220 likes | 491 Views
The Evolution of Simulator Data Packages and QTG’s. David Pepper Manager, Training Device Engineering United Airlines. CONTENTS:. Traditional flight test data & acceptance testing. New sources of simulator data. Simulation of computer controlled aircraft. Multiple engine fits.
E N D
The Evolution of Simulator Data Packages and QTG’s David Pepper Manager, Training Device Engineering United Airlines JAA/FAA Working Group - Hoofddorp Netherlands
CONTENTS: • Traditional flight test data & acceptance testing. • New sources of simulator data. • Simulation of computer controlled aircraft. • Multiple engine fits. • QTG proposal. JAA/FAA Working Group - Hoofddorp Netherlands
Traditional Flight Test Data • Parameters are recorded during aircraft flight testing. • Manufacturer develops simulation model, aided with wind tunnel data. • Simulation model validated by aircraft data spot-checks (ATM), & simulator model proof-of-match to aircraft (QTG). • Training simulator validated by regulator comparing QTG tests to actual A/C data. JAA/FAA Working Group - Hoofddorp Netherlands
Traditional Acceptance Tests • Manual fly-out • QTG aero tests • Flight Control statics & dynamics. • Ground handling. • Takeoff. • Longitudinal maneuvers. • Lateral maneuvers. • Landings. • QTG Latency, Motion, Sound, Visual Tests. JAA/FAA Working Group - Hoofddorp Netherlands
“Antique” QTG comparedto a new CCA QTG JAA/FAA Working Group - Hoofddorp Netherlands
New Sources of Simulator Data • Better flight test techniques & data recorders. • Use of engineering computer “predicted data.” • More sophisticated simulation models. • Availability to run Flight Control algorithms in lab. • Stimulation/Simulation of actual flight computer boxes. JAA/FAA Working Group - Hoofddorp Netherlands
Simulation of Computer Controlled Aircraft • Additional requirements for Normal, and Degraded/Direct Law tests. • Traditional tolerances may not be appropriate when validating/repeating tests. • Primary Flight Control response easily changed - swap in a new black box or rehost/emulate! • How to revalidate after a PFC computer changed? JAA/FAA Working Group - Hoofddorp Netherlands
Critical Eng Fail on T.O. Dynamic Eng Fail after T.O. Power change Dynamics Flap change Dynamics Speedbrake change Dyn. Gear change Dynamics Longitudinal Trim Longitudinal Man. Stab. Longitudinal Static Stab. Buffet & Stall Phugoid Short Period Vmca Roll Step response Spiral Rudder response Dutch roll Normal landing Go around Envelope Protections Which “CCA Off” Tests Do We Measure? JAA/FAA Working Group - Hoofddorp Netherlands
What Should “CCA on” Tolerances be ? • 5%? • 10%? • Depends on the control laws? JAA/FAA Working Group - Hoofddorp Netherlands
Multiple Engine Fits • QTG tests have become complicated by data packages which baseline a given engine, and then difference another engine. • Engine “derivatives” may be simple thrust changes, or can be much more involved with entirely different accel/decel characteristics. • Manufacturer’s are reluctant to conduct two or more complete flight test programs! JAA/FAA Working Group - Hoofddorp Netherlands
Multiple Engine Fits (continued-2) • Thrust matching may NOT be good enough for different engines whose performance effects the aerodynamic characteristics of the airplane. Validating these tests may require additional flight test data, or engineering analysis work. • Thrust matching MAY be acceptable for thrust variants of the same engine model. Unless both engine models are simulated, however, manual fly-outs of these maneuvers may not be possible. JAA/FAA Working Group - Hoofddorp Netherlands
QTG Proposal • Keep the total number of tests at a reasonable amount. To validate a simulator, we don’t need every test at every conceivable flap condition! • Direct Law tests will not normally change over the life of the aircraft. Put these in their own separate section of QTG. • Direct Law plots should include: (1) original flight test data; (2) Manufacturer’s simulation model proof-of-match; (3) Training simulator results. Ideally, all 3 plots should overlay. JAA/FAA Working Group - Hoofddorp Netherlands
Proposal(continued-2) • Drive aircraft surfaces to validate Direct Law tests. Manual fly-outs of Direct Law tests should be possible, provided the pilot practices the maneuver enough to duplicate the surface movement. • Traditional tolerances are OK for Direct Law tests, but they should be re-examined for Normal Law tests. JAA/FAA Working Group - Hoofddorp Netherlands
Proposal(continued-3) • End-to-end testing is better accomplished in Normal Law, where tests should be driven via pilot control inputs. • Normal Law tests should be kept in their own separate section of the QTG. • For each Normal Law QTG test, baseline plots of original black box performance should include 3 parameters: (1) Aircraft flight test; (2) Manufacturer’s engineering simulator model proof-of-match; (3) Training Simulator Results. JAA/FAA Working Group - Hoofddorp Netherlands
Proposal(continued-4) • Normal Law tests will also probably evolve over time. Each time a black box is updated, the manufacturer should specify which QTG tests will be affected, and using the new black boxes, run new proof-of-match tests on their engineering simulator. • Normal Law tests should not be held to higher tolerances than what is acceptable on the aircraft. JAA/FAA Working Group - Hoofddorp Netherlands
Proposal (continued-5) • The manufacturer’s new black box plots then become the baseline for the training simulator to match. Only 2 parameters will be plotted on these tests: (1) manufacturer’s engineering simulation response; (2) training simulator results. • To support incorporating these new tests, a list of effective part numbers, and revision levels/standards should be generated. Every new test should list any new initial conditions or test procedures required. JAA/FAA Working Group - Hoofddorp Netherlands
Proposal (continued-6) • The manufacturer should generate a descriptive rationale to explain the changes between the original flight test results and the new proof-of-match data. • Normal Law validation is then a comparison to either: (1) Aircraft data; or (2) The engineering simulation proof-of-match. • It is also suggested that the regulatory authorities “approve” a manufacturer’s simulator data package, and that the operator not be held responsible for the “quality” of individual tests. JAA/FAA Working Group - Hoofddorp Netherlands
Proposal (continued-7) • Finally, it is suggested that every test in the QTG be re-examined, to determine: (1) If another test covers the same information? (2) If there is an associated “training need?” (3) Exactly what are we validating? (4) Are the Normal Mode tolerances reasonable? (5) Is there a better way to prove the objective? JAA/FAA Working Group - Hoofddorp Netherlands
Other Issues • Electronic ATG. • Harmonization • When and how to apply “Engineering Judgment?” • Gray box vs. black box vs. simulation validation. • Climb Data. Snapshot or Time History? • The RAe Airplane Flight Simulator Handbook specifies recorded parameters for each test. Should these be included in the new FAA/JAA regulations? JAA/FAA Working Group - Hoofddorp Netherlands