Experiment study and fatigue life prediction of torsion shaft of self-propelled gun based on stochastic response surface method

Abstract

The randomness of low cycle fatigue life of torsion shaft of self-propelled gun is the key factor affecting the reliability of the running system. A self-propelled gun experiment was carried through the field test to achieve the load spectrums of torsion shaft. The conclusion that the maximum tensile stress is the failure stress of torsion shaft was drawn. In order to predict the fatigue life of torsion shaft accurately, the stochastic response surface method was introduced to study low cycle fatigue life. The stochastic parameters of low cycle fatigue life were expressed to stochastic variables, and the structural stochastic response was fitted based on Hermite polynomials. By solving the coefficients of stochastic polynomials, the function of the low cycle fatigue life of torsion shaft was achieved, and the drawbacks of the classical response surface method which approaches the structural stochastic response using a group of general polynomials yet can not guarantee the convergence were overcome. The results show that the stochastic response surface method can achieve the statistical information of the structural stochastic response rapidly and accurately.