https://ntrs.nasa.gov/search.jsp?R=19960012281 2019-01-15T07:34:25+00:00Z NASA-TM-4703 NASA Technical Memorandum 4703 [_q[00 £)[o]C_ I The NASA Landing Gear Test Airplane John F.Carter and Christopher J.Nagy FORREFERENCE NOTTOBETAKENFROMTHISROOM LI''-"_'__;'r'v COPY 'I _ Jl J J i ii - ,w,,: i 8 P% LANGLEYRESEARCHCENIER June 1995 LIBRARNYASA NASATechnicalLibrary i 3 117601423 7045 The NASA Landing Gear Test Airplane JohnF.Carter NASA Dryden Flight ResearchCenter Edwards, CA ChristopherJ.Nagy PRC Inc. Edwards, CA Technical Memorandum 4703 June 1995 ABSTRACT A tire and landing gear test facility has been developed and incorporated into a Convair 990 aircraft. The system can simulate tire vertical load profiles to 250,000 lb, sideslip angles to 15degrees, and wheel braking on actual runways. Onboard computers control the preprogrammed test profiles through a feed- back loop and also record three axis loads, tire slip angle, and tire condition. The aircraft to date has pro- vided tire force and wear data for the Shuttle Orbiter tire on three different runways and at east and west coast landing sites. This report discusses the role of this facility in complementing existing ground tire and landing gear test facilities, and how this facility can simultaneously simulate the vertical load, tire slip, velocity, and surface for an entire aircraft landing. A description is given of the aircraft as well as the test system. An example of a typical test sequence is presented. Data collection and reduction from this facility are dis- cussed, as well as accuracies of calculated parameters. Validation of the facility through ground and flight test is presented. Tests to date have shown that this facility can operate at remote sites and gather complete data sets of load, slip, and velocity on actual runway surfaces. The ground and flight tests have led to a successful validation of this test facility. NOMENCLATURE deg degrees DFRC Dryden Flight Research Center, Edwards, California Hz Hertz KGS knots ground speed KIAS knots indicated airspeed KSC John F. Kennedy Space Center, Florida LSRA Landing Systems Research Aircraft n mi nautical miles psi pounds per square inch STS Space Transportation System INTRODUCTION Tire and landing gear development and testing for aircraft are usually done by ground test facilities due to the expense and hazards associated with testing on aircraft. Tire dynamometer and sled tire track are the two facilities used mainly for dynamic tire testing. Exist- ing facilities have limitations in simulating the landing surface, time varying vertical loads, and tire slip angles. Tire dynamometer facilities roll aircraft tires against a metal drum at any combination of velocity, vertical load, and slip angle. These facilities have the advantage of long run times, very good load and speed control, and good control of the slip angle of the tire. However, dynamometers have disadvantages for dynamic tire testing, such as 1) the dynamometer rotary drum surface does not accurately simulate a runway surface, 2) the curvature of the contact area of the drum causes incorrect radial tire deflection during the test, and 3) heat build up of the drum causes the temperature of the test tire to be abnormally high. Because of these problems, dynamometer data are used primarily to measure the strength and endurance of tire carcass material, not the tire surface forces or wear. Appendix A shows a tire dynamometer at Wright Patterson Air Force Base in Dayton, Ohio. An example of data obtained from this type of dyna- mometer is given in reference 1. Tire sled-type facilities mount the tire on a carriage and move the carriage down a straight path. 2 A test surface can be constructed which simulates an aircraft runway, but the process can be time consum- ing and may not accurately represent the surface. Existing facilities also have problems due to their limit- ed run times, limited capability for time varying vertical load, speed, and tire slip angle control. Because of limited track length, simulations of complete aircraft landings typically are completed in segments, with a single landing test requiring as many as five test runs. In addition to the inconvenience of multiple runs, cooling of the test tire between runs can cause inaccurate results. The unique design of the Space Shuttle Orbiter landing gear with its highly loaded tires, hazards asso- ciated with tire failure, as well as limited opportunities for landing test data from the vehicle