SPIN REPRODUCTION, CONTROL SYSTEM TESTING AND CONTROL ROOM TRAINING USING X-PLANE

SPIN REPRODUCTION, CONTROL SYSTEM TESTING AND CONTROL ROOM TRAINING USING X-PLANE Dr. Panagiotis A. Vitsas Research Aerospace Engineer International Test Pilot School, 1 Blair Boulevard, London, Ontario N5V 3Z9, Canada 1 Tel: +3 69 93 81, email: info@vitsasaero.com Abstract Flight simulation has been seeing a growing application in flight test operations and has proven to be an effective tool in training, risk reduction and test planning improvement. This paper examines three different cases of flight simulation application to flight testing using X-Plane 1 from Laminar Research. The first of the cases examined considers the video reproduction of spins executed at the National Research Council Canada (NRC) in the CT-133 aircraft. The spins performed were reproduced in outside and inside views from various angles and combined with actual footage as well as with views of airdata and inertial data instrumentation. A short analysis on CT-133 spin radius and helix angle was also performed. The second case examined in this paper considers the testing of a pitch- and a roll-hold control system of a Boeing 77 on X-Plane. The state-space model of the airplane was used to design the system in MATLAB /Simulink which was then inserted by UDP connection to X-Plane. The time response of pitch and bank angle on X-Plane was compared to the time response of the state-space model showing satisfactory agreement and proving X-Plane s ability to test autopilot systems and aid in flight control system training. Finally X-Plane and MATLAB were used to set-up a flight test control room environment at the International Test Pilot School (ITPS Canada). Driven from an outside simulator, numerous stations and displays were connected through X-Plane and TCP/IP protocol. Overall X-Plane flight simulator was found to be an excellent aid for various phases of flight testing. Its implementation has been proven relatively easy, time efficient and cost effective and it can serve as a useful tool to both small and more extensive flight test projects. 1 INTRODUCTION It s been more than half a century that flight simulation has been used as a tool in flight testing. Since then it has been seeing a growing application [1]. The growth of computer power and the knowledge being built in the aerospace community are combined and yet flight simulators become more and more accurate improving planning, risk reduction and training in flight testing []. Flight simulators have been used successfully in many manned air vehicles flight tests [3] and are currently applied successfully in the flight testing of UAVs [, 5]. 1 present address: Vitsasaero Aerospace Consulting, Morfou 11, Patras 61, Greece 13, Dr. Panagiotis A. Vitsas 1

The simulators used for flight testing purposes are either in-house built by the various companies or ready off-the-counter software packages that are of smaller cost and can be easily adapted to the specific flight test project. In the latter category belongs X-Plane 1 by Laminar Research. X-Plane 1 is a PC-based flight simulator which uses an engineering approach called Blade Element Theory to generate the simulated environment and has been widely used as an aerospace research tool. X-Plane can reproduce flight data and allows third parties to hack inside and alter the system or input and output selected flight parameters. This paper examines the application of X-Plane in three different areas related to flight testing as they are currently used in International Test Pilot School (ITPS Canada). The first part deals with the reproduction of spins performed in the National Research Council Canada (NRC) CT-133, based upon data recorded by onboard instrumentation. The second part examines X-Plane ability to accurately test a pitch and a roll-hold system for the Boeing 77, which systems are modeled in MATLAB /Simulink. The third and final part describes the set-up of a control room training facility using X-Plane and MATLAB. SPIN REPRODUCTION ITPS Canada has an incremental approach to spin training comprising theoretical lectures and spinning on a variety of aircraft with the current progression being DA-, Pitts, Extra, T-33 with BU-131 as a qualitative evaluation. Post spinning flight, a discussion of the analysis of the spin characteristics is conducted which is backed up by X-Plane reproduction videos allowing the students to examine carefully the spin characteristics and the flight dynamics involved. The first part of this paper is based on the data collected during the CT-133 spins at the NRC in Ottawa, on 1th November 1. Some of the recorded data is presented and also used for spin analysis..1 CT-133 Research Jet The Canadair CT-133 is the result of a 1951 contract to build T-33 Shooting Star Trainers for the Royal Canadian Air Force (RCAF). The power plant would be a Rolls-Royce Nene 1 turbojet instead of the Allison J33 used by Lockheed in the production of the original T-33. A project designation of CL-3 was given by Canadair and the name was changed to Silver Star. The appearance of the CT-133 is very distinctive due to the large fuel tanks usually carried on each wingtip (Figure 1). The NRC CT-133 research airplane is equipped with a customized air data sensing system, and the NRC Flight Research Laboratory (FRL) integrated inertial reference and navigation system (FIRNS), which integrates IMU-derived and GPS-derived inertial data. The data acquisition system acquires air and inertial data at 6 Hz, optimized for atmospheric research [6]. The inertial parameters that are directly recorded are: GPS time, aircraft position and altitude, Euler angles of roll, pitch and yaw angles, angular rates, linear accelerations and velocities. Ground speed and track are derived from the aforementioned parameters. A GoPro camera was also mounted in the front cockpit above the instrument panel to record the spins.

