TIR 

Spectroscopy S tudies”, P olish C hitin S ociety, M onograph X II, 95 -108 

(2007). 

Briggs, D. E . G., R . P. E vershed, B . A. S tankiewicz, “ The M olecular 

Preservation of F ossil Arthropod C uticles”, Ancient B iomolecules, 2, 

135-147 (1998). 

Brugnerotto, J., J. L izardi, F . M. G oycoolea, W. Argüelles-Monal, J . 

Desbrières, and M. Rinaudo, “An Infrared Investigation in Relation with 

Chitin and Chitosan Characterization”, Polymer, 42, 3569-3580 (2001). 

Gow, N. A. R., G. W. Gooday, J. D. Russell, and M. J. Wilson, “Infra-red 

and X -ray Diffraction D ata o n C hitins of Variable S tructure”, 

Carbohydrate Res., 165, 105-110 (1987). 

Grimaldi, D . a nd M . S . E ngel, “Evolution of t he Insects”, C ambridge 

University Press, Cambridge, pp. 42-49, 60-65, 357-380, 2005. 

Kolb, V. M. and W. Zhu, “Complexes of Ribose with Silicates, Borates, 

and C alcium: Implications t o Astrobiology”, i n “Instruments, M ethods, 

and Missions to Astrobiology VIII”, R. B. Hoover, G. V. Levin, and A. Y. 

Rozanov, Eds., SPIE Vol. 5555, pp. 70-77(2004). 

Kolb, V. M ., a nd P. J. Liesch, “ Role of amino acids and th eir M aillard 

mixtures w ith r ibose in th e b iosilicification p rocess”, in “ Instruments, 

Methods, and Missions for Astrobiology IX” R. B. Hoover, G. Y. Levin, 

20

and A. Y. Rozanov, Eds., SPIE, Vol. 6309, 630 90T(1-8) (2006), and the 

references therein. 

Kolb, V. M . and P. J . Liesch, “Models f or S ilicate F ossils of O rganic 

Materials i n t he Astrobiological C ontext”, C hapter i n t he book “ From 

Fossils to Astrobiology”, M. Walsh and J. Seckbach, Editors, Springer, 

Publ., 2009 (Published in November of 2008), pp. 69-88. 

Lambert, J. B. G. Lu, S. R. Singer, and V. M. Kolb, "Silicate Complexes of 

Sugars i n Aqueous S olutions", J . Amer. C hem. S oc., 126, 961 1-9625 

(2004). 

Liesch, P . J . a nd V. M . K olb, “ Importance of the interaction between 

sodium s ilicate a nd o rganic ma terials to astrobiology: Alcohol-based 

organo-silicates as p otential b iosignatures”, in “ Instruments, Methods, 

and Missions for Astrobiology X” R. B. Hoover, G. Y. Levin, and A. Y. 

Rozanov, Eds., SPIE, Vol. 6694, 669405(1-10) (2007). 

Stankiewicz, B. A., and D. E. G. Briggs, “”Preservation of Chitin in 25- 

Million-year-old fossils”, Science, 276, 1541-3 (1997). 

Voet, D., J. G. Voet, and C. W. Pratt, “Fundamentals of Biochemistry, Life 

at the Molecular Level”, 2 nd Edition, J. Wiley & Sons, Inc., 2006. 

21


Part Ten 

Education and Public Outreach

Combining Writing Across the Curriculum Strategies in Community-Based 

Programs to Teach Core Scientific Concepts 

James Kramer 

Executive Director 

Simpson Street Free Press 

Madison, Wi 

Abstract: Persistent achievement gaps affect many Wisconsin school districts. Correlations 

between income and achievement greatly increase the challenges facing local schools. A recent 

Harvard Family Research Project study demonstrates that learning support systems outside of 

school hours are a critical element in long-term student success. Many communities now 

recognize that partnerships with community-based non-profit groups are essential. The Simpson 

Street Free Press is a solid academic program, with outstanding community support, and a 

proven record of success. For 17 years the Simpson Street Free Press has successfully 

demonstrated that academic success is attainable for all kids. Our student reporters, ages 11-18, 

learn practical academic and vocational skills through an elaborate process of writing and 

publishing. We use proven writing across the curriculum strategies to produce thoughtful, wellresearched 

articles on topics ranging from the latest NASA missions to ancient civilizations; 

from plate tectonics and climate change, to Wisconsin’s changing economy. These nationally 

recognized best practices effectively bridge the achievement gap and as a result, 90% of our 

students increase their core GPA within two semesters and 92% of our program graduates have 

gone on to college. 

1. Organization History and Mission 

Wisconsin’s own Simpson Street Free Press is now officially designated “one of America’s best 

youth programs.” The prestigious 2008 national Coming Up Taller Award was recently 

presented to our organization at a White House ceremony in Washington, D.C. The award was 

presented by the President’s Committee on the Arts and Humanities, the National Endowment 

for the Arts, and the National Endowment for the Humanities. Committee members closely 

examined SSFP curriculum methods and cited our organization for “a thoroughly innovative 

approach to fostering academic achievement” and “pioneering new and exciting strategies to 

integrate core subject curriculum into community-based youth programs.” 

Across the country, communities search for innovative ways to promote achievement and 

engage young people in civic life. We do just that. For 17 years the Simpson Street Free Press 

has successfully demonstrated that academic success is attainable for all kids. Our student 

reporters, ages 11-18, learn practical academic and vocational skills through an elaborate process 

of writing and publishing. In turn, the publications produced by these hardworking kids reach 

thousands of their peers with powerful messages of achievement and success. It is this multimission 

approach that makes our programs so efficient and so effective. Simpson Street Free 

Press, Inc. is an organization built on ideas and innovation. We have honed an approach to 

community-based academics that really works. And, in turn, our publications influence young 

readers on a massive scale. Our core curriculum approach builds academic self-confidence, in 

particular for students from low-income backgrounds. The work our students produce is widely 

1

ead and very popular. Our student reporters produce thoughtful, well-researched articles on 

topics ranging from the latest NASA missions to ancient civilizations; from plate tectonics and 

climate change, to Wisconsin’s changing economy. This sort of coverage draws young readers 

into academic subjects in new and interesting ways. With print circulation at 23,000 and the 

addition of a new online version in 2009, the Simpson Street Free Press reaches thousands of 

young people with powerful messages of academic achievement and healthy life choices. The 

students of the Simpson Street Free Press are role models in the truest and most accurate 

meaning of that oft-used term. 

Our organization has two missions: 

Mission #1: Provide a challenging academic experience for our teen writing staff. 

Simpson Street Free Press students pursue their craft in a challenging and authentic newsroom 

atmosphere. Intensive academic lesson plans are the backbone of all Simpson Street Free Press 

programs. These lesson plans are developed and executed by an experienced, professional 

teaching staff. We teach young writers to pull out main ideas, write good lead sentences, and 

organize their writing effectively. Our core strategy is to teach across the curriculum. Students 

learn science, geography, history, and current events while practicing the basics: writing, 

reading, researching, critical thinking, and using computers. Simpson Street Free Press students 

acquire practical academic skills, and quickly learn to apply these skills in the classroom. Our 

writers are required to revise each article many times prior to publication. The assignments are 

challenging and the work demanding, but the rewards are tangible and practical. Tomorrow’s 

community leaders are training today at the Simpson Street Free Press. 

Mission #2: Spread a positive message of youth achievement, academic success, and 

community service throughout Madison and the surrounding area. 

The student writers of the Simpson Street Free Press are well known and very influential among 

their peers. They are role models in the truest sense of the term. There’s no achievement gap in 

the Free Press newsroom and our students send that message, in very clear terms, to thousands 

of local kids. Our reporters write clearly and poignantly about achievement and success. It is 

symbolically important that this influential publication is written and produced in south Madison 

– largely by south Madison kids. The messages of the Simpson Street Free Press are clear: 

drugs, alcohol, and smoking are bad; core academics and community service are cool. 

Achievement can be, and is, for all kids. These messages resonate with our young readers. Free 

Press writers are effective role models because they are real and because they are local. They 

seem “just like us” to kids who read the paper. In our southside newsroom, tomorrow’s 

community leaders are spreading positive and timely messages today. 

2) Program/project description 

Simpson Street Free Press Program Description 

Solid coverage of core academic subjects is the trademark of the Simpson Street Free Press. The 

35-40 students who work at the Simpson Street Free Press receive a valuable academic 

experience that is unduplicated anywhere. Subjects like science, history, and geography come to 

life on the pages of our award-winning newspaper. Our programs are designed to complement 

school curriculum and support student performance. Recent research provides evidence that 

achievement gaps and writing proficiency are related. And writing has taken on new importance 

since 2005 when a writing section was added to SAT college entrance tests. A 2007 report by 

the National Center for Educational Statistics cites modest national gains in writing proficiency. 

2

But the study also reported a consistent achievement gap: a 20-point gap between students of 

color and their white counterparts. The report also recognized significant gains in school 

districts that increased their reliance on writing across the curriculum strategies. 

This is exactly what we do at the Simpson Street Free Press. All Free Press lesson plans are 

constructed around our core, writing across the curriculum philosophy. Through writing-based 

lesson plans our students explore core subject areas, polish practical skills, and engage in 

community service. And, perhaps most importantly, they engage their community. Through 

writing our students learn to think critically, act decisively, work efficiently and work as a team. 

Through writing our students gain critical academic self-confidence. We conduct trimester 

student evaluations and track report cards to monitor each student’s progress. Our success rate is 

consistently high. These nationally recognized best practices effectively bridge the achievement 

gap and as a result, 90% of our students increase their core GPA within two semesters and 92% 

of our program graduates have gone on to college. Quite simply, our lesson plans focus on basic 

academic skills: writing, reading, using technology, and conducting core curriculum research. 

3) A Focus on Core Science Curriculum 

The work our students produce is widely read and very popular. Our student reporters produce 

thoughtful, well-researched articles on topics ranging from ancient civilizations to the latest 

NASA missions; from plate tectonics and climate change, to Wisconsin’s changing economy. 

This sort of coverage draws young readers into academic subjects in new and interesting ways. 

Simpson Street Free Press student writers explore and discover the fascinating world of science 

through a series of writing-based lesson plans. Working with our professional teaching staff, our 

students (and their readers) learn to think critically about a range of science-related topics. Using 

current research and local professionals as resources, these young science reporters explore our 

environment, our planet, our solar system, and the fascinating field of science. 

In turn, young readers of the Simpson Street Free Press embark on explorations of their own. 

Simpson Street Free Press science coverage is designed to complement the Madison 

Metropolitan School District’s school curriculum. Science-focused curriculum guides, produced 

by our teaching staff, complement each publication of the Simpson Street Free Press and 

facilitate its use as a classroom teaching tool. Our students make regular community and media 

appearances, reading samples of their science writing work. Their work is also regularly 

published in The Capital Times, Wisconsin State Journal, and other area newspapers. 

4) Expanding our Message and Vision 

With print circulation at 23,000, the Simpson Street Free Press now prints more pages, more 

often, and reaches more young readers than ever before with powerful messages of academic 

achievement and healthy life choices. In an effort to include more students, the Simpson Street 

Free Press is expanding. We are working to expand our current newsroom facility as well as 

launch an entirely new online publication. 

New online publications will allow us to include more students in our programs and reach 

thousands of additional young readers with stories that engage and educate. Expanding our 

menu of programs and lesson plans is an investment in proven strategies. Planning is already 

3

underway to implement these new “by kids, for kids” online publications. They will operate in 

concert with our print versions. Using online technology will enhance our award-winning 

approach to community-based academics and help us expand our lesson plans. Young readers 

across the state will be able to freely access topics that interest them by browsing the Free Press 

archives. 

These expansion efforts will allow more students than ever to polish practical academic and reallife 

work skills. A newsroom is an excellent place to gain practical experience. The Free Press 

newsroom is a dynamic cauldron of learning. Because our reporters focus their research and 

writing efforts in the core subject areas of science, geography and history, they acquire the skill 

sets that really matter. 

The Simpson Street Free Press would like to thank Wisconsin Space Grant for its 

ongoing support of our program. 

4

Educators’ Aerospace 

Workshop for the 21 st Century 

Jack W. Blake 

Brookwood Middle School 

Genoa City, Wisconsin 

Program Focus. The main goal of this aerospace workshop was to expand the abilities of the 

participants in the use of aerospace materials for their classroom instruction. The workshop also 

focused on the resources available to teachers to give them the expertise to develop and 

implement aerospace activities and materials into the curriculum. Regarding NASA directorates, 

the Space Operations Mission Directorate was applicable to this specific program. Additionally, 

allowing for a range of grade levels, 14 science standards were addressed as part of this program. 

Participants entered the workshop with a variety of background knowledge. They were 

challenged further with some additional concepts at an in-depth level. 

Teachers participating in this program were from the kindergarten through high school level and 

one individual from the college level. There were five teachers from the K through 4th grade 

level, fifteen teachers from the 5 th through 8 th grade level, four teachers from the high school 

level and one from the college level for a total group of 25 participants. Demographically, the 

group consisted of 20% male, not underrepresented, 4% male, underrepresented, 64% female, 

not underrepresented, and 12%, female, underrepresented. 

We met as a group for two Saturdays, once in December, 2008 and once in February, 2009. 

