UWE Bristol Engineering showcase 2015

Patrick Busch
MEng Aerospace Engineering (Systems Engineering)
Project Supervisor
Dr. Chris Toomer
Correction, calibration and evaluation of real measured data of
aeroplane behaviour
Introduction
This study is the third in a series of masters´ theses
on the topic of position, attitude and control input
measurement during flight, particularly during spin
of aircraft. This master´s thesis is designed to assist
the PhD thesis of Steffen Schrader (University of
Applied Sciences Osnabrueck) on the topic of spin
prediction and simulation.
Spinning
Spin is a flight mode that can be referred to as
autorotation around a vertical axis on a downward
path. In order to enter a spin
the aeroplane must be stalled
and at the same time have a
sufficient yaw rate or sideslip
angle. When a stall occurs,
the critical angle of attack
is exceeded and the airflow
over the wing is separated
and becomes turbulent. The
lift produced decreases
Fig. 1: Spin
considerably and hence the equilibrium of forces is
unbalanced. Due to the mentioned sideslip angle
or yaw rate, one wing will be stalled more than the
other, resulting in a lower lift force and hence the
aeroplane will start to roll in the direction of the
more stalled wing. As a result from this rolling
motion, a pitching and yawing moment will be
induced. The drag on the more stalled wing will
increase and support the rotation around the
vertical axis as the aeroplane descends rapidly.
Principle of Measurement
To obtain position and attitude information of an
aeroplane during every moment of its flight, the
acceleration, measured on three axes by
accelerometers, is integrated twice and the
angular rate, measured around three axes by a
gyroscope, is integrated once.
Using the equations of motion the velocity and
displacement can be calculated from the
acceleration:
v=∫a dt
r=∫v dt
And attitude is calculated from angular rate:
θ=∫α dt
Sentio32
The measurement system implemented to record
accelerations and angular
rates basically consists of
the Sentio32 platform and
a Motion Sensor Board,
which contains the actual
gyroscopes and
accelerometers. The Mid Fig. 2: Sentio32
Sweden University designed the Sentio32
hardware platform to provide a compact wireless
sensor network platform that is IEEE802.15.4
standard compatible and provides a high
performance on local processing power. The
sensors used during the measurements regarding
this investigation are the capacitive 3D
accelerometer KXSD9-2520 manufactured by
KIONIX and two gyroscopes made by InvenSense,
the IDG500 and the ISZ500.
Statistical Analysis
In order to obtain optimum results a universe of
50 measurements was build and a statistical
analysis was performed. From this data calibration
curves were elaborated.
Fig. 3: Statistical Analysis
In order to verify the capabilities of the sensor in
the second phase of the thesis flight test were
conducted in two airplanes.
Fig. 4: Test Aircraft, Citabria and Cessna C150
Basic flight test were conducted in the C150, those
consisted of rolling, pitching and yawing the
airplane. Finally actual spin tests were conducted
in the approved Citabria Campion.
Examples of the recorded data are shown in the
following figures.
The cockpit
instuments were
captured during
the basic flight
tests to have a
refference when
evaluating the
sensor data. For
the spin tests an
outside
reference was
required.
Fig. 5: Inflight recordings
The result, a complete recording of a spin can be
seen in Fig. 6 below.
Fig. 3: complete spin recording
Project Summary
This thesis deals with the improvement of a low-cost
inertial measurement unit (IMU) for aeroplanes,
which is already implemented in the Sentio32
hardware platform and mounted to a 4-seated
general aviation aeroplane. The correction as well as
the calibration of the measured data, is the main
topic of this project. The basic errors of the
accelerometers and gyroscopes are elaborated based
on a static test and can be traced back to a random
offset of the accelerometer data and a linear drift in
the angular rates. Different correction methods are
applied based on a calibration curve derived from a
set of measurements or a trend prediction from the
measurement itself. Errors induced my motion and
rotation are examined and corrected in the same way.
In the end the elaborated corrections are applied
flight tests that were undertaken during the course of
this investigation.
Project Objectives
The Main Objectives are:
- comparison of four similar sensors
- statistical analysis of the universe
- generation of a calibration curve
- evaluation of the functioning of the calibration curve
- conduction of flight tests to verify sensor data
Project Conclusion
In conclusion one can say that in the course of this
investigation findings were made that did not match
the desired outcome. Nevertheless the work done
was not in vain. The results found exclude the
Sentio32 sensor for use in the exact determination of
aircraft position and attitude during flight and
especially during spin. This is a valuable result since it
indicates that another sensor has to be selected and
installed before continuing with the inflight
measurements of spin characteristics.