Creating an Aircraft Handling Qualities Simulator for the USAF Test Pilot School

Transcription

1 U.S. Air Force T&E Days February 2009, Albuquerque, New Mexico AIAA Creating an Aircraft Handling Qualities Simulator for the USAF Test Pilot School William R. Gray, III USAF Test Pilot School, 220 S. Wolfe Ave, Edwards AFB, CA, 93524, USA Jay Christian Kemper Calspan Corporation, 220 S. Wolfe Ave, Edwards AFB, CA, 93524, USA Test pilot school graduates must have a solid understanding of aircraft flying qualities and handling qualities. The USAF Test Pilot School has traditionally used classroom instruction followed by in-flight demonstration of test techniques to provide graduates with this understanding. Flying qualities simulation has traditionally seen little use at test pilot schools because high-quality simulators are exclusively designed to re-create particular aircraft. This type of simulator, while valuable for pilot training, adds little to a TPS curriculum except a reduction of fidelity and elimination of motion cues it can only show how a particular aircraft behaves. Recent advances in desktop computer technology and electrically-driven inceptors have allowed the authors at the USAF TPS to create a handling qualities simulator that can used as a flying qualities laboratory, a handling qualities research device, and a force multiplier for increasingly valuable curriculum sorties. Built by the authors with off-the-shelf hardware, USAF-owned visual display technology, and opensource software, the TPS flying qualities simulation laboratory can provide everything from complex demonstrations on variable stability platforms to unsupervised operation of almost unlimited aircraft types. It can also be linked with control rooms for test team training. This low-cost platform is revolutionizing flying qualities instruction at the USAF TPS and may be a valuable addition to other academic institutions. I. Introduction Simulators are unquestionably effective training devices. Long before visual displays, motion cues, and high model accuracy, simulators were reducing the need for flight training hours by helping pilots practice cockpit procedures and basic flying skills. Simulators have also provided the ability to realistically train pilots on emergency procedures that simply cannot be safely duplicated or practiced in-flight. 1 Aircraft simulation now spans the typical pilot s experience, from home PC-based simulators to FAA-certified Level D airline simulators, and are very heavily used in every modern military. In spite of this pervasive penetration, test pilot schools remain relatively simulator-free. This certainly doesn t imply that simulators are not used in flight test quite the opposite is true! Aircraft development projects rely heavily on aircraft simulation and it is fair to say that modern test pilots conducting envelope expansion testing on current generation military aircraft spend much more time in the simulator than in the air. The particular requirements of a test pilot school, especially the need to get the student test pilots as much varied in-flight experience as possible, and the wide variety of aircraft typically flown during the course make it impractical to obtain and maintain the aircraft-specific simulators that are the norm in the operational environment. Test pilot schools are also much less interested in procedural training and basic flight training than a typical flight training organization. Indeed, test pilot schools are much less training environments than educational environments. Their students already know how to fly so the school must concentrate on improving their academic understanding of the aircraft and pilot as a system. Finally, the precision flying and feel of the aircraft required for a valid TPS demonstration can only be practiced and experienced in actual flight motion bases and mathematical approximations are simply too limited or inaccurate for developing the student s Chief Test Pilot, USAF Test Pilot School, 220 S. Wolfe Ave, Edwards AFB, CA, 93524, AIAA Member. Flight Research Engineer, Calspan Corporation, 220 S. Wolfe Ave, Edwards AFB, CA, of 10 This material is declared a work of the U.S. Government and American is not subject Institute to copyright of protection Aeronautics in the and United Astronautics States.

