Configuration Effects On Departure Prediction Parameters Used In Representative Stall Models For Flight Training Simulator

Over the past decade, loss-of-control (LOC) accounts for 22 accidents and 1,991 total fatalities out of 91 accidents and 4,970 total fatalities from all causes, making LOC by far the largest category of fatal aviation... [ view full abstract ]

Over the past decade, loss-of-control (LOC) accounts for 22 accidents and 1,991 total fatalities out of 91 accidents and 4,970 total fatalities from all causes, making LOC by far the largest category of fatal aviation accidents. [1] Of all the LOC accidents that occurred worldwide in recent history, the leading cause was aerodynamic stall. Aviation authorities recognize the need to train pilots on stall recovery techniques. However, in-flight recovery training with large aircraft is expensive and unsafe. In contrast, the use of ground-based simulators for pilot training is safe, inexpensive, accessible, and has reached a high level of technological maturity. Current flight simulators are considered inadequate for the simulation of aerodynamic stall, due to the lack of accurate models. Conventional methods such as wind-tunnel tests for generating the data used in aerodynamic models become significantly more expensive to perform for high angle-of-attack scenarios. [2] Therefore, there exists a need for cost-effective methods of simulating stall.

A method for developing representative stall models is being explored for the class of twin turbo-prop commuter aircraft and for the class of twin-engine regional jet. The model will be able to capture representative post-stall behaviours (e.g. g/pitch break, reduction in lateral stability, reduction in control effectiveness, and roll-off tendencies) essential for stall prevention and/or recovery training. Despite having lower fidelity than specific stall models, the representative stall model is cheaper to implement both in time and money, thus providing an accessible training environment for pilots for upset recovery. The methodology for developing a representative stall model will depend on a pre-existing specific aircraft stall model, with added terms to capture changes in aerodynamic forces and moments due to configuration changes (e.g. swing sweep, nacelle location, tail height, etc.) from the baseline aircraft to a target aircraft. Terms in the model are thus broken down into a sum of basic effects and configuration effects in the stall regime.

In order to identify and establish a basis of key aircraft configuration changes, it is necessary to determine the impact and sensitivity of those changes on departure tendencies. This is key because LOC is not due to stall per se, but due to departure from controlled flight that results from stall. Hence the focus of this study is to explore relationships between stability/controllability parameters that predict departure tendency and aircraft configuration including but not limited to tail height, wing sweep, and fuselage size. Particularly, focus will be placed on lateral-directional stability parameters such as Cnß,dyn and LCDP, which have been shown in literature to give good prediction to aircraft departure tendencies in stall. [3] Additional existing departure susceptibility prediction criteria in literature are reviewed and summarized. Analysis of lateral stability criterion Cnß,dyn for T-tailed regional jets and turboprop commuter types is performed. Prediction parameter(s) variations due to tail and wing configurations are presented. Methods of obtaining or estimating Clp and Cnß needed for this analysis are described. Finally, potential methods for validating representative stall models with the help of departure prediction criterion are proposed.