Multiscale lattice discrete particle modeling of steel-concrete composite column bases under pull-out and cyclic loading conditions

Steel-Concrete Composite (SCC) connections in steel buildings are inherently complex in terms of internal stress distributions and failure modes that are important to characterize for effective design and performance assessment. In this context, numerical results obtained from a multiscale lattice discrete particle model are presented to examine its efficacy in characterizing the response of SCC connections (specifically embedded column base connections in steel moment frames) subjected to earthquake-like cyclic loading. In the numerical model, mesostructural information on the concrete base is described at the level of the constituent materials, allowing to capture initiation and propagation of fracture resulting from cyclic loading. To address mesh sensitivity at the macroscopic level, an energy regularization approach, based on a generalization of the crack band theory, is proposed and validated. The proposed regularization technique mitigates mesh sensitivity in the simulations, while significantly reducing the overall computational cost. A multiscale validation of the model is presented by comparing the numerical results with experimental data obtained from independent cyclic tests on SCC connections, indicating reasonable accuracy across a range of test parameters. Limitations of the approach are discussed.