Figure 1: NRC CT-133 used by ITPS for spin testing.. X-plane Flight Data Recorder X-Plane is sometimes used in accident investigation or re-creation, and in that black box recorded flight data must be extracted and re-formatted for X-Plane use. That format is the Flight Data Recorder (or.fdr) format. FDR file is text which means that users can make their own FDR files as easily as possible from whatever data they have and then re-create these flights in X-Plane. The file contains date and aircraft details as well as 1 data values for each time instant excluding accelerations. The flight data is used by X-Plane either for airplane movement or for instrument indications reproduction. X-plane can also create universally readable movies (.mov). The downside of these files is that they record exactly what you see when you record them. The recording specifications like frame rate and resolution can set by the user. For the FDR video however the maximum frames per second number that could be achieved was, which however was considered of satisfactory smoothness..3 Flight Data Weather data were provided by NRC and were used for the derivation of CAS [7] and indicated altitude from the GPS recorded values in order to reproduce accurately the cockpit instrumentation as seen by the student test pilots performing the spins. The resulting derived values were very close to the ones reported. In order for the flight data file to be easily manipulated in X-plane, the initial NRC file was down sampled to 6Hz. Low pass filter was not needed to be applied as no aliasing effects were noted and its application adversely affected the accuracy of some sensitive parameters such as inertial position. Attitude angle data used in the spin reproduction are seen in Figure showing an oscillatory, steep spin, while recorded rates and accelerations used in the following spin analysis are seen in Figure 3. The four spins executed tested different type of recoveries but this aspect of the tests is not examined in the present study. 3

Acceleration(g) Rotation rates(deg/sec) Angle(deg) 3 Theta Phi Psi Spin 1 3 Spin 3 Spin 3 3 Spin 1 1 1 1-1 -1-1 -1-1 3 Spin time(sec) - 1 3 Spin time(sec) - 1 3 Spin time(sec) - 1 3 Spin time(sec) Figure : Attitude angle data used in spin reproduction (Euler angles in earth axes, Theta = pitch angle, Phi = roll angle, Psi = yaw angle). 15 1 Spin 1 Roll rate Pitch rate Yaw rate 15 1 Spin 15 1 Spin 3 15 1 Spin 5 5 5 5-5 -5-5 -5-1 -1-1 -1-15 1 3-15 1 3-15 1 3-15 1 3 3 Ax Ay Az 3 3 3 1 1 1 1-1 1 3 Spin time(sec) -1 1 3 Spin time(sec) -1 1 3 Spin time(sec) -1 1 3 Spin time(sec) Figure 3: Recorded rotation rates and linear accelerations used in spin analysis, in body axes.. X-Plane movie creation The accuracy of the reproduced video s playback speed and view orientation was checked by comparing the in-cockpit ( inside ) view to the available GoPro footage. The in-cockpit X-Plane camera was adjusted in order to have the same view-orientation with the GoPro recorded image. This adjustment between the two views, took place mainly while the aircraft was on ground using as reference the horizon and other fixed objects. X-Plane was found to provide slightly greater field-of-view than the GoPro camera and this was taken into account in the scaling of the two views (Figure ). In order to synchronize the two recorded views, some moments were selected during the spin videos with a characteristic aircraft orientation relative to the horizon. It was found that initially the

