AN APPROXIMATE METHOD FOR LARGE-SIZE STRUCTURAL PART RELIABLE LIFE PREDICTION

Abstract

In the viewpoint of probability theory, An large-size structural part is a system consisting of many "elements", i.e. high stress zones, since all the high stress zones have their related individual probabilities to cause the structural part failure. Such a part should be taken as a serial system if any high stress zone failure means to part failure. As a random variable, the fatigue life of a complex structural part with many high stress zones does not equal to the fatigue life of any one high stress zone of the part. A probabilistic fatigue life prediction model for large-size structural part is developed by means of system reliability analysis principle, the concept of extreme statistics of the lives of all the high stress zones, as well as the Tailor expansion of Weibull reliability function. By this model, the fatigue life of a large-size structural part can be predicted according to the material’s P-S-N curve of the part, and the effect of the dependent failures of the high stress zones can be intrinsically reflected.

Keywords

• Large size structural part;Fatigue;Reliable life;Common cause failure;Extreme statistics