Probabilistic Process Simulation and In-Situ Process Prediction during Composite Manufacturing as Contribution to Industry 4.0

In contrast to metallic parts the final geometry and superior structural properties of composite structures evolve during the manufacturing process itself. Thus, the process parameters and process boundary conditions have a significant influence on the final component properties. To ensure a high component quality narrow tolerances during the entire component development process are currently defined. Yet, due to manufacturing-induced deformations as well as material uncertainties and varying process conditions the required geometric tolerances cannot be met in all cases. In case of unauthorized process or part deviations additional time-consuming and costly assessment processes are initiated in present industrial environment. In order to develop new manufacturing processes more efficiently an enhanced simulation strategy for a priori and in-situ prediction of manufacturing-induced deformations and residual stresses of composite structures is developed. Within a sequential temperature-displacement finite element analysis strategy an anisotropic viscoelastic material model is applied to consider temperature and time dependent relaxation effects. In order to take into account fluctuations in process and material parameters the deterministic process simulation is extended by a probabilistic process simulation. For this, fast and efficient surrogate models are derived in order to perform probabilistic simulations in acceptable time. This enables to evaluate the robustness of different curing processes. Furthermore, the derived surrogate models allow for predicting the in-situ properties of components as full 3D field information. For instance, structural properties, such as residual stresses, deformations or stiffness as well as thermal hot-spots and thermo-chemical properties such as degree of cure and glass transition temperature are calculated in real time. In context of Industry 4.0, this enables to evaluate the process and component quality during manufacturing and, therewith, results in an improved process control and enhanced quality management. The method is demonstrated for an automotive car suspension blade