26.12.2012 Views

Scientific and Technical Aerospace Reports Volume 38 July 28, 2000

Scientific and Technical Aerospace Reports Volume 38 July 28, 2000

Scientific and Technical Aerospace Reports Volume 38 July 28, 2000

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Footage shows the night vertical takeoff of the Viking Hollan hot air balloon. The crew is shown participating in survival<br />

technique training, boarding the plane to depart to Canada, <strong>and</strong> preparing for the vertical takeoff in the hot air balloon across the<br />

Atlantic Ocean. Scenes also include the making of the capsule for the balloon, some flight activities, <strong>and</strong> the l<strong>and</strong>ing of the balloon.<br />

CASI<br />

Vertical Takeoff; Balloon Flight; Vertical Flight; Climbing Flight; Vertical L<strong>and</strong>ing; Crash L<strong>and</strong>ing<br />

<strong>2000</strong>0064110 NASA Glenn Research Center, Clevel<strong>and</strong>, OH USA<br />

Issues <strong>and</strong> Effects of Atomic Oxygen Interactions With Silicone Contamination on Spacecraft in Low Earth Orbit<br />

Banks, Bruce, NASA Glenn Research Center, USA; Rutledge, Sharon, NASA Glenn Research Center, USA; Sechkar, Edward,<br />

DYNACS Engineering Co., Inc., USA; Stueber, Thomas, DYNACS Engineering Co., Inc., USA; Snyder, Aaron, NASA Glenn<br />

Research Center, USA; deGroh, Kim, NASA Glenn Research Center, USA; Haytas, Christy, Clevel<strong>and</strong> State Univ., USA; Brinker,<br />

David, NASA Glenn Research Center, USA; May <strong>2000</strong>; 14p; In English; 8th; 5th; Materials in a Space Environment, 4-9 Jun.<br />

<strong>2000</strong>, Arcachon, Arcachon, France, France; Sponsored by Centre National d’Etudes Spatiales, France<br />

Contract(s)/Grant(s): RTOP 632-6A-1E<br />

Report No.(s): NASA/TM-<strong>2000</strong>-210056; E-12258; NAS 1.15:210056; No Copyright; Avail: CASI; A03, Hardcopy; A01, Microfiche<br />

The continued presence <strong>and</strong> use of silicones on spacecraft in low Earth orbit (LEO) has been found to cause the deposition<br />

of contaminant films on surfaces which are also exposed to atomic oxygen. The composition <strong>and</strong> optical properties of the resulting<br />

SiO(x)- based (where x is near 2) contaminant films may be dependent upon the relative rates of arrival of atomic oxygen, silicone<br />

contaminant <strong>and</strong> hydrocarbons. This paper presents results of in-space silicone contamination tests, ground laboratory simulation<br />

tests <strong>and</strong> analytical modeling to identify controlling processes that affect contaminant characteristics.<br />

Author<br />

Oxygen Atoms; Silicones; Contaminants; Mathematical Models<br />

<strong>2000</strong>0064622 Virginia Polytechnic Inst. <strong>and</strong> State Univ., Multidisciplinary Analysis <strong>and</strong> Design (MAD) Center for Advanced<br />

Vehicles, Blacksburg, VA USA<br />

Study of a Global Design Space Exploration Method for <strong>Aerospace</strong> Vehicles<br />

Baker, Chuck A., Virginia Polytechnic Inst. <strong>and</strong> State Univ., USA; Watson, Layne T., Virginia Polytechnic Inst. <strong>and</strong> State Univ.,<br />

USA; Grossman, Bernard, Virginia Polytechnic Inst. <strong>and</strong> State Univ., USA; Mason, William H., Virginia Polytechnic Inst. <strong>and</strong><br />

State Univ., USA; Cox, Steven E., Florida Univ., USA; Haftka, Raphael T., Florida Univ., USA; February <strong>2000</strong>; In English; See<br />

also <strong>2000</strong>0064579; No Copyright; Abstract Only; Available from CASI only as part of the entire parent document<br />

In the early stages of the design process of aerospace vehicles, the search for optimal configurations is wide open, <strong>and</strong> the<br />

use of local optimization tools may risk missing the best designs. Therefore, global optimization methods are attractive for the<br />

early design stage. Unfortunately, global design optimization usually requires the evaluation of a very large number of designs,<br />

a formidable computational challenge. The present work demonstrates the use of massively parallel computers for h<strong>and</strong>ling this<br />

computational challenge. A variety of load balancing methods are used to ensure efficient utilization of the computer nodes.<br />

Global optimization was applied to the High Speed Civil Transport (HSCT) aerospace configuration design problem with <strong>28</strong><br />

design variables <strong>and</strong> 68 nonlinear constraints in a multidisciplinary design optimization (MDO) environment. Even with the use<br />

of simple engineering analyses, a thorough design space search is computationally expensive due to the large number of designs<br />

that need to be evaluated. The more expensive constraints, such as range <strong>and</strong> takeoff distances, are performance related. Previous<br />

work with this design problem has shown that the design space contains disconnected isl<strong>and</strong>s of feasible points, representing different<br />

available design concepts. These complexities make it difficult for an optimizer to successfully find the global optimum.<br />

The global optimization method used is a Lipschitz algorithm that (essentially) uses all possible values of the Lipschitz constant.<br />

By using all possible values of the constant, equal emphasis is placed on local <strong>and</strong> global search by the optimizer. This causes the<br />

optimizer to continue searching globally for the isl<strong>and</strong>s of feasible space, while converging on designs in the promising regions<br />

that have already been discovered. For each optimization cycle in the algorithm, a large set of new designs that need to be evaluated<br />

is generated. Since the objective function <strong>and</strong> constraint evaluations for each design can be computed independently, parallel computers<br />

can be easily used to concurrently evaluate all of the new designs. When there is enough variance in the evaluation time<br />

of these designs it is beneficial to have some form of dynamic load balancing. The load balancing schemes investigated for this<br />

study are: static load balancing, dynamic load balancing with a master-slave processor organization, fully distributed dynamic<br />

load balancing, <strong>and</strong> fully distributed dynamic load balancing via threads. The parallel optimization runs were conducted for a fixed<br />

number of iterations of the modified Lipschitzian algorithm on an SGI Origin <strong>2000</strong>. Around 10,000 aircraft designs were evaluated<br />

in the design space exploration before finding the global optimum, with the optimizer sampling many of the promising regions<br />

of the design space in the process. The variation in the evaluation times of the designs investigated was small. The parallel efficien-<br />

35

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!