Weld distortion prediction of the Cargo Transfer Vessel key structure by fuzzy finite element model and inherent strain theory
Abstract
This study introduces an innovative fuzzy finite element model (fFEM) based on inherent strain theory, optimized with artificial intelligence, to predict welding deformations in critical components of Cargo Transfer Vessel (CTV). By integrating a genetic algorithm with the commercial finite element software ABAQUS, optimal inherent strain coefficients are identified. The fFEM significantly reduces computation time, achieving approximately 60% efficiency compared to traditional thermo-elastic-plastic models. The model demonstrates high accuracy, with an average prediction error of less than 10%. The efficacy of this method is validated through comparative analyses against traditional models and experimental results, highlighting its precision and effectiveness in predicting welding deformations in large marine structures. This approach provides a robust framework for addressing welding deformations, ensuring structural integrity and precision in marine engineering applications.