Bayesian inversion algorithm for estimating local variations in permeability and porosity of reinforcements using experimental data

A novel Regularising Ensemble Kalman filter Algorithm based on the Bayesian paradigm was applied to RTM processes estimate local porosity and permeability of fibrous reinforcements using measured values resin pressure flow front positions during injection. The algorithm allows detect locations defects in preform. It tested virtual experiments with two geometries, a two-dimensional rectangular preform more complex 3D shape, as well laboratory experiments. In both experiments, it demonstrated that proposed methodology is able successfully discover good accuracy. also provides confidence intervals for predictions estimations defect probabilities, which are valuable analysis process.