Search for: All records

attributed to loss of load carrying capacity of the individual members. Dominant failure modes in structural steel members include interactions between inelastic lateral torsional buckling, global buckling, and local buckling (referred to as Interactive Buckling). Accurate performance assessment of steel moment frames highly relies on the accuracy of the model-based simulations of such limit states. Commonly used concentrated hinge and fiber-based models fail to address the physics of this response leading to inaccurate performance assessment of structures. A nonlinear displacement-based fiber element [named Torsional Fiber Element (TFE)] to simulate monotonic and cyclic interactive buckling in steel members is proposed and implemented on OpenSees (an open-source finite element software). The element includes St. Venant as well as warping torsion response that are essential for lateral torsional buckling response in a wide-flange I-section, through enriched displacement fields and strain interpolation. Response of local buckling is represented in a quantitative manner using a novel multi-axial constitutive relationship with calibration of an effective softening behavior in the post-buckling response. Mesh dependency issue related to the softening material model is also discussed and addressed through a proposed non-local strain measure. The efficacy of the model is assessed through several continuum finite element simulations and experimental data. 

Abstract Mitigating fracture in welded column splices is an important challenge for the safety of existing steel moment‐resisting frames. While models to predict splice fracture have recently been developed, suitable approaches are not available to simulate the response of frames after splice fracture. Motivated by this, a two‐dimensional displacement‐based fiber element construct, termed the Splice Fracture Element (SFE), is presented. The SFE includes numerous features: (1) representation of the loss of strength in any fiber at a critical stress determined from fracture mechanics, (2) the ability to simulate the loss of shear strength of the cross‐section when the entire section is severed – a phenomenon not readily simulated in conventional fiber elements, and (3) the ability to track the kinematics of the severed parts of the column to represent transfer of compressive stresses on contact. This formulation is implemented into an open‐source software (OpenSees) and applied to conduct Nonlinear Response History Analysis (NLRHA) of two demonstration problems, including a 1‐story frame and a 20‐story frame. Benchmark simulations that do not simulate splice fracture or represent it without the loss of shear strength are also conducted. The results indicate that the SFE element can successfully simulate the key phenomena associated with splice fracture and post‐fracture response. 

A framework to assess the fracture fragility of partial joint penetration (PJP) welded column splices in steel moment frames constructed before the 1994 Northridge earthquake is presented. These pre-Northridge splices feature low flange penetration of the PJP welds, and low-toughness weld materials, such that they are considered susceptible to fracture with possible catastrophic consequences. Estimating their fracture risk is especially important, given that retrofitting them is highly disruptive to building operations. The presented framework addresses shortcomings of previous research and performance assessment guidance that does not consider key mechanistic or statistical effects. To accomplish this, three-dimensional fracture mechanics finite-element simulations are conducted to assess fracture toughness demands. These demands are then interpreted through a master curve–based approach that rigorously considers spatial randomness and weakest-link sampling of weld toughness properties, along with the uncertainty in estimation of these properties. The framework is implemented in a tool which automates the entire process, facilitating application in a professional setting. The tool (and the underlying framework) is demonstrated on a range of splice configurations to examine the effects of configuration, loading, and material parameters. Limitations are outlined.