Simulation of residual stress heterogeneities using cellular automaton

点击翻译按钮获取中文摘要

The proposed research is a continuation of a previous discovery grant which was dedicated to the simulation of the cold creep / fatigue interaction behavior of titanium alloys using Cellular automata (CA). In the present proposal, CA will be used to investigate the residual stress distribution and localization at the microstructural scale in manufactured material. This will help the development of statistical based models for fatigue life prediction of surface treated parts. CA which are fully discrete, spatial-distributed dynamical systems which serve as an alternative framework for mathematical descriptions of physical systems. In addition to providing computational advantages, CA also offer a relevant framework for mathematical analysis, however, they have not been extensively used in the framework of material science and mechanical engineering. The principal objective of the proposal is to developed CA models able to estimate the residual stress heterogeneities induced by surface treatment processes. Secondary objective will include the development of experimental technique to validate the developed models and the adaptation of microstructure based fatigue models to take into consideration the statistical aspect of surface residual stress. Surface treatment such as machining, surface peening (hammer or shot), or induction hardening will be investigated. Experimental validations will also be covered by the present proposal. The distribution of the residual stress as measurement by Xray diffraction techniques provides a first validation parameter. However, the precise experimental validations of the residual stress distributions and their local correlation represent a big challenge. The scale at which the measurements have to be done will require the development of original and precise techniques. One of the ways to achieve such data is the use of a in-house correlation image software developed with the research institute of Hydro Quebec. Finally, the statistical results gathered by the CA models will be integrated to various microstructural based model for fatigue life prediction.