Our meeting place was the 8 th grade science lab at Brookwood Middle School in Genoa City, 

Wisconsin. Our sessions ran for five hours (or more) during each Saturday meeting. Participants 

were given the option of attending the first session in December, the second session in February, 

or both. Participants were immersed in numerous space related activities, predictions, and some 

calculations. 

Both Saturday sessions introduced several concepts regarding flight and aerospace. Introduced 

and studied by demonstration and/or activity, participants were involved in a variety of concepts. 

For the most part, participants worked individually. However, teams of two or cooperative 

learning groups of three or more were formed, as needed. Teachers found the hands-on activities 

supported the concepts in an enjoyable and enlightening manner. Teachers were given the 

opportunity to share with others in a give and take format. This proved to be very valuable as 

teachers learned from colleagues around them in a quickly formed network. Using an ‘adapt and 

adopt’ mind set, teachers shared how something could work in their classroom at their particular 

grade level. Hopefully, a stronger interest in space science as a whole was generated or enhanced 

by way of the workshop. Teachers returned to their respective classrooms with a variety of 

methods and lessons to present aerospace ideas, pique interest, and enhance learning for students 

of all levels. 

Note: Acknowledgement and thanks to WSGC and Brookwood Middle School for grant award 

and percent of matching funds. 

5

Activities and labs were provided by many sources. Specific activities and providers: Hot Wheels 

–It’s A Drag activity, Mattel [and J. Blake/self-developed]; Outer Space Chemistry, Fire Arrows 

Kit, SK/ Boreal Labs; Living on the Shuttle, Lunar Survival Test, NASA Inquiry Stations, Toys in 

Space, NASA; Fuel Cell Challenge, [J. Blake/self-developed]; C.O.G.bottle activity, Arizona 

Aerospach Outreach; Microgravity Demo/Bottles,[J. Blake/self-developed; Space Station Magna, 

AIMS [and J. Blake/self-developed]; Mars Mission/miscellaneous materials, Evan Moore Corp. 

and Aerospace Education Services Project-Oklahoma State University; Shuttle & Orbiter outline 

sketches & associated drawings/miscellaneous activities, Wisconsin Aviation Careers – 

Wisconsin DOT, Bureau of Aeronautics. Science lab materials housed in the science lab at 

Brookwood Middle School, consumables like paper, glue, string, etc. as well as small hand tools, 

and so forth were made available, as needed. 

Evaluation. At the end of each session, participants were asked to respond to an evaluation 

form. It consisted of seven parts. The first five statements required a ranking of the individual 

statement, one to five. A rank of five indicated a strongly agree (SA) response while a rank of 

one indicated a strongly disagree (SD) response. The sixth statement listed several activities 

using nine capital letters, A through H. Participants were asked to circle the letters of the 

activities they enjoyed the most with the understanding that they could circle all letters, if 

desired. The seventh statement allowed the teachers a chance to write any additional comments 

about the workshop. The results of this evaluation, including comments from participants can 

be found in pages that follow. 

An instructor’s journal was kept regarding each Saturday session. Reflecting on what went 

well, what needed more in-depth explanation or coverage and what could have been done 

better was the purpose of the journal. Further study of these notes, immediately following each 

session, served to help improve any future workshops presented related to aerospace and flight. 

References – cited in last paragraph, Program Focus 

AIMS & self [Jack Blake] Space Station Magna 

Arizona Aerospace Outreach C.O.G. 

Evan Moore Corp. & Aerospace Education Services-OSU Mars mission/miscellaneous activities 

Mattel Toys & self [Jack Blake] HotWheels-It’s A Drag Activity 

NASA, Living on the Shuttle, Lunar Survival Test, NASA Inquiry Stations, Toys in Space 

Science Kit and Boreal Labs Outer Space Chemistry, Fire Arrows 

Self [Jack Blake] Fuel Cell, Microgravity Demo/Bottle 

Wisconsin DOT, Bureau of Aeronautics Shuttle & Orbiter sketch/miscellaneous activities 

6

2008-2009 Evaluation 

Educators’ Aerospace Workshop for the 21 st Century 

Please circle the correct number for the statements below by using the following scale: 

5 = Strongly Agree 4=Agree 3=Uncertain 2=Disagree 1=Strongly Disagree 

The Workshop: SA A U D SD 

1) Increased my knowledge about aviation, 5 4 3 2 1 

aerospace concepts and principles. 

2) Increased my interest in this subject. 5 4 3 2 1 

3) Increased my confidence in doing science 5 4 3 2 1 

activities. 

4) Provided me the opportunity to interact and 5 4 3 2 1 

work in cooperative groups. 

5) Provided many hands-on aerospace activities. 5 4 3 2 1 

6) Which of the following activities did you enjoy the most? 

{Please circle as many letters as you would like} 

A. Microgravity Demo/Bottles 

B. Toys in Space 

C. Lunar Survival Test 

D. NASA Inquiry Stations 

E. Fire Arrows/Rockets 

F. Fuel Cell challenge 

G. HotWheels-It’s A Drag 

H. Outer Space Chemistry 

I. C.O.G. 

7) On the back of this form, please write any additional comments about things you liked 

or disliked about the workshop. 

7

2008-2009 Evaluation 

Educators Aerospace Workshop for the 21 st Century 

Please circle the correct number for the statements below by using the following scale: 

5 = Strongly Agree 4=Agree 3=Uncertain 2=Disagree 1=Strongly Disagree 

The Workshop: SA A U D SD 

1) Increased my knowledge about aviation, 4.8 avg 5 4 3 2 1 

aerospace concepts and principles. 

2) Increased my interest in this subject. 4.9 avg 5 4 3 2 1 

3) Increased my confidence in doing science 4.8 avg 5 4 3 2 1 

activities. 

4) Provided me the opportunity to interact and 5.0 avg 5 4 3 2 1 

work in cooperative groups. 

5) Provided many hands-on aerospace activities. 4.9 avg 5 4 3 2 1 

6) Which of the following activities did you enjoy the most? 

Please circle as many letters as you would like. 

(Number in parentheses is number of times letter was circled) 

A. Microgravity Demo/Bottle (21) 

B. Toys in Space (23) 

C. Lunar Survival Test (20) 

D. NASA Inquiry Stations (19) 

E. Fire Arrows/Rockets (24) 

F. Fuel Cell challenge (18) 

G. HotWheels-It’s A Drag (22) 

H. Outer Space Chemistry (18) 

I. C.O.G. (23) 

7) On the back of this form, please write any additional comments about things you liked or 

disliked about the workshop. 

8

Some comments from participants 2008-2009 

Educators’ Aerospace Workshop for the 21 st Century 

“Very well organized. Great activities-both big ones & small ones!”. 

“Good activities for having students use critical thinking skills”. 

“I really enjoyed the fact that we got to build several of the items presented. Taking home these 

materials was something that I appreciate as making them by myself would/could be frustrating. 

Thank you”. 

“Great workshop. Found some quick hands-on activities that will allow students to understand 

how many things interact with their own daily lives”. 

“Loved learning things that are readily applicable to my classroom and easy”. 

“Very well done, easy to understand, great hands-on activities. Thank you”. 

“It was great to get the opportunity to share with others, keeping it calm & casual. Nice mix of 

activities”. 

“Very nice workshop-could you do it longer? ” 

“Great stuff-appropriate for elementary, MS and HS”. 

“I liked the activities which made it fun for a Saturday . Keep it up!”. 

“It was great to interact with other science teachers who have been in education for a long time. 

The activities seem easy enough to tweak to the grades I teach and make them relevant & fun!”. 

“Very good- these are activities I can (and will) use”. 

“Thank you so much for the opportunity to participate in this workshop. Everything was great”. 

9

Abstract 

Background 

Teaching the Teachers 

An Outreach Program of Crossroads at Big Creek 

Coggin Heeringa 

Crossroads at Big Creek Environmental Learning Preserve 

Sturgeon Bay, Wisconsin 

In the project, Teaching the Teachers, Crossroads at Big Creek offered stipends 

to teachers who participated in classes and workshops pertaining to spacerelated 

science. 

The project was based on the belief that when elementary school teachers 

develop confidence and content knowledge of astronomy by participating in 

experience-based activities. This increases the quantity and quality of spacerelated 

instruction at the elementary level. 

The target audience included elementary teachers who had little or no 

experience or prior interest in astronomy because these people are in the position 

to either inspire or inhibit learning of young people. 

Even astronomers have difficulty comprehending and explaining the wonders of 

the universe. If professionals find the abstract concepts of astronomy mindboggling, 

elementary school teachers are understandably daunted at the task of 

introducing this material to their young students. Many teachers, lacking content 

knowledge, dread their astronomy units. Others grasp the continuously evolving 

concepts, but have difficulty developing age-appropriate lessons. 

As a part of a previous WSGC grant, Crossroads at Big Creek, collaborating 

with the Door Peninsula Astronomical Society, developed a one graduate credit 

class which now has been offered five times through the Office of Outreach and 

Extension, University of Wisconsin-Green Bay. Those who have taken the class 

have been enthusiastic and have infused the space-related material into the 

curriculum. 

Alas, in offering these classes, we were “preaching to the choir.” The teachers 

who enrolled in our astronomy classes were the individuals who already 

exhibited a strong interest in science. 

This project was made possible with a grant from the Wisconsin Space Grant Consortium. 

11

The cost of a one credit graduate class from the University of Wisconsin system 

in 2008 was $342.30. Most teachers found this cost excessive and 

consequently, they met their certification requirements by taking just about any 

class that offers a price break…..often classes having nothing to do with their 

personal interests. We also knew that cash-strapped teachers will attend 

workshops on almost any topic if stipends are awarded. 

In conversations with teachers, we discovered that a significant number of 

teachers were intimidated by the idea of taking a graduate level class in 

astronomy. Furthermore, an increasing number of teachers now are meeting 

certification requirements by creating portfolios rather than completing graduate 

level classes. 

Procedures 

To meet the needs of our target group….teachers with little or no experience or 

prior interest in space-related science…. we offered two after-school 

workshops. A teacher could attend one or both workshop sessions, following 

which they were given the option of arranging a space-related visit to 

Crossroads or a classroom visit from either the naturalist or a member of the 

Door Peninsula Astronomical Society. Each teacher who participated and 

returned an evaluation received a $50 stipend and a free supper for each 

workshop and $200 scholarship enabling them to take the class. 

Flyers headed “Crossroads Is Giving Money to Teachers” and “Planetarium, 

Planning, and Pizza” were given to all elementary teachers (public and private 

schools) of four school districts, announcing that the stipend and a free pizza 

supper would be offered to teachers who participated. Pre-registration was 

required. 

During the workshops, we emphasized hands-on, experience based activities, 

but the workshops focused on content learning, with the objective of helping 

teachers answer the questions of their students with accuracy and confidence. 

Another objective was to introduce teachers to our newly acquired StarLab 

Planetarium. 

The graduate level one credit hour class offered through the Education Outreach 

program of the University of Wisconsin-Green Bay, “Teaching Astronomy in 

the Elementary School” drew from workshop participants. This class delved 

more deeply into presenting concepts through hands-on activities. 

12

Evaluation 

By trading evaluation forms for the stipend checks, we had an excellent return. 

The comments were very positive. However, the real test of our project was the 

increase in field trip requests to Crossroads Environmental Center and the 

improved level of preparation of the elementary students who visited our 

astronomy facilities. 

Members of the Door Peninsula Astronomical Society, who served as presenters 

for planetarium shows, knew (almost immediately) if a teacher had attended our 

workshops because of the eagerness and background knowledge exhibited by 

their students. 

Further Work 

We plan to continue this program during the 2009/10 school year and to 

continue to offer stipends to teachers to participate in workshops, classes and to 

attend lectures. We are convinced that the stipends encourage teachers that 

otherwise would not attend our programs and we are absolutely positive that 

teachers who attend our workshops are better prepared to present astronomical 

concepts to their students. 

13

EAA WOMEN SOAR – YOU SOAR 

Dr. Lee J. Siudzinski 

Experimental Aircraft Association (EAA) 

AEROSPACE OUTREACH PROGRAM 2008 – 2009 

SYNOPIS: The Experiment Aircraft Association, Inc. (EAA) used their annual Fly-in, 

(AirVenture) attended by over 700,000 aviation enthusiasts, to help spark women’s interest in 

STEM content through aviation. The EAA Women Soar – You Soar Conference was held July 

26 – 28, 2009 for high school age girls in grades 9 – 12. The event was attended by over 100 

girls and held both at the EAA grounds, Whitman Airport, and the University of Wisconsin, 

Oshkosh. 

PROGRAM DETAILS: During the EAA AirVenture the girls participated in various hands-on 

workshops, air show viewing with MIT and Embry Riddle performers, and mentor sessions. The 

activities included building and designing their own aircraft, investigating the first supersonic 

flight, understanding basic flight controls, weather and flight planning, aviation science and 

space careers in the 22nd century, the challenge of aerospace engineering and maintenance, 

building a glider to better understand control surfaces, and presentations by the Milwaukee 

School of Engineering. 

