Simulation of damage evolution during hot ring rolling

Abstract

Ring rolling provides a cost-effective process route for manufacturing seamless rings. The bearing rings, i.e. the inner race and the outer race of a rolling bearing, are typically manufactured by the hot ring rolling process. Bearing steels have a relatively high alloy content for improved hardenability in large cross sections and they are more susceptible to problems of porosity. To reduce material waste and improve product quality, a better understanding of the relations between parameters in the hot ring rolling process and the occurrence of porosity is needed. In this thesis a numerical model is developed to simulate the hot ring rolling of bearing steels and to predict the damage evolution in the bulk of the ring. Hot ring rolling tests including preform forging were conducted to provide experimental evidence. Coupled thermo-mechanical multi-stage finite element analysis was performed with a damage indicator based on integration of positive stress triaxiality. In spite of the suggestion of a more careful process when a low ring growth rate is used in hot ring rolling, experimental and numerical studies demonstrate that with a low ring growth rate there is a higher susceptibility to damage than when a high ring growth rate is applied. To evaluate damage evolution of the material in a controlled laboratory experiment, the standard Gleeble MAXStrain unit was modified to allow axial elongation during multi-axial compression tests. The tested specimens were inspected by means of X-ray micro-computed tomography to reveal the presence of voids after the test. Depending on the depth of applied hits, different triaxiality histories were obtained from the finite element analysis. An estimate for a parameter that determines a negative triaxiality as the threshold for healing under compressive stress in Oyane’s damage evolution model was derived, based on numerical and experimental results. A demonstration of metamodeling-based optimization is presented to find proper process settings for the hot ring rolling process including the preform forging. The metamodel is used as an approximation of the process model with a much lower solution time. The maximum damage in the interior of the cross section is minimized by an optimization solver.