Intelligent orientation of parts based on defect prediction in Multi Jet Fusion process

Multi Jet Fusion technology allows the rapid production of high-quality polymeric parts. This process has attracted the interest of several companies involved in rapid prototyping, mass customisation and small-batch production. The components produced via Multi Jet Fusion can exhibit some characteristic defects depending on the part geometry and orientation, the process parameters and the environmental conditions. First of all, the most characteristic defects related to the process and their main causes are analysed. Five numerical descriptors are proposed to qualitatively predict the impact of these defects based on the digital representation of a part. Then, an intelligent system is developed to replicate the decision-making process of a human expert orienting parts before nesting. Particularly, a genetic algorithm-based approach is used to identify a pseudo-optimal part orientation under a set of objectives defined by the user. The fitness function is obtained as a weighted sum of the above-mentioned descriptors. Finally, a case study demonstrates the effectiveness of the proposed method in overcoming the characteristic defects of the process via part orienting.