The girls also had an opportunity to walk around and discuss aircraft of women air show 

performers, fly in a Ford Tri-motor aircraft, build a wing rib, experience the ropes low challenge 

course, complete flight simulation exercises, and participate in KidVenture. (KidVenture is a 

series of hands-on aviation learning activities attended by over 10,000 young people during the 

week long fly-in) The program concluded with a final program and presentation of various 

scholarship awards and recognitions. 

EVALUATION: Follow-up evaluation with participants, mentors, presenters and parents who 

accompanied the young women, resulted in positive comments and appreciation for making this 

experience possible. 

A special thanks and acknowledgement to the Wisconsin Space Grant Consortium for their 

generous financial support that helped make this annual event possible. 

15

Synopsis: 

EAA Space Week 2008 

Lee Siudzinski 

Chrissy Paape 

Experimental Aircraft Association (EAA) 1 


Aerospace Outreach Program 2008-2009 

The Experimental Aircraft Association, Inc. (EAA) enhanced its successful Space Discovery 

Week, a school- and community-focused experiential program, through an exciting new 

partnership with Space Explorers, Inc. Space Discovery Week had traditionally included handson, 

aviation-related programming designed to encourage school participation and to offer key 

new educational challenges. The collaboration with Space Explorers, Inc., an organization 

devoted to bringing the excitement and challenges of space exploration into classrooms through 

innovative educational curricula, enriched the Space Discovery Week experience with the 

addition of the Mars Explorer Simulation program. 

Program Details: 

EAA integrated-education programs provided a continuum of learning for all ages. These 

programs, including the newly added Mars Explorer Simulation, provided academic 

reinforcement to students and offered educators resources that combined standards-based 

curricula with hands-on techniques and challenges that amplified learning in the classroom and 

at home. 

Through the addition of the Mars Explorer Simulation to the Space Discovery Week experience, 

EAA enabled the following program goals: 

� Offered educators an additional opportunity to participate in programming to augment 

classroom learning via the Internet. 

� Expanded program impact through the Mars Explorer Simulation in order to provide 

students and educators the opportunity to participate in a year-long, hands-on learning 

experience. 

� Provided educators with the resources and support needed to integrate the EAA 

experience with classroom learning objectives. Relative to the Mars Explorer Simulation, 

educators received an educator guide; standards-based lesson plans with activities such as 

scale-modeling; a study guide and answer key; an exclusive video of a planetary 

geologist working on the MER mission; ongoing mission information; and completion 

certificates. 

1 The authors would like to extend special thanks and acknowledgement to the Wisconsin Space Grant Consortium 

(Aerospace Outreach Program) for their generous financial support. 

17

� Incorporated the technology resources of NASA, EAA and Space Explorers, Inc., to 

maximize the renewed level of interest among youth for space travel and exploration. 

� Provided a multi-faceted experience in which youth participated in events and learning 

experiences with their classmates at an off-campus school field trip, and where both 

students and educators experienced the internet-based Mars Explorer Simulation during 

Space Discovery Week, as well as at home and in the classroom. 

Students and teachers took part in a variety of workshops as part of EAA’s Space Week. Space 

Explorers involvement included leading activities centered on comparing similarities and 

differences between Earth, Mars, and our Moon. At the conclusion of the session, teachers and 

students received membership cards to continue learning through the Mars Explorer Simulation. 

This program engaged students in problem-solving activities in that participants encountered the 

same issues NASA experiences in its exploration of Mars, e.g., dust storms, solar energy lapses 

and navigating rocky terrain. Students were exposed to many fundamentals of space science 

including Newton’s Laws, Kepler’s Laws, gravity and orbits, and they examined the possibility 

of sustaining life on Mars by completing tasks such as taking panoramic spectral images, 

analyzing the magnetic field, and taking measurements of rocks and soil samples using a variety 

of scientific instruments. 

The project additionally met state and national science content standards for space/science 

content, hands-on science activities, demonstration of space/science principles, and application 

of space/science principles. EAA has demonstrated success in the development of activities and 

programs that are aligned with state and national standards. Pre- and post-learning materials, 

along with planned teacher’s guides helped educators plan for and maximize the experience. 

Results: 

EAA Space Discovery Week served approximately 1,700 school youth, educators and 

chaperones from local and regional schools, as well as a public audience of 1,000 adults and 

youth together for the featured weekend activities. Space Week was be held at the EAA 

AirVenture Museum in Oshkosh, Wisconsin. 

Evaluation: 

Follow-up evaluation with parents and teachers, who accompanied the students, resulted in 

positive comments and appreciation for bringing space to their children and students. Follow 

through by teachers with their students involving Space Explorers was excellent. EAA has 

experienced many return visits by students and family members. 

18

Space Travel Simplified - Part 1 

Bradley J. Staats 

Spaceflight Fundamentals, LLC 

New London, Wisconsin 

Synopsis: Space Travel Simplified - Part I was a teacher workshop1 that focused on the 

history, math, science, and technology of spaceflight. This workshop offered a unique 

approach to teaching by incorporating "real world" applications into the classroom. 

Goals and Project Value: How does an orbit work? How did our astronauts get to the Moon? 

How does a spacecraft safely reenter the atmosphere? How does a spacecraft rendezvous with 

the International Space Station? What benefits has the world gained from space exploration? In 

this day and age of modem spaceflight, these are just a few of the fundamental questions that 

students may ask an instructor regarding space. Can teachers comfortably field these and other 

questions without dispensing any misconceptions? Space Travel Simplified - Part 1 was a oneday 

workshop that accurately answered these and many more questions while focusing on the 

history, math, science, and technology of spaceflight. The course offered a unique approach to 

teaching and answering the age old question "Why do I have to learn this?" by elegantly 

incorporating "real world" applications into the classroom. Instructors experienced a unique 

approach to teaching math, science, and technology standards by tackling real world issues in 

inspiring classroom experiences. Space Travel Simplified - Part 1 utilized award-winning 

approaches to advance an educator's knowledge base by employing a fun, hands-on approach to 

learning. 

For this inaugural workshop, the Milwaukee area schools (Cooperative Educational Service 

Agency # 1 district) were the target audience. The workshop was setup to provide 20 instructors 

with a full day of instruction and over $50 per person in materials and resources to take back to 

their respective classrooms. A lunch was also provided to all participants. After advertising 

extensively for this 3-12 workshop, a total of 22 instructors pre-registered for it. 110% 

enrollment in the workshop! 32% were elementary instructors, 32% were middle level 

instructors, while 36% were high school instructors. 

Due to the broad range of teaching levels present for this workshop, the workshop was broken 

into three basic segment/components - elementary, middle, and high school level topics. The 

workshop was co-facilitated. The AM sessions were devoted to elementary discussions while 

the PM session was devoted to high school topics. During these sessions the following 

topics/activities were covered: perception / paradigm shift; center of gravity; Bernoulli 

implications; Newton's universal law of gravitation; various orbital shapes (conic sections); 

circular orbits; geosynchrous orbits; elliptical orbits; spaceflight mathematics; history of human 

space exploration; and technologies of space exploration. 

1 The main financial support for this workshop was provided by the Wisconsin Space Grant Consortium. Additional 

support was provided by the following sponsors: Spaceflight Fundamentals, LLC; Science Kit & Boreal Labs; and 

the University School ofMilwaukee. 

19

Evaluation Results: At the conclusion of this workshop, the following questions were asked on 

an evaluation fonn. The results, of this evaluation, are based on a 100-point scoring system with 

100% strongly agreeing with the provided statement, 80% = agreeing with the statement, and 

so on. 

I-My exposure to this project has increased my knowledge/understanding in space, aerospace, 

space-related science, design, and technology. Score = 94%. 

2-Student exposure to this project should increase an interest in space, aerospace, space-related 

science, design, and technology. Score = 93% .. 

3-This project has the potential to increase secondary (pre-college) student recruitment, 

experience and training in the pursuit of a space or aerospace related science, design, or 

technology profession. Score = 79%. 

4-The project has self-sustaining/replicable qualities due to the fact that the participants are 

trained and supplied with the basic materials to go out and duplicate in their classrooms the work 

that was incorporated in this workshop. Score = 93%. 

5-The project meets the goal of Teacher Training which is defined as successfully educating, 

training, and exciting teachers about the math, science, technology, and history pertaining to 

spaceflight. Score 93%. 

6-This project has the ability to expose and prepare the next generation of scientists to aerospace 

related fields. Score = 83%. 

7-1 have a better understanding of how our spacecraft and flight support work together. 

Score 83%. . 

8-The instructors were knowledgeable about the subject matter that was being taught. 

Score 100%. 

9-The workshop was well organized. Score = 96%. 

10-The instructors' presentation style was well suited for the audience in attendance. 

Score =91%. 

II-As a workshop participant, I feel significant professional growth by having attended this 

workshop. Score = 88%. 

12-The pacing ofthe workshop was well suited for the audience in attendance. Score 83%. 

13-1 am pleased with the infonnation and materials that I received as part of this workshop. 

Score 99%. 

14-What is the total number of students that you teach per day? 

*Based on the number of instructors present and their teaching assignments, a total of 1302 

students will be positively impacted by this workshop. 

15-How do you plan to implement this material into your classroom curriculum? 

*The following are highlights of responses to this question: I will use these activities to help 

them to understand gravity and various spacecrafts. I plan to work with other instructors and use 

materials for demos and inspiration. I will use these activities in my lessons about gravity and 

the solar system. 4th grade does a unit on space, also some of the science/math concepts can be 

placed easily with other units. Grade 8 science / grade 8 math. I will use the lab activities, 

materials, etc. The hands-on materials will work well in class. I plan to use the materials to 

demonstrate the different stages and I will definitely use the center of gravity activities with my 

students. Earth science astronomy class. Definitely will use in 3 rd grade space & solar system 

curriculum. Will incorporate with large groups in grade 2 through standards applying to science 

and technology. Demos/projects with freshman physical science/space units. 

20

16-Please express any additional comments regarding the workshop and/or instructors. 

*The following are highlights of responses to this request: Most beneficial workshop I've been to 

in a decade! Great workshop. They were an excellent team - one complimented the other well. 

Instructors did a super job. The workshop was very worthwhile - I learned lots and have some 

great activities to incorporate into my science lessons. Thank you for the opportunity to be part 

of this grant. Very well done. Thank you! Excellent. Pleased with the overall workshop. Had 

lots of fun. Great materials. Great, knowledgeable instructors. Was great - loved the activities, 

thanks for all of the free sturn! 

Evaluation Analysis: Based upon the positive evaluations and comments of these grades 3-12 

teachers, there should be a definite increase for interest in space, aerospace, space-related 

science, design, technology, and its potential benefits for their students in the Milwaukee area. 

Invariably, based on our evaluations, this project should allow secondary (pre-college) students 

the opportunity to increase their interest, recruitment, experience and training in the pursuit of 

space or aerospace related science, design, or technology in the Milwaukee area. 

The project has self-sustaining/replicable qualities due to the fact that the instructors that were 

trained were supplied the basic materials to go out and duplicate the work that we incorporated in 

our workshop. The goal is for teachers to go back to their classroom and replicate this work to 

their students - The "multiplier effect" is then engaged. Through this effect, each teacher is able 

to provide their students with exposure to this exciting curricular approach. For the 2009-2010 

school year, 1302 students will have the opportunity to be exposed to this worthwhile 

curriculum. Based upon the amount of grant money received from the Wisconsin Space Grant 

Consortium (WSGC) and the number of students each registered instructor has, it only cost 

WSGC an average of $1.77 per student to run this workshop - This is an amazing investment! 

The goal for this pilot program is to have it offered in various regions around the state. Based 

upon the workshop's evaluations, the project certainly met this year's specific goal of Teacher 

Training. The whole purpose of the project is to educate, train, and excite teachers about the 

math, science, technology, and history pertaining to spaceflight - This workshop definitely and 

successfully accomplished this feat. 

Alignment With NASA Directorate, Center Goals, and/or Educational Standards Science 

Mission Directorate: The scientific investigation of the Earth, Moon, Mars and beyond with 

emphasis on understanding the history of the solar system, searching for evidence of habitats for 

life on Mars, and preparing for future human exploration. This workshop definitely exposes and 

prepares the next generation of scientists to aerospace related fields. As we prepare for future 

human exploration, we will need many new scientists and engineers to accomplish this endeavor. 

Our project's goal was to train teachers who in turn can train these future scientists and 

engineers. Our proposal, therefore, provided assistance in this area. 

Space Operations Mission Directorate: Through the three themes of the International Space 

Station, Space Shuttle Program and Flight Support, provide critical enabling capabilities that 

make possible the science, research and exploration achievements of the rest of NASA. This 

provided as part of its focus the ISS, Space Shuttle program and required Flight Support. The 

focus allowed the participants to better understand how they work together and provided 

enabling capabilities as a whole. 

21

Educational Standards: The National Research Council's (NRC) Science Education Standards 

were addressed throughout the workshop. Special emphasis was put on the following Standards 

areas: The Teaching Standards: Guiding and Facilitating Learning & Building Learning 

Communities; The Professional Development Standards: Learning Science Content, Learning 

To Teach Science, & Learning To Learn; and The Content Standards: Scientific Inquiry, 

Technological Design, & Science and Technology. 

