Tailorable and Adaptive Connected Digital Additive Manufacturing

The TACDAM project will remove the the final hurdles for the adoption of additive manufacturing in automotive applications. Although the biggest single contributor to product cost in Additive Manufacturing (AM) has in the past been build time, costs associated to pre- and post-processing are now becoming relatively much more significant. Of these, one significant issue in post-processing is the removal of surplus powder from the component. In Powder Bed Fusion approaches to AM, successive layers of powder (eg. metal powder) are laid out on the manufacturing bed and a high powered laser is used to sinter the powder together in the appropriate regions of the build. At the end of the process, this leaves surplus powder trapped in the interior of the component which has to be removed. This is typically achieved by a number of mechanisms, including vibrating the component at high frequency to fluidise the powder allowing it to flow out under gravity. This process has rarely been analysed in any detail, but is a vital aspect of the manufacturing process as a whole. It is particularly critical with the complex components being developed by organisations within this research team, such as the compact heat exchangers developed by HiETA; any residual powder within the thousands of tubes in the heat exchanger will substantially degrade performance. At the same time the complex geometries developed mean that the flow of the fluidised powder is not straightforward.

The objective of our contribution to the overall project will be to develop a methodology to model the flow of the residual powder within the component in order to be able to identify problems in the powder removal, and optimise powder removal strategies.

We will start by reviewing the existing state of the art in fluidised powder flow. This has not typically been applied to this type of problem so we will identify areas where the physical modelling may need further development to cope with the specifics of this problem. Based on this we will develop a non-Newtonian formulation for the flow of the fluidised powder, implementing this into the OpenFOAM CFD code library. As the powder flows out of a component, air voids will develop, so we will need to formulate the fluidised powder flow within a free surface flow model; this will be accomplished using the standard Volume of Fluid formulation as implemented within OpenFOAM, and validated against experimental results from the literature and from other parts of the project (ASDEC). The modelling will also need to account for the vibrational modes of the component, possibly through coupled modelling of the solid component, which will be investigated as a separate task. Specific characteristic geometries will be identified for investigation using the new modelling; these will be geometries such as angled bends, manifolds and constrictions which either occur frequently in AM or exhibit particular problems with powder removal. In identifying these geometries we will take particular input from the industrial partners in the collaboration (particularly HiETA, accessing their general knowledge of AM). We will simulate these geometries to identify problems with the flow, particularly issues such as dead spots where powder is not being removed; and attempt to correlate this with empirical knowledge of powder removal. Finally, we will examine possible ways in which this modelling could be made more widely used, for example through embedding into an expert system for AM manufacturing.

Surplus particle removal is an area of the AM process which has significant impact on the manufacture of complex automotive parts, particularly ones with complex internal structure such as heat exchangers. At the same time, little research has gone into identifying problems with the powder removal. We aim to develop CFD-based models for the flow of fluidised powder to better understand the powder removal process. This will enable us to examine in great detail common and/or problematic geometries and provide feedback to the commercial partners to develop their understanding of the process. This will enormously aid the overall aim of the project which is to remove the final pre- and post-processing hurdles to the adoption of AM technology into automotive applications, with the potential to generate significant impact on this part of UK manufacturing. In addition, we will look in more general terms at how we might embed the results of such flow modelling into the AM process more generally, through both detailed and reduced order modelling and expert systems; this could potentially affect the application of AM in all areas of manufacture, not just the automotive sector