Cyclic Nonlinear Modeling Parameters for Unconfined Beam-Column Joints

Nonlinear macro-models of concrete components are routinely used for performance-based seismic assessment of existing "non-ductile" buildings. Accurate test-calibrated models improve cost and time effectiveness of rehabilitation. ASCE 41-17, the primary standard in the United States for seismic assessment and rehabilitation, offers nonlinear modeling parameters for concrete frame beams and columns that were recently updated to reflect median esti-mates (50% probability of exceedance). However, the ASCE 41-17 beam-column joint nonlinear modeling provisions are still based on lower-bound estimates shown by many researchers to be highly conservative. Moreover, existing literature backbone models are not suitable for shear-critical "non-conforming" unreinforced joints as they are mostly deterministic, non-sensitive to failure mode, designed for reinforced joints, or calibrated based on small unreinforced joint datasets. This paper proposes new probabilistic nonlinear modeling parameters for shear-critical unconfined (unre-inforced) beam-column joints based on a database of cyclic tests to update the current ASCE 41-17 modeling parameters. Three alter-natives for deformation modeling parameters are proposed, along with two alternative expressions for residual shear capacity ratio. The proposed models offer the analyst to select the desired prob-ability of exceedance to enable consistent modeling criteria with beams and columns in concrete frames to avoid nonlinear analysis bias that is currently rendering joints unrealistically as the weakest link in non-ductile frames, leading to unnecessary costly retrofits. All the proposed models exhibited very good correlation with the test database and verification dataset. The proposed models offer immediate input to reduce the conservatism of the current seismic assessment standards for unconfined beam-column joints, which resulted from test scarcity at the time of their development.