Numerical optimization of transverse flux induction heating systems

Abstract

Transverse flux induction heating (TFH) of flat metal products is one of the most effective induction technologies. This method provides very high efficiency in combination with unique technological flexibility and extremely low floor space required. It makes TFH beyond competition to be applied in continuous strip production and processing lines. However, to realise all potential advantages of TFH concept is possible only by optimal design and control of heating installations. Experience of last years shows that successful creation of TFH systems can be only based on numerical modelling. At the same time even advanced 3D simulation codes do not provide exactable engineering solution by themselves. Additional application of automatic optimisation techniques is only the way to overcome this situation. This paper is devoted to application of most effective optimisation algorithms for atomised design of TFH systems. Genetic algorithms of optimisation are more and more often used in engineering science because of their unique possibilities. Use of genetic algorithms in combination with 3D electromagnetic and thermal analysis of TFH systems allows to make optimal 3D shape design of inductors. This problem can not be solved in effective way by conventional design methods.