Parallel Computing of Helicopter Response to Turbulence Toward Real-Time Implementation

Turbulence and blade response to turbulence are cyclostationary and require blade-fixed sampling. The state-of-the-art methods of turbulence simulation via sample functions and response prediction via stale-transition and input-covariance matrices are sequential; they take excessive execution time. This precludes their application to large-dimension and comprehensive models of design analysis and flight dynamics simulation as well as to real lime implementation. Accordingly, parallel methods for turbulence simulation and response prediction and approximate sequential methods for turbulence simulation are developed. The covariance and spectral density statistics of turbulence and response from the slatc-of-the-art methods are compared with those from the developed methods. The developed methods accurately generate the statistics, take far less execution time and thereby offer considerable promise to large-dimension and comprehensive models and to real time implementation.