Active control of the RTM process under uncertainty using fast algorithms

The project achieved two main goals. First, the feasibility study demonstrated, in virtual and lab experiments, that a novel Bayesian Inversion algorithm (BIA) can successfully estimate local permeability and porosity of a preform using in-process information. In particular, the algorithm was able to determine locations and shapes of defects in fibre preforms. This outcome is important for making non-destructive evaluation (NDE) of composites faster and more robust, which in turn can deliver more reliable and cheaper manufacturing of composites. The project also demonstrated feasibility of an Active Control System (ACS) based on the BIA to ensure that the RTM process satisfies one of the key requirements of the composite industry: to have repeatable production cycles.