Contribution to robust resin transfer molding

Resin transfer molding (RTM) is an established process to manufacture high quality composite parts with thermosetting matrices. In spite of the economically interesting properties of this process, there are critical issues concerning process reliability. For instance, the quality management and cycle time should be improved in order to make the process more economic. During mold design, injection gates and vents could be placed optimally using simulation software and optimization algorithms. The flow control is critical during the injection. However, the permeability of the fabric preform is sensitive to handling and can differ in each injection. To achieve an optimal injection, those disturbances need to be considered in simulations. A finite element software for fill simulations (sLIP) has been developed which considers major effects such as anisotropy, gas entrapments, resin curing. A major enhancement of the simulation is a parametric simulation model to include stochastic disturbances. Particular attention is given to the initialization of the stochastic variations, and the evaluation of the resulting process variations. To optimize process parameters, evolutionary algorithms are coupled with this simulation software. As free parameters, gate and vent locations are considered as well as injection pressures, volume flows, and their timing. Topics like model parametrization and the definition of the optimization objectives are discussed. Optimizations considering process reliability are compared to deterministic optimizations. In a flow visualization approach, the real permeability distribution can be estimated for each single injection. By comparing simulation values with actual values of sensors, the differences between assumed and real permeability distributions can be determined iteratively. Thus, a model of the permeability distribution can be generated. This procedure is known in structure analysis as ‘model update’ and is applied if assumptions for unpredictable values have to be made. Only a few sensors are required to get precise monitoring of the flow in the closed mold. Several benefits can come from this visualization technology. In this work, a continuous permeability measurement method has been derived, which allows to measure the saturated in plane permeability in a single experiment. A further possibility is to estimate the quality of a molded part by simulations based on the updated permeability model. Incomplete filling or intolerable joints of flow fronts can be recognized immediately after the injection. Also, it becomes possible to establish a feedback control system for better injection results.