Expert elicitation to assess real-world productivity gains in laser powder bed fusion
Tracey Ziev,Parth Vaishnav
DOI: https://doi.org/10.1108/rpj-01-2024-0006
2024-11-28
Rapid Prototyping Journal
Abstract:Purpose Laser powder bed fusion (LPBF) additive manufacturing (AM) enables the design of complex parts using materials that are otherwise difficult to fabricate. Due to the high cost of machines, the parts produced by LBPF are expensive. Both researchers and industry are therefore focused on lowering costs by improving productivity while ensuring part quality. The purpose of this study is to quantify the productivity gains from using laser beam shaping, multi-laser printing and the use of large build chambers to print larger size parts. Design/methodology/approach This paper performs an expert elicitation with 18 experts. Findings This paper finds that experts believe that larger parts are less likely to print successfully. Increasing the part footprint is more detrimental to print success than increasing part height. Experts also believe that beam shaping is expected to provide limited print time improvement (median 4% reduction, 90% CI: 2%–25%) while improving part quality, whereas going from one to two lasers is expected to provide a median of 25% (90% CI: 10%–45%) print time improvement but degrade part quality. Through cost analysis of a representative part, this paper shows that the uncertainty in build success rates for large parts dominates expected cost reductions from laser beam shaping or multi-laser printing. Research limitations/implications The study has three key limitations. First, it is possible that the sample of experts who agreed to take the survey biases the results. By definition, these are individuals who are willing to share what they know. There may be other experts who have a different view of the efficacy of the technologies evaluated here, but that view might be based on proprietary knowledge, which those experts are unable to share. Second, an elicitation captures what is known at a moment in time. As technology improves and as widespread deployment results in learning, the most consequential finding − that experts believed that success rates for large builds are likely to be low − may become less valid. Third, the overarching goal of this study is to assess technologies to improve AM productivity for high performance metal parts. A single study can only partially achieve this goal. The selection of technologies is constrained by both the desire to keep the study tractable and the suitability of expert elicitation as a method. For example, expert elicitation is not appropriate to assess the efficacy of technologies where sufficient empirical data or analytical techniques exist. Practical implications The results show that AM research and policy initiatives, including standards and regulatory schemes, must support efforts to improve the repeatability and reliability of the technological innovations that are needed to deploy AM in cost-critical or high throughput applications. These results also reinforce the criticality of workforce development components of existing (and future) AM policy initiatives. The elicitation revealed a significant number of factors that must be considered and potentially managed to ensure successful builds. Notably, no experts interviewed discussed all factors. While this may be a consequence of availability bias, it suggests that inexperienced AM users and nonexpert decision-makers, including managers, who would like to adopt new AM technologies, may be unaware of the myriad mechanisms by which build failure can occur and may fail to take mitigating action. This result contradicts a common belief that complicated parts can be fabricated with little to no expertise (assuming access to a design file for the part). Workforce development programs will be essential to help AM users develop the knowledge required to successfully implement metal AM. Originality/value Several strategies, including increasing the build volume to print larger parts or more parts at a time, using multiple lasers and beam shaping are proposed to improve the productivity of AM. However, the real-world efficacy of these strategies is not known. This work pools the judgment of experts to give decision-makers some insight into the current, real-world efficacy of these approaches.
engineering, mechanical,materials science, multidisciplinary