Using Parallel Processing Tools to Predict Rotorcraft Performance, Stability, and Control

This paper discusses the development of the High Performance Computing (HPC) Collaborative Simulation and Test (CST) portfolio CST-03 program, one of the projects in the Common HPC Software Support Initiative (CHSSI) portfolio. The objective of this development was to provide computationally scalable tools to predict rotorcraft performance, stability, and control. The ability to efficiently predict and optimize vehicle performance, stability, and control from high fidelity computer models would greatly enhance the design and testing process and improve the quality of systems acquisition. Through this CHSSI development, the US Navy Test Pilot School performance, stability, and control test procedures were fully implemented in a high performance parallel computing environment. These Navy flight test support options were parallelized, implemented, and validated in the FLIGHTLAB comprehensive, multidisciplinary modeling environment. These tools were designed to interface with other CST compatible models and a standalone version of the tools (FLIGHTLAB-ASPECT) was delivered for use independent of the FLIGHTLAB development system. Tests on the MAUI Linux cluster indicated that there was over 25 times speedup using 32 CPUs. The tests also met the accuracy criteria as defined for the Beta trial.