Numerical Predictions of Fibre Preform Permeability for Enhanced Process Modelling of Advanced Composite Structures

Abstract

Liquid Composite Moulding [LCM] process simulations are increasingly used as process design tools to select optimal manufacturing parameters for fibre reinforced polymer [FRP] composite materials. Accurate permeability data of textile preforms is an essential input for these simulations. For maximum accuracy it is desirable to capture both the preforms’ permeability and their spatial variation. For this, a comprehensive tool has been developed to predict the permeability of reinforcing textiles, reducing the reliance on empirical permeability data. An automated tool has been developed for the generation of permeability predictions for multi-layered unit cells utilising textile modelling techniques. A reinforcing textile is initially scanned and its geometric parameters determined using image analysis techniques. Textile models of single layers are then created and combined to replicate the stacking step during preform manufacturing. Compaction simulations are applied to these stacks, capturing changes in geometry as well as nesting. Voxel meshes representing the volume of resin around the compacted geometries are then generated. The voxel meshes are automatically cleaned, deleting any floating elements, and the boundary regions defined based on the unit cell size. By executing flow simulations on these meshes, the permeability characteristics of the preform are obtained. This tool has been used to predict the permeability tensor of two woven textile architectures. Single layer predictions were carried out and the predicted permeabilities obtained were in close agreement to the permeability behaviour captured experimentally. It is shown how the tool can capture the effects of textile variability on its permeability. Additionally, the tool is used to predict the permeability maps of single layer woven textiles. These permeability maps were used to execute the process simulations and compute the predicted flow front positions, which were then compared with those observed in experiments. The tool has also been used to study the effect of preform structure on its permeability, including consideration of the number of layers, stacking sequence, ply shift and applied compaction. The analysis was verified by computing the complete permeability tensor, and comparing this to in-plane and through-thickness permeability data obtained experimentally. The developed permeability prediction tool provides an easy to use, efficient and fast alternative to obtain permeability characteristics of reinforcing textiles. Minimal input data is required, making it an attractive alternative to characterising permeability experimentally. Providing accurate permeability predictions ensures that LCM filling simulations will accurately represent the manufacturing process, and that process parameters may be confidently selected to manufacture highest quality FRP parts in a robust and cost effective manner.