Development of Novel Surrogate Models for Stress Concentration Factors in Composite Reinforced Tubular KT-Joints

Circular hollow section (CHS) joints are among the most critical components in offshore jackets, often requiring rehabilitation to maintain structural integrity. The structural stress approach, based on stress concentration factors (SCFs) for hot-spot stress (HSS) calculations, is commonly used to estimate the fatigue life of critical structural elements such as CHS joints. Various empirical models exist for the rapid estimation of SCF in composite-reinforced CHS joints; however, most studies focus on SCF at the crown and saddle positions under uniplanar loading. This limitation reduces their applicability to multi-planar loading conditions, potentially leading to the underestimation of HSS. This study investigates the use of fiber-reinforced polymer (FRP) composites to strengthen CHS KT-joints under complex loading, focusing on reducing SCF and improving fatigue life. A total of ۵,۴۲۹ finite element simulations were conducted to examine the effects of geometric and reinforcement parameters on SCF. The simulation data were used to train artificial neural networks (ANNs), which were incorporated into a computational tool for the rapid approximation of hot-spot stress in FRP-reinforced KT-joints. The application of composites to CHS joints significantly reduces SCF, particularly with an increased number of reinforcement layers, a higher elastic modulus, and an orthogonal fiber orientation to the weld toe. This study presents a novel methodology for developing efficient models to estimate SCF in composite-reinforced CHS joints under complex loading, addressing a key gap in fatigue design for such joints. The developed computational tool enables the rapid calculation of hot-spot stress in CHS joints. Doi: ۱۰.۲۸۹۹۱/CEJ-۲۰۲۵-۰۱۱-۰۴-۰۱۲ Full Text: PDF

کلیدواژه ها:

Circular Hollow Section Joints ، Composite Reinforcement ، Stress Concentration Factors (SCF) ، Artificial Neural Networks (ANN) ، Structural Rehabilitation ، Surrogate Modelling.