produced X-Plane movie was faster than the actual GoPro video and it had to be decelerated at 8% of the initial speed in order to match it. The 8% deceleration was applied to all X-Plane movies created for this paper regardless of the viewing position. This time inaccuracy is of unknown reason and it is possibly a limitation of the X-Plane movie creation feature. The resulting video synchronization was found to be of very good quality in terms of aircraft attitude and rotation rates. Moreover, beside the created movie s view agreement to the actual horizon position, X-Plane was found to accurately reproduce other outside references including sun position and geographical characteristics such as lakes and rivers beneath the aircraft (Figure 5). A number of X-Plane movies were created reproducing the spin from various view points and combining different views; for example, an external view combined with some instrumentation parameters aiding in the detailed understanding and monitoring of the speed changes and altitude loss during the spin, is seen in Figure 6. In Figure 7 a capture of a mid-fuselage camera is seen as preferred by some pilots for a more comprehensive spin attitude understanding. The final videos successfully show the inharmonic oscillatory transfer of rotational momentum between roll and yaw, back and forth and aid in the CT-133 spin flight dynamics understanding..5 Spin analysis Except from the visual reproduction, the recorded data were used in ITPS data reduction routines in order to calculate several spin parameters and compare them to those depicted in the X-Plane reproduced movie. ITPS provides the students with a complete spin data reduction routine built in MATLAB which is used post flight to calculate all the relevant spin parameters part of which are described in [8,9]. Moreover, in order to compensate for the offset of the sensor from the aircraft CG the following matrices [1] are used: where the rotational accelerations are approximated by. For the CT-133 case, the CG offset was [-6.7, -1.,.] inches and was found to have an effect of about 1% in the derived spin parameters. 5

Figure : GoPro camera was used as reference in order to adjust the reproduced video view orientation and playback speed. Figure 5: X-Plane spin movie managed to accurately reproduce visually geographical details like lakes below the aircraft. 6

Figure 6: Movie combining an external view with in-cockpit instrumentation (Indicated airspeed, altitude and attitude indicator). Figure 7: Mid-fuselage camera view preferred by some pilots for a more comprehensive spin attitude understanding. 7

The effects of earth centripetal acceleration and the Coriolis acceleration on the recorded accelerations were taken into account but were found to be of minor importance. Also the movement of the CG during flight (due to fuel burn) and its effect on the instrumentation CG offset was considered to have an insignificant effect on the calculated parameters and was thus ignored. While the spins flown in the CT-133 are not steady, the formulas deriving the spin parameters can be used to provide an average approximate value. The derived spin radius of the CG and the spin tube helix angle for the four spins performed can be seen in Table 1 (radius values range from.37 to.73 of the wing span). The obvious difference in the values between left and right spins is attributed to the rotation of the turbine engine. Spin No Spin direction Spin radius (m) Helix angle (deg) 1 L.9 3.8 R 9.5 5.6 3 L 6.. L 6.7.3 Table 1: Derived average spin parameters. Reproducing the spin in X-Plane using wingtip traces, the spin helical motion can be seen from the outside view, as in Figure 8 for spin 3. Figure 8: Spin helical path as reproduced by X-Plane (Spin No 3). The spin path parameters are estimated geometrically from the above movie capture and are found to be of a value of 3.7 deg for the spin helix angle and.8 m for the spin radius for Spin No.3. These values are considered to be close to those derived by the analysis (less than 1 degree for the helix angle and less than 1.5 m for the spin radius) taking into account the averaging included in the nonsteady state spin analysis and the limited accuracy of the geometric estimations based on the movie capture. Generally, the reproduced spin movement as seen from the outside view, in terms of helical motion geometry, is considered to be representative of the actual spin path. 8

