No abstract

Air carriers will use simulators to train pilots to recover from fully developed stalls. Flight simulator models will have to portray an airplane's dynamics satisfactorily to achieve the associated training objectives. That imposes new simulator requirements. This study evaluated several full stall simulator models to meet those requirements. Three new stall models were tested in a Boeing 737-800 simulator. One model used the conventional approach by matching flight test data to within a tolerance. Another model did not rely on flight test data, but instead combined computational aerodynamics, scaled wind tunnel data, and expert opinion from a test pilot who had stalled the actual aircraft. The third model added a roll asymmetry to the unmodified simulator model as a simple way to possibly meet the training objectives. The test had two phases. In the first phase, test pilots who had stalled a 737 airplane evaluated the models by performing typical flight test stall maneuvers in the simulator. The second phase used airline pilots type-rated in the 737 but who had not stalled a 737 airplane. The airline pilots were placed in groups, and each group trained with one of the models. Each airline pilot was then checked on the model developed from flight data, which represented the truth model. The second phase also included a surprise stall scenario with each airline pilot having to recover from a stall using the model they would train with. The results revealed wide ranges in the subjective evaluations of the test pilots, as well as in the objective performance of the airline pilots across the models. However, many of the averages did not show significant differences. All airline pilots agreed or strongly agreed that they were surprised by the surprise stall scenario. In that scenario, less than one quarter of the airline pilots strictly followed the proper stall recovery procedure on which they had been briefed. Less than half maintained a nose-down input until the stall warning stopped. For situations when developing a stall model based on flight data is not practical, the alternative approach of developing a model based on computational aerodynamics, wind tunnel data, and subject expert opinion appears feasible.

This paper describes a quasi-transfer-of-training study in the NASA Ames Vertical Motion Simulator (VMS). Sixty-one general aviation pilots trained on four challenging commercial transport tasks under one of four different motion conditions: no motion, small hexapod, large hexapod, and VMS motion. Then, every pilot repeated the tasks in a check with VMS motion to determine if training with different motion conditions had an effect on task performance. New objective motion criteria guided the selection of the motion parameters for the small and large hexapod conditions. Considering results that were statistically significant, or marginally significant, the motion condition used in training affected 1) longitudinal and lateral touchdown position; 2) the number of secondary stall warnings in a stall recovery; 3) pilot ratings of motion utility and maximum load factor obtained in an overbanked upset recovery; and 4) pilot ratings of motion utility and pedal input reaction time in the engine-out-on-takeoff task. Since the training motion conditions revealed statistical differences on objective measures in all the tasks performed in the VMS motion check, with some in the direction not predicted, trainers should be cautious not to oversimplify the effects of platform motion. Evidence suggests that the new objective motion criteria may offer valid standardization benefits, as increases in the training motion fidelity, as predicted by the two conditions covered by the criteria, resulted in expected trends in pilot ratings and objective performance measures after transfer.

No abstract

This paper describes a quasi-transfer-of-training study in the NASA Ames Vertical Motion Simulator (VMS). Sixty-one general aviation pilots trained on four challenging commercial transport tasks under one of four different motion conditions: no motion, small hexapod, large hexapod, and VMS motion. Then, every pilot repeated the tasks in a check with VMS motion to determine if training with different motion conditions had an effect on task performance. New objective motion criteria guided the selection of the motion parameters for the small and large hexapod conditions. Considering results that were statistically significant, or marginally significant, the motion condition used in training affected 1) longitudinal and lateral touchdown position; 2) the number of secondary stall warnings in a stall recovery; 3) pilot ratings of motion utility and maximum load factor obtained in an overbanked upset recovery; and 4) pilot ratings of motion utility and pedal input reaction time in the engine-out-on-takeoff task. Since the training motion conditions revealed statistical differences on objective measures in all the tasks performed in the VMS motion check, with some in the direction not predicted, trainers should be cautious not to oversimplify the effects of platform motion. Evidence suggests that the new objective motion criteria may offer valid standardization benefits, as increases in the training motion fidelity, as predicted by the two conditions covered by the criteria, resulted in expected trends in pilot ratings and objective performance measures after transfer.