Improving seismic performance of welded joints in CFST Truss towers: Effects of concrete fill and ring stiffener configurations

This study presents a novel approach to enhance the longevity and seismic resilience of ‎ Concrete-Filled Steel Tubular (CFST) frame towers commonly used in ‎onshore wind turbines. Fatigue-related failures at welded joints pose significant challenges to the durability of these towers. This research investigates the impact of optimized ring stiffener ‎configurations and concrete infill on reducing stress concentrations and improving ‎fatigue resistance under simulated seismic loading. The study utilizes with samples including CFST-gusset plate joints and non-concrete-filled joints, with different configurations of concrete infill and ring stiffeners. ‎These samples were evaluated under static loading to assess the influence of stiffener width, thickness, and spacing on Stress Concentration Factors (SCFs). Subsequently, cyclic loading tests were conducted to observe crack initiation and progression. The results show that ‎optimized configurations of both concrete fill and ring stiffeners significantly reduce ‎SCFs, improve stress distribution, and delay crack initiation, enhancing the seismic ‎fatigue performance of CFST joints. Moreover, the study integrates Artificial Intelligence ‎‎(AI) techniques, specifically Support Vector Machines (SVM), to predict the fatigue ‎behavior of these joints under different loading conditions. The SVM model was trained on experimental data and demonstrated high predictive accuracy in forecasting the fatigue life of CFST joints, achieving an R² value of 0.95, an RMSE of 0.11, and an MAE of 0.08. These results validate the AI model's ‎ability to optimize design parameters and predict fatigue life, contributing to more ‎efficient and reliable CFST joint design. This integration of AI enhances design ‎optimization by providing precise predictions based on complex variables like stiffener ‎geometry and concrete filling. Overall, the findings suggest that incorporating ‎optimized stiffener configurations, concrete infill, and AI-based design tools can ‎significantly improve the seismic resilience and fatigue life of CFST frame towers joints, ‎providing valuable insights for future engineering practices.