3 CONTROL SYSTEM TESTING X-plane is currently used by ITPS for the training of two types of automatic flight control systems (FCS): (i) trimmed flight holding systems (autopilot systems) and (ii) stability and command optimization systems. The current paragraph presents X-plane s application in the first of the aforementioned categories which regards the testing of flight holding systems. More specifically a pitch-hold and a roll-hold control system were implemented through MATLAB /Simulink in X-Plane and the generated response was compared against the airplane s known state-space response. Regarding the second of the above categories, it deals with the training of stability and command optimization systems (fly-by-wire) and uses X-Plane for visualization purposes combined with linear and nonlinear fighter aircraft flight models built in MATLAB /Simulink. This latter part is outside the scope of the present paper. 3.1 Boeing 77 and Flight Test Conditions In the present study the airplane used to test the FCS was the Boeing 77- which is a default aircraft of X-plane. For the Boeing 77 longitudinal control is obtained through four elevator segments and a movable stabilizer. The lateral control employs five spoiler panels an inboard aileron between the inboard and outboard flaps and an outboard aileron which operates with flaps down only on each wing. The five spoiler panels on each wing also operate symmetrically as speedbrakes in conjunction with the most inboard sixth spoiler panel. Directional control is obtained from two rudder segments [11]. The flight conditions that the FCS tests took place are seen in Table. These conditions were selected as data for the stability and control derivatives were available [1] and they could also be accurately reproduced in X-Plane (with test tolerance at ±1 knot and ±1 ft). A similar study for a pitch hold system of the B77 can be found in [13]. Flight Characteristic Value Weight 636,636 lbs Altitude, ft TAS 399 knots (673 ft/s) CG.5 Initial Pitch ~.5 deg Table : Boeing 77 flight conditions examined. 3. FCS and State-Space Simulink models The Boeing 77 state-space model was built on the available data for the 77-1 version. However, versions 77-1 and 77-, the latter of which is used for testing in the X-plane flight simulator, have some differences in terms of aerodynamic improvements, wing geometry and engines and thus some difference in the FCS responses is expected. 9

The flight control systems to be tested are shown in Figure 9 and Figure 1 and are typical pitchand roll-hold systems using the respective attitude angle as a single feedback. For this evaluation a zero throttle deflection input is considered for the pitch-hold control system test and a zero rudder deflection for the roll-hold control system test. The vertical and roll gyro transfer functions are considered to be equal to 1. Moreover, the servo and control mechanism blocks are modeled as a system with time response of representing the first order lag factor in the transmission of signal to the aircraft control surfaces. Figure 9: Pitch-Hold Control System functional diagram [1]. Figure 1: Roll-Hold Control System functional diagram [1]. Testing of the FCSs took place for different values of gain of the servo and control mechanisms ( ). These values were selected based on the root-locus technique in order to test the system for various stability responses. The state-space models used for the analysis were built according to [1] and are described next: 1

