Complex physical-model based dynamic system safety analysis of Aviation Piston Engine considering hybrid uncertainty of fault

Abstract

Aviation piston engines constitute an essential segment of general aviation, whilst safety issues are becoming increasingly prominent by high system complexity. However, the conventional methods are gradually becoming insufficient to support the safety analysis in complex systems with multi-physical during the lifetime. Therefore, an innovative analysis that can depict the dynamic safety and potential coupled faults of the entire engine with multi-physical systems considering the effects of hybrid uncertainty faults is proposed in this study. Consequently, safety feedback is provided to the design in a more timely, accurate, and efficient manner based on the safety distribution throughout the lifecycle during the design stage. The analysis involves an integration of the nominal physical model and stochastic fault model via functionalized process variables. Firstly, a deterministic nominal model of the multi-physical domain of the engine is developed to characterize the coupled relationship between different domains by differential equations. Secondly, a stochastic fault model with hybrid uncertainty is quantitatively integrated into the nominal model for safety analysis by affecting process variables. Specifically, hybrid uncertainty encompasses the uncertainty of occurrence and severity. Finally, a simplified dynamic safety variation of the entire-engine is investigated by the approach. Result of Monte Carlo simulations of safety highlighted the coupled failures of different components in the respective safety ranges could be discriminated from the entire-engine perspective. The identification of potential coupled faults demonstrates the effectiveness and advantages of the proposed method in the design stage.