2 powers of observation sufficiently for open-air testing. In spite of rarely using ground simulators, test pilot schools have long incorporated a special type of simulator; variable-stability research aircraft. These in-flight simulators (essentially airborne motion platforms) can create very flexible and high-fidelity simulations and are without question among the most cost-effective though costly tools in a test pilot school s curriculum. The cost structure of simulators has changed dramatically since their first use before World War II. The most important change, of course, has been the advent of the digital computer followed by the exponential rise in computing power accompanying a rapid decline in cost. Sometime in the late 20th Century a dramatic change occurred the cost of creating the information presented by the hardware fell below the cost of creating the hardware to present the information. When the principle author attended pilot training in the early 1980 s it took a roomful of computers to create the then cutting-edge visual displays for each cockpit. The cost of the simulator cockpits, instruments, and structure was high but was relatively small when compared to the cost of the processing power needed to create a real-time pilot experience. a But once the cost of creating the aircraft model and visual displays fell (along with the cost of all digital computing) it became more expensive to create the realistic instruments than the system to drive them. Many simulators have now completely jettisoned hardware-realistic instruments in favor of rendered instruments the realism just isn t worth the cost for many training situations, and touch-screen displays provide sufficient interactivity. This revolution in simulator technology has completely changed the landscape for ground-based simulation at the USAF Test Pilot School. It has allowed the authors to create a flexible simulation capability that, at a cost of less than 5% of the estimates of just a decade ago, can provide a single tool to simulate multiple types of aircraft, provide academic demonstrations of aircraft stability and control, allow students to experience and explore changes in an almost unlimited number of aircraft characteristics, and practice flight test techniques (FTTs) with realistic control room interaction. While the simulator is not intended to replace any aircraft sorties, early experience strongly indicates that the simulator acts as a sort of force multiplier, making the already limited flight exercises much more effective. II. Simulator Requirements The USAF Test Pilot School has long recognized the advantages simulators provide and dreamed of effectively incorporating them into the curriculum. Indeed, limited simulation has been a part of the TPS student experience almost as soon as it was available at Edwards AFB. Sometimes the only way that students can be exposed to aircraft is through simulation for many years students were flown to Denver, Colorado specifically to spend a few hours in the United Airlines training simulators. Over the years, the TPS has had a few small flying/handling qualities simulators that allowed the staff and students to change a basic aircraft model to investigate various equations of motion or control characteristics. While effective, these simulators were typically difficult to work with and required a level of instructor expertise that was not sustainable. They were also quite limited, having little or no inceptor feedback and flexible but difficult-to-modify aircraft models. After being hired at the TPS as the first civil service flight instructor, the lead author began an initiative to create a locally-designed PC-based simulator. He hoped that the available technology would allow an open-source or government-source simulation capability that could meet the specific requirements of the Test Pilot School. Definition of those requirements was the critical step because any attempt to create a realistic simulator in the commonly-accepted sense would be prohibitively expensive and far too limited. The USAF TPS does not need an F-16 simulator or a T-38 simulator or, really, a simulator of any particular airplane. It needed a simulator platform that could duplicate as wide a variety of aircraft characteristics as possible. It needed a variable-stability, variable-feel, variable-interface simulation laboratory that could be operated by a student or instructor without the aid of a simulator operator. It needed something so unique to the test pilot school environment that it had to be built from the ground up. Requirements definition began with imagining a variety of uses for the platform. Each use brought its own set of characteristics but the most important characteristic was, and remains, flexibility. Equations of Motion Laboratory: Direct experience is a powerful tool to help students gain a more intuitive understanding of the aircraft equations of motion (EOM) and how changes in the aircraft characteristics effect stability and control. This laboratory would use a simple graphical a This analysis does not account for the cost of a motion platform. Although the cost of motion platforms has fallen, their expense has not fallen in parallel with the rest of the simulator technology. Building the infrastructure and purchasing a motion base would increase the cost of a TPS simulator by an order of magnitude and is not currently under consideration. 2 of 10

