A methodology for the development of functionally graded powder spreading in laser powder bed fusion process using discrete element method

Abstract The use of multi-material components offers customization of physical properties, weight reduction, effective thermal management, and the creation of material-compatible buffer components to join two material with ease. These features surpasses the capabilities of single-material compositions.When multiple materials are used, their sharp interfaces can separate and lead to mechanical failure. To prevent this, functionally graded materials with diffuse interfaces can be employed. Functionally graded materials (FGM) possess preferred spatial variation of properties aligned in specific directions. However, producing complex FGM components through conventional methods is challenging, as the conventional manufacturing methods are part and tool-specific. Components made using additive manufacturing, such as powder bed fusion (PBF), can create functionally graded materials (FGM) with intricate geometric features and precision at the micron scale. This opens up new avenues for innovative design possibilities with FGM components. The methodologies developed to create FGM by PBF are still in their infancy and require further attention to realise defect-free components. By employing high-fidelity mathematical models, new methodologies can be developed and minimise expensive trial-and-error experimental development strategies.The discrete element method (DEM) is a suitable numerical approach for modelling discontinuous media, such as powder particles in PBF. In this study, a spreading procedure in a powder bed fusion process is developed so that the desired distribution of material composition can be obtained before laser melting. A partition-based approach is adapted to achieve functional gradation along the spreading direction. The role of recoater speed on the evolution of the distribution of the material was studied through a parameter called gradation index (GI). A unique experimental setup was developed to analyse the prediction of the developed model. Results show that an angular partition at the dispenser can generate a customised functionally graded spreading in the build platform, and the obtained graded spreading is found to vary as a function of the recoater speed, partition angle, and spread layer thickness.