Structural rheological model for calculating fatigue life according to a corrected linear hypothesis of damage accumulation

The paper proposes a method for calculating the fatigue life under multi-cycle loading. The method is based on the use of a structural rheological model. The paper considers multiparameter random loading when random changes in stress tensor components are independent. The structural model is characterized by a set of yield surfaces (spheres) in the space of the strain deviator. To develop it, we decribed the material deformation diagram with the Ramberg - Osgood equation. The error of the material deformation diagram approximated by this dependence usually does not exceed 5 %. The structural model was used to calculate microplastic deformations that cause an accumulation of damages under high-cycle loading. The paper presents the basic relations of the method and the technique for identifying the damage accumulation model. The identification of the model is based on the usual fatigue characteristics of the material. To improve the accuracy of the estimated durability, it was proposed to use a corrected linear hypothesis of damage accumulation. The paper presents an algorithm for plotting a loading block and determining a correction coefficient based on the calculation results according to the structural model. The results of test calculations are given. It was shown that for the case of one-parameter random loading, the results of the proposed method are consistent with those of traditional approaches. For the case of multi-parametric loading, we compared the calculation results with experimental data for a plane stressed state. The difference in the resource evaluation does not exceed 15-20 %. Given the large natural characteristic spread of materials, such a difference is quite acceptable. The paper proposes an algorithm for determining the number of cycles of different amplitudes in a random process. The algorithm is based on counting the number of reverses during the displacement of yield surfaces of the structural model. This algorithm may be of interest as one of the ways of schematizing random processes.