A genetic algorithm for optimizing gravity die casting’s heat transfer coefficients

Numerical simulation of solidification has improved our understanding of casting processes significantly 16
over the last two decades. One of the most desirable features in the design of casting of high strength 17
components is directional solidification. Generally, expertise from skilled foundry men is required during 18
the design of casting-mould assembly interrogation in order to achieve a satisfactory thermal control, 19
thus directional solidification. This process is not only costly, both financially and temporally to found- 20
ries, it also heavily rely on foundry men’s experiences. Our main aim in this project is to explore a novel 21
and fully automated computer scheme that ties the geometric features of the casting with evolutionary 22
algorithms to achieve thermal control. By extracting the medial axes of the casting geometry and corre- 23
late it with the interfacial heat transfer coefficient via evolutionary algorithm, we are able to perform 24
non-exhaustive search of the optimized solution. Preliminary results from our computer experiments 25
showed favourable results. In this paper, the focus is sharpened on the convergence and optimality of 26
the developed GA.