Participants: The workshop was limited to 20 science and/or math instructors. It was made 

available on a first come, first serve basis. Once the 20 slots were filled, two additional 

registrations were also accepted. Spaceflight Fundamentals, LLC fully complies with the 

Americans with Disabilities Act of 1990 (ADA), Section 504 of the Rehabilitation Act of 1973, 

and its amendments, all of which prohibit discrimination on the basis of disability in the 

admission, access to, or participation in programs or activities. 

Location of Project: The workshop was advertised to science/math (Grades 3-12) teachers in 

the CESA #1 district (Milwaukee area. The workshop was located at the University School of 

Milwaukee, 2100 West Fairy Chasm Road, Milwaukee, WI 53217. Phone: (414) 352-6000. We 

coordinated the workshop advertisement with school districts in the Milwaukee / CESA #1 area. 

The target audience was science / math classroom instructors (Grades 3 -12). Based upon future 

funding, follow-up (Part 2) workshops and additional (Part 1) workshops could be set around the 

state. 

Work Plan: Our work plan involved an eight-hour workshop. In those eight hours, we focused 

on the concept of spaceflight via a variety of hands-on activities (labs, simulations, 

computations, etc.) and discussions. High emphasis was placed on cooperative work and 

constructivistic approaches being fueled through facilitator lead Socratic dialogue. The goal was 

to allow the instructors to have the chance to infuse their new knowledge of Part 1 into their 

respective curriculums with the hopes that a follow-up workshop can be funded in order to 

further our focus. 

General Information: Spaceflight Fundamentals, LLC is a small but dedicated company to the 

advancement of aerospace in the classroom. For the past eight years, our company has been 

authoring and publishing educational materials on aerospace education in the state of Wisconsin. 

Also, in those eight years, we have had the opportunity to organize and instruct several teacher 

graduate course workshops. The workshops have always been well received and have made 

definite positive impacts in our state's classrooms. With that said, our hope is that workshops 

like these will allow further opportunities to educate and motivate the current and next 

generation of instructors/students on aerospace education in the state of Wisconsin. We look 

forward to creating future proposals / activities for teacher aerospace workshops and further 

broadening our ability to work with other state organizations with the same goals. We feel that 

the state of Wisconsin has benefited from our activities, and that we hope to continue being a 

positive force in the WSGC's community outreach efforts while helping to nurture and grow the 

aerospace industry in our state. 

22

Background 

Expanding GIS Across the Curriculum 

Jennifer Johanson 

Physical Science Department; Alverno College 1 


Alverno College is a women’s college located in Milwaukee. About one-third of our students 

are minorities, and a high percentage of our students are the first generation in their families to 

attend c ollege. O ur s tudent m akeup t argets m any of t he unde rrepresented g roups t hat N ASA 

hopes t o r each. A lverno’s a bility-based c urriculum f osters de velopment of e ight s pecific 

abilities; pr oblem s olving, a nalysis, c ommunication, de veloping a global perspective, effective 

citizenship, s ocial interaction, va luing i n de cision m aking, a nd a esthetic e ngagement. T hese 

eight abilities a re explicitly ta ught in d ifferent c ourses across th e c urriculum a nd v alidated 

through assessment at four developmental levels. 

Geographic Information Systems ( GIS) i s pa rt of a group of related geospatial d ata co llection 

and management technologies, including remote sensing and global positioning systems (GPS), 

which routinely use both Earth and satellite-based data. GIS technology allows a user to look at 

data in a d ifferent way, placing spatial and/or temporal data on a map. Visually mapping data 

makes it easier to pick out spatial and/or temporal patterns. As such, GIS is a natural tool to help 

teach logical thinking, allowing students to see spatial patterns rather than looking only at graphs 

and tables. Geospatial technology has been widely used in urban planning, environmental and 

earth s ystem s ciences, a nd m ore r ecently i n o ther f ields. H ealthcare, s ocial s cience, b usiness, 

and education have begun to routinely use GIS, as well as emergency planning and emergency 

response. G IS i s us ed i n t racking t he s pread o f di sease, l ocating crime patterns, f inding s afe 

travel routes in natural disasters, etc. 

Alverno College f aculty w anted t o i ntegrate t he us e of G IS t echnology into bot h s cience an d 

non-science co urses i n p art t o i ncorporate a n ap preciation o f t he impact o f aer ospace 

technologies in widely varied majors in many different disciplines. GIS technology is a natural 

tool for teaching many of the abilities, as well. F or example, looking for trends and patterns is 

part of t he a nalysis a bility, and e ffectively pr esenting i nformation on a m ap i s pa rt of t he 

communication a bility. However, w ithout f aculty t raining or p rofessional de velopment 

opportunities in G IS, f aculty w ould not b e ab le t o p rovide ef fective l earning ex periences f or 

students. 

Wisconsin Space G rant C onsortium ( WSGC) provided A lverno C ollege a n in itial two-year 

Higher E ducation Initiatives gr ant f or a pi lot p roject t o i ntroduce GIS technology into our 

curriculum in 2005, which allowed us to establish a new course in GIS, promote the GIS course 

as an elective or required course in selected majors, and use the course to train faculty and staff 

in the use of GIS and its role in different disciplines. In 2007, WSGC awarded Alverno a second 

two-year Higher E ducation Initiatives gr ant, w hich ha s a llowed A lverno t o f und G IS-related 

1 Funding for this project provided by Wisconsin Space Grant Consortium 

23

curriculum development, to provide additional professional development opportunities to faculty 

in various disciplines, and to train additional faculty. The funding from this grant allowed faculty 

at A lverno C ollege t o w ork out side t heir a reas of e xpertise a nd t o l earn a bout, de velop a nd 

integrate some of the many possible ways of using spatial analysis in their courses using GIS. 

Faculty members were able to participate in GIS training and professional development, and to 

identify and design curriculum activities that integrate GIS concepts and applied skills. This has 

fostered multiple experiences for students in diverse majors to learn more about the applications 

of space and geospatial technology using GIS across the institution. 

The W SGC H igher E ducation Initiatives G rant is i ntended t o f und “ undergraduate e ducation 

projects w hich s upport t he m ost i nnovative i deas on how t o i ncrease t he space an d aerospace 

content of unde rgraduate uni versity and c ollege of ferings”, and t o a lign w ith t he g oals a nd 

objectives o f t he N ational S pace G rant C ollege and F ellowship P rogram. This pr oject 

specifically meets the following goals: (1) to enable the development of a diverse workforce of 

future scientists, engineers, technology professionals, and educators; (2) to stimulate and nurture 

innovative programs to assure the development and transfer of practical applications in aerospace 

research and education; (3) to provide access to the excitement, knowledge, and technology from 

America's E arth, Air a nd S pace p rograms; and ( 4) t o educate s tudents at al l l evels b y 

encouraging and s upporting i nterdisciplinary a nd m ulti-disciplinary r esearch ex periences an d 

education programs. In addition, it fits with both the Goddard Center activities to develop and 

maintain a dvanced i nformation s ystems f or t he display, a nalysis, a rchiving a nd di stribution of 

space and Earth science data and to develop National Oceanic and Atmospheric Administration 

(NOAA) satellite systems that provide environmental data for forecasting and research; and the 

Stennis S pace C enter A pplied S ciences P rogram t hat s eeks t o b ridge t he g ap b etween E arth 

science r esearch r esults an d t he u se o f d ata t o help i ts p artner ag encies make b etter i nformed 

decisions. 

Project Activities and Accomplishments 

The f aculty and s taff feel t hat t his pr oject ha s be en ov erwhelmingly successful, and i t ha s 

received support from the administration to continue to expand its impact and reach across the 

campus. T he f ocus of t his pr oject w as t o provide A lverno f aculty and staff w ith a n 

understanding of how G IS t echnology i s us ed i n m ultiple di sciplines, a nd then to pr ovide 

training and professional development for those faculty and support staff interested in learning 

more about GIS. In turn, the faculty and staff would develop curriculum using GIS, to afford 

Alverno s tudents w ith a n unde rstanding of how G IS i s us ed i n t heir di sciplines, a nd pr ovide 

learning opportunities to student using GIS. Essentially, we hoped to make geospatial science 

and technology an integral component in our curriculum content and instruction, but we w ere 

starting from scratch, with no specific geospatial content in any area of the curriculum. 

Alverno ha s s trategically used t he grant f unds a warded b y W SGC t o pr ovide t he m ost 

advantageous oppor tunities to p repare b oth f aculty a nd s tudents. Faculty me mbers in itiated 

curriculum research, planning, and development in collaboration with a program consultant. Our 

developing und erstanding of t he pot ential us es of G IS ha s be en r egularly s hared w ith t he 

Alverno f aculty a nd s taff a t faculty i nstitute s essions, w hich oc cur pr ior t o t he s tart o f e ach 

semester an d at t he end o f s pring s emester, an d ar e a f ormalized t ime f or f aculty t o s hare 

24

important learning concepts. Through these sessions, additional faculty became interested in the 

prospect of including this technology as a tool for their students’ learning, expanding the use of 

the technology in courses be yond social science and environmental science to such courses as 

business and management, adult education, psychology, global studies and healthcare studies. 

GIS steering committee. The steering committee was set up to bring GIS to Alverno, and to 

determine the most effective ways to utilize finds to incorporate GIS into the curriculum. It was 

initially comprised of members of the Social sciences and Environmental sciences in the School 

af A rts and S ciences. It h as s ince ex panded t o include r epresentatives from each o f t he four 

schools a t A lverno; t he School of A rts and S ciences, t he S chool of Business, t he S chool o f 

Nursing, and the School of Education, along with support staff from the library and academic 

computing. Faculty, staff, and administration are all represented on t he committee, as well as 

most of t he a bility de partments. W e ha ve be en hi ghly s uccessful i n obt aining uni versal 

acceptance of the potential for this technology across the curriculum. 

Networking with other institutions. A strong network has been formed between Alverno 

and t he U niversity of W isconsin-Milwaukee ( UWM), w hich o ffers a G IS c ertificate p rogram. 

Faculty at UWM have been part of the Alverno GIS steering committee, and have shared their 

expertise b y p resenting at a G IS s eminar f or t he en tire A lverno f aculty i n S eptember 2 007. 

Alverno has also formed ties with several consulting firms, and their staff have also been part of 

the GIS steering committee. 

Initial GIS training for faculty and staff. The Initially, t he g oals of t he G IS A cross t he 

Curriculum pr oject were to a pply G IS t echnology i n t wo s pecific di sciplines; Environmental 

Science, and Social Science, by restructuring select course content in geography, environmental 

geology, and earth science courses. T o do t his, we formed a steering committee composed of 

key faculty members from Physical/Earth Science, Technology, and Social Science departments 

along w ith out side c onsultants, t o r eview t he c urrent c ourse c ontent a nd i dentify a core of 

courses from each discipline area that can integrate GIS concepts and infuse geospatial science 

and t echnology content. Within e ach o f t hese t wo de partments, i nterested faculty t hen were 

selected to be trained in GIS. With this training, the selected faculty were to develop curricular 

activities for students using GIS, so that students had a chance to use GIS in different settings, to 

work with it as an analytical tool, and to see its applications in multiple disciplines. If faculty in 

other disciplines b ecame interested in using G IS, we hoped to be able to train them at a later 

date, w ith a ne w s ource of f unding. The co sts of pr ofessional de velopment i n G IS a re f airly 

high, s o w e a ssumed t hat onl y a f ew hi ghly m otivated f aculty m embers w ould be a ble t o be 

trained in GIS. 

However, we found that we co uld train our f aculty more cost-effectively b y hiring a G IS 

professional to teach a 1- to 2-credit GIS course at Alverno. Alverno obtained 25 licenses for the 

GIS software (enough for a full computer lab at a time to work on GIS). This software was not 

funded by the grant, but as in-kind matching funds from the College. We selected an instructor 

from a pool of well-qualified applicants. The course was first offered in the spring of 2006, and 

has been offered each spring since then, with the potential for students to repeat doing different 

project activities. This c ourse provides an ov erview of G IS uses, and was built on l aboratorybased 

exercises and assessments designed to introduce students to the major concepts, features, 

25

and applications of GIS technology allowing students to learn to use the practical applications of 

the software. As part of these courses faculty and students worked collaboratively to learn the 

software and apply GIS technology to various case scenarios across multiple disciplines. 

Over the four years of the grant, this course has provided for initial GIS training for thirteen selfselected 

faculty, four support s taff, and 30 s tudents. The faculty trained t hrough t his c ourse 

represent t he f ollowing d isciplines; p hysical s cience, b iological s cience, m athematics, 

psychology, a dult e ducation, c omputer s cience, and nur sing. T he support s taff m embers ar e 

from t he l ibrary an d t he m ath r esource cen ter, and t he e ducational r esource c enter, and a lso 

represent t he Q uantitative Literacy s ubcommittee o f th e C ommunication A bility d epartment. 

These key support staff have been wonderful student resources and ambassadors for GIS, helping 

students w ho ha ve s elected i ndividual r esearch projects a bout G IS o r u sing G IS. O ur i nitial 

focus on f aculty training did not include support staff, but through the development of the GIS 

course, and the support of the WSGC grant, we have been able to provide training for them with 

very little additional cost, and the benefit to students of having these staff resources cannot be 

quantified. 

