Influence of stochastic variations in the fibre spacing on the permeability of bi-directional textile fabrics

To predict the range of probable outcomes of resin injections in LCM processes for non-uniform textiles, a stochastic injection simulation procedure is implemented. Continuous random fields describing the distances between fibre tows in bi-directional non-crimp fabrics are generated based on a spectral expansion implementing trigonometric functions with the random parameters frequency and phase. Maximum values of the random frequencies are determined from the visually observed fibre angle variations and the textile structure of actual fabrics. From the fibre spacing, continuous distributions of the local porosity are calculated based on geometrical considerations, and local permeability values are estimated using the Kozeny-Carman approximation. Series of radial resin injections in a flat mould are simulated and evaluated statistically. Probable outcomes of resin injections and the global permeability value and its variance are deduced. The geometrical scale of permeability inhomogeneities increases with increasing average fibre tow distance. The ratio of in-plane fibre tow dimensions and average distance between the fibre tows, which characterises the fibre-void distribution in the fabric, affects both the average permeability values and the relative permeability variations. The fibre angle variation imposes a boundary to the local variability in the textile structure. The global permeability variations derived from the series of simulations are in the order of magnitude of typical experimental results. (c) 2005 Elsevier Ltd. All rights reserved.