The liquid composite moulding (LCM) processes are widely used to manufacture high quality fibre composite parts for aerospace or sports applications. Typical processes are the resin transfer moulding or the vacuum infusion. In the LCM process, the dry textile is infiltrated by a polymer matrix due to a pressure drop between the inlet and outlet. The impregnation of the dry textile is modelled by Darcy’s Law, which describes the flow velocity in dependence on the permeability (textile parameter), the viscosity (resin parameter) and the pressure drop (process parameter).

The focus of this work is on the permeability. The permeability is a complex parameter: fhe fibre volume content, architecture of the textile and stitching patterns show great influence on this parameter. Because of this complexity, no current model can describe the permeability in a sufficient way and therefore valid experimental measurements for each lay-up is necessary. Different measurement methods are available to estimate the permeability. A very effective method is the three-dimensional radial flow method. Due to the analytical assumptions, the approach is very restricted on the geometry and process parameters. The need of much material is also a disadvantage. The use of machine learning methods to estimate the in-plane permeability (2D) shows good results for a restricted permeability interval. Combining the three-dimensional measurements with a machine learning method, which is based on finite element method (FEM) simulations, the restrictions of the analytical approach can be overcome. In this study, a new method is presented to measure and evaluate the three-dimensional permeability in an effective and reliable way.