Conferences and further professional development. As part of this initiative, faculty and 

students have presented at and/or attended professional conferences related to GIS. In addition 

to the presentation made on t he first two year grant, “GIS across the Curriculum” at the 2007 

WSGC annual meeting, two additional presentations at professional conferences have been given 

by Alverno faculty. D r M arilyn R eedy p resented at a r egional co nference o f t he M idwest 

Institute for Students and Teachers of Psychology on using GIS in teaching of psychology, and 

Dr Jeanna Abromeit presented at a national conference of the American Sociological Association 

on using GIS across the disciplines in a liberal arts college. 

The GIS steering committee also prepared and presented a seminar, titled “Teaching the Abilities 

with a P lace in M ind: G IS A cross th e C urriculum”, h eld a t Alverno’s f aculty in stitute in 

September 2007. This was presented to the entire faculty, and included a presentation b y our 

outside c onsultants, our U WM s teering committee m embers, a nd A lverno f aculty who ha d 

already i ncorporated G IS r elated a ctivities i nto c urriculum. It also i ncluded a h ands-on 

exploration of GIS related activities and resources. 

Additional professional development in the form of conference attendance has been encouraged, 

and faculty and students have attended four regional conferences of GIS users, sponsored by the 

Wisconsin Land Information Association, and the ESRI Wisconsin Users group. 

Based on faculty and student interest in specific disciplines, invited speakers have been set up for 

several Alverno groups, and plans have been made to set up additional speakers. Paul Vepraskas 

has been invited to speak at the department meeting for Business and Management, and also for 

graduate s tudents i n t he S chool of N ursing, where a p roject r elated t o developing a G IS for 

documenting slip and fall incidents in a nursing home was developed. 

Dr Reedy has continued with her GIS related work, and spending her sabbatical year working on 

a G IS r elated pr oject on s ocial geography, a nalyzing r ural a nd ur ban ne eds us ing on -line G IS 

data. She wrote two hands-on GIS curriculum units from this study, one for grades 5 - 12 social 

26

studies/civics a nd on e f or a n a dult e ducation p rogram n eeds assessment. T he uni ts pr ovide 

interactive lessons focusing on t he questions of what is GIS data, how one uses GIS software, 

and how one infers need from GIS data. The unit also discusses issues of ethics and GIS data, 

and is matched with state standards. 

Jennifer Johanson has expanded her GIS by attending UWM for graduate work in a combined 

GIS and environmental geology field. T his work is being fully funded by Alverno to support 

expansion of the GIS efforts at Alverno. 

Curricular changes. Our initial goal was to design and implement 2-4 new or redesigned 

courses that integrate discipline-specific content with GIS related content, and to add from there. 

In the first two year grant period we implemented the new GIS skills course, discussed above, 

redesigned an environmental geology co urse t o i nclude exercises in G IS t hat r elate t o 

environmental s cience t opics, a nd a dded G IS e xercises a nd c oncepts t o t he geography course. 

These course teach GIS in a d iscipline specific setting, using exercises specifically designed for 

geology and geography, respectively. 

In addition t o t hese t hree i nitially p lanned cu rricular ch anges, ad ditional co urses h ave b een 

modified by the instructors trained in GIS technology to integrate GIS-supported demonstration 

projects and/or research projects to expose students to the power of the tool and to some of the 

possible us es of t his t echnology i n t heir ow n f ields. T hese c ourses i nclude e ducation c ourses, 

global effective citizenship courses, graduate nursing courses, and database course. One or more 

of t he f ollowing G IS r elated pr ocesses ha ve be en a dded t o t hese a nd ot her c ourses: da tabase 

design, relational database use, computer mapping production techniques, cartographic design, 

communication pr operties of t hematic m aps, da ta s election a nd qua lity, a nd t he pr oblems of 

graphic display in print and electronic formats. Here are examples of what has been done: 

• A d atabase co urse h as i ncluded a research c omponent w ith G IS i ssues a s one of t he 

possible topics. Student groups in the class researched G IS applications, and pr esented 

their f indings to th e entire c lass a s p art o f th eir f inal a ssessment. T he students w ere 

amazed at the extensive range and applications of GIS into so many disciplines. 

• An education class project involves solving a problem based on a scenario with specific 

criteria. S tudents w ere t o ch oose an ar ea t o es tablish a n ew d ay c are t hat s ervices t he 

needs of Hispanic, women homeowners with an average income below $30,000 t hat is 

within f ive bl ocks o f an e lementary s chool a nd is on a bus l ine. T he s tudents us ed an 

existing on-line GIS (Milwaukee COMPASS) to analyze data and solve this problem. 

• The g raduate nur sing s tudents us ed G IS i n a p roject w here t hey co llected d ata o n 

occurrences of slips and falls. These data were then tied in to a floor plan of the nursing 

home to see if there area patterns of areas that have many occurrences of falls. 

• Mini GIS lessons that are potentially applicable in a number of different courses have 

been prepared, focusing on the use of demographic data and physical geography to 

evaluate similarities and differences between countries. This lesson was piloted in fall 

2007 in GEC 302, one of a series of general education courses required of all students for 

validation in the abilities of effective citizenship and developing a global perspective. 

This lesson will be modified and presented to the GEC instructors for potential inclusion 

into the entire series. This would expose all Alverno students to GIS, as this course is 

required of all students. This course is generally taken in a student’s junior year. 

27

• A significant new outcome for this project is the development a field data-gathering 

component, training students and faculty in the use of global positioning systems (GPS) 

technology to collect GPS locational data and build a GIS dataset. Costs for purchase of 

ten GPS units was covered by Alverno through a separate grant, and the development is 

underway, with a geocaching exercise currently being developed by an environmental 

science student as part of an internship and independent study project with Jennifer 

Johanson. The project is scheduled to be completed in 2009, and be piloted in 2010. 

This activity will be used at new student orientation, exposing incoming students to the 

power of GIS related technology, and hopefully inspiring students to consider a career in 

sciences. 

• A GIS exercise has been developed for a required math concepts course. This activity 

combines maps with data, first requiring students to make maps showing the data, and 

then allowing them to use computer generated versions of the same data to see how 

powerful this tool can be. This course is taken by a majority of Alverno students, and is 

taken in their first year of study. 

• A curricular unit using GIS for grades 5-12 social studies has been developed for the K-8 

education students, so they can see how to apply GIS in a classroom setting. 

• An example of GIS technology was implemented in an Education Psychology course, 

using a hands-on online GIS application. 

• An adult education course implemented a GIS application to solve a needs assessment 

problem using demographic data in a GIS setting. 

Student internships. Three A lverno s tudents ha ve pu rsued and obt ained G IS related 

internships. The first was in 2006, following the first GIS course offered. She used GIS in her 

internship w ith t he U S Bureau of Land M anagement. T his i nternship was a rranged t hrough 

contacts with the outside consultants in the GIS steering committee. The other two students are 

currently in internships where they use GIS to identify and manage urban forest resources for the 

City of Milwaukee. 

Additional GIS Accomplishments 

Other major accomplishments in incorporating GIS across the curriculum include development 

of una nticipated G IS r elated phot o ba nk r esources, a cquisition of da ta a nd ha rdware t hrough 

additional grant funds from outside sources, and inclusion of a field data collection component to 

enhance the lab based GIS exercises. 

Faculty initiated a d atabase development project showcase GIS technology to other faculty and 

students using a real need on c ampus that involves Global Studies photos (goal 9). This project 

was initiated in the first year of the grant, and was presented to the entire faculty at the Alverno 

faculty institute last August. Additional work on t he project has been completed and the results 

will again be presented at August Institute in 2007. The project involved developing a GIS t o 

catalog photographs t aken b y f aculty on i nternational t rips. T he GIS development i nvolved 

determining information that will be included with each photo, and methods to easily search the 

database so that faculty not familiar with GIS systems will be able to access and use the photos 

in the database. The current database configuration allows faculty to add photos to the database 

using a data input page with keywords for location, subject, faculty, and other information which 

28

can be used as search terms by faculty accessing the database. This project has furthered interest 

in GIS on the part of faculty and students. 

The social sciences department identified and procured a database of demographic information to 

be used with GIS in a number of their courses (goal 8). The library, with the help of the GIS 

trained s taff, i s i n t he p rocess of w orking w ith t he s ocial s ciences a nd ot her de partments t o 

develop the most appropriate means to store and access GIS data sources. 

Alverno faculty will continue to seek membership in national organizations for GIS users to keep 

abreast of t echnology c hanges, a s w ell a s t hrough l ocal m eetings a nd n etworking s essions t o 

establish collaborations, a nd a ttend annual c onferences r elating t o p rofessional 

skills/development in Geospatial S cience an d T echnology. W e w ill a lso c ontinue t o de velop 

curriculum and seek grant funding to add to the already numerous GIS curricular experiences on 

campus. The existing courses have become self supporting, and therefore the funding provided 

for the pilot has been successful at launching GIS into the curriculum at Alverno. 

Summary 

The i ncreasingly pervasive us e of GIS t echnology i n bot h t he publ ic and pr ivate s ectors ha s 

created a de mand f or G IS e ducation a nd t raining. T he i ntuitive pow er of m aps o ften r eveals 

trends and patterns that cannot be determined by data alone. As GIS technology has expanded, 

spatial analysis and application of databases to analytical activities such as the determining the 

spread of disease, evaluating changes in population, trends in membership or enrollment, etc in 

areas such as public health, environmental and social sciences, business and marketing, military 

science, l aw, and e ngineering s purred i nterest i n i ncorporating t his t echnology i nto t he 

instructional content of disciplines beyond just earth and space science. The use of GIS, along 

with i nformation a bout how G IS i s us ed i n v arious f ields, br ings a b etter ap preciation f or t he 

widespread u ses o f s cience and t echnology, and es pecially t he application o f aer ospace 

technologies. GIS technology is readily usable as a tool to teach the problem solving, analysis, 

and c ommunication a bilities i n a ne w w ay, t herefore e xpanding curricular oppor tunities. 

Exposure t o t his ve rsatile, e xciting t echnology, e specially a t t he e arly s tages of unde rgraduate 

education, may inspire students to enter scientific and technological fields of study. 

Alverno faculty has been able to provide curricular experiences where students learn to collect 

and a nalyze da ta us ing this t echnology in mu ltiple d isciplines to s olve p roblems and ma ke 

decisions. Geospatial Science and Technology engages students and promotes critical thinking, 

integrated learning and analysis, and develops multiple intelligences. This approach will enable a 

student t o understand r eal-world pr oblems us ing da ta a nalysis a s t hey pr epare f or pos itions i n 

business, he althcare, e ducation, s cience, a nd g overnment. A lverno of fers a uni quely 

advantageous context f or pr eparing s tudents i nterested i n G eospatial S cience and T echnology. 

The i nclusion of us ing GIS analysis, t heory, a nd s oftware a pplication i s a ne w di mension o f 

learning for Alverno students and a new mechanism for teaching analysis and problem solving 

skills f or f aculty fitting well in to A lverno’s a bilities-based m odel of e ducation. S haring 

information a cross m ultiple di sciplines a nd us ing a hol istic a pproach to l earning, applying 

knowledge, and developing related skills are the foundation of abilities-based instruction. 

29

The design of this project and application of the funding provided by WSGC has helped us meet 

and e xceed all of ou r i nitial out comes f or t he pr oject, a nd s ignificantly complete our goal of 

integrating G IS across the c urriculum. It ha s pr ovided f or s tudent l earning oppor tunities 

concurrently with faculty training right from the first year, and has thus provided for interest in 

GIS at both faculty and student levels. We have trained seventeen faculty and staff, developed 

and i mplemented a G IS s kill c ourse, i ntegrated G IS c oncepts and t echnology i nto m ultiple 

discipline courses across the campus, collaborated with other institutions to support workforce 

initiatives and expand educational advancement in the field. W e are encouraged by the support 

of local institutions, organizations and business leaders who use geospatial technology. As we 

shared our progress with other Alverno faculty members, the interest level to participate in this 

project escalated more than we envisioned and the integration of this technology into additional 

courses in diverse disciplines as well the eagerness to participate in the professional development 

provided amazed our team. We continue to have requests to assist additional faculty members 

with basic and more advanced ways to enhance their ability to integrate GIS concepts and skills 

to a higher level of application. Thanks to the WSGC funding, Alverno has been able to integrate 

GIS in to th e c urriculum in a s elf-sustaining way, a nd t he f aculty, s taff, a nd s tudents ha ve a ll 

benefitted as a result. 

30

Synopsis 

ASTRONAUTICS COURSE 

Harald Schenk 

Physics/Astronomy Department; UW-Sheboygan 

Sheboygan, Wisconsin 

The Wisconsin Space Grant Consortium (WSGC) provided UW-Sheboygan a 

Higher Education Initiatives grant for organizing a new interdisciplinary course in 

ASTRONAUTICS for first and second year students in the Sheboygan area. It was the 

first such course offered on any UW-C campus. It may be adapted for other UW-C 

campuses. The new course included a High-Altitude balloon component. This was to give 

the students a chance to collect first-hand scientific data. 

Project Goals 

The focus of this class was to create engaging and scientifically compelling 

experiences for first and second year students in the Sheboygan area. With Sheboygan’s 

selection as a mid-west Spaceport, it was felt that the content of this course would blend 

in with the objectives of our new Great Lakes Aerospace Science and Education Center 

(GLASEC). The strategic focus of GLASEC is the same as the objective of the National 

Space Grant Program. UW-Sheboygan currently offers three Astronomy courses during 

the year. A course in ASTRONAUTICS would give students another option to continue a 

possible career in one of the space sciences. It was also hoped that it would attract some 

of the new engineering students that were scheduled to start attending the campus starting 

in 2008. 

