PSI - Issue 78

Tahir Ahmad et al. / Procedia Structural Integrity 78 (2026) 631–638

632

1. Introduction In steel storage rack systems, collapse often initiates from localized damage caused by unexpected loads. These loads can be categorized based on their origin, including impact loads (e.g., from forklifts), dynamic forces (such as earthquakes), or material-related incidents (e.g., fires). Such localized initial damage can escalate through the structure, leading to further failure of surrounding components and potentially resulting in a disproportionate, system-wide collapse. Case studies of progressive collapse in racking systems show that key structural elements - especially uprights - typically fail first due to impact forces. This triggers a change in the load-bearing path. Because the remaining structure often cannot redistribute internal forces efficiently, adjacent sections may also fail, causing a chain reaction or "domino effect" collapse. These events can cause significant injuries, fatalities, and financial losses, highlighting the urgent need for understanding and mitigating progressive collapse in rack structures. Most current research into the dynamic behavior of storage racks focuses on seismic and forklift-induced impacts. Montuori et al. (2019) predicted the seismic collapse modes of steel storage pallet racks. A wide body of experimental, numerical, and analytical studies has been developed, and their findings are incorporated into design guidelines like FEMA 460 (2005), FEM 10.2.08 (2011), and EN 16681 (2016). Forklift impact scenarios have also been examined, as they are known to cause full structural collapse. According to studies on shelving racks impacted by forklifts (McConnell and Kelly, 1983), rack failures can be broadly categorized into three types: no collapse, localized (limited) collapse, and progressive collapse. However, many of these investigations concentrate on overall collapse mechanisms and often overlook the specific processes associated with progressive collapse, particularly the vital role of connections and how internal forces are redistributed, including the formation of catenary action. Moreover, current design codes for storage racks do not sufficiently address progressive collapse scenarios. Although a substantial amount of research has been conducted on the static and cyclic performance of beam-to-upright connections in storage racks, there is still limited work focused specifically on how these connections behave during progressive collapse. In such scenarios, three distinct aspects - largely absent from previous studies - require careful evaluation: the effects of load reversal; the interaction of tensile axial and bending forces on beams and their connections; elevated axial forces in columns adjacent to a removed column. When a primary load-bearing column fails, beam-to-upright connections can experience reversed loading. Due to the nature of their design, these connections are prone to initial loosening under reversed loads, increasing the likelihood of root cracking at the tabs, disengagement from column perforations, and eventual failure. Under typical static or seismic loading conditions, connections are primarily subjected to bending, and the influence of axial tension in beams is often neglected. However, after a column is removed, these connections must withstand both axial and bending forces, making the contribution of axial tension in beams a critical factor in evaluating the performance of beam-to-upright connections in progressive collapse scenarios. When a column fails, the columns close to the removed one suffer a dramatic increase in the axial force that may influence the behavior of the columns and base-plate connection under biaxial bending. This paper presents the progressive collapse assessment of a typical steel storage rack under different column removal scenarios. The behavior of members and connections subjected to the typical forces following the failure of a column were derived using tests, FEM and effective section properties due to local buckling defined according to EN 15512 (2020). Two column failure scenarios were considered, trying to reproduce the more unfavourable forklift impact loads. Both nonlinear static and dynamic analyses were carried out to evaluate the progressive collapse behaviour of the steel storage rack. 2. Modeling of steel storage racks under progressive collapse scenarios involving column removal 2.1. Beam-column connection subjected to tensile axial forces The nonlinear behavior of beam-to-column connections was characterized through a combination of experimental testing and detailed finite element modeling (FEM). Initially, an experimental program was carried out on representative beam-column connection specimens, as shown in Fig. 1a, to directly observe and quantify their nonlinear response. In the test setup, the column was fixed at both the base and the top, while a displacement-controlled load was applied at the tip of the beam. Failure occurred at the beam – column interface. These tests provided essential baseline data on the moment – rotation capacity of both the beam cross-section and its connection. Subsequently, a detailed finite element model of the tested beam-column connection was developed in ABAQUS (2024), as illustrated