Finite Element Analysis is a powerful computational tool for stress analysis. Although the roots of the finite element method can be traced back to WW2, it has been the advent of the computer and the rapid increase in, and availability of, computational power that has lead to its widespread adoption.

At FAC technology we use FEA to model tests which would otherwise have been conducted experimentally. This greatly reduces the product development time and cost. Of course these models are only as valid as the data and mathematical assumptions made in producing them. Modern FEA software programs often come with user friendly graphical user interfaces (GUIs), that allows a user to build and interact with the finite element method. In fact, many users may never interact with a finite element solver through any other method. However, proficiency in a using a GUI is by no means correlated to the accuracy of the model. To help explain this, analogies will be drawn between FEA and the popular spreadsheet application, Microsoft Excel.

The initial stage of FEA is called pre-processing. During pre-processing, a user would define his inputs and what calculations are to be performed. In FEA this would include creating the part geometry and element mesh, inputting the material properties and selecting the material model, defining the loads and boundary conditions, etc. In Excel, a user would also put in some data and then tell Excel what to do with it, e.g. find the mean of a set of numbers.

The next step in the FEA workflow would be to solve the model. The FEA method essentially involves setting up an enormous system of equations which are then solved by the FEA program. FEA can readily output solutions but the validity of such solutions is dependent on the parameters defined during pre-processing being correct (geometry, material model, load conditions etc.) In Excel if one wanted to find the mean of a set of numbers, Excel will solve the calculation. However, if one told Excel to calculate a value using the wrong method, by selecting the median or mode instead, then one may well get an answer that is similar to the mean, but it would still be fundamentally incorrect.

Post-processing is where the user visualises and interprets the results calculated. In Excel this might comprise of plotting graphs. In modern FEA, this may involve interpreting visualisations of the stresses within, and deflections of, a component. However, the amount of data that a FEA run can produce is enormous and an understanding of the mechanics of the materials being simulated is required to usefully analyse the results. Further, errors of various guises are fundamentally intrinsic to the FEA method and require assessment, mitigation and validation.

To ensure that FAC Technology’s FEA approaches are valid, we carry out extensive experimental testing to ascertain both material properties and the accuracy of our FEA models. FAC Technology also has a close, collaborative, relationship with Imperial College, as the company’s founder, Dr Alex Fergusson, maintains visiting academic status in the Mechanical Engineering Department. This allows FAC Technology to stay at the forefront of research into the design, testing and analysis of composite materials and structures. Dr Fergusson also advises on ongoing research projects at Imperial; the development of blast resistant composite sandwich structures for the US Office of Naval Research is one example.

It is perfectly possible for someone with no understanding of the finite element method, the mechanics of composite materials, or the issues associated with apparent properties of composite material to perform FEA of a composite structure using modern software. Using these programs, it would be fairly easy to produce engrossing visualisations and animations, but whether such output would be representative of reality may be an entirely different matter.