For the longitudinal motion the state-space arrays considering a steady throttle input are: ; ; ; ; The above state-space formulation also takes into consideration the parameter represents the forward acceleration per unit change in speed as described in [1]., which For the lateral-directional motions, the state-space arrays due to ailerons only are the following: ; ; ; ; The above state-space models and flight control systems were inserted in Matlab /Simulink in order to produce the FCS responses to be used as reference for comparison to X-plane s response. The Matlab /Simulink model for the pitch-hold system is seen in Figure 11. The roll-hold system model was built in a similar manner. 3.3 X-Plane setup In order to test the FCS in the X-Plane, the state-space block of the Simulink model is removed and is replaced by X-Plane. The communication between Simulink and X-Plane is established via UDP protocol, which is a network protocol that both software support for their communication. The UDP transfer rate was set at 1 packets per second using the xpc Target toolbox blocks that automate UDP packing and unpacking. The resulting Simulink model is seen in Figure 1 for the pitch hold case, while the roll hold case is built in a similar manner. The setup involved two desktop computers connected in a LAN, one running the MATLAB /Simulink model and the other the X-Plane. 11

57.3 de rad to deg1 Elevator deflection (deg) u u+u1 -K- u_response Speed Bias ft/s to knots Speed time response (knots) 1/57.3 Reference theta (deg) deg to rad -1 s+1/.1 FCS Transfer Fcn x' = Ax+Bu y = Cx+Du 77- State-Space model u theta Output Saving to Workspace Vertical Gyro Gain 1 theta 57.3 rad to deg theta_response Pitch time response (deg) Figure 11: Simulink pitch hold control system. 57.3 de rad to deg Elevator deflection (deg) Ref pitch commaned (deg) Manual Switch1 -Kdeg to rad -1 s+1/.1 FCS Transfer Fcn UDP -Krad In1 Out1 Send to defl ratio Pack X-Plane UDP signal Send to Binary X-Plane Constant1 UDP In1 Receive Binary Receive from X-Plane Terminator Vtas q theta alpha Time time Unpack X-Plane UDP signal Mssn time (sec) U Speed response (ktas) q q response (rad/s) alpha theta AoA response (deg) 1 Trim pitch (deg) theta Ref Pitch response (deg) Figure 1: Simulink model coupled with X-plane for the pitch hold case. 1

The control surface deflections were sent to X-Plane via the artificial stabilizer input channel so that they did not interfere with the joystick commands. Also the input commands were in sync to the surface deflections and their deflection ratios were the same throughout the tests. A difference between the state-space model and X-Plane model that had to be resolved was that the commanded aileron deflection angle ( ) for the state-space model was the total differential deflection of right and left inboard ailerons. With that in mind, the aileron roll command sent to X-Plane was divided by two in order to match the aileron deflection of the state-space model. Moreover, the default X-Plane Boeing 77- does not model the roll control realistically, as all available surfaces are used for all flight conditions. This issue was corrected by editing the existing airplane in Plane Maker. Plane Maker is a program bundled with X-Plane that lets users design their own aircraft or modify an existing one and was used in order to set only the inboard ailerons and outboard spoilers for high speed roll control, as in the real case. (Recently a corrected version of B77- has been available in the X-Plane community [15].) 3. Results and Analysis The results of the FCS tests and a short analysis are presented in this paragraph. For both the pitch- and roll-hold systems three different gain values were tested known to provide stable and unstable responses according to root-locus analysis. For the pitch-hold case the gain values tested were -1, 1 and -5, while for the roll-hold case the gain values were 5, -5 and respectively. For the stable gain case, the commanded reference angles were, and 6 deg. Figure 13 and Figure 1 refer to the pitchhold case with stable and unstable gains respectively, while Figure 15 and Figure 16 refer to the roll-hold case. As it can be seen from the plots, for the stable gain cases, the X-Plane transient response is in very good agreement for both pitch and roll hold control systems. The steady state error was also less than.3 deg for all cases examined. While the X-plane response sensitivity on the initial trim conditions was significant, all the results could be perfectly reproduced with accurate trimming. The difference seen in the pitch response curves of Figure 13 could be attributed to the nonlinear way X-Plane models the flight dynamics taking into account the speed and altitude change, while the state-space model s linear behavior considers steady flight parameters throughout the test. For the unstable gain cases, the agreement between the two responses is not that good. For the exponentially unstable cases of Figure 1(a) and Figure 16(a) the X-Plane lags in relation to the statespace response, while for the oscillatory unstable cases of Figure 1(b) and Figure 16(b) X-Plane does not manage to show the state-space predicted behavior. This disagreement is attributed to the control surface deflection limitations posed by X-Plane and which are not taken into account by the state-space model. The control surface deflection limitations set in X-Plane are 1 deg down and deg up for the elevator and the ailerons and deg for the operating spoilers. These limitations do not let the airplane get into growing instabilities and result either in a stable response or in continuous steady amplitude oscillations. 13