3 user interface (GUI) to allow the student or instructor to adjust variables within the model so the student could directly experience how these changes effect the aircraft. Requirements: A baseline aircraft simulator (controls/model/display) with acceptable handling qualities. Run-time modifiable EOM variables. Straightforward GUI. Sufficiently detailed displays to create clear visual cues and head-up displays. Variable Feel System Laboratory: The controllability of an aircraft is highly dependent upon the feel system, whether provided artificially (called non-reversible controls) or through aerodynamic forces (called reversible controls). A variable feel system laboratory must simulate a wide variety of control characteristics and these characteristics should be easily changed or adjusted by the student or instructor. Additional requirements: Actuated controls (stick, rudder pedals, throttles) adjustable from outside or within the aircraft model and capable of simulating stable and unstable behavior. Provision for an actuated side stick and center stick. At least two actuated throttles. Select-an-Aircraft Playground : One of the most important elements of a test pilot school is the so-called qualitative-evaluation program, designed to directly expose students to a wide variety of aircraft characteristics. This is done in the best way possible by putting them in the cockpits of a wide variety of real aircraft. But these flights are limited by cost, availability, schedule, and safety. There are many aircraft that have a lot to show a student test pilot but simply cannot be scheduled, such as lifting bodies, spacecraft, and many other historically and/or technically important examples. An easily reconfigurable simulation platform would give the students the opportunity to explore some of these aircraft. New requirements: A collection of simulations. The ability to interact with a variety of programming languages. Cockpit controls with as much flexibility as possible. A simple GUI to minimize the effort in changing models. Flying and Handling Qualities Test Technique Trainer: A test pilot must be able to successfully evaluate the handling qualities of an aircraft. Learning this discipline requires more than an academic understanding of the subject it requires experience. Prior to 2008, almost all hands-on training for these flight test techniques was accomplished in-flight including all training on the very simplest FTTs meaning that precious flight time was wasted learning techniques that could easily be taught in a simulator. More importantly, there was not enough flight time for most students to fully appreciate complex handling qualities FTTs. New requirements: Instructors and students must be able to easily progress through lesson plans encompassing a wide variety of aircraft characteristics. The ability to show data in real-time so the student can see the results of the FTT. Make minimization of time delay the first priority input-to-response delay should be no greater than the display refresh rate. Flight Test Team (Control Room) Trainer: Simulation has become an important part of premission preparation. In the flight test world, the simulator does not end with the cockpit. In a flight test environment the simulation often includes the control room sometimes the actual control room that will be used for the flight test. In spite of the extraordinarily effective real-world practice of using simulators to prepare for complex test points, control room simulators are not used to train the flight test engineers attending the USAF TPS. And these engineers that will one day run real-world control rooms. Modern control room technology (including the control rooms at the USAF TPS) make it relatively easy to connect a simulator to a control room. New requirements: 3 of 10

4 Connect the simulator and the TPS control rooms. The ability to use the actual control room setups that will be used for TPS flight test missions. There is no requirement for a motion base for the TPS simulation facility there insufficient justification for the substantial cost. Although motion has always been associated with the very best simulators, there is very little indication that the motion cues provided by even the best hardware improves pilot training. After conducting a literature survey on the effectiveness of simulator motion in 2006, McCauley concluded:...that there is a substantial body of data to support the training effectiveness of flight simulation in general; that there is virtually no evidence to support the training effectiveness of motion platforms; that motion contributes to in-simulator performance, particularly for experienced pilots; that motion cues may be beneficial for flight training in unstable aircraft and in tasks involving disturbance cues, although the evidence is weak; and that motion, noise, and vibration contribute to the realism of the simulation and, therefore, strongly influence the acceptance of a simulator by the pilot community. There is no reliable evidence that a motion base prevents simulator sickness. Instructional design is more important than physical fidelity for training effectiveness. 