Concrete Uniaxial Nonlocal Damage-Plasticity Model for Simulating Post-Peak Response of Reinforced Concrete Beam-Columns under Cyclic Loading

Abstract

Rigorous predication of localized deformations in reinforced concrete (RC) structures under cyclic loading is critical from the standpoint of seismic performance assessment. Among the available predictive tools, the fiber-discretized frame model is an attractive option for RC components because it captures the spread of plasticity and the interaction between the bending moment and axial force in a structural member, and can be generalized to different cross-sections from uniaxial material-level calibrations. However, in the presence of constitutive softening, this type of model suffers from pathological sensitivity to the mesh size of the finite element simulation, leading to nonphysical member response. A nonlocal methodology is presented to address these issues for RC beam-columns subjected to a combination of axial and cyclic lateral loads. The methodology is based on a uniaxial nonlocal constitutive model that is formulated in the combined framework of the theory of plasticity and damage mechanics. The model captures the observed strength and stiffness degradation of the concrete material under uniaxial compressive and tensile loading in addition to tension-compression transition effects. The model incorporates a length scale parameter that enforces interactions between neighboring material points, thereby overcoming the mesh sensitivity associated with the presence of constitutive softening. The scope of this study includes (1) developing the uniaxial damage-plasticity formulation for a fiber-based frame model, (2) developing a nonlocal damage formulation and proposing a simplified implementation approach to reduce the associated computational cost, and (3) interrogating the effect of different model parameters and making recommendations for characterizing all the associated parameters for RC frame member simulations. The performance of the nonlocal model is thoroughly assessed, and its predictive capability is demonstrated against experimental test data of 24 RC beam-columns subjected quasi-statically to reversed loading cycles. The limitations of the nonlocal methodology are discussed, and future research directions are highlighted.