The accuracy of the chemical composition for the production of steel alloys casting parts has great importance. This is due to the dependence that the desired mechanical properties, the occurrence of liquid or solid inclusions and behavior during heat treatments from this accuracy. Its adjustment becomes a challenge in the loading step, when little is known about the composition of the scraps and one deals with various raw materials and prices. Thus, the authors, in partnership with the casting industry ecoplan aços especiais rs, developed models of metallic yield and of charge using mass and fuzzy sets, along with thermodynamics. Its implementation was made in C#. Compared with the literature, the findings proved to be technically equivalent and/or better economically. Compared with the industry, the model reached up races for the tested alloy, sae 1045, with expected chemical composition within specification and with economy for the step. The authors also proposed indicators of cost-benefit and supply of alloying elements. Finally, using anova, this article compares the indicators obtained with the model results and those of the industry. The indicators ζ and ζFe were found smaller and larger than those of the industry, respectively, both being statistically different. the indicator τ showed statistical equivalence.
steel; charge; foundry; model; optimization; planning