Theta (deg) Theta (deg) Theta (deg) Theta (deg) Theta (deg) 6 SS X-plane 6 SS X-plane 5 1 15 5 3 Time (sec) (a) 5 1 15 5 3 Time (sec) (b) 6 SS X-plane 5 1 15 5 3 Time (sec) (c) Figure 13: Pitch angle time responses with a stable FCS gain of deg and (c) 6deg. =-1 for reference angle of (a) deg, (b) 1 K ct =1 SS X-Plane 6 5 K ct =-5 SS X-Plane 3-1 1-6 8 Time (sec) (a) -1 5 1 15 5 Time (sec) (b) Figure 1: Pitch angle time responses with unstable FCS gains for reference angle of deg [(a) =-5]. =1 and (b) 1

Phi (deg) Phi (deg) Phi (deg) Phi (deg) Phi (deg) 6 SS X-Plane 6 SS X-Plane 1 3 5 Time (sec) (a) 1 3 5 Time (sec) (b) 6 SS X-Plane 1 3 5 Time (sec) (c) Figure 15: Bank angle time responses with a stable FCS gain of deg and (c) 6deg. =5 for reference angle of (a) deg, (b) 5 K ct =-5 SS X-Plane 6 SS X-Plane K ct = -5-1 -15 - - 5 1 15 Time (sec) (a) - 6 Time (sec) (b) Figure 16: Bank angle time responses with unstable FCS gains for reference angle of deg [(a) =]. =-5 and (b) 15

CONTROL ROOM TRAINING The third and last part of this study deals with control room training. It is well known that telemetry control room is a basic part of many manned vehicle flight test operations and of all UAV flight tests. A control room is defined as any facility on ground or airborne that establishes a two-way communication with the tested flight vehicle and monitors in real time the flight and the information collected. Training of the control room personnel is of great importance and has proven to be a necessary part of many recent flight tests like the NASA Hyper-X program [16]. ITPS provides control room training during its flight test courses aiming to get the students familiar with control room crew resource management, communication terminology and procedures as suggested by both military and civilian documentation [17, 18]. Publicly available documentation of the Edwards AFB on control room personnel training and evaluation is also considered in the course [19]. The control room set-up currently used by ITPS employs an F-16C modified X-Plane simulator which communicates through TCP/IP protocol with a separate control room, driving a variety of visual displays and strip charts. The control room set-up and connectivity concept is seen in Figure 17. Voice communication is also included on a dedicated network with the instructor of the control room operating a push-to-talk microphone for transmission to the flight simulator room. Figure 17: Control room setup and connectivity using X-Plane and MATLAB. 16

