Carnegie Mellon University
Additive manufacturing of metal parts is experiencing strong growth as powder bed machines in particular become more widely available. Although parts are being successfully made there are few tools available that allow the user to predict what the microstructure and properties of the parts will be. This project will focus on a commonly used nickel-based superalloy and develop some of the models that are needed. Porosity is a particularly important component of microstructure because fatigue typically starts from the largest flaw in the loaded volume. We will predict the pore structure that can arise from lack of fusion in additive parts based on the process conditions and scan geometry. We will further predict the pore structure that arises from the pore structure of the powder particles themselves and the way in which such pores can be trapped in the melt pool, thus persisting into the part. In order to address strength, we will develop a strength model based on the known thermodynamic data for the IN718 alloy that predicts the occurrence of precipitation reactions and the thermal histories as a function of location in a part. Experimental data for strength in additive parts will be used to refine and validate the model. Accordingly the overall objective of the project is to implement microstructure prediction models that can be used by the additive manufacturing community.