Abstract: Several potential failure modes may occur when mechanical components fail. For that all the failure modes have a certain impact on the reliability sensitivity of the components, it is necessary to take each failure mode into account as a reliability sensitivity estimation is performed. A reliability sensitivity analysis method for mechanical components with multiple failure modes is proposed. Concerning the unknown distributions of random variables, the random perturbation technique and the fourth moment method are employed to respectively calculate the first four moments and the reliability index of each performance function of the failure mode. The potential failure modes, only a few existed in the component, are regarded in series, and the failure of the component is thus the failure of a series system. The calculating formulas of reliability sensitivity with respect to the mean and variance of random variables are derived with the gradient method. The numerical example illustrates that the proposed method could quickly and effectively estimate the reliability sensitivity of the mechanical components with multiple failure modes, and is also applicable for nonlinear performance functions.

Key words: Arbitrarily distributed variables, Fourth moments, Multiple failure modes, Random perturbation, Sensitivity estimation