2 The simulator was developed in a modular fashion using ethernet networking and off-the-shelf hardware and software. Custom code was created only when necessary. At the onset of the project, it seemed unlikely that there would be a full-time developer on staff so the guiding ethic was keep it simple, stupid! The first iteration was a simple desktop simulator built around an MATLAB R Simulink R aircraft model b with a cheap desktop game control stick and 17 LCD display. This simulator was used to prove the value of a variable-stability simulator for instruction in the equations of motion. The current version now nearly meets the initial goals of the project and the path to meeting them is clear. III. Configuration A. Hardware The simulator is built from a variety of off-theshelf components connected via an ethernet network through a single hub. Figure 1 depicts the basic setup and connections. This architecture has proven quite flexible. Hardware may be added with relative simplicity; for instance, the control room software used for the data displays at the engineer s station came out-of-the-box ready for an ethernet connection. Getting the data to the data display was simply a matter of formatting the data correctly at the simulation and packaging it as required for the control room software. A back-up system was also added using a network connection and open-source configuration control software. Building the system in a modular fashion has allowed and encouraged a spiral development process, where basic capabilities are achieved and additional capabilities are added when time and money is available. Each module is itself under constant revision as better tools become available or school requirements change. In the next year, the authors hope to add another visual display channel (center, above the primary window ) and a fully reconfigurable touch-screen LCD panel display. Figure 1. Hardware Configuration b Starting with the Navion model as included in the Aerosim Blockset provided without charge by Unmanned Dynamics, LLC to academic or non-commercial users. ( 4 of 10

5 B. Computers The USAF Test Pilot School seems to have a glut of desktop computers. Between regular hardware refresh cycles and staff/student turnover there are always a few fully functional computers awaiting disposition. With the exception of the visual display system, the processor power required for each simulator computer is significantly less than that needed for a current desktop computer concurrently running multiple Microsoft R applications on Windows XP R. With the exception of the visual display generators, every computer on the simulator network is at least one generation behind 2008 desktop standards but none of these computers are running anywhere near their processor or memory capacity. This arrangement is not particularly standardized or attractive, but the operating system (typically Linux) and software is hardware-agnostic and the differences have proven completely superficial. This arrangement is also quite inexpensive. Of the seven desktop PCs in use, only three were specifically purchased the rest were re-used surplus. The three computers purchased for the visual display system were required to run high-end graphics cards. These cards produce the full-color, full pixel resolution and full rate visual display on 30 inch LCD computer monitors. The resulting images are very smooth and very realistic. The simulator was designed primarily as a flying and handling qualities instructional device so refresh rates and pixellation were of critical importance. The refresh rate is currently set at 50 Hz (at the model refresh rate). This is less than the monitors are capable of but much better than television (30 Hz) or cinema (24 Hz). C. Software The simulator software is built around a few guiding principles: 1) Use open-source or government source where possible, 2) Do not gold-plate, the basic problem is difficult enough, and 3) Meet doubt with flexibility we don t know what future users will want to change in the model so, within reason, make everything as flexible as possible. In its original incarnation, almost everything was written in MATLAB R Simulink R thanks to the limited skill-set of the programming team of one the principle author. With the addition of Mr. Kemper many software doors were opened and C++ became the primary programming language for creating the interfaces necessary between the various software packages run on each machine. As depicted in Figure 2, the simulation takes inputs from the controls and returns feel system characteristics. The Figure 2. Software Interactions simulation system simultaneously sends aircraft position and attitude to the visual display system, which returns the altitude above the surface. (This value is best computed at the visual display system to ensure that there is no error for computing takeoff and landing surface interactions.) The center visual display computer is used as a controller for all of the visual display computers all scene generation is created by the SubrScene IGS c. Selected aircraft parameters are sent to the control room software for data display creation. The data display software is an outstanding example of the school s reuse of suitable software. The Air Force Flight Test Center (AFFTC) standard for control room processing and displays is Symvionics, Inc. s IADS R software d. This software is free for the TPS under the AFFTC IADS R contract and ensures that students are trained in the very displays they will be using at the school and that most will be using after graduation. It is also far more powerful than the Test Pilot School requires, giving curriculum developers and instructors many potential tools to enhance the educational value of the simulator time. Employing open source and government furnished equipment (GFE) software (in the vernacular: free ) has allowed software purchases to be limited to a single installation of MATLAB R Simulink R. The authors are currently looking at other GFE options for the visual display system and are in the process of purchasing the software necessary to create the heads-down displays. c d 5 of 10

