IDENTIFICATION OF PARAMETERS FOR CHARACTERIZATION AT HIGH-TEMPERATURE OF 38MnVS6 STEEL USED IN HOT FORGING PROCESSES

Abstract

Finite element reliability is highly sensitive to the input data quality and constitutive equations used. In this article, two constitutive equations for modeling the plastic flow of metals are used. Two equations commonly employed in metal forming finite element analysis were selected, e.g., the Hensel-Spittel and Johnson-Cook. Compression tests were conducted at various temperatures and strain rates to obtain the plastic flow stress curves. These experimental results were used as a target to fit the constitutive equation parameters. The parameter identification procedure was conducted by solving the so-called inverse problem. An objective function gives a value (error) that indicates how well a parameter set makes the equation fits the experimental results. This error value is minimized by applying optimization techniques; in this case, by using a Genetic Algorithm followed by a gradient-based strategy. Two objective functions are used; the first is based on the well-established least square approach, and the second is proposed. The proposed objective function shows a faster performance than the least square. In addition, the proposed objective function also gives the best fit for the Hensel-Spittel equation. The Johnson-Cook equation is fitted using only the proposed function. The Hensel-Spittel equation shows several local minimums, whereas the Johnson-Cook shows a global one. Keywords: Hot forging, Parameter identification, constitutive equations, finite element analysis.