ISBN 9783843927864

978-3-8439-2786-4, Reihe Ingenieurwissenschaften

Rhena Helmus
Out-of-Autoclave Prepregs: Stochastic Modelling of Void Formation (Band 22)

159 Seiten, Dissertation Technische Universität München (2016), Softcover, A5

Zusammenfassung / Abstract

Out-of-Autoclave (OoA) prepregs are a more cost and energy efficient counterpart to traditional autoclave-cured prepregs due to low pressure consolidation and, to that effect, comply with the increasing demand for composite materials as well as the request for increased manufacturing sustainability. But the lack of pressure during the vacuum-bag-only processing is accompanied by an enhanced susceptibility to porosity. To counteract a quality degradation, thermoset OoA materials are designed with a special, breathable microstructure that allows for void evacuation, instead of void dissolution in the resin, as possible with autoclave processing. Nevertheless, the production of low porosity parts requires a careful definition of processing parameters based on a thorough understanding of void formation. This is complicated by inherent material uncertainties. Even though OoA prepregs are machine-made, they are accompanied by fluctuations in process parameters, handling and environmental conditions. Small variations within input parameters may have a strong impact on the final part’s void content, given the already narrow processing window of OoA processing. Consequently, contingencies in input parameters have to be considered in realistic void simulations. Over the past decades, process modelling mainly focused on deterministic models. The following work attempts to improve void simulations through the quantification of the effect of stochastic variability in input parameters, which is a prerequisite to reduce the risk of manufacturing induced defects. Various input variabilities are allocated to three types of voids, intra-ply, resin and inter-ply voids, which are present in OoA prepregs. Each of them is modelled individually, using different stochastic approaches: First, a deterministic 2D consolidation model is embedded in a stochastic environment to account for random effects within the breathable structure. Second, the distribution and size of void nuclei within the resin are determined by point processes, which serves as an input for modelling bubble growth due to absorbed moisture. Finally, air entrapment between randomly formed prepreg surfaces is studied experimentally before stochastic models, that represent the material’s surface, are introduced.