6 Figure 3. Some Visual Display Depictions Creating the simulator has been much more than just choosing and assembling components. The authors have had to solve numerous communications issues, create GUIs that bridge the gap between complexity and ease-of-use, create and/or modify aircraft models to meet the needs of the TPS, and occasionally invent new methodologies. The TPS curriculum has a flight control system design project in which the students modify a few important gains in a pitch feedback system, fly the system in a very rudimentary simulator, then fly their design in a Calspan variable-stability Learjet. The ground simulation portion of this project is being transferred to the new simulator. While working on that problem, Mr. Kemper devised a method to use an external model for the control laws that would allow the students to design their system on their laptop computers then connect their computer to the simulator network and run their Simulink R block diagram in real time on the simulator hardware. This exciting capability has created the opportunity for students and instructors to create and test models on their desktop or laptop computer then easily drop their model into the simulator for evaluation. Figure 4. Current Simulator Layout 6 of 10

7 D. Aircraft Models One of the most important capabilities of the USAF TPS handling qualities simulator is the flexibility to be quickly reconfigured to simulate a wide variety of aircraft. This capability has been demonstrated by making several models available, including an F-15, F-16, T-38, variable-stability T-38, and Blanik glider. All of the models except the Blanik are simulating aircraft with artificial control feel that changes very little with the aircraft state. The glider has a reversible control system the control feel is generated as part of the model using hinge moment theory. This particular feature is very unusual in a simulator it may be unique to the USAF TPS simulator and creates numerous additional educational opportunities since the all of the hinge moment coefficients may be changed while the simulation is running. IV. Curriculum Development The most important reason to create a simulator at the USAF Test Pilot School was to support the curriculum. Although the simulator was used in several student test projects, experimental curriculum events, and research projects during initial development, the first application in the full curriculum was added in mid The simulator is exceeding expectations. Prior to creating a curriculum, the simulator had to have a few critical tools. First, an aircraft model with sufficient variability was necessary. A copy of the current T-38 model used in USAF simulators was acquired at no cost through AF Research Laboratory channels (the simulator is written in ADA, a language that has reached the status of venerable ) and a C++ interface to the model was written. The model was also modified to allow the real-time modification Figure 5. Student Test Pilot and Flight Test Engineer Stations of all stability derivatives, inertial states, and configurations. All of these parameters may be adjusted through a simple GUI while the model is running in real-time. An additional GUI was developed to allow the instructor to easily step through initial conditions so that curriculum events could be designed in advance and changes made with minimal interruption. With these tools in place, the variable-stability T-38 could be adjusted until the desired effect was achieved and the configuration and initial condition saved for inclusion in the simulator. Most instructors for student simulators will be quite incapable of going beyond the GUI created for the simulator session but, should they be so inclined, may still adjust the aircraft model as required for the student s understanding. The TPS does not have dedicated simulator instructors so the simulator must be as instructor-friendly as possible. A. Flying Qualities A TPS curriculum must provide its graduates with a solid understanding of basic aircraft stability and control, starting with the equations of motion and major stability derivatives. Some of the most important work a test team is expected to conduct is verification of the safe flying qualities during initial envelope expansion. The first curriculum application of the simulator was the flying qualities introduction simulator (also known as FQ1 ). The hour-and-a-half simulator event was created to introduce the students to openloop flying qualities flight test techniques, demonstrate the effects of changing the most critical stability derivatives, and introduce the flight test engineer students to control room displays. The simulator follows a basic pattern of teaching how to accomplish an FTT then using that FTT to investigate how the aircraft 7 of 10