Completed Activities 

It was decided to use a team-teaching approach for this class. Professor Dennis 

Crossley teaches all Physics classes on the campus. Harald Schenk teaches all Astronomy 

classes at UW-Sheboygan. By having two instructors, it would give students a look at 

how the subject is seen from different perspectives. 

Prof Crossley volunteered to handle the administrative job of getting the course 

approved by the UW-C curriculum committee. It was thought that by making the course 

interdisciplinary, we would get a larger variety of students. The course was initially 

approved to start in the fall of the 2008 semester. Due to conflicts with other classes, the 

start was delayed until the spring of 2009. 

In the meantime, Harald Schenk attended a High-Altitude balloon workshop at 

Taylor University in Indiana. It was paid for, in part, by the WSGC grant. Taylor 

promised attendees of the workshop a $200 stipend toward future balloon launches that 

were to be made from their home campus. The workshop took place during May, 2008. It 

was one of two High-Altitude workshops being held at Taylor that year. 

31

Activities at the workshop involved hands-on activities by assembling various 

team payloads. Instruments included a temperature sensor, a pressure sensor, light & 

humidity sensors, a Geiger counter, plus still and video cameras. Canisters containing 

these components were then attached to a balloon, and launched from the campus in the 

afternoon of the first day of the workshop. This first launch was tracked by GPS to an 

altitude of 93,900 feet. After a trip of 65.08 miles, it was safely recovered when it landed 

in a farm field in Ohio. 

The next morning, a second flight was made using the same recovered 

instruments. It took off at 4 AM. GPS signals showed it to have reached an altitude of 

97,000 feet. This payload also made it into Ohio, but upon descending it landed in a large 

pond. The instruments were safely recovered. 

On the final day of the workshop, attendees and staff talked about collaborating 

on upcoming flights from each campus. These would be conducted in the fall of 2008, 

and in the spring of 2009. Taylor would supply staff to support each launch. Since our 

campus had a delay in the start of the ASTRONAUTICS class, we decided to try a spring 

2009 launch. 

Campus Activities 

Our first ASTRONAUTICS class was scheduled to start in January of 2009. 

Initial enrollment was slow. The reason eventually became apparent. Students use a 

system called PRISM to sign-up for classes. Our class had been classified as a PHY 291. 

This is the same number as ANY optional topic in Physics. The description of the course 

in the online catalogue in PRISM had nothing to do with the actual content of our course. 

After this was changed, students started to sign up. By that time, we were within one 

week of the start of classes. As a result, the initial enrollment was only 23 students. We 

hope that this total will increase the next time that the course is offered. 

Our syllabus was based on a text that I had found helpful in graduate school. It 

was called ‘UNDERSTANDING SPACE’. We had used it in our Orbital Mechanics class 

at the University of North Dakota. This book is also used by the U.S. Air Force. The 

PHY 291 syllabus also included a number of guest speakers, plus the High-Altitude 

balloon activity. Speakers would give the students a variety of topics outside of the text. 

Topics included a presentation on Spaceport Sheboygan, a talk on high-powered 

rocketry, and a talk on our annual Rockets-for-Schools event. Students were offered 

opportunities for extra credit by helping at any one of these. 

As a Solar System Ambassador with JPL, I was also able to incorporate news on 

some of the latest space missions. Discussions with students revealed that they enjoyed 

the dual perspective that the team-teaching provided. 

Future classes 

By the time that our spring mid-term had arrived, the State budget was forcing 

changes to our class. With payments to staff being cut, it was decided that future 

ASTRONAUTICS classes at UW-Sheboygan would be limited to only one instructor. In 

addition, plans to add a future lab were halted. A lab would have allowed the 

ASTRONAUTICS students to work on a balloon payload each semester. As it stands 

32

now, they have no input on what is launched, nor do they get experience through handson 

work. Due to the budget cuts, Prof. Crossley will be returning to his Physics classes. 

We are also going to ask for permission to switch the PHY 291 class to an AST 291 

designation. The thought is that students may have been intimidated by having it listed as 

a Physics course. Astronomy is always one of our more popular offerings. 

Eventual High-Altitude balloon studies 

High altitude balloon studies can bring an element of excitement into the 

classroom. They can also provide an opportunity to do real science. From an altitude of 

20 miles, students can study areas that are too high for airplanes, and too low for 

satellites. Our future goal is to investigate the possibility of doing imaging in the UV and 

IR parts of the spectrum. If this is feasible from a stabilized balloon platform, then we 

will conduct a number of astronomical observations. We would also like to establish a 

database with balloon flights throughout the year to see how the upper atmosphere 

interacts with solar particles. This relationship is important, because it has an effect on 

the decay rate of Earth-orbiting satellites. 

By the mid-term of the spring semester, a date had been set for our first balloon 

launch. It would take place on the weekend of April 25-26 th . The staff at Taylor 

University had supplied us with a URL which can predict the path that a balloon would 

take. It can be found at “nearspaceventures.com”. The staff at Taylor had suggested that a 

late-spring launch would produce a shift in upper-atmospheric winds. We could then 

expect the payload to drift to the west. Since Sheboygan is located along the western 

shore of Lake Michigan, we would not want to lose a payload by having it travel east. 

The alternative would be to launch it from a western location, and then chase the payload 

east. I had discussed this possibility with a staff member at UW-Fond du Lac. This sistercampus 

is about 35 miles west of Sheboygan. It would make a good launch site, and we 

could store balloon supplies on campus. Predictions on the “nearspaceventures” URL are 

good for a few days, at best. The exact launch location would not be known until a few 

days before the actual launch was to take place. 

The promise of a launch was already producing excitement in the Sheboygan 

area. Our local newspaper, the SHEBOYGAN PRESS had published several articles 

about the new ASTRONAUTICS class, and about UW-Sheboygan having joined the 

WSGC. When they were told about the balloon plans, they suggested sending along a 

reporter. Middle-school science teachers from District 12 had also shown an interest in 

involving students in our area in such a launch. Several UW-Sheboygan instructors had 

agreed to give their students extra credit for taking part in the activity. What had been 

intended as a ‘test launch’, to simply see if such an activity was feasible from our 

campus, was being blown out of proportion. I hoped that we would be able to deliver. 

Among the attendees of the Taylor workshop was another group from Wisconsin. 

They were from the College of the Menomonee Nation. This group is also a member of 

the WSGC. During the workshop, we agreed to have our students meet for occasional 

joint balloon activities. As a result of that, Dan Hawk suggested that we launch TWO 

balloons on the scheduled weekend in April. 

As the designated launch date approached, Taylor University informed me that 

our campus would not qualify for support personnel. In order to qualify, we would have 

33

to give our students a series of PRE and POST exams in STEM education. I had taken 

these exams while attending the workshop at Taylor. I felt many of the questions to be 

irrelevant. Without a lab, in which the students actually worked on payload components, 

there would be no opportunity to pick-up the knowledge that the Taylor staff had 

intended to impart. Taylor University was willing to send us a balloon package. This 

contained the same components that are sold as a ‘kit’ by StratoStar for $7,000. The 

items arrived on the day before the intended launch. There was no time to prepare the 

items for launch. In addition, the weather was predicted to consist of high-winds. 

Personnel at Taylor University had warned me that the prediction software was 

indication that a balloon launch would take the payload due east, and land about 67 miles 

away. If the launch would take place on the Sheboygan campus, then the payload would 

cross Lake Michigan. It would be lost in the State of Michigan. If it would be launched 

from the Fond du Lac campus, then it would reach the center of the Lake. In either case, 

we would most likely never see the $7,000’s of equipment again. 

Permission was obtained to display the items during Rockets-for-Schools 

activities in Sheboygan. I was hoping that staff, and various students might develop a 

future interest in the activity. 

Results from the 291 class 

With the problems in the Wisconsin State budget threatening class cancellations, 

Professor Crossley and I hoped for the best in getting a chance to continue future 

Astronautics classes. The Curriculum Committee approved the switch from a Physics to 

an Astronomy designation. We received approval to hold another class in the fall of 2009. 

Final exams were given by the end of May, 2009. During this spring semester, 

students were scheduled to assess all instructors on UW-C campuses. This gave us 

valuable feedback. Students in the 291 class indicated that they appreciated the dual 

perspective that was provided by the team-teaching approach. Since many of the students 

were incoming freshmen, they had no prior knowledge of many of the topics presented. 

But several students indicated that they would take one of the astronomy classes in the 

future. 

My goal for the fall is to divide the students into teams on the first class session. 

Each team will be assigned a topic that would have to be completed by the end of the 

semester. Topics will include: Cheap Access to Space; Design of a Cis-Lunar supply 

craft; Design of a lunar space station to serve as a repository for lunar oxygen that could 

re-fuel a supply craft; Design of a lunar base to make use of ice deposits at the poles, and 

to mine oxygen in the lunar soil; Design of an Inter-Orbital tub that could transfer 

supplies between planets in an un-manned fashion, and the Design of a station on the 

Martian moon Phobos to serve as a supply depot for Mars exploration. 

Instead of having the students complete the usual four exams during the semester, 

each team will give the class an update on their work. Completed topics can be used for a 

possible poster display in the Madison Rotunda. 

34

Abstract 

New Directions in Astrobiology Education 

VeraM. Kolb 

Department of Chemistry 


Kenosha, WI 53141 

In this paper we report on the new directions we are taking in the project on the infusion 

of astrobiology in the organic chemistry curriculum. There are two major initiatives. The 

first is on the connection between astrobiology and green chemistry. The second is the 

diversity initiative, in which we examine some key astrobiology discoveries made by the 

scientists from diverse ethnic and racial groups. 

Introduction and Background 

The author has received several grants from the Wisconsin Space Grant Consortium on 

the subject of the introduction of astrobiology in the organic chemistry courses. Much 

has been achieved (Kolb 2004, 2006 a, 2006 b, 2008). However, more remains to be 

done. 

The first obstacle of introducing astrobiology in the chemistry curriculum at our 

institution is that our chemistry program is not amenable to any major changes. Our 

chemistry curriculum is rather lengthy and rigid. Not much can be added, and very little 

can be deleted. This is due to the American Chemical Society accreditation 

requirements. We have somewhat bypassed this obstacle by substituting some chemical 

examples with the astrobiological ones, when they served the same purpose. For 

example, the Lewis structures are a major topic in the Organic I course, and require that 

representations of various atoms and molecules include the dots for the electrons. Instead 

of using standard examples of the Lewis structures from the textbook, we were able to 

substitute these with the structures of the interstellar molecules. However, the 

possibilities for such substitutions of material are limited. 

The second problem in addressing astrobiology within the chemistry curriculum is on the 

inclusion of the newest and most exciting astrobiology discoveries without need to 

rework the entire teaching module on the required chemistry topic. This problem can be 

overcome by assigning students a special topics paper on the subject of astrobiology. We 

have assigned such papers in both regular and advanced organic chemistry courses. The 

students liked these projects and most of the students produced high quality papers. 

Another way to introduce the newest astrobiology topics into the organic chemistry was 

to tap into and utilize the requirement for the library research via the Science Scholar 

Finder search engine (which includes the Chemical Abstracts, and is thus mandated by 

the American Chemical Society for the accreditation). This approach was also 

successful. However, a need still exists for a broader inclusion of the hot astrobiology 

topics into the chemistry curriculum. 

35

In the past we have not focused on the specific achievements of diverse groups of 

scientists, either chemists or astrobiologists, from different ethnic and racial groups, and 

have thus missed on directly addressing diversity issue. This issue is critical for our 

campus, which is the most diverse in the UW -System. 

The new initiatives 

The first new direction we took is to link astrobiology to chemistry in a novel way, which 

would be on the forefront of both astrobiology and chemistry. This is the link between 

astrobiology and green chemistry. Green chemistry is chemistry which is 

environmentally friendly. The author was extremely fortunate to have been accepted into 

the NSF -sponsored workshop on the green chemistry at the University of Oregon, 

Eugene, July 18-24, 2009. In this workshop both the theory and the practice were 

covered. The participants were able to do many experiments in the lab which 

demonstrated the principles of green chemistry. These experiments were developed to the 

point that they could be easily adapted in the organic lab teaching. In addition, several 

such experiments were amenable to the undergraduate research. We have seen 

immediately a clear connection between the green chemistry and astrobiology, and have 

developed this idea for a presentation and a future publication, by including many 

references and concepts, which are ready for the classroom applications (Kolb, 2009). 

Here we list an abbreviated list of topics. 

