Modeling Adiabatic Shear Banding in Martensitic Low Carbon Steel

Resolving adiabatic shear banding computationally has long been a challenge. With heating, the mobility of dislocations increases, and mechanical softening is realized within the shear zone material. A new thermodynamically based model to describe the thermal softening and dynamic recrystallization process in a martensitic low carbon steel is presented. This model employs both dislocation density and grain boundary density as primary structural state variables. Shear band initiation for deformation rate-dependent materials is takes place when the mechanisms of structural hardening and softening rates become equal in magnitude. These models are implemented into a computational framework designed to alleviate systemic mesh sensitivity in simulations of adiabatic shear banding. This computational framework embeds an adiabatic shear band within a computational element in a weak sense as a jump condition with an adapted Gauss quadrature rule which does not restrict shear band placement. A non-local level set technique is also employed to enable spatially uniform shear band growth so that mesh imprinting is significantly minimized. Experimental and simulation results are presented which describe both development of adiabatic shear banding but also structural evolution during dynamic recrystallization during deformation of this martensitic low carbon steel.