Quantifying variability within glass fibre reinforcements using an automated optical method

This paper presents a novel optical technique to quantify in-plane geometric variations within dry glass fibre reinforcement materials. Samples of up to 290 × 450 mm can be examined. Three different reinforcement structures have been studied; a random mat, a plain weave and a stitched bi-axial fabric. Using an empirically derived function, reinforcement areal weight has been predicted locally from a single reinforcement photograph. It was found that areal weight predictions were typically within 5% of experimentally obtained values for 25.4 mm square samples. For periodically structured (woven or stitched) reinforcements, local information describing the tow orientation and geometry has been collected automatically. Manual verification of the reinforcement geometry showed good agreement with the automated technique. 3D textile models have been created within a textile modelling software that include the measured variability.