A numerical method for the determination of parameters of the strain softening creep model

The trends of decreasing of weight of machines and increasing of their quality, and intention to the fullest use of mechanic properties of materials demand the development of numerical methods for analysis of the stress-strain state of materials undo the terms of creep. The article discusses the development of new numerical method for determining the parameters of the strain softening creep model. The base of new method is generalized regression model, which was built on the basis of difference equations for describing the creep. The relations between coefficients of difference equation and parameters of the strain softening creep model allow reduce the problem of parametric identification to an iterative procedure for RMS of coefficients of regression model, which is linear. The approbation of numerical method with five creep curves of aluminum alloy is accomplished. The approbation confirms scientific credibility of built relations and efficiency of new numerical method.

strain softening creep model, difference equations, generalized regression model, nonlinear regression, root mean square evaluation