Application of layout optimisation to the design of additively manufactured metallic components

Additive manufacturing (`3D printing') techniques provide engineers with unprecedented design freedoms, opening up the possibility for stronger and lighter component designs. Such designs can be developed using structural topology optimisation, a method that computationally designs minimum mass components. Whilst many methodologies now exist this thesis focuses solely on a method known as layout optimisation where components are designed to meet a static strength target using a minimal amount of material. Although layout optimisation offers several advantages over the more predominant topology optimisation methods in the field such as the ability to optimise for strength rather than stiffness, high computational efficiency and the clarity of solutions, it hasn't received anywhere near as much research attention. And as with all topology optimisation methods there has been very little literature on validating the structural performance of solutions that have been additively manufactured, despite this manufacturing route being widely cited as an enabler of these methods.

A methodology for producing practical and structurally efficient component designs using layout optimisation was developed and applied to several 3-D problems including one real world problem from the Bloodhound land speed record project. Structural performance was assessed though physical load testing of specimens additively manufactured from titanium Ti-6Al-4V using the Electron Beam Melting (EBM) process. Once all the appropriate design and manufacturing considerations were included into the methodology the resulting load test specimens successfully achieved their strength target. Errors in dimensional accuracy and the presence of internal porosity were highlighted from X-Ray Computed Tomography (XCT) and laser scanning. The main source of the dimensional errors was identified experimentally and suitable process parameters that significantly reduce these errors were established. These process parameters are not just relevant to truss structures but to any reasonably complex component designs.