Product Reliability Due to Random Load Variations at Elevated Temperatures

Abstract

The survival capability of mechanical equipment often depends upon creep life at elevated temperature due to irregular load variations in a random service environment. The product reliability is related to the probability distribution of the creep damage, which depends, in turn, upon the statistical behavior of the stress waveform. The general stochastic connection between them is derived in this presentation for linear damage accumulation in a material with a power law of failure. The creep damage is related to the damage rate by a stochastic integral, while the damage rate is a stochastic function of the stress intensity. The statistical behavior of the damage rate depends upon the standard deviation of a stationary, Gaussian stress variation about a zero mean stress level. This paper illustrates the specific analysis of a time-invariant, linear system for a broadband, stationary, Gaussian excitation with a zero mean value and a uniform spectral density.