For a compound helicopter with sufficient control redundancy, this study presents a knowledge based method for estimating the set of controls required to maintain trim as a function of additional controls (main rotor RPM, auxiliary thrust, and stabilator pitch). Trim analyses with parametric sweeps through the additional controls are simulated in RCAS for a compound helicopter model based on a UH-60A. The resulting data sets are used to construct quadratic regression fit models, which represent the response of the six classical trim controls subject to variation in additional controls. In hover these models predict states that aren’t fully trimmed, but the force and moment residuals are low. At forward flight speeds the response of the controls are no longer quadratic, and these force and moment residuals increase. Eliminating outlier trim states from the training data set gives the greatest improvement to the predictive capabilities of the model, but sacrifices some of the range of controls over which the model can apply. Kriging interpolation retains the full range of controls, but produces multiple local minima. For application to estimating the best controls for minimization of power it is shown that both truncating the data set and Kriging can provide good estimates of the minimum power trim state controls without excessive force and moment residuals. Allocating the additional controls as given by the response surfaces and solving for trim returned a slightly lower power than was found in the parametric trim sweeps.
Proceedings of the 72nd American Helicopter Society Annual Forum, West Palm Beach, Florida, May 17–19, 2016.