One obvious way for the organic chemistry to be environmentally friendly is to use water 

as solvent, instead of more toxic organic solvents. Another approach is to run the 

reactions between the solid components without the solvent, in a so-called solventless 

reactions mode. As the solid materials are mixed together, the melting point of the 

mixture is lower than the melting points of its individual components (the principle of the 

mixed-melting point). In some cases the entire mixture may melt upon mixing. The 

reactions would then occur in a viscous state. These and some other known examples of 

green chemistry have a great potential for astrobiology. Conversely, astrobiology can 

inspire green chemistry also. The astrobiological reactions typically occur in water, 

which models the prebiotic soup. While much is known about such astrobiological 

reactions, they were never optimized for the industrial use. Now is may be the right time 

to do it, since the industry is moving in the green direction. Although we do not know the 

details about the mechanisms of the reactions in the solid mixtures on the asteroids and 

meteors, we know what kinds of products are obtained. We know this from the diverse 

list of chemicals that were identified in the carbonaceous chondrite meteorites. The 

structures of these products may inspire new applications for the industrial solventless 

reactions. The connection between the green chemistry principles and astrobiology 

represents a new pedagogical approach for infusion of astrobiology into the organic 

chemistry. 

The diversity initiative has been conceived while the author was fortunate to participate 

in the Summer Institute at UW-Parkside, which was devoted to diversity issues. As a 

part of introducing diversity into the organic chemistry curriculum the author has decided 

to cover some key chemical discoveries that were made by a diverse group of scientists. 

Included are African Americans, Indians, Japanese, Mexicans and others. This teaching 

material has been developed during the summer, and is ready now for the classroom 

36

application. We are now developing the analogous teaching material about some key 

astrobiology discoveries that have heavy chemistry content, and which were made by the 

astrobiologists of diverse ethnic and racial groups. 

Conclusions 

Two new pedagogical initiatives on the introduction of astrobiology into the chemistry 

curriculum are being developed. The first is on the astrobiology-green chemistry 

connection. The second is on the key discoveries by the astrobiologists from diverse 

ethnic and racial groups. 


WSGC is acknowledged for a steady support of our higher education efforts. 

References 

Kolb, V. M. "Introduction of Astrobiology into the Undergraduate Curriculum", 

in "Making Space for Life", Proceedings of the 10 th Annual Wisconsin Space 

Conference, held June 9-10, 2000, in Milwaukee, WI, A. Yingst and S. D. Brandt, 

Editors. Published in 2004 together with the Proceedings of the 11 th Annual 

Wisconsin Space Conference by the Wisconsin Space Grant Consortium, Green 

Bay, WI. Part 5: Chemistry. 

Kolb, V. M. "Teaching Astrobiology from a Leamer-Centered Perspective", in 

"Continuing the Voyage of Discovery", R. A. Yingst, S. D. Brandt, J. Borg, S. 

Dutch, M. Gustafson, M. Rudd, and A. Roethel, Eds., Proceedings of the 15 th 

Annual Wisconsin Space Conference, Wisconsin Space Grant Consortium, Green 

Bay, WI, 2006a. Part Nine: General Education and Outreach. 

Kolb, V. M. "Astrobiology at Parkside: A Web site Dedicated to Teaching", 

2006b, ibid. 

Kolb, V. M. "Permanent infusion of astrobiology into the organic chemistry 

curriculum", submitted for publication for the proceedings of the 18 th Annual 

Wisconsin Space Conference, held in Fox Valley, WI, August 14-15,2008. 

Kolb, V. M. "Astrobiology and green chemistry: A new pedagogical connection", 

ibid, oral presentation, SPIE, paper 7441B-3, San Diego, August 4-6, 2009; 

Proceedings to be published. 

37

Providing High School Students with Earth Imaging Tools * 

Thomas C. Jeffery 

University of Wisconsin – Whitewater 

Abstract: Science education constitutes a significant segment of the core curriculum in primary 

and secondary schools. Traditionally, courses on Biology, Chemistry, and Physics constitute the 

majority of science education in most schools. Earth imaging, utilizing air photos or satellite 

images, can contribute to each of the disciplines above, however, without providing a foundation 

devoted to understanding earth imagery, the ability to extract meaning from the information 

provided is limited. This inability to incorporate earth imagery in primary and secondary schools 

had much less significance five years ago since the cost of the data made it prohibitive. 

However, the recent availability of Google Earth and the entire archive of Landsat imagery at no 

cost make earth imagery a frequently used tool in spatial problem solving in both the public and 

private sectors and also make it possible to offer instruction on analyzing and interpreting earth 

imagery in primary and secondary education. 

Remote Sensing Value 

Earth imaging. Acquiring images o f t he earth’s s urface p redates t he i nvention of t he 

airplane, but it was the innovation of controlled flight and much later, unmanned satellites, that 

enabled a erial phot ography t o grow i nto t he va luable da ta s ource t hat i t h as b ecome (Jensen, 

2000). The ability to view large portions of the surface of the planet as it exists, and not as the 

result of i nterpretation or i nterpolation a llows t he vi ewer t o vi sually analyze f eatures a nd 

relationships. T he introduction in the mid and latter portions of the 20 th century of the satellite 

based i maging pl atform i ncreased t he t ype a nd vol ume of da ta t o i nclude m ultispectral 

information (Lillesand, et a l., 2004 ). T his h as led to e ven greater u tility o f th e ima gery for 

methods of compositional analysis. 

University level. Due to the value of aerial photography and satellite imagery as a means of 

studying the form, structure, and composition of the earth’s surface, it was only natural that the 

study of e arth i magery be c ontained w ithin t he s cope of Geography and E arth S cience 

departments. Increasingly, ot her d epartments and di sciplines ha ve i ncorporated i magery a s a 

means of investigating and analyzing earth surface features and relationships. B iology, botany, 

environmental s tudies, chemistry, a nd ot hers are i ncreasingly r elying on i magery i n t heir 

investigations. Universities and research scientists were some of the initial users of this data, as 

they could afford the computing equipment and the cost of the data, however, very little of this 

technology has been adopted by primary and secondary schools. 

* This project was made possible by funding from the Wisconsin Space Grant Consortium and the University of 

Wisconsin-Whitewater College of Letters and Sciences, along with the UW-W Geography & Geology department 

39

Private and public sector. The beginning of the 21 st century has brought changes in earth 

imaging that have increased the utility and adoption of this data by businesses and agencies that 

heretofore ha d l ittle ex posure t o i magery. Initially, o nly th e milita ry a nd h igher le vel 

governmental agencies such as the National Oceanic and Atmospheric Administration (NOAA), 

the National Aeronautics and Space Administration (NASA), and the Soil Conservation Service 

(SCS) had t he eq uipment a nd e xpertise t o i ncorporate i magery i nto t heir r esearch an d w ork. 

However, a s w e s ee t oday, vi rtually all l evels o f g overnment, dow n to and including c ities, 

towns, and villages have begun adopting and incorporating image data. 

The incorporation of technology in the private sector is typically dependent on t he value added 

or profitability associated with inclusion. Due to the high front-end cost of the equipment, data, 

and training, there was, until recently, limited utilization of earth imagery for most businesses. 

Primary and secondary level. The discipline of geography was part of the core curriculum 

in ma ny primary a nd s econdary schools i n t he l ate 19 th and ear ly 2 0 th centuries. H owever, a 

current evaluation of curriculum would suggest that the discipline of geography has been divided 

and either discarded or absorbed by other courses. D ue in part to this, and also the traditional 

cost o f e arth ima gery, it is unlikely th at th is d ata is not c urrently be ing incorporated in mo st 

public school systems. 

Google Earth and Landsat imagery. Initially, the cost of earth imagery was the prohibitive 

factor for inclusion in the public sector, private sector, and education. In large part, both primary 

and s econdary institutions have avoided t hese da ta due t o t he c ost. However, r ecent 

developments h ave m ade av ailable h igh resolution imagery t o an yone w ith a co mputer an d 

connection to the internet. Unlike most technology, which trickles down from high-end to lowend 

users as cost declines over the life of the technology, two events within the last five years 

have m ade high r esolution satellite ima gery available f or n o c ost, w ithout t he l ong w ait for 

technological improvement. 

In 2005 Google E arth released its v irtual globe, which m ade available co mplete p lanetary 

coverage of imagery. In the years following, the imagery has been updated and improved so that 

the current version carries high resolution color data for nearly the entire planet. That in itself is 

remarkable, b ut combined with the fact that access is only limited b y a computer and internet 

connection, with no a dditional fee, makes this data set truly r evolutionary. T he onl y possible 

downside to this data is that it is not available for multispectral analysis. However, in 2009 the 

US g overnment m ade Landsat i magery available f or no c ost. This h as made n early 3 7 years 

worth of multispectral data available for anyone with computer access. T hese two events have 

opened image interpretation and analysis to the world. 

Objectives 

The pur pose of t his r esearch w as t o de velop a m ethod of i ntroducing the to ols a nd s kills 

necessary for analysis of earth imagery to primary and secondary students. Due to the likelihood 

40

that th e current generations of s tudents will find ear th i magery t o b e u biquitous a nd a lso be 

required t o unde rstand a nd ut ilize i magery, i t i s our educational responsibility t o pr ovide this 

information. 

Procedures and Methods 

Bringing earth imagery to secondary education. In order to incorporate earth imagery in 

secondary school curriculum, it was decided that an outreach program that brought high school 

students and teachers to the university for a training session and exposure to the software and 

data would be the most effective method of introducing this information. This was not intended 

as a complete or thorough education in earth imagery, but rather as a means of introduction that 

could and should be expanded upon within the high school curricula. 

Even a t a n i ntroductory l evel, t he vol ume of t he m aterial m andated n early a f ull s chool d ay 

worth o f in struction. Within th at ti me f rame, th e ma terial w ould consist o f a n introductory 

section de voted t o t he b asic s kills of vi ewing and m anipulating phot ographic qua lity i mages. 

The primary concept focuses on viewing the surface of the earth from a p erspective heretofore 

seldom seen, if ever, by the student. The basic viewing section was followed by an exercise that 

required feature identification. T his section drew upon t he visual analysis and interpretation of 

features through the use of shape, color, size, and other interpretative elements. 

The exercise that followed directed the students to search for and identify objects about which 

visual c lues w ere pr ovided. T his s ection w as a m ore r igorous f ollow-up t o t he i ntroductory 

identification s ection, a nd introduced que stions f or w hich a nswers were n ot r eadily ap parent 

from the image alone. Questions in this section typically included queries that would necessitate 

critical thinking and problem solving skills. One example of this was an image of the Fairbanks, 

Alaska airport, which depicts the traditional runways and terminal. H owever, this airport also 

contains l arge ponds a djacent t o t he r unways. T he que stion w as po sed a s t o t he r eason 

necessitating t hese ponds , c onsidering t hey would l ikely be a h azard a t m ost a irports. The 

students would be required to solve this conundrum by looking for clues as to the purpose of the 

ponds and also consider the geographic context of the airport. ( The ponds are landing/takeoff 

areas for sea planes). During both of these exercises undergraduate student assistants would be 

present to aid the students in using the software and data. 

The f inal e xercise o f t he s ession i nvolved t he a nalysis of m ultispectral Landsat i mages. T he 

focus of this section was to introduce the students to energy bands outside the visible spectrum 

and de monstrate how t hese b ands of e nergy are us eful i n s cientific analysis o f ve getation. 

Specifically, using near infrared energy to evaluate vegetation health and biomass. Interspersed 

throughout al l o f t he exercises w as a d ialogue and co mmentary f rom t he f aculty m ember i n 

charge of the session. T his would not be considered a true lecture, due to the fact the material 

was p resented l ess f ormally. H owever, t he n arrative w as comprehensive an d w as l inked t o 

specific sections of the exercise. 

41

Results and Observations of the Session 

The high s chool c lass t hat pa rticipated i n t he s ession he ld a t U W-Whitewater’s g eography 

laboratory, numbered 13 students and one teacher. These students were concurrently enrolled in 

their s chool’s e arth s cience course a nd h ad p reviously don e a va riety of geological a nd 

geographical exercises created by their teacher. None of the work they had done previously had 

included any use or discussion of aerial photography or satellite imagery. 

An informal v erbal poll indicated that non e of the students had previously used Google Earth 

imagery for any coursework, and only a few had ever heard of it prior to that da y. Since the 

exercises were designed from the perspective of no pr ior knowledge, the comprehension level 

was appropriate. After a brief introduction, the students were paired with undergrad geography 

students e xperienced in b oth G oogle E arth and s atellite ima ge a nalysis. W hile th e w ritten 

exercises were provided as stepwise instruction, it soon became clear that the interaction with the 

undergraduate as sistants was a v aluable resource. A ll of the high school students soon began 

relying on their college colleagues for additional insight into not only the exercise, but also other 

features on the imagery that commanded their attention. 

Throughout all of t he exercises i t w as r eadily apparent t hat the hi gh school s tudents w ere 

fascinated b y t he i magery and c ould of ten b e f ound de viating f rom t he e xercise que stions i n 

order to follow their own curiosity concerning features all around the globe. This was certainly 

the type of response that an educator relishes, however it was necessary to curb the exploration 

in order to maintain the limited time frame. 

Conclusions 

The goal of this project was to provide high school students with an introduction to a palette of 

tools and data that they will be able to draw upon both in future education opportunities as well 

as i n t heir ow n pe rsonal a nd pr ofessional e xperiences i n l ife. N ot onl y i s i magery n early 

ubiquitous in the private and public sectors, with the public release of Google Earth in 2005 it 

has become mainstream for anyone with an internet connection. F or these reasons it is vitally 

important th at in terpreting and u nderstanding earth imagery is regarded equally t o other ba sic 

tenets of education. 

