A framework for fused deposition modelling process evaluation using multivariate process capability analysis

• Published in: PloS one 2024-12, Vol.19 (12), p.e0308380
• Main Authors: Alatefi, Moath, Al-Ahmari, Abdulrahman M, AlFaify, Abdullah Yahia, Saleh, Mustafa
• Format: Article
• Language: English
• Subjects: 3D printing; Accuracy; Additive manufacturing; Algorithms; Analysis; Capability indices; Data reduction; Deposition; Design; Engineering and technology; Evaluation; Fused deposition modeling; Innovations; Investigations; Literature reviews; Manufacturing; Manufacturing industry; Methods; Models, Theoretical; Multivariate Analysis; Performance evaluation; Physical Sciences; Process capability analysis; Product design; Quality Control; Research and Analysis Methods; Sensitivity analysis; Specifications
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0308380

The rapid advancement of additive manufacturing (AM) requires researchers to keep up with these advancements by continually improving the AM processes. Improving manufacturing processes involves evaluating the process outputs and their conformity to the required specifications. Process capability indices, calculated using critical quality characteristics (QCs), have long been used in the evaluation process due to their proven effectiveness. AM processes typically involve multi-correlated critical QCs, indicating the need to develop a multivariate process capability index (MPCI) rather than a univariate capability index, which may lead to misleading results. In this regard, this study proposes a general methodological framework for evaluating AM processes using MPCI. The proposed framework starts by identifying the AM process and product design. Fused Deposition Modeling (FDM) is chosen for this investigation. Then, the specification limits associated with critical QCs are established. To ensure that the MPCI assumptions are met, the critical QCs data are examined for normality, stability, and correlation. Additionally, the MPCI is estimated by simulating a large sample using the properties of the collected QCs data and determining the percent of nonconforming (PNC). Furthermore, the FDM process and its capable tolerance limits are then assessed using the proposed MPCI. Finally, the study presents a sensitivity analysis of the FDM process and suggestions for improvement based on the analysis of assignable causes of variation. The results revealed that the considered process mean is shifted for all QCs, and the most variation is associated with part diameter data. Moreover, the process data are not normally distributed, and the proposed transformation algorithm performs well in reducing data skewness. Also, the performance of the FDM process according to different designations of specification limits was estimated. The results showed that the FDM process is incapable of different designs except with very coarse specifications.