resulted in a strong reliance on tire test facilities. Because of high landing speeds, high vertical loads, long roll out dis- tances, and unusually rough runway surfaces, existing tire test facilities have had difficulties in accurately simulating the tire wear and forces of an entire shuttle landing. The Landing Systems Research Aircraft (LSRA) is a unique addition to complement existing aircraft dynamic tire testing facilities. Its capabilities are compatible with the Space Shuttle Orbiter requirements. The design goal of the LSRA is to conduct dynamic tire testing on an actual surface while simulating ver- tical loading, tire slip angle, and speed of an entire aircraft landing simultaneously. Computer control of a tire test fixture allows for precise control of vertical load and slip angle of the test tire. The computer con- trol software also provides a speed advisory to the pilot. These capabilities make it possible for the LSRA to recreate a realistic combination of run distance, runway surface, vertical load, tire slip angle, and ground velocity for aircraft landings. The LSRA is the result of a cooperative effort of the Dryden Flight Research Center (DFRC), Lyndon B. Johnson Space Center (JSC), John F. Kennedy Space Center (KSC), Langley Research Center (LaRC), Ames Research Center (ARC), and many industry and military organizations. Flight test has been conducted on runways at Edwards Air Force Base and KSC. This paper describes the systems and capabilities of the LSRA vehicle. In addition, this paper discuss- es ground calibration and flight tests used to validate the LSRA as a test facility. 2 AIRCRAFT DESCRIPTION The NASA Convair 990 (SN 10-29, tailNo. 810) is ahigh-speed, medium range, low-swept-wing jet transport (fig. 1). This aircraftis equipped with four wing-pylon mounted General Electric® CJ805-23 aft fan turbojet engines and a fully retractable tricycle landing gear (the main gear can no longer be re- tracted with the LSRA modification). The aircraftis controlled by dualwheel andcolumns located in the cockpit. The control surfaces are moved using acombination of mechanically driven flight tabs and hy- draulics. The basic control system is augmented with ayaw damperwhich drives the rudder. The LSRA underwent significant structuralmodification toprovide space forthetest gear and also to react the test gear loads into the aircraft.Normal aircraftstructuralfactors of safety were maintained for all the original structural design conditions plus the additionalloading conditions forlanding gear testing as defined inthis report. The primary components of the landing gear test system added to the LSRA are shown in figures 2 and 3. Figure 2 identifies the test pallet system elements within the aircraft.The hydraulic power of the gear test pallet is provided by accumulators which use compressed nitrogen gas. Onboard hydraulic pumps are used to pressurize the accumulators. The test pallet system is controlled by a test conductor console which contains hardware switching capability and system monitoring capability. Included in the system is acomputer which controls the motionof the testgear pallet. In additionto the vertical load, the test pallet system can apply braking to the test tire. Aircraftperformance specifications before and after the LSRA modification are presented intable 1. Figure 3 shows thepallet which is theinterface pointbetween the testfixture andthe landing gear test system. The pallet is attached to the aircraftthrough a pair of parallelogram swing links which restrain the test gear inpitch, roll, andyaw. The topof thetest pallet is attached totwo hydraulic actuators which provide the vertical reaction load. The vertical loads are reacted into the airframe through atruss system located inside the cabin. ®TheCJ805-23engineisaregisteredtrademarokfGeneralElectricL, ynn,MA. Table 1. Aircraft operational limits before and after LSRA modifications. Original LSRA CV990 Aircraft Max. taxi weight, lb 255,000 250,000 Max. takeoff weight, lb 245,000 245,000 Max. landing weight, lb 202,000 225,000 Max. landing speed, kgs 195 230 Max. range, n mi 3000 600 Max. ceiling, ft 41,000 13,000 Max. velocity, KIAS 520 250 Empty weight, lb 115,000 177,000 NOTE: data taken from operations manuals of the CV990/LSRA. SYSTEM CAPABILITIES Many landing gear test fixtures can be attachedto the LSRA test pallet. Currently two attachments havebeen designed. One is amodified shuttle mainlanding gear strutwith dual tires, the other is