Based on the response by the high school students who participated in the outreach session, the 

experience would have to be judged a success. With no real prior experience at the beginning of 

the da y, t he s tudents c ommented t hat the a ctivities w ere very i nformative a nd t hat t hey w ere 

looking f orward t o further exploration of t he i magery. G iven t hese f actors i t a ppears t hat a 

single day session is a valid and effective method of extending an introduction to earth imaging. 

It would also be advised that a single session does not provide a complete and comprehensive 

education i n e arth i maging. A s s uch, i t w ould be advisable to pr ovide t he pa rticipating hi gh 

school teachers with data, software, and exercises that would enable them to offer more imaging 

42

instruction a s pa rt o f t heir ow n curriculum. In or der t o f acilitate a b roader exposure t o t his 

information, UW-Whitewater Geography & Geology department plans to host similar sessions 

throughout the school year. 

Refernces 

Jensen, J.R. 2000. Remote sensing of the environment. In: Remote Sensing of the 

Environment: An Earth Resource Perspective, Prentice Hall, pp. 1-28. 

Lillesand, T.M., Kiefer, R.W., and Chipman, J.W. 2004. Earth resource satellites operating in the 

optical spectrum. In: Remote Sensing and Image Interpretation, John Wiley & Sons, 

Inc., pp. 397-490. 

43


Appendix A 

2009 Program


& 


Present: 

the Nineteenth ANNUAL 

WISCONSIN SPACE CONFERENCE 

“The Cassini Encounter with the Gem of the 

Solar System” 



Thursday, August 13 – Friday, August 14, 2009 

___________________________________________________________________________________ 

CONFERENCE 2009 PROGRAM 

___________________________________________________________________________________ 

1

Thursday, August 13, 2009 

7:30-8:45 am Registration Todd Wehr Conference Center 

Continental Breakfast 

Posters (formal poster session at 2:45 p.m.) 

8:45-9:15 am Welcome and Introductions 

*** Plenary Session *** 

Tom Bray, Dean of Applied Research, Milwaukee School of Engineering (MSOE); 

Associate D irector f or S pecial I nitiatives, Wisconsin S pace Grant C onsortium 

(WSGC) 

Dr. Herman Viets, President, Milwaukee School of Engineering 

R. Ai leen Yi ngst, Director, W isconsin S pace G rant C onsortium, U niversity of 

Wisconsin-Green Bay; NASA Mars Rover Exploration Mission Scientist 

9:15-10:15 am Session 1: Keynote Address 

Dr. Ellis Miner, The Cassini Encounter with the Gem of the Solar System I, NASA 

Science Manager for the Cassini Mission to Saturn 

10:15-10:45 am Morning Break and Refreshments 


10:45-11:45 pm Session 2: NASA Programs Student Panel Todd Wehr Conference Center 

Moderator: R. A ileen Y ingst, Director, W isconsin S pace Grant C onsortium, 

University of Wisconsin-Green Bay; NASA Mars Rover Exploration Mission Scientist 

Aaron Olson, NASA Academy-Goddard Institute of Space Studies, UW-Madison 

Cheryl Perich, NASA Academy-Marshall Space Flight Center, Marquette University 

Nathan Wong, NASA Academy-Marshall Space Flight Center, UW-Madison 

Kenion Blakeman, Undergraduate Student Research Internship-Marshall Space Flight 

Center, Carthage College 

Matthew Kallerud, LARSS Internship-NASA Langley Research Center, MSOE 

Vanessa Peterson, LARSS Internship-NASA Langley Research Center, UW-Madison 

Isa Fritz, NASA Reduced Gravity Program, Carthage College 

2

11:45-1:00 pm Lunch 

*** Concurrent Sessions -- Research Streams *** 

1:00-2:30 pm Session 3R: Astronomy/Physics/ Todd Wehr Conference Center 

Atmospheric Sciences 

Moderator: Kevin C rosby, Chair, D ivision of Natural S cience, C arthage C ollege; 

WSGC Institutional Representative 

Jordan G erth, Realizing a Better Hydrostatic Response in NWP with MODIS 

Products, Undergraduate, UW-Madison 

Malanka Riabokin, Gravitational Heating in Galaxy Groups, Undergraduate Student, 

UW-Madison 

Cyrus V andrevala, Understanding the Evolution of Supernova Progenitors, 

Undergraduate Student, Marquette University 

Justin Madsen, Validation of Novel Rigid Body Frictional Contact Algorithms Using 

Tracked Vehicle Simulation: a Stepping Stone for Billion Body Dynamics, Graduate 

Student, UW-Madison 

E. A lec Joh nson, Fast Reconnection in Fluid Models of Pair Plasma, Graduate 

Student, UW-Madison 

Kerry K uehn, Reflection and Refraction of Vortex Rings, Associate P rofessor, 

Wisconsin Lutheran College 

*** Concurrent Sessions -- Education Stream *** 

1:00-2:30 pm Session 3E: K-12 Education & General Public Outreach Campus Center 130 

Moderator: Sharon Brandt, Program Manager and Associate Director for Outreach, 

WSGC 

James K ramer, Combining Writing Across the Curriculum with Community-Based 

Programs to Teach Core Scientific Concepts, Executive Director, S impson S treet 

Press 

Coggin H eeringa, Teaching the Teachers about Astronomy, Director, C rossroads a t 

Big Creek 

3

Lee Siudzinski, Women Soar: Girls Grades 9-12, Manager, Education Relationships, 

Experimental Aircraft Association 

Lee Si udzinski/Chrissy P aape, Space Discovery Week, Manager, E ducation 

Relationships, E xperimental A ircraft A ssociation/Vice P resident, S pace E xplorers, 

Inc. 

Brad Staats, Space Travel Simplified – Part 1, President, Spaceflight Fundamentals, 

LLC 

2:30-3:00 pm Afternoon Break and Refreshment Todd Wehr Conference Center 

3:00-3:45 pm Poster Session 

Facilitator: R. Aileen Yingst 

Chelsey J elinski E rickson representing t eam m embers: N eal B itter, Adam H arden, 

Narwhal -First Place Rocket Team in the Engineering Division, Undergraduates, 

MSOE 

Isa Fritz representing team members: Bradley Frye, Samantha Kreppel, Erin Martin, 

Joseph Monegato, Caitlin Pennington, Angle of Repose of Lunar Simulants in Reduced 

Gravity, NASA Reduced Gravity Program, Undergraduates, Carthage College 

Matthew K allerud, Dynamics Characterization of the Electron Beam Freeform 

Fabrication System, NASA LARSS Internship at Langley, Undergraduate, MSOE 

Aaron Olson, Shock Heritage for Detectors, Undergraduate, UW-Madison 

Vanessa P eterson, Modeling Wake Vortices for the Use of Ground-Based Lidar, 

NASA LARSS Internship at Langley, Undergraduate Student, UW-Madison/ 

Bradley R entz, A Comparative Study of Type IIb Supernovae, Undergraduate, 


Harrison Skye, Mixed Gas Joule Thomson Cycles with Precooling, Graduate Student, 

UW-Madison 

*** Concurrent Sessions -- Research Stream *** 

3:45-5:15 pm Session 4R: Chemistry/Biology Todd Wehr Conference Center 

4

Moderator: David B lock, A ssociate P rofessor of E nvironmental Science, C arroll 

University 

Dan Hawk, Extremophile Research, Undergraduate Student, College of Menominee 

Nation 

Cheryl P erich, Vapor Compression Distillation Flight Experiment and Urine 

Processor Assembly Brine Extraction, NASA A cademy I ntern at M arshall, 

Undergraduate Student, Marquette University 

Megan C hristenson, Understanding Risk Determinants of Chagas Disease in 

Southern Peru, Graduate Student, UW-Madison 

Kenion B lakeman, Toxic Off-gassing Analysis at Marshall Space Flight Center, 

NASA Internship at Marshall, Undergraduate Student, Carthage College 

Nathan Won g, Techniques for Oxygen Production from Lunar Materials, NASA 

Internship at Marshall, Undergraduate Student, UW-Madison 

*** Concurrent Sessions -- Education Stream *** 

3:45-5:15 pm Session 4E: Education and Public Outreach Campus Center 130 

Moderator: Joh n B org, Professor, D epartment of M echanical Engineering, 

Marquette University; WSGC Associate Director for Higher Education 

Bill F arrow, Freshman Engineering Video-Rocket Design Project, Assistant 

Professor, MSOE 

Jennifer Joh anson, Expanding GIS Across the Curriculum, Assistant P rofessor, 

Alverno College 

Harald Schenk, A Course in Astronautics, Associate Lecturer, UW-Sheboygan 

Thomas Jeffery, Providing High School Students with Earth Imaging Tools, Assistant 

Professor, UW-Whitewater 

*** Adjourn for the Day *** 

5


Friday, August 15, 2008 

8:00-9:00 am Registration Todd Wehr Conference Center 

Breakfast 

8:00-8:45 am Undergraduate Workshop CC130 


9:00-10:00 am Session 5: Keynote Address Todd Wehr Conference Center 

Dr. Ellis Miner, The Cassini Encounter with the Gem of the Solar System II, NASA 

Science Manager for the Cassini Mission to Saturn 

10:00-10:30 am Morning Break and Refreshments 

10:30-12:00 am Session 6: Student Satellite Program 


Moderator: Bill Farrow, WSGC Associate Director for Special Initiatives, Assistant 

Professor, Milwaukee School of Engineering 

Balloon Payload Team: 





Victorialynn Salas, Marquette University 


Balloon Launch Team: 

Eric Deering, Milwaukee School of Engineering 

Brian Nguyen, Milwaukee School of Engineering 

Nathan Sward, Milwaukee School of Engineering 

Sounding Rocket Teams: 

First P lace, N on-Engineering, Schmitt Triggers, Brad H artl r epresenting t eam 

members: J acob W ardon, S teven W elter, M ark W itte, U W-LaCrosse and W estern 

Technical College 

6

Second Place, E ngineering, Drew and Crew, Wesley L arrabee representing t eam 

members D rew Falkenburg, C aleb V arner, Undergraduates, M ilwaukee S chool of 

Engineering 

12:00-12:15 pm Group Photograph As Directed 

12:15-1:15 pm Awards Luncheon Todd Wehr Conference Center 

1:15-1:45 pm Awards Ceremony 

R. Aileen Yingst, Director, Wisconsin Space Grant Consortium 

Sharon D. Brandt, Program Manager, Wisconsin Space Grant Consortium 

Karen Valley, Chair, WSGC Advisory Council 

Program Associate Directors 

1:45-2:00 pm 2010 Conference 

2:00 pm Conference Adjourned 

*** Adjournment *** 

2:00-4:00 pm Advanced Undergraduate Research Workshop Computer Lab, S-130 

7

WSGC Members and Institutional Representatives 

Lead Institution 


Scott Ashmann 

Aerogel Technologies, LLC 

Stephen Steiner 

AIAA - Wisconsin Section 

Marty Gustafson 

Alverno College 

Paul Smith 

Astronautics Corporation of America 

Steven Russek 

BioPharmaceutical Technology Center Institute 

Karin Borgh 

Carroll University 

Michael Schuder 


Kevin Crosby 

College of Menominee Nation 

Norbert Hill 

Experimental Aircraft Association (EAA) 

Lee Siudzinski 

Great Lakes Spaceport Education Fdn. 

Carol Lutz 

KT Engineering 

David B. Sisk 

Lawrence University 

Megan Pickett 


Christopher Stockdale 

Medical College of Wisconsin 

Danny A. Riley 


William Farrow 

Orbital Technologies Corporation 

Eric E. Rice 

PLANET LLC 

Thomas Crabb 

Ripon College 

Mary Williams-Norton 

Saint Norbert College 

Terry Jo Leiterman 

Space Education Initiatives 

Jason Marcks 

Affiliates 


George French 

Spaceflight Fundamentals, LLC 

Bradley Staats 

University of Wisconsin-Fox Valley 

Martin Rudd 

University of Wisconsin-La Crosse 

Eric Barnes 


Gerald Kulcinski 

University of Wisconsin-Milwaukee 

Ronald Perez 

University of Wisconsin-Oshkosh 

Nadejda Kaltcheva 


David Bruning 

University of Wisconsin-River Falls 

Glenn Spiczak 

University of Wisconsin-Sheboygan 

Harald Schenk 

University of Wisconsin-Stout 

John Rompala 

University of Wisconsin-Superior 

Richard Stewart 

University of Wisconsin-Whitewater 

Rex Hanger 

Western Technical College 

Michael LeDocq 

Wisconsin Aerospace Authority 

Tom Crabb 

Wisconsin Association of CESA Administrators 

James Larsen 

Wisconsin Department of Public Instruction 

Shelley A. Lee 

Wisconsin Department of Transportation 

Karen Valley 

Wisconsin Lutheran College 

Kerry Kuehn 

See www.uwgb.edu/wsgc for up-to-date contact information

Wisconsin Space Grant Consortium ■ University of Wisconsin-Green Bay ■ 2420 Nicolet Drive ■ Green Bay, WI 54311-7001 

Phone: 920.465.2108 ■ Fax: 920.465.2376 ■ E-mail: wsgc@uwgb.edu ■ Web site: www.uwgb.edu/wsgc

Space Grant Consortium - University of Wisconsin - Green Bay

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?