The ability to model the cyclic behavior of low yield point steel accurately is essential for its application to seismic design. In this study, uniaxial tests of monotonic and strain controlled cyclic loading are conducted to investigate the mechanical properties of the low yield point steel BLY160. As observed in the test, work-hardening is an important hysteretic characteristic, which significantly influences the energy dissipation capacity of the steel. However, the cyclic softening in both stiffness and stress evolve as the plastic strain accumulates, which is manifested in the decrease of both the stress amplitude and the elastic domain of the stress–strain hysteresis loops. This change in the shape of the cyclic loops is called the “flattening effect” in this paper. To describe the cyclic hardening and softening effect by using one unified set of material parameters, a more sophisticated model based on the concept of internal variables is proposed as an extension of the combined isotropic-kinematic hardening model. The novel idea of a “transformation zone” with the introduction of an additional internal variable is proposed to synchronize the updates of isotropic and kinematic hardening such that the simultaneous change in the stress amplitude and the elastic region can be reproduced. Note that the cyclic softening develops in three stages: incubation, growth and the stabilization period. The logistic function, which presents a curve with an “S” shape, is first used to trace the evolution of the softening. Finally, the model parameters are identified by the test data, and the accuracy of the proposed model is verified by comparing the simulated performance and the experimental response under different loading histories. The results demonstrate that the extended model can elaborately describe the evolutional processes of cyclic hardening and softening by incorporating characteristics, such as the nonlinear transient behavior from elastic to plastic status, early yield under reversed loading and strain history dependence.
Read full abstract