Aplicação de um método de otimização no projeto de fornos para tratamento térmico

Abstract

The design of a heat treatment furnace can be carried out by trial and error method. However, this practice can lead to inaccurate results, as well as warping and cracking on the surfaces of materials to be heat treated. Thus, in this work a three-dimensional cavity filled with transparent medium is used to simulate the operation of an annealing heat treatment furnace. The problem is reversed, as installation positions for the heaters and their respective heat flows must be found, these being the entry conditions. So, it is modeled using the Radiosity Method, with a system of linear equations that is solved by the Gauss-Seidel method. To perform the optimization, the proposed solution is the application and implementation of an algorithm based on the Fortran programming language, an optimization method known as the Generalized Extremal Optimization Method (GEO). Some configurations, considering the number of heaters, were tested and the results obtained, analyzed for the distribution of flows on the design surface. After the implementation of this algorithm, reductions were found in the deviations presented by the conventional method. There was also a satisfactory average computational time, considering the parameters to be calculated (spatial distribution of the heaters and their respective heat flows). The results are visible both graphically and numerically, so that it demonstrates the effectiveness of the optimization method.