This paper presents fatigue experiments of single-sided V-shaped butt welds in very high strength rolled and cast steel plates in as-welded condition. The rolled plates were of grades S460, S690, S890, and S1100 (nominal yield stress between 460 MPa and 1100 MPa) and the cast plates were of grades of similar strength. The fatigue resistance of the specimens with cracks initiating at the weld root was lower than that of those with cracks initiating at a weld toe. A positive but weak correlation was observed between ultimate tensile strength and fatigue resistance. This positive correlation was further confirmed by evaluating large databases of butt welds. The evaluation also confirmed the hypothesis that a stronger correlation occurs for welds with smaller notch effects, i.e., ground flush butt welds. Some specimens with casts delivered in Quality Class II according to ASTM E 446 failed in the cast plates, and these initiated from cast imperfections such as shrinkage cavities. Additional experiments delivered in Quality Class I did not fail in the cast base metal, or failed only after repairing prior cracks that grew from the welds. Unexpectedly, some failures were observed in the rolled steel plates away from the V-shaped butt welds (and away from the grips). Metallurgical examination revealed that the cracks in these specimens initiated at ground flush welds applied in the rolling direction, in delivered condition, likely to repair rolling flaws. Therefore, repair by welding of rolling flaws should not be allowed if relying on the high fatigue resistance of the base metal.
Cold-formed steel (CFS) shear walls with steel sheathing have gained widespread adoption in multi-story buildings owing to favorable seismic performance. However, existing design approaches such as the effective strip method (ESM), tabulated methods, and numerical simulations suffer from inherent trade-offs among prediction accuracy, computational cost, and applicability. To address these limitations, this paper develops a machine learning (ML) framework for reliably predicting the shear strength for such shear walls. A hybrid database combining 230 experimental and 100 validated finite element (FE) results was developed, with input features standardized via Z-score normalization and stratified sampling applied to preserve data representativeness. Six ML models, including Ridge Regression (RR), Lasso Regression (LR), Support Vector Regression (SVR), Decision Tree (DT), Random Forest (RF), and eXtreme Gradient Boosting (XGB), were trained and evaluated against current methods in AISI S400. Results indicate that ensemble models, especially XGB and RF, achieve superior predictive accuracy and robustness compared to linear models. Notably, the proposed XGB model consistently outperforms existing codified methods across U.S., Canadian, and other regional datasets, while quantifying the contribution of the blocking to shear strength. Reliability analysis further yields a resistance factor of 0.82 for the XGB model, with a divisor of 1.33 to maintain consistency with the AISI design framework.
This study addresses the dual challenges of seismic resilience and environmental sustainability by investigating a novel, reusable, and replaceable hybrid Steel-Nitinol (NiTi) Shape Memory Alloy (SMA) device developed for coupling beams in shear wall systems. Unlike conventional reinforced concrete coupling beams, which are typically characterized by reinforcement congestion, pinched hysteresis, and significant residual drift, the proposed system concentrates inelastic action within mechanically spliced, superelastic NiTi oval dampers that function as replaceable structural fuses.The research progressed from material-level characterization to full-scale experimental testing of a coupled wall frame. Utilizing the FEMA 461 cyclic loading protocol, the experimental program evaluated the influence of different damper configurations, including single (1OD), double (2OD), and triple (3OD) oval SMA arrangements, on the global cyclic response. A bare system without SMA dampers was also tested to establish a baseline for the inherent damping and energy dissipation behavior of the structural system. Experimental results demonstrate that the SMA-integrated system exhibited a stable, flag-shaped hysteretic response and improved residual displacement control up to the final loading amplitude of 55 mm. This maximum displacement corresponded to approximately 25% strain within the dampers, aligning closely with material-level characterization. At this ultimate demand level, individual oval devices resisted peak forces of 11 to 13 kN depending on the loading rate, with final failure governed by SMA rupture. Increasing the number of SMA elements led to a consistent enhancement in lateral force capacity and energy dissipation. The 3OD configuration delivered the highest performance, demonstrating superior re-centering behavior driven by the heightened restoring forces relative to the baseline system. Furthermore, the capacity contribution of each additional damper was found to be approximately additive, confirming a predictable, modular structural response.To complement the experimental work, a computationally efficient numerical model was developed in a FEM software (SAP2000) utilizing a parallel multilinear link assembly. The numerical results showed close agreement with the experimental data, validating the model's ability to accurately capture transformation plateaus and unloading paths without the need for empirical, system-level parameter tuning. Additionally, a three-dimensional ABAQUS finite-element model was employed to verify the backbone response of the tested configurations. Ultimately, the proposed SMA coupling device offers predictable strength scaling, stable hysteretic damping, and effective residual drift control, serving as a practical, low-damage alternative for high-performance steel-concrete hybrid systems that aligns modern seismic design with Net-Zero sustainability goals.
Available research on high-strength steels after fire exposure has focused mainly on residual static properties, whereas their post-fire fatigue response remains insufficiently documented. To address this gap, constant-amplitude fatigue tests in the finite-life regime were conducted on S600MC high-strength steel specimens at room temperature and after natural cooling from 600, 800, and 1000 °C. Based on the test results, S-N curves and corresponding curves with a 95% survival probability were established, and the effects of thermal exposure on fatigue strength were evaluated in conjunction with fractographic observations and comparisons with current design-code references. The results show that the fatigue strength at 2 × 106 cycles increases after exposure to 600 °C, reaching a value 21.56% higher than that at room temperature, but decreases at higher exposure temperatures; after exposure to 1000 °C, it is 32.86% lower than the ambient-temperature value. The fracture surfaces exhibit typical crack-initiation, crack-propagation, and final-fracture regions. Specimens exposed to 600 °C show a denser and more uniform dimple morphology in the final-fracture zone, consistent with their comparatively improved fatigue resistance. The fracture observations and fatigue-damage analysis indicate that damage develops gradually over most of the fatigue life and accelerates near final failure. Fatigue-redundancy-index evaluation further shows that, within the investigated dataset, the residual fatigue strength of S600MC base metal remains above the selected code-based design references even after severe thermal exposure. These findings provide experimental evidence for post-fire fatigue assessment of S600MC steel and establish a basis for subsequent studies on welded details, heat-affected zones, and fire scenarios involving coupled thermal and mechanical actions.
