Characterization of 3D fiber preform permeability tensor in radial flow using an inverse algorithm based on sensors and simulation

In Liquid Composite Molding (LCM), the fiber preform is placed in a mold and resin is injected through a gate to fill the empty spaces within the mold. LCM processes are modeled as resin flow through fibrous porous media in which if one knows the permeability values, one can determine the arrival times of the resin at any location. A mold with radial injection with 192 flow arrival detection sensors along 16 radial lines are mounted flush with the top and the bottom mold surface with an additional sensor on the top surface opposite to the injection hole of the bottom surface. The inverse problem addressed here is from the recorded resin arrival times at sensor locations, how accurately can one determine the permeability of the preform? The proposed method uses correlation between the experimentally recorded resin arrival times and 3D flow simulation of the experiment. The optimization routine varies the permeability components in the simulation to achieve the best possible match with the experimental arrival times at all the sensor locations. Currently, all in-plane permeability components and through-the thickness permeability are characterized from a single experiment with the potential to evaluate the cross-thickness off-diagonal terms as well. In addition, the technique also demonstrates how one can reduce the variation in through thickness permeability by the use of a distribution media at the injection hole to avoid blockage of the inlet by the fiber tows. The optimization routine sequentially optimizes the values of the individual components of the permeability tensor using golden search method, and then repeats the entire sequence until the best match is found. The validation and sensitivity of this method is explored and it has been shown that this technique is promising for characterization of all permeability components from a single radial flow experiment.