8 static and dynamic stability changes with stability derivative variations. One particularly important part of the simulator is the initial demonstration of each FTT by back-driving the flight controls so the student may experience a perfect maneuver and see the data that result from that maneuver. The simulator is designed for a two-student team; a pilot at the controls and a test engineer at the control room station. 5 depicts how the displays are arranged so that the both students can see their partner s displays and the engineer can directly observe the pilot s inputs and out-the-window display. (The students switch seats for the last part of the session so the engineer can practice the inputs and the pilot can concentrate on the control room displays.) Prior to the addition of this simulator session, students had to be taught how to conduct these FTTs during their initial flying qualities demonstration sorties. Initial feedback from these flight instructors has been exceptionally positive because they can now concentrate on the characteristics of the aircraft, not teaching students how to correctly conduct FTTs that are quite simple once experienced. Student understanding of the implications of the equations of motion has also improved. The success of this initial curriculum simulator has solidified support in the school for the simulator as a tool to maximize the effectiveness of curriculum sorties. Figure 6. Variable Stability T-38 Instructor/FQ Simulator GUI Windows B. Handling Qualities One of the more complex facets of the TPS curriculum is aircraft handling qualities evaluation. Testing and evaluating the feel of an aircraft and its suitability for the many tasks it was designed to accomplish through direct pilot control, requires a solid foundation of stability and control and a considerable amount of exposure to different handling qualities characteristics and flight test techniques. Since handling qualities may be evaluated for any number of things that do not fly cars, video games, and bicycles for instance using a simulator to instruct the basics of handling qualities evaluation can be very effective. As with flying qualities, flight hours are precious and any preparatory work that can be done on the ground makes the flight hours far more effective. With the basics out of the way, students can concentrate on the peculiarities of aircraft handling qualities. e e There is very little appreciation amongst the pilot community of the extent to which the most difficult flying tasks are completed almost entirely with unconscious actions. After sufficient training, most pilot control inputs result from motion cues 8 of 10

9 The handling qualities simulator instruction will occur in two blocks totalling about four hours. The first simulator will concentrate on helping the students understand basic concepts, like how pilot effort or reduced tolerance for error can destabilize the closed-loop pilot/aircraft system and how to conduct basic handling qualities evaluations. In order to keep the concepts relatively simple, aircraft and flight control models are used that do not employ stability and control augmentation. The first handling qualities simulator will also prepare the students for their first NF-16D VISTA demonstration sortie. The second handling qualities simulator will concentrate on the various ways that stability and control augmentation can create unexpected and/or hazardous situations; this session will help prepare the students for their second NF-16D VISTA sortie. C. Research Projects and Test Management Projects The USAF Test Pilot School conducts a variety of research projects either as student education through the Test Management Project (TMP) portion of the curriculum or through independent staff research projects. The first incarnation of the simulator a simple desktop system was used in 2004 to study the boundaryavoidance tracking hypothesis. 3 After incorporating electrically driven cockpit controls the simulator lab supported a TMP that used a glider to study dynamic soaring and the coefficients of drag for control surface displacement. 4 This particular project involved using data from a previous TMP to create a Blanik L-33 model then designing an ethernet connection for a handheld flight test aid so the students could practice their maneuvers and data gathering techniques in the simulator. A TMP currently in the planning stages will use the simulator to evaluate a variety of strategies intended to use actuated inceptors to provide limit protection in lieu of flight control system limiters. This project will include modelling a Calspan variablestability Learjet and feel system. Several other projects, including unmanned aerial vehicle handling qualities optimization, high fidelity model-following flight control systems, and boundary-avoidance tracking research, are in the concept development or planning stages. D. Areas for Growth With a highly flexible and extensible configuration, the USAF TPS Handling Qualities Simulator has become a touchstone for curriculum development and adding direct experience to the students learning environment. Most efforts are currently focused on moving legacy curriculum events onto the simulator and supporting student test management projects, but there are a few important long-term goals for the simulator. 