[en] In the design of components for the power industry in general and the nuclear industry in particular, the engineer is called upon to perform inelastic analyses for the satisfaction of certain code and regulatory requirements. In other situations inelastic analysis is used to investigate the cause of failure of a component or to gain further insight of test results. In all these cases a design limit or a failure criterion is set for evaluating the analytical results, but in most cases it is left up to the analyst to choose the appropriate material constitutive relations. Ideally, the constitutive relations should model the material behavior in the mode of failure or deformation under consideration. For example, yield stress, strain rate dependence could be of major importance in dynamic plasticity at high strain rates. Similarly, deformation plasticity theory was found to give better results than incremental theory in plastic buckling. In many cases however, the analyst, especially if the uses general purposes programs, has a very limited choice in the constitutive relations for use in the analysis. In addition to the above considerations, the analyst has to make sure that his analysis results could be used easily with the failure criteria adopted. For example, in evaluating creep-fatigue damage, currently there are several damage criteria, some of them make use of plastic and creep strains separately, while others use strains and strain rates without definite distinction between creep and plastic strains. In these cases the analyst must make sure to adopt the appropriate constitutive relation