Limit analysis of plates using the EFG method and second-order cone programming

• Published in: International journal for numerical methods in engineering Vol. 78; no. 13; pp. 1532 - 1552
• Main Authors: Le, Canh V., Gilbert, Matthew, Askes, Harm
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 25.06.2009; Wiley
• Subjects: Computational techniques; Curvature; Displacement; EFG method; Exact sciences and technology; Fundamental areas of phenomenology (including applications); limit analysis; Limit load; Mathematical analysis; Mathematical methods in physics; Mathematical models; nodal integration; Optimization; Physics; Plates; Programming; second-order cone programming; Solid mechanics; Static elasticity (thermoelasticity...); Structural and continuum mechanics; second-order cone programming; nodal integration; conic programming; Stabilization; Euclidean theory; limit analysis; Minimization; Optimization; Dimensioning; Galerkin-Petrov method; Finite element method; Smoothing; Meshless method; Kinematics; EFG method; Modelling; Least square fit; Curvature; Plastic hinge; Plates; Ultimate load
• ISSN: 0029-5981; 1097-0207
• DOI: 10.1002/nme.2535

The meshless element‐free Galerkin (EFG) method is extended to allow computation of the limit load of plates. A kinematic formulation that involves approximating the displacement field using the moving least‐squares technique is developed. Only one displacement variable is required for each EFG node, ensuring that the total number of variables in the resulting optimization problem is kept to a minimum, with far fewer variables being required compared with finite element formulations using compatible elements. A stabilized conforming nodal integration scheme is extended to plastic plate bending problems. The evaluation of integrals at nodal points using curvature smoothing stabilization both keeps the size of the optimization problem small and also results in stable and accurate solutions. Difficulties imposing essential boundary conditions are overcome by enforcing displacements at the nodes directly. The formulation can be expressed as the problem of minimizing a sum of Euclidean norms subject to a set of equality constraints. This non‐smooth minimization problem can be transformed into a form suitable for solution using second‐order cone programming. The procedure is applied to several benchmark beam and plate problems and is found in practice to generate good upper‐bound solutions for benchmark problems. Copyright © 2009 John Wiley & Sons, Ltd.