Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Additive manufacturing (AM) is a group of processing technologies which has the potential to revolutionize manufacturing by allowing easy manufacturing of complex shapes and small series. One AM-method which is of high interest for processing of metals is laser powder bed fusion (LPBF), where a powerful laser is used to melt powder in a layer-by-layer fashion into a consolidated part.Many defects can form during LPBF processing, but copper, copper alloys, and similar metals are particularly prone to the formation of pores.The reason why these metals tend to have these defects is the high reflectivity decreasing the effective energy input, and the high thermal conductivity transporting the absorbed energy away. Several ways of decreasing the pore formation tendency are known: increasing the power, increasing the absorptivity, and alloying to decrease thermal conductivity.The relative effectiveness of these strategies are however not known, making it difficult to choose the correct strategy for the different industrial applications of these metals.In this thesis a simple analytical model for predicting the occurrence of these defects is proposed, allowing a lowest possible processing power given the material properties to be estimated. The model agrees well with experiments, and can be used to predict the transition from so called balling mode processing to conduction mode processing, and for qualitative analysis of different processing strategies. Two strategies of decreasing the porosity are proposed and tested: coating of the powder with an absorbing layer and processing with a green laser, which is mor e strongly absorbed by copper than the conventional IR-lasers.The copper powder was coated with thin layers of tin and nickel using a simple and cheap immersion deposition method, and it is shown that the amount of porosity is decreased more than can be expected of the alloying alone, showing both that the method is working and that laser interaction with the solid is important for the heat balance of the system.The interaction with the solid is shown through computational fluid dynamics simulations to arise from periodic fluctuations of the molten metal which cause laser light to be reflected forward, pre-heating the powder bed.The preheating is larger the higher the solid absorptivity is, explaining the porosity decrease when using coated powders, and shows that processing using a green laser increase the laser coupling more than the absorptivity of the liquid would.Single line tracks made from fine 5 µm pure copper with a green laser source show that the desired conduction mode melting can be achieved at a power below 70 W, an order of magnitude lower than what is needed for processing of the conventionally used 45 µm powder with an IR-laser.
,