GLOBAL SENSITIVITY ANALYSIS METHOD FOR MODEL-BASED SYSTEM SAFETY ASSESSMENT ON AVIATION PISTON ENGINE

Aviation piston engines provide an important source of power for general aviation, thereby the identification of safety-critical influences during the design phase is an important requirement for ensuring aircraft safety. Model-based system safety assessment have been developed to overcome the limitations of traditional analytical methods in objectively decomposing and identifying failure in complex coupling problems. Nevertheless, the method is limited research for aviation piston engine, especially for determining safety-critical influences from the global perspective of the entire-engine structure with high complexity. In this paper, a new method for rapidly analyzing entire-engine safety sensitivity analysis from the global perspective is proposed, by introducing model-based system safety assessment for the analysis of aviation piston engine. Compared with local sensitivity analysis results, the proposed RSM-Sobol global sensitivity analysis could not only supply additional information on the interactions between parameters but also accommodate rapid iterative complex system safety analysis in the design stage. A methodical case of sensitivity analysis of the structural parameters of a simplified aviation piston engine is given using the proposed method in this paper. Results show that the importance of each parameter can be ranked by the sensitivity index through global sensitivity analysis. Therefore, the safety criteria could be satisfied by reducing the uncertainty of critical parameters and adjusting the input range based on the analysis, allowing an efficient and intuitive safety analysis during design.