Rule-based expert system application to optimizing of multiscale model of hot forging and heat treatment of TI-6AL-4V

Abstract

Nowadays, multiscale methods are increasingly used in modeling of processes and designing of materials. However, there are two serious obstacles to wider application of such methods. One is a demand of computing power, which is present in almost all applications. Increasing of available computing power is not a solution. Exponential growth of computational complexity with increase of accuracy will always lead exceeding of available resources. Therefore, adaptation of multiscale models is important point of interest. The second obstacle is a difficultness of multiscale model design. Model is an abstraction of reality. Therefore, some assumptions have to be done – which phenomena are important and which do not. There is an additional point of interest in multiscale modeling – in which scale the particular phenomena can be modeled with a good balance of accuracy and computing power demand. Usually, there is no single good answer. Designing of a model can be seen as a kind of an optimization process, where the goal is a function of efficiency and reliability. The need of a such process is pointed out in many publications (e. g. [1]), but in the most of cases it have to be done by a researcher himself.