Estimating the stress distribution within MERO joint using (FEM-ANN) hybrid technique

Abstract

A three-dimensional roofing structure known as a space truss comprises steel pipes joined together by spherical joints known as MERO joints, typically behave as hinged nodes. Only a few studies have examined the integrity and stresses within the spherical joint; most literature on specs trusses focuses on member arrangement and design. As a result, this study examines the stress distribution within the sphere, and how they is affected by the joint dimensions (sphere and bolts diameters, threading length and inclination angle of diagonals), in addition, the study aims to develop practical (AI) based predictive model to estimate the stress values at the critical points of the sphere. A numerical parametric study utilizing a FEM model calibrated by literature experimental was used to accomplish these goals. Thirty-four spherical steel bolted joints were developed using ANSYS workbench software to determine the stress values at center, top surface, bottom surface and at thread contact area. The outcomes of the parametric study were used to train and validate the developed ANN predictive model. The research results indicated that the average accuracy of the ANN productive model is about 85%, the maximum stresses are always located at the threading area, then at the sphere surface, while it is minimum at the sphere center. Finally, the diagonal inclination angle has almost no impact on the stress distribution.