Modeling and simulation of an expert heat treatment system for plain carbon steels

In this paper, modeling and simulation of an expert heat treatment system for plain carbon steels has been presented. The aim of this paper is to model and simulate an expert system for an electric muffle furnace for optimizing and controlling the simple heat treatment processes like annealing and normalizing in case of plain carbon steels. A heat treatment process consists of three stages-heating, soaking and cooling. This paper deals with the heating stage only. The heat treatment system described here consists of three parts-expert system, furnace temperature controller and muffle furnace. The expert system takes decision regarding the temperature set-point or the desired furnace temperature based upon the inputs provided by the user and the in-built knowledge base. The furnace temperature controller maintains the desired temperature within the muffle furnace and is based on adaptive neuro-fuzzy inference system. The mathematical modeling of the furnace was also done. The simulation of this whole heat treatment system was done on MATLAB/ SIMULINK 2007b. The results obtained after simulation were compared with the results obtained by experimentation on an actual electric muffle furnace. The two results were found to be in good agreement with each other. It was also found that the expert system gave the correct values of desired furnace temperature for different types of annealing processes for plain carbon steels.