Numerical Prediction of Stiffness and Strength of a Highly Complex Topology Optimized Thermoplastic Part designed for 3D Printing

摘要

Additive manufacturing or 3D printing has revolutionized the way companies manufacture products by allowing the manufacturing of shapes considered impossible to manufacture until now. This gives the advantage of designing the parts with materials positioned in exact locations as is required from its performance loading conditions. However, while this technology is beneficial from weight-to-performance perspective, this results in highly complex shapes which are difficult to discretize for finite element analysis and performance predictions. This difficulty is even more pronounced while predicting failure strength of the topology optimized 3D printed part in addition to part stiffness. Typically, explicit solvers are used to predict highly non- linear events such as component failures. Traditionally, hexahedral elements are used in explicit solvers due to the higher accuracy and lower cost of simulations compared to tetrahedral elements for similar element dimensions. However, highly complex shapes generated through topology optimization are prohibitively difficult to model using hexahedral elements. This paper deals with prediction of stiffness and failure strength of a part of highly complex geometry arrived through topology optimization using advanced higher order tetrahedral elements in LS-DYNA explicit solver. Use of tetrahedral elements allows the discretization to be almost fully automated. Equivalent analysis without explicit material failure is also performed using Abaqus implicit solver for benchmarking. The predictions are compared with actual physical test results for the 3D printed parts which exhibit good correlation.