PSI - Issue 18

Matthias Hell et al. / Procedia Structural Integrity 18 (2019) 823–836 Hell et al. / Structural Integrity Procedia 00 (2019) 000–000

3

825

Fig. 1. Fatigue life assessement with the elasto-plastic notch base approach.

After the experimental material characterization, the fatigue properties of the material are transferred to the component geometry using an elasto-plastic finite element model, which comprises the component geometry, the cyclic stress-strain behaviour and the load assumption, i.e. a load-time function and the boundary conditions. With the local strains, which have been determined by the finite element analysis, the cycles to crack initiation for constant amplitude loading may be determined from the fatigue-life curve, e.g. the strain-life curve according to Basquin Manson-Coffin-Morrow [2,3,4,5] or a continuous fatigue-life curve according to Wagener [6]. For the consideration of the effect of mean stresses on fatigue life, different damage parameters, e.g. according to Smith-Watson-Topper or Bergmann [7, 8], have been proposed. However, for variable amplitude loading, a piecewise calculation of the stress strain path may be necessary in order to account for load sequence effects on the stress-strain state during elasto plastic loading due to memory effects. Regarding transient effects within the stress-strain behaviour, an evaluation of the actual stress-strain state due to the cyclic elasto-plastic deformation may also require material parameters for the description of the evolution of the cyclic stress-strain behaviour during cyclic hardening or softening [1]. Furthermore it is important, to evaluate the influence of varying boundary conditions, i.e. changing local strain constraints, on the cyclic material behaviour. In addition, the experimental investigation should feature an evaluation of the slip behaviour, which affects the stress-strain behaviour depending on the load-time function [9].

2.1. Experimental material characterization

The material characterization with strain-controlled tests on un-notched specimens is the experimental basis for the estimation of the fatigue life of components and provides the material parameters, upon which a decision about the fatigue potential of the material is possible. Because the stress-strain relation affects the numerical assessment of the local stress-strain response within the component to an external load, the quality of the whole fatigue assessment strongly depends on the validity of the experimentally derived stress-strain relation and the strain-life relation.