A common requirement for investigating the micromechanics of textile based composites is to define the general mathematical representation for fibre and tow paths, and then to determine specific values for the material in question. Such approaches can vary from requiring only two data points that can be readily measured for a material, through to the use of digital image processing to determine tow paths and fibre orientation distributions within real world samples
The first step in any engineering project is to perform preliminary calculations to size the structure and check that it meets the design requirements. In order to complete this stage successfully, the engineer needs accurate material property data. This is comparatively easy for isotropic materials as there is a plethora of experimental test data, established using well defined test methodologies, already available. However this same task is nontrivial for composites as these materials are, in reality, structures unto themselves. Consequently, composite laminates are typically designed for the specific application. Further, it is not merely the properties of the constituent materials used that defines the properties of a composite lamina – how the fibres are woven together, the manufacturing processes used, how the material is laid over a mould, the temperature etc. – all affect the properties of the composite material produced. Due to the vast range of variables that determine the properties of a composite, it’s rarely possible to simply look up all the data needed for in depth analysis.

One option is to use generic properties for a similar composite from already published data. This is fine if one is considering a fairly common composite, such as an aerospace certified unidirectional carbon/epoxy prepreg. But what if one is intending to use a combination of a dry textile, e.g. woven carbon fibre, with a certain resin system and one of various closed mould production processes? The various permutations and combinations of material and processing variables mean that the likelihood of finding a sufficiently broad and relevant source of data is very low.

In the frequent scenario where one does not have sufficient experimental data multiscale modelling of the material may provide an effective starting point. Experimental testing has an obvious appeal, but is time consuming and expensive. On the other hand, mutliscale modelling allows for ‘virtual’ experimental testing. Of course, the validity of any virtual test is in turn base on the validity of the models, material properties and methods used therein. FAC Technology approaches this task from both perspectives. Extensive and ongoing testing of a range of materials and processes provides us with a database of properties. We also use this data to benchmark various modelling approaches to establish their efficacy and limitations. This work allows us to reduce the amount of experimental testing needed, especially in the early design stages.

Back To Technology