Tubular K-joints are critical load-transfer components in offshore space-frame structures, and their ultimate resistance can govern structural integrity under severe axial brace actions. Although external ring reinforcement has been studied for some tubular joint types, no experimental, numerical, or analytical study has previously examined externally ring-reinforced tubular K-joints. This paper investigates the initial stiffness, ultimate capacity, and failure mechanisms of externally ring-reinforced tubular K-joints under axial brace loading. A three-dimensional finite element framework is developed by incorporating material and geometric nonlinearities and explicitly modelling the welds connecting the braces to the chord. The modelling strategy is validated against available experimental evidence, and a numerical database of 168 finite element models (including 160 reinforced configurations) is established. The results demonstrate substantial strength enhancement due to external rings, with strength ratios reaching up to 2.21, and the influence of ring width (wr) on ultimate capacity being more pronounced than that of ring thickness (τr). Also, the higher β and lower γ and θ generally increase the enhanced joint capacity. However, the effect of gap ratio (ζ) on the capacity is slight. The reinforced joints predominantly exhibited local chord-wall deformation, while the external rings effectively reduced deformation localisation and improved the joint resistance. Based on the numerical database, a single unified formulation is proposed to predict the ultimate strength ratio, achieving R2 = 0.992 and satisfying the UK DoE acceptance criteria, while the maximum and mean prediction errors across the reinforced models remain below 7% and 3%, respectively.
Recently, steel-concrete-steel sandwich composite walls (SCSSCWs) have been widely used in high rise buildings as the lateral-shear resisting members. When subjected to wind or seismic loads, combined axial force (N)-bending moment (M) actions usually act on such SCSSCWs. This paper conducted numerical and analytical studies on N-M interaction behaviours of T-shaped SCSSCWs (TSCSSCWs). Firstly, a finite element (FE) model was developed to simulate the ultimate strength behaviour of TSCSSCWs. Using the validated FE model, the effect of concrete grade, flange width, steel strength, and steel-plate thickness on the N-M interactions of TSCSSCWs were in-depth investigated. With the variation of axial forces, the shape of N-M interaction curves for TSCSSCWs was governed by three typical failures, i.e., compression-controlled, tension-controlled, and balanced failure. Finally, a fibre element model and two corresponding simplified models were developed based on the plastic stress distribution method. The extensive validation of these model against 17 test data and 162 FE results proved their reasonable predictions on the N-M interactions of SCSSCWs with different cross sections.
This study evaluates the joint quality and shear performance of riveted joints fabricated using induction heating (IH) through experiments and simulations. Riveting specimens fabricated using three IH processes (IHA, IHB, and IHC) were compared with coke-heating (CH) specimens. Temperature measurements revealed that the IHA process provided more uniform heating of the rivet shank than the IHB and IHC processes, resulting in improved hole filling and joint formation. Metallographic observations indicated that the higher riveting temperatures achieved by the IHA process promoted the formation of bainitic microstructures, contributing to enhanced joint strength. Shear experiments on single-rivet specimens showed that all specimens exceeded the design shear capacity. A systematic analytical framework for evaluating the shear performance of IH-riveted joints was developed, including a thermo-mechanical coupled finite element model for the riveting process and a quasi-static finite element model for the shear process. The numerical results agreed well with the experimental results in terms of temperature history, residual stress distribution, and shear failure behavior. The results further indicate that IH riveting did not adversely affect the base plates. Residual stress analyses further suggested that the IHA process produced a residual stress state comparable to that of the CH method, indicating a negligible influence on the fatigue performance of the joints. In addition, the shear strength of the best-performing IHA specimens approached that of the CH specimens.
This paper presents a comprehensive experimental and numerical study focusing on the block shear failure behaviour of stainless steel staggered bolted connections. The experimental results showed that although austenitic stainless steels possessed slightly higher ultimate tensile strengths, their ultimate block shear resistances were lower than those of duplex stainless steels, primarily due to earlier shear cracking near bolt holes. Two distinct failure modes were identified for austenitic connections, whereas duplex connections only exhibited a single failure mode. The developed finite element (FE) models, validated against the test results and considering material anisotropy and fracture behaviour, accurately captured the observed behaviours and enabled an extensive parametric study. The parametric study results indicated that gauge distance, pitch distance, end distance, bolt spacing, and the number of bolt rows significantly improved the block shear resistance of stainless steel staggered bolted connections, while plate thickness and bolt hole diameter had only marginal influences. A comparison between the existing design predictions and the experimentally validated FE results showed that the current specifications (AISC 360, AS 4100, EN 1993, CSA S-16) cannot adequately account for the pronounced strain-hardening behaviour and failure characteristics of stainless steel staggered bolted connections. To improve the predictive accuracy for stainless steel staggered bolted connections, a modified design method was developed by incorporating the tensile-to-yield strength ratio of stainless steels. Additionally, advanced machine learning approaches, grounded in a rational understanding of the structural mechanisms, were employed to further enhance the accuracy of prediction.
