Theoretical studies on CFT-column frame SPSW/BRSPSW linked by partial welded connection forms

Steel plates, being slender structural elements, are susceptible to early failure caused by localized buckling. This buckling significantly reduces their overall capacity and stiffness, compromising their structural performance. Therefore, this article presents a comprehensive computational framework to accurately predict the inelastic behavior of steel plate shear walls (SPSWs) and buckling-restrained steel plate shear walls (BRSPSWs) with various partially welded connections. Detailed finite element (FE) modeling techniques were developed and validated against an extensive experimental database, systematically accounting for factors such as material constitutive models, mesh density, contact interactions, and geometric imperfections to ensure high accuracy predictions. Using the validated FE approach, comprehensive parametric studies were carried out with an efficient pushover loading scheme to examine four representative cases: four-corner/double-side welded connections for SPSWs and BRSPSWs (FC/DS-SPSW/BRSPSW-W). The results highlight that the SPSW/BRSPSW-W are significantly influenced by the height-to-thickness and length-to-height ratios. Besides, they are gradually impacted by the connection's dimensions, while they are influenced minimally and negatively by panels thickness and axial ratios. Accordingly, a novel analytical model was derived to predict the buckling coefficients for all four connection cases, showing that the buckling models of both FC/DS-SPSW-W are impacted by length-to-height while the FC-DS-BRSPSW-W are impacted by length-to-height and stiffness buckling restrained. Furthermore, a modified plate frame interaction (MPFI) method was then formulated by integrating parametric FE data and analytical buckling model to enable accurate calculations of yield/peak bearing capacities and yield/post stiffnesses. Fundamentally, the MPFI predictions were comprehensively validated against a combined database of experimental test data and FE modeling.