SUBSURCFACE MATERIAL CHARACTERISATION BY SIMULATION ON CARBURIZED ROLLING CYCLE FATIGUE TEST SPECIMENS

The aim of the research is to determine the metallographical and mechanical properties (residual stresses, hardness, phases and texture) of carburized surfaces using finite element method and study the failure of surface during relative motion, and the possibilities of the test methods, which can give us information about the behaviour of these surfaces. This work focuses on contact fatigue tests and properties. Nowadays it can be established that application of finite element method is a reliable way to simulate carburization. It is possible to determine exact approximation for the hardness and carbon content values near the surface after carburization and quenching, but take into account that the simulation of tempering is still an issue, if the material contains retained austenit, the hardness values of the test will be higher near the surface, because of the retained austenit-bainit transformation, so the simulations had to be validated by carburized cylindrical (disc-like) specimens, made of low alloy case hardening steel (16MnCr5). These specimens were tested with glow discharge optical emission spectroscopy, to measure the chemical composition of the surface layer. It is important to know what kind of residual stresses can be observed in the surface after carburization, therefore it was extensively studied during the simulation. Furthermore the values of the hardness in the surface of the specimens have been determined. Having compared the simulation and experiments, good agreement has been found between the results, so the finite element method was validated successfully. For high cycle rolling fatigue test rollers were manufactured. Two disc fatigue tests were carried out on five rollers. Four tests were stopped due to trigger sign by surface damage, while one of those was stopped due to break of the shaft after one millions of rotation. Microstructural investigations have been started on the rollers.

Abstract licence: CC BY-NC

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.