For the current set-up the control room contains an Instructor Operating Station (IOS), a number of computers as ground control stations running X-plane and MATLAB simultaneously and two overhead displays. The IOS controls the whole flight being able to change any conditions or introduce failures and emergencies. It also drives one of the overhead displays showing a global real-time interactive map. The data transfer from the flight simulator to the ground control stations is done through X-Plane which outputs the data in a text file and MATLAB reads and plots the desired parameters in real-time strip-charts (Figure 18). Each ground control engineer can select before the beginning of the flight which parameters to be shown on the respective ground station. The total number of parameters exported by X- Plane is 8 at a transfer rate of 1Hz, which is considered enough for training purposes and basic post flight data analysis. An additional number of derived parameters are calculated real-time using MATLAB routines and plotted on strip-charts along with predefined limits. Future development in MATLAB will include a greater variety of data display types like gauges, digital counters and scales in order to expose the ground personnel to different visualization options. Figure 18: MATLAB strip-chart sample seen on the ground control stations. The second overhead display can be used for various flight visualizations including HUD replay and flight external views. The present X-Plane also set-up allows the simulator to be started at any point in a flight profile in order to focus on a particular phase of the mission. The only limitation encountered regarded a refresh rate in the MATLAB plots of around.hz which was due to filled CPU resources by running X-Plane and MATLAB simultaneously; however this limitation will be surpassed in future development of the set-up, where UDP protocol will be used instead of TCP/IP and this will not require X-Plane running on the ground station computers. In conclusion, the above described set-up proved to be a cost-effective and efficient training tool for ITPS, which successfully introduced the student test pilots and flight test engineers to control room operations. 17

5 CONCLUSIONS X-Plane flight simulator by Laminar Research is known to be a useful aerospace research tool. Throughout this paper some of its applications regarding flight test aspects have been demonstrated, such as they are currently implemented in ITPS training, applicable to both manned and unmanned air vehicles. The first part of this paper examined the spin reproduction in X-Plane using NRC CT-133 data. Building up a Flight Data Recorder file with all the required flight parameters by downsampling the NRC data, high-resolution movies could be created from various view angles. Synchronizing the resulting movies with actual in-cockpit spin footage available from an onboard GoPro camera, the agreement on attitude, rotation rates and outside references during the spins was found of excellent quality. Outside view movies were found very helpful in the understanding of the spin dynamics and the representation of the spin radius and helix was found to be in acceptable qualitative agreement to the mathematical analysis. The present study shows that with the available instrumentation, X-Plane can be used successfully to accurately reproduce spins and aid in spin understanding. Flight control system testing of attitude-hold systems was the second part examined in this paper. A first order lag pitch-hold and a roll-hold FCS modeled in MATLAB /Simulink were coupled with X- Plane though UDP protocol. The set-up was tested for various gains and reference angles on the Boeing 77 for a high speed cruise condition. The resulting time responses were compared to the analytic statespace model of the aircraft that was built on known aerodynamic coefficients. X-Plane flight model was found to predict satisfactory both the transient response as well as the steady state for steady gain values for both the pitch and roll case. For the unsteady gain values, the responses showed partially a qualitative agreement, while their differences were attributed to the restriction of the control surface deflections modeled in X-Plane which is ignored in the state-space models. Further testing is required to enforce the findings for other flight conditions and aircraft types. However, initial findings show that X-Plane can provide realistic responses of FCS performance, contribute in autopilot design and aid students in practical understanding of FCS design principles. The last part of this paper described the set-up of a control room training facility with the aid of X- Plane and MATLAB currently under operation at ITPS Canada. An F-16C flight simulator was connected through TCP/IP protocol to a control room driving several ground station strip-charts and overhead displays with map position and flight visualization, while an instructor station controlled all flight conditions and failure/emergency cases. The set-up worked satisfactorily introducing test pilot and flight test engineer students successfully to control room operations. Overall X-Plane 1 by Laminar Research was found to be an excellent aid for various phases of flight testing. Its implementation was found easy, time efficient and cost effective proving that it can serve as a useful tool to both small and more extensive flight test projects. 6 ACKNOWLEDGEMENT Credits are to be given to Laminar Research for granting permission to use X-Plane and publish the results of the present study. The author would also like to express his appreciation to the NRC and research test pilot Anthony Brown for providing all the requested data and constructive guidance in the spin study. 18

