Parameter Indentification for a Viscoplastic Selfconsistent Model based on Analytical Derivatives of an Objective Function

J.W. SIGNORELLI, ; R. LOGÉ; Y. CHASTEL; R. LEBENSOHN

MODELLING AND SIMULATION IN MATERIALS SCIENCE AND ENGINEERING

IOP Publishing

Año: 2000 p. 193 - 209

0965-0393

An inverse method for automatic identification of the parameters involved in apolycrystalline viscoplastic selfconsistent (VPSC) model is presented. The parameters of theconstitutive viscoplastic lawat the single-crystal level, i.e. the critical resolved shear stresses (CRSS)of slip and twinning and the micro-hardening coefficients, can be identified using experimentaldata at the polycrystal level, i.e. stress?strain curves and deformation-induced textures. Theminimization problem is solved by means of a Gauss?Newton scheme and the sensitivity matrixis evaluated by analytical differentiation of the direct model equations. As a particular case, theoptimization procedure for the Taylor full constraints (FC) formulation is also presented. Theconvergence and stability of the identification scheme are analysed using several validation testsfor different deformation paths imposed to a polycrystal of hexagonal structure. As an exampleof application of this inverse method, the relative CRSS of the active deformation systems of aZircaloy-4 sheet are identified, based on several textures measured for different reductions androlling directions.