Methods for optimising the Electron Beam Melting process simulation towards multilayer analysis

The electron beam melting (EBM) is one of the most promising metal powder-bed based additive manufacturing technologies. In recent years, a greater awareness of the advantages and limitations of this technology has been acquired, which have allowed its diffusion in the manufacturing world. In this context, process simulation plays a fundamental role in the process parameters optimisation for material development. The thesis work begins with an overview on the EBM process simulation. The overview highlights the main advantages and limitations according to the current computational resources. Some guidelines for the implementation of simulation models are outlined and the main degrees of freedom according to the EBM process are presented. The most used models have been grouped in according to the scale of the modelling that mainly concerns the powder modelling. Additionally, an overview of the current state of the art of the process development of 48Ti–48Al–2Cr–2Nb alloys is provided. Most of the models simulated the powder bed as a continuum, and usually the local scale finite elements (FE) models are implemented. In this works a new computational analysis approach for the EBM process based on FE method is presented on the basis of a recent tool widely used in simulation. The new approach attempts to reduce the calculation time by the use of progressive activation of the elements. Criteria for the activation elements are proposed according to the physics of the EBM process. All activation criteria have been verified by analysing the activation front, the activation zone size and the calculation time. Thanks to the time reduction, a multilayer simulation has been implemented. The model is initially validated by simulating the melting of a single layer of 48Ti–48Al–2Cr–2Nb alloy and comparing the model outputs with the output provided by a more detail model already validated in literature. Differences in terms of melt pool size, temperature field, calculation time has been analysed . The multilayer simulation has been then validated against experimental results.