7 REFERENCES 1. Simulation in support of flight testing, RTO-AG-3-V19, NATO Science and Technology Organization, September.. Cottling, M. C., McCue, L. S., and Durham, W. C., Simulator-based flight test engineering as a capstone to the dynamics and control curriculum, 5th AIAA Aerospace Sciences Meeting and Exhibit, January 7, Reno, Nevada. 3. Evans, M. B., and Schilling. L. J., The role of simulation in the development and flight test of the HiMat vehicle, NASA Technical Memorandum 891, NASA Dryden Flight Test Research Facility, Edwards, California, 198.. Sorton, E. F., and Hammaker S., Simulated flight testing of an autonomous unmanned aerial vehicle using Flightgear, Infotech@Aerospace, Arlington, Virginia, September 5. 5. Ribeiro, L. R. and Oliveira, N. M., UAV autopilot controllers test platform using MATLAB /Simulink and X-Plane, th ASEE/IEEE Frontiers in Education Conference, Washington, October 1. 6. Brown, A. P., Beyers, M., and Bastian, M., Dynamic stall flight data, effects of pitch rate, surface roughness and ground effect, 6 th AIAA Aerospace Sciences Meeting and Exhibit, January 8, Reno, Nevada 7. The standard handbook of aeronautical and astronautical engineers, edited by Mark Davies, McGraw Hill, New York, 3. 8. Soule, H. A. and Scudder, N. F., A method of flight measurement of spins, NACA Report no. 377, Langley Field, 193. 9. Phillips, W. F., Mechanics of flight, Second Edition, Wiley, New Jersey, 1. 1. International Test Pilot School, Performance and Stability Notes, 1. 11. Heffley, R. K. and Jewell, W. F., Aircraft Handling Qualities Data, NASA CR-1, December 197. 1. Roskam, J., Airplane Flight Dynamics and Automatic Flight Controls; PART I: Chapters 1 Trough 6, Kansas, USA: Roskam Aviation and Engineering Corporation, 63-6, 198. 13. Jenie, Y. I., and Indriyanto,T., X-Plane-Simulink Simulation of a Pitch-Holding Automatic Control System for Boeing 77, nd Regional Conference on Aerospace Science, Technology and Industry, Taiwan, 7. 1. Jenie, S. D., and Budiyono, A., Automatic Flight Control Systems, Bandung Institute of Technology, 6. 15. XP1 77- Corrected 1., http://forums.x-plane.org/index.php?app=downloads&showfile= 1817, accessed on nd September 13. 16. Lux-Baumann, J. R., Dees, R. A., and Fratello, D. J., Control Room Training for the Hyper-X Program Utilizing Aircraft Simulation, NASA/TM-6-13685, November 6. 17. Test Control and Conduct, AFFTC Instruction 99-5, Edwards Air Force Base, California, 1 May. 18. Telemetry Control Room and Radio Communications, SFTE Reference Handbook, 3 rd Edition, 13. 19. Flight Test Control Room Personnel Training and Evaluation, AFFTC Instruction 99-8, Edwards Air Force Base, California, 1 May. 19

8 BIOGRAPHY Dr. Panagiotis (Panos) Vitsas was born and raised in Agrinio, a small town of Western Greece feeding his passion for aviation through aircraft model building and early flight simulators. Since then Panos has received an engineer diploma in Aeronautics from the University of Patras in Greece; an MSc in Human Factors and Safety Assessment in Aeronautics from Cranfield University in UK; a PhD in Aeronautics from the University of Patras and is an International Test Pilot School graduate. During his academic studies he has received numerous sponsorships and awards and he has published papers in several journals and conferences. Former employment positions include A-7 Corsair engineer in the Hellenic Air Force, research engineer in European rotorcraft projects, academic lecturer on airvehicle design and aerodynamics and design consultant of an unmanned airship. Today he is the owner and a senior consultant of Vitsasaero- Aerospace Consulting providing research and consulting services to a wide range of aerospace areas. In the same time Panos is an assistant lecturer at the University of Patras and an instructor at International Test Pilot School. He is as SFTE Member, an AIAA Senior Member and a private pilot.