asingle tire fixture that contains a rotary actuator which can be turned for desired slip angle. Table 2 presents the maximum load and steering capability of the LSRA for these two fixtures. Figure 4 shows a model of the single tire fixture. The fixture frame is attached to the test pallet. This frame houses the rotary actuator which turns the test tire axle assembly. Tire braking is applied through the braking assembly. This test fixture was used exclusively for all the testing described inthisreport. A test computer controls the test tire vertical load and slip angle using feedback loops, and sends a discrete signal to activate thewheel brake. Table 3 shows the capabilities of the control system. During a test,the computer also displays tothe pilot the difference between thecurrent measured ground speed and the desired speed profile for the test.The test tire vertical loadfeedback is provided directly from vertical Table 2. Structural load capabilities ofthe LSRA. Main gear Variable yaw fixture dual tire single tire Vertical load, lb 250,000, -50,000 150,000, -25,000 Drag load at tire contact point, lb + 100,000 + 50,000 Side load at tire contact point, lb +40,000 +40,000 + 800,000 Brake torque, in-lb N/A - 250,000 Steering torque, in-lb N/A 380,000 Table 3. Performance of theLSRA test system with the single rotationaltire fixture. Load control system max rate, unloaded ---t-5irdsec Load control system max rate, max load 7 irdsec Steering control system max rate 35 deg/sec Maximum steering angle + 20 deg Load control system bandwidth 2 Hz Steering control system bandwidth 3 Hz Maximum error from commanded profile, load + 3000 lb Maximum error from commanded profile, slip .25 deg Typical error from commanded profile, speed + 10kts 4 load cell measurements while the slip angle is computed, as seen in figure 5, from a combination of an angular displacement sensor on the steering fixture and two optical ground velocity sensors which pro- vide aircraft slip angle across the runway. SAFETY SYSTEMS The test pallet system includes a fail-safe feature which retracts the test pallet to its stowed position. Pallet retraction can be caused by fault detection in hardware or software. The gear control system fault detection software performs comparisons between redundant input signals, compares input signals to minimum and maximum output values, and compares steering and extension values to simulated predic- tions. Hardware fault switches detect over extension, ground contact, and over rotation of the test pallet. If the test pallet cannot be retracted due to a mechanical failure, the hydraulic actuators can be separated from the test pallet by explosive bolts. The tire retraction and the actuator separation can be performed manually by either the test conductor orthe pilot. In addition tothe test pallet retraction system, there are two fire suppression systems associated with the test pallet system. A water deluge system was installed which can spray water directly on the CV990 main landing gear tires, brakes, and the test tire. A halon fire suppression system was placed in the cargo bay near the hydraulic pumps to extinguish any fires in that area. TEST OPERATIONS Flight planning and data analysis are performed with the aid of a six-degree-of-freedom simulation resident on a desktop workstation. This simulation was programmed using the FORTRAN® computer language and executes at 100 Hz, with a400-Hz execution for landing gear dynamics. Aerodynamic data used in the simulation were obtained from wind tunnel models, and then refined using data obtained dur- ing early NASA flight test of the CV990. The workstation is interfaced with a gear control computer which is a duplicate of the aircraft gear control computer. This configuration allows for production and hardware-in-the-loop simulation testing of new time history profiles, as well as verification and valida- tion of flight software revisions. The workstation and duplicate aircraft gear control computer configura- tion were designed to be portable so that simulation, analysis, verification, and validation functions would be retained at remote testing sites. The time history profiles of load, slip, and speed are produced using output from this simulation, and then converted to a binary format which is loaded onto a data diskette. After testing the profiles on this disk using the hardware-in-the-loop configuration, this data diskette is then used to load the profiles onto the aircraft gear