Artificial Intelligence driven topology optimisation of Additively Manufactured Composite Components

Additively Manufactured fibre reinforced composite (AMC) components manufactured via fused deposition modelling (FDM)rapidly find applications within the European aerospace and transport industry , due to their well-known advantages mainlyrelating to less machine, material and labour costs, less manufacturing waste, and usage of more efficient materials.

A majordrawback of AMC components is their usually complex and in cases tessellated geometry; this gives rise to combined (e.g.,fibre pull-outs and matrix cracking) and quasi-brittle damage mechanisms that deviate from the usual “high strength andductile metal” design paradigm.

Such a “complexity”, if controlled, can result in components of tailored mechanicalproperties, e.g., of increased fracture toughness and pseudo-ductile post fracture response.

Unfortunately, current analysisand design methods lack the necessary level of refinement, or the underlying theoretical framework indeed, to efficientlyaddress this critical issue.AI2AM aims to deliver a holistic approach to additively manufacture topologically optimum composite components ofincreased fracture toughness.

It will achieve this by developing a state-of-the-art fracture simulation framework for compositestructures harnessing the fidelity and computational advantages of phase field modelling for fracture and scaled boundaryfinite element methods.This high fidelity physics based ""continuum toolbox"" will be used to train surrogate models based on machine learningmethods.

The surrogates will then be deployed within a novel topology optimisation framework to deliver optimal and 3Dprinted geometries.

The envisaged methodology crosses the boundaries of computational mechanics, optimisation, andmachine learning and brings together a talented academic with world-class experts in topology optimisation, composites,and additive manufacturing.