A precise prediction of the cure-induced shrinkage of an epoxy resin is performed using a finite element simulation procedure for the material behaviour. A series of experiments investigating the cure shrinkage of the resin system has shown a variation in the measured cure-induced strains. The observed variation results from the thermal history during the pre-cure. A proposed complex thermal expansion model and a conventional chemical shrinkage model are utilised to predict the cure shrinkage observed with finite element simulations. The thermal expansion model is fitted to measured data and considers material effects such as the glass transition temperature and the evolution of the expansion with the degree of cure. The simulations accurately capture the exothermal heat release from the resin and the cure-induced strains across various temperature profiles. The simulations follow the experimentally observed behaviour. The simulation predictions achieve good accuracy with 2–6% discrepancy compared with the experimentally measured shrinkage over a wide range of cure profiles. Demonstrating that the proposed complex thermal expansion model affects the potential to minimise the shrinkage of the studied epoxy resin. A recommendation of material parameters necessary to accurately determine cure shrinkage is listed. These parameters are required to predict cure shrinkage, allow for possible minimisation, and optimise cure profiles for the investigated resin system. Furthermore, in a study where the resin movement is restrained and therefore able to build up residual stresses, these parameters can describe the cure contribution of the residual stresses in a component.
A small experimental setup for in-situ measurement of the load-transferring strains during the curing process of thermosets is proposed. Combining the output from an unconstrained and a kinematically constrained setup, it is possible to design a cure profile for the first time, lowering the residual stresses in the final product while keeping the cure time short based on the output from a few simple experiments. The stress relaxation during the curing process under a kinetically constrained condition is accounted for by comparing the final cure-induced strain during a kinetically unconstrained and constrained cure experiment. The constrained polymer is curing between two laminates where the constraining layer is removed after finalizing the cure profile, making it possible to measure the final cure-induced strain for that case as well. The temperature at which the load-transferring point is reached is found to be a key process parameter from which the final cure-induced strains can be predicted for the unconstrained case. From the corresponding constrained cure experiments, the final residual stresses can be measured.