Analysis of the impact of plate element selection on the application of the finite element method in modeling large concrete slabs with holes

Abstract

This study addresses the challenge of stress prediction under forced deformation for large concrete slabs with holes (e.g., industrial factory floors, nuclear power plant equipment foundations) by analyzing the influence of plate element type selection on modeling accuracy and computational efficiency. Based on experimental data from a 4*4*0.25 m large slab with holes, the performance and characteristics of various plate elements (including triangular, quadrilateral, and polygonal types) in simulating the shrinkage stress field are compared, elucidating the influence of element type on stress concentration around holes, the behavior of zero shear stress points, and computational efficiency. The research results indicate that using higher-order quadrilateral elements (such as eight-node isoparametric elements), combined with a contact interface spring model to calculate stress around holes, can satisfy engineering calculation requirements, with stress calculation errors at hole edges within 8%; however, the computational speed is considerably slower than that of neural network models. Based on the above research findings, this paper proposes an element selection strategy for plates with complex geometries and introduces a hybrid computational framework combining finite element coarse calculations with neural network refinement.