An experimental and numerical investigation of the effect of residual compressive stress on the high cycle fatigue life of notched low carbon steel test specimens is presented. Experimentally determined cyclic stress strain curves for S355 low carbon steel are utilized in a finite element analysis plasticity modelling framework incorporating a new cyclic plasticity material model representative of cyclic hardening and softening, cyclic mean stress relaxation, and ratcheting behaviors. Fatigue test results are presented for standard tensile fatigue test specimens and novel double notch specimens. Double notch specimens are tested with and without compressive residual stress prior‐induced through tensile overload. It is shown that cyclic plasticity phenomena have a significant influence on the induced residual stress distribution and also on material behavior when fatigue tested in the high cycle regime. It is observed that higher initial compressive residual stresses magnitude does not necessarily lead to a longer fatigue life. Finite element analysis using the new cyclic plasticity material model shows this behavior is due to combined residual stress redistribution under fatigue test cyclic loading and cyclic hardening effects. A fatigue life methodology based on the stress‐life approach augmented by a critical distance method is proposed and shown to give good agreement with experimental results for test specimens with no induced residual stress. The results obtained for specimens with induced residual stress are more conservative, but the degree of conservatism is significantly lower than that in the conventional stress life approach. The proposed methodology is therefore suitable for analysis and design assessment of components with pre‐service induced compressive residual stress, such as autofrettaged pressure components.

No abstract

The second part of the study presents development of the Dirac delta functions framework to modelling of cyclic hardening and softening of material during cyclic loading conditions for the investigated in Part I low carbon S355J2 steel. A new criterion of plastic strain range change is formulated. This provides more certainty in the cyclic plasticity modelling framework compared to classical plastic strain memorization modelling. Two hardening parameters from the developed kinematic hardening rule are written as functions of both plastic strain range and previously accumulated plastic strain. This representation of hardening parameters is able to accurately match experimental results with different types of loading programs including random loading conditions and considering initial monotonic behavior with yield plateau deformation. Ratcheting behaviour is simulated by the developed cyclic plasticity framework by considering an approximated form of the Dirac delta function for modelling the deviation effect and introducing an additional supersurface for better prediction of ratcheting rate. The proposed cyclic plasticity model requires up to 21 material constants, depending on application. A clear and straightforward calibration procedure, where sets of material constants are determined for each plasticity phenomenon considered, is presented. Application of the model to different materials under various tension-compression and non-proportional axial-torsion cycles shows very close agreement with test results.

No abstract