1. Computer-Based Training The USAF has embraced computer-based training (CBT) very aggressively it can now take up more than half of a trainee s academic instruction time. The USAF TPS, on the other hand, has no academic hours devoted to CBT in the entire 49-week course. The simulator might offer the opportunity to shift some academic hours into CBT. Certain parts of the aircraft trim and stability course rely on the instructor s descriptions and the students imagination for visualizing the various facets of aircraft static stability. A simulator could teach much through direct demonstration, using the reconfigurable nature of the controls and models to step students through a variety of examples while giving instantaneous feedback on how different parameters are measured and perceived. 2. Student Free-Play The simulator is capable of almost instantaneously changing from one aircraft into another including performance, flying qualities, control feel, forward visibility, and cockpit instruments/displays. With two control sticks, it will be possible to simulate multiple control systems at the same time, such as the control surfaces and reaction jets on an aircraft like the X-15. As the TPS acquires, creates, and integrates different aircraft models it will become possible to allow the students to sit at the simulator, select an aircraft and an initial condition, and be flying in just a few seconds. The touch-screen LCD panel display will also serve as the that are simply not consciously accessible. Duplicating these motion cues for even the simplest and least dynamic motions is very expensive the cost-per-accuracy ratio is quite high for motion-base simulators. Thus simulators not only cannot duplicate the flight experience, they may actually be ineffective or counterproductive for learning tasks that require a finely-tuned subconscious feedback system such as close formation flight. The only simulators that come close to producing true aircraft motions are variable-stability aircraft. 9 of 10

10 simulator control so students can operate it by themselves, including changing configurations and (when available as with the current variable-stability T-38) play with the aerodynamic model and flight control system. V. An Opportunity for Collaboration The USAF TPS simulator has become a remarkable educational environment. With very few exceptions, the software is free for government users and the hardware provides a very high fidelity/cost ratio. It was designed with education in mind and should be very well-suited for any academic environment. (Several visiting Master s Degree students spent some time with the principle author on an early desktop version to help them prepare for their future Test Pilot School research projects. They remarked that their one hour on the simulator gave them more understanding of the aerodynamic stability derivatives than ten hours of classroom time.) The simulator was created around the idea of simulating variable stability aircraft assets that are routinely described by graduating TPS students as being among their most important educational experiences. As of the publication of this paper, the simulator is a one-of-a-kind collaboration between a test pilot with basic programming skills and an aeronautical engineer with real programming talent. A few of these systems scattered around other academic institutions would create outstanding opportunities for collaboration and sharing and significantly increase the potential for finding new and exciting uses of the simulation laboratory. VI. Conclusion In spite of the well-known advantages of aircraft simulators in the flight training environment, simulators specifically designed to support the academic curriculum at the USAF Test Pilot School have historically been discounted as insufficiently flexible or financially untenable. Recent advances in computer technology have made it possible to create a reconfigurable handling qualities simulator using surplus parts, government furnished software, and a few relatively inexpensive hardware/software elements. The resulting simulation laboratory, started in the principle author s spare time and rapidly improved at the hands of the collaborating author, has exceeded initial expectations by providing a highly flexible system that can be used to model many different aircraft, create high-impact curriculum events, interface with external systems such as control rooms, and serve as a aircraft development platform. The authors sincerely hope that other organizations might be able to use their work or, much better, collaborate to improve the laboratory at several different locations. Acknowledgments The authors would like to thank the USAF Test Pilot School leadership for their trust and financial support, and the Plans and Programs Division for their patience and ingenuity. References 1 Smith, J., Experience With Flight Simulators Training Effectiveness and Future Developments, Tech. Rep. AFHRL- TP-81-41, USAF Human Resources Laboratory, Brooks Air Force Base, TX, McCauley, M. E., Do Army Helicopter Training Simulators Need Motion Bases? Final Technical Report 1176, Naval Postgraduate School, Monterey, CA, Gray, III, W. R., Boundary-Avoidance Tracking: A New Pilot Tracking Model, AIAA Atmospheric Flight Mechanics Conference and Exhibit, Reinhard, R., Celi, S. A., Geraghty, J. T., Stah, J. W., Glover, V. J., and Bowman, G. G., A Limited Evaluation of Full Scale Control Surface Deflection Drag (HAVE FUN), Technical Informaton Memorandum AFFTC-TIM-07-04, USAF Test Pilot School, Edwards AFB, CA, of 10