Predictive modelling of automated fibre placement processes

Composite materials are being used in the design of mechanical parts of increased complexity. This has put the industry under growing pressure to speed-up the rates and decrease their cost of production. Automated fibre placement (AFP) is at the heart of the recent effort of the composite manufacturing community to meet this goal. The method consists in automatically depositing onto a mould narrow strips of carbon fibre tows pre-impregnated with uncured resin (prepreg) using a computer-controlled poly-articulatedrobot. AFP has already allowed to considerably decrease the production time of relatively simple parts with low curvature, but setting-up the machine to reliably manufacture a part with a minimum number of defects is heavily dependent on engineering experience, and time-consuming trials. Being able to predict the as-manufactured geometry and fibre paths in parts made by AFP would significantly help reduce the empiricism inherent to current state-of-the-art AFP manufacturing. Looking forward such models could also be very beneficial in the context of recent trends towards manufacturing 4.0 where the manufacturing process is heavily instrumented with in-process sensing and the deposition parameters adapted on the fly to mitigate the formation of defects.
The group at the Technical University of Munich (TUM) has already started the development of a numerical tool for the simulation of AFP processes. The framework captures very well all the phenomena related to thermal exchanges but is lacking a proper material model for the incoming prepreg tape. The team at Bristol Composite Institute (ACCIS), on the other hand, has developed some world-leading capabilities for the modelling of the mechanical behaviour of prepreg tapes in their uncured state. Putting together the expertise of both teams will accelerate the development of a proper optimisation tool for AFP processes that the industry is striving for. In addition, TUM is also working very closely with the Airbus spin-out company InFactory and are able to produce some state-of-the-art in-proces data for the validation of UoB models. TUM and UoB have sucessfully collaborated in transnational research in composite manufacturing through a G8-2012 initiateive (EP/K025023/1) have recently exchanged ideas on AFP processing. In September 2018, a joint workshop on AFP technology was organised in Munich with with the aim to identify avenues for potential collaboration. Manufacturing 4.0 was singled out as a good area were a joint DFG (i.e. German Research Foundation) - EPSRC grant proposal could be submitted.