A sensitivity study of the process model for predicting the distortion during heat treating

Distortion during heat treating processes is mostly troublesome. The ability to predict the distortion relies upon the accurate calculation of the complex thermal, mechanical and metallurgical changes in response to the volatile heating/quenching cycles during the heat treatment processes. The state of art in predicting the distortion is the use of computer modeling systems based upon the finite element method (FEM). The success of the process model seems to be a matter of chance; very often, the predictions do not agree in a reasonable range with the measurements. The uncertainties could weaken the motivation and real benefit of using the FEM program for process modeling. In this paper, we present a study of the process modeling system for predicting the part distortion in the quenching stage of a heat treating process. The objective is to identify the critical factor(s) which dictate the predictions. The sources contributing to the variations in modeling results are broadly divided into three categories: deterministic, statistical and stochastic. We examine some of these affecting factors from each category. As a first attempt, we use a material system which has no microstructural changes in the quenching process. Distortion due to the subsequent material removal is also examined.