control computer. A new time history profile can be developed in approximately one hour. A new gear control software version can be qualified for flight in approximately three hours. Figure 6 shows a typical landing test sequence. The CV990 aircraft makes a final approach. After touchdown and derotation, the pilot calls for test initiation and uses spoilers, thrust reversers, and brakes to follow the pilot speed advisory. The test gear is extended and controlled to match the preprogrammed test profiles of vertical load, slip angle, and braking on the test tire. Upon completion of the test, the test gear is automatically retracted. If a problem occurs during the test, either the computer or the hardware fault detection system will command a retraction of the test gear. If a retraction does not occur, the test ®FORTRANisaregisteredtrademarkofInformationProcessingTechniquesCorp.,PaloAlto,CA. conductor or pilot can unload the test fixture by exploding the bolts connecting the test gear assembly to the hydraulic actuators, thus relieving the vertical load to the test gear assembly. The LSRA has performed approximately 100 test operations at Edwards AFB and KSC. During these operations, all flight test profile preparation, data reduction, and analysis were performed at the test site. DATA REDUCTION The LSRA can collect onboard data or telemetered data. The data rates for the parameters range from 25 to 200 Hz. The test pallet has been instrumented with load cells in three axes. Appendix B presents the equations for calculations and corrections for vertical, side, and drag loads. Accuracies of the measured loads for the Shuttle Orbiter tire tests are + 3000 lb vertical load, + 500 lb side load, and + 300 lb drag load. In addition to the onboard and telemetered data, the LSRA has video cameras which can provide five different views of the tire fixture. These cameras allow for real-time monitoring of tests, as well as post flight analysis using video tape which is synchronized with the other data. High-speed film oftests is also available. TEST VALIDATION / FLIGHT TEST RESULTS Calibration of the LSRA load cells was performed at DFRC. This was done by attaching static test equipment to the test pallet and loading itto known values of vertical, side, and drag loads. This calibra- tion effort provided information to validate the LSRA gear control software calibrations, provided infor- mation on elastic deformation of the test fixture, and verified post flight data measurements. The LSRA has performed two validation landing simulations; one was performed at the Edwards AFB concrete runway, the other at KSC. The purpose of these tests was to validate the LSRA as a tire testing facility by simulating an actual Space Shuttle Orbiter landing and comparing the test tire wear from the LSRA to the tire wear of the Space Shuttle Orbiter. While both tests were successful, only the KSC test will be discussed in detail to illustrate the process. The Space Shuttle Orbiter landing chosen for the comparison was the STS 51-D landing. On this landing, the Space Shuttle Orbiter landed on Runway 33 with approximately 8 knots of crosswind from the right-hand side of the vehicle. The weight of the Space Shuttle Orbiter was approximately 200,000 lb. Inertial platform data as well as strain gage data recorded from this landing were used to derive the load, slip, and speed profiles for the left inboard main gear tire of the Space Shuttle Orbiter. The local tilt angle of this tire during the Orbiter landing was simulated by raising the right-hand strut of the LSRA until the test tire tilt angle was approx- imately -1.6 deg (left wing down). The LSRA performed the profile shown in figure 7 on Runway 33. A load "spike" was placed at the beginning of the load time history profile to create the initial load of 70,000 lb to simulate initial tire touchdown. After the initial load, the average load control for the time history stayed within +3000 lb of the target value. The slip controller held the slip angle to within approximately .40 deg until the speed fell below 50 knots, at which point the resolution of the optical sensors caused some steering oscillations. The steering system exhibited an oscillation of approximately .2 deg at 2 Hz. Subsequent slip controller im- provements have eliminated these two anomalies. Figure 8 shows a time history of the achieved slip an- gle plotted against the commanded slip angle after the improvements were made. The steering system 6