A new material model for thermo-mechanical analysis of steels in fire

Abstract

Recently developed multi-hazard frameworks for performance based engineering require material models capable of incorporating the unloading/reloading of structural materials at elevated temperatures. These include models for fire-following-earthquake events and the travelling fire methodology (which highlighted the prevalence of material’s unloading/reloading). This paper proposes a new combined isotropic-kinematic hardening material model that has been developed to assess the unloading/loading behaviour of steel materials for thermo-mechanical analysis with fire, accounting for the Bauschinger effect and transient hardening behaviour. It works within the classical rate independent plasticity framework by exploiting the latitude of the two yield-surface model. The proposed material model integrates temperature effects on the yield surfaces through the concept of a shrinking/expanding bounding surface, and on the Bauschinger effect using an exponential growth function of plastic internal variables (PIVs). Transient hardening behaviour is modelled by incorporating a second non-linear kinematic hardening variable using an exponential decay function of the PIVs, and a corresponding discrete memory parameter tracks any abrupt changes in the direction of plastic loading. Describing the Bauschinger effect and its associated transient hardening behaviour using an opposite pair of exponential functions is a novel solution that simplifies the computing effort required by classical material models developed purposefully for cyclic loading. Hence it makes the proposed material model appropriate and efficient for structural analysis with fire. The proposed material model has been implemented in Abaqus using the Umat subroutine, and verified against experimental data where good agreement was observed. Its application in analysing structures subjected to complex, realistic building fire is also demonstrated, indicating that it is suitable for incorporating within performance based engineering frameworks for structures in fire.