Optimization-heuristic of mechanical properties of acicular ferrite steel

This paper presents two algorithms, Simulated Annealing and Iterated Local Search. Both metaheuristics use a neighborhood hybrid structure to evaluate their effectiveness and maximize the mechanical strength of microalloyed steel. Tests show that the best metaheuristic for this type of problem, which makes use of a neighborhood structure and a chemical composition, is Iterated Local Search because it gives a better mechanical strength than Simulated Annealing. Acicular Ferrite was developed in the laboratory using the best mechanical properties obtained by the heuristics in computational tests. Then the mechanical strength of the created steel was evaluated. The experimental results show that the yield strength obtained in the laboratory is comparable to that obtained in computational tests.