Host Institution: Cranfield University
Lead Investigator: Alex Skordos
Co-Investigators: James Kratz, Jonathan Belnoue, Konstantinos Tifkitsis, Lawrence Cook, Ric (Xiaochuan) Sun
The feasibility study funded by the EPSRC Future Composites Manufacturing HUB entitled ‘Layer by layer curing (LbL)’ was carried out in the period from 20/11/2017 to 01/06/2018. The study aimed to test if the LbL process is feasible by carrying-out computer simulation of LbL curing, designing the process to achieve an adequate compromise of interlaminar properties and efficiency, implementing physically the whole layer variant of the process and assessing the quality of laminates produced. The investigation was carried out using a highly exothermic prepreg system (Hexcel 913/glass). The cure and consolidation behaviour of the prepreg were characterised and captured in appropriate constitutive models and the coupled simulation was executed for scenarios of the whole layer and ATL variants of the process. The simulation was used to optimise the curing of a 40 mm thick laminate for both the LbL process and conventional processing subject to the same temperature overshoot constraints to allow designing the LbL process conditions and comparing against conventional processing. The cure using the LbL process can be achieved within about 1 h 20 min in comparison to about 3 h for the optimised conventional cure process. This significant benefit is accomplished within the actual stage of deposition of the material, which assuming a consolidation stage duration similar to curing means that the overall LbL process can be completed in less than one quarter of the duration of the conventional process. Furthermore with the LbL process, there are profiles that can complete the cure of the 40 mm thick laminate with moderate or low temperature overshoot. The LbL process can exploit process time savings found by eliminating thermal mass heat up/down of tooling by processing at isothermal conditions. This is not possible with conventional curing at this level of thickness with reactive thermosetting polymers. The process was implemented using a servo-hydraulic machine equipped with heated plates. The setup produced specimens of 40 mm thickness with acceptable temperature overshoot during the process in agreement with the simulation. Furthermore, the temperature distribution evolution measured during the process and the thickness of the consolidated part matched those predicted by the simulation proving the validity of the modelling analysis and optimisation results. The quality of the LbL laminates was investigated using microscopy and was found to be adequate with porosity kept at low levels, microstructure/morphology similar and finer than in conventional processing. The mechanical behaviour of material produced using the LbL route was compared with that of conventional laminates. Interlaminar shear strength (ILSS) and double cantilever beam (DCB) mode I toughness testing showed that laminates produced using LbL curing match those of conventional laminates. Overall the feasibility study delivered a 75% reduction in process time without degrading product quality by achieving equivalent interlaminar properties and porosity.