Independent assessment and benchmarking of no/low cost finite element analysis software for linear and non/linear static structural analysis.

Abstract

The aim of this research was to determine if the development of low-cost (less than 500 USD) or no-cost analysis software, specifically in the area of computational structural mechanics through finite element analysis (FEA), has advanced to the point where it can be used in place of trusted commercial FEA software packages for linear and non-linear static structural analyses using isotropic materials. This was done by conducting an initial market research study and identifying a range of available no-cost/low-cost FEA packages. Eighteen suitable packages were identified and a preliminary analysis was done to identify analysis capabilities, presence of internal modules, extent of available user documentation, and end user support. The packages underwent a process of systematic elimination from the preceding phases of the research if they were unable to meet the minimum imposed criteria. Six packages were deemed suitable and were further investigated. From these, three packages were chosen to be subjected to performance benchmarking, namely: Code_Aster/Salome Meca; Mecway and Z88 Aurora. SimScale, a browser-based analysis package was included as well because it met all the baseline criteria and has the potential to offer a completely cloud-based approach to computer aided engineering, potentially reshaping the way an engineering business views its operational capabilities. Performance benchmarking assessed the ability of a package to generate a model and obtain accurate solutions relative to industry accepted benchmark publications, trusted analytical solutions found in reputable engineering text, as well as experimental results obtained in this work. The benchmarking process was also done on commercial FEA packages so that a comparison can be made between the no-cost/low-cost packages and those considered to be the premium FEA software packages available. It was found that the no-cost/low-cost options were able to perform adequately for most of the test cases. SimScale and Z88 Aurora had difficulties with generating suitable meshes which meant that compromises in model generation approaches needed to be made. Overall, the results yielded by the low-cost/no-cost options showed good correlation with test case target values as well as exhibiting many capabilities and tools found in the high-cost, trusted commercial packages investigated. It is therefore concluded that there are no-cost/low-cost FEA packages that can be used in place of high-cost commercial packages for linear and non-linear static structural analyses of isotropic materials.