Evaluating fatigue life of materials and structures under low-cycle loading

Abstract

The problem of evaluating strength and fatigue life of critical engineering objects, the operational properties of which are characterized by multi-parametric nonstationary thermal-mechanical effects, is discussed. The main degradation mechanisms of structural materials (metals and their alloys) characteristic of such objects are considered. The main requirements for mathematical models of fatigue damage accumulation are formulated. To evaluate the stress-strained state (SSS) and fatigue life of critical engineering facilities (CEF) in the framework of modern achievements of mechanics of damaged media (MDM), a mathematical model is developed which describes the processes of cyclic thermal-plastic deformation and fatigue damage accumulation in structural alloys subject to multiaxial non-proportional paths of combined thermal-mechanical loading. The MDM model consists of three interconnected parts: thermal plasticity constitutive equations, accounting for their dependence on the failure process; an evolution equation of damage accumulation and a strength criterion of a damaged material. To qualitatively and quantitatively assess the constitutive relations of MDM for low-cycle modes of thermal-cyclic loading, the numerical and experimental results for the fatigue life of a notched compact sample in a non-uniaxial stress state under block modes of thermal-cyclic loading were compared, which supported the validity of the constitutive relations of the MDM model for thermal fatigue. An approach based on the Palmgren-Miner rule of linear summation of damage, when used for evaluating the fatigue life of structural elements, may result in both conservative and non-conservative estimates. This conclusion is confirmed by the experimental and theoretical studies of foreign and domestic authors. The results of numerical modeling of the fatigue durability of a compact specimen with an obtuse notch are presented for two loading histories. It is shown that several cycles of more intensive loading at the beginning of the strain history are the reason why the rule of linear summation of damage can overestimate or underestimate the value.