Geometric Accuracy Prediction and Improvement for Additive Manufacturing Using Triangular Mesh Shape Data

Abstract

One major impediment to wider adoption of additive manufacturing (AM) is the presence of larger-than-expected shape deviations between an actual print and the intended design. Since large shape deviations/deformations lead to costly scrap and rework, effective learning from previous prints is critical to improve build accuracy of new products for cost reduction. However, products to be built often differ from the past, posing a significant challenge to achieving learning efficacy. The fundamental issue is how to learn a predictive model from a small set of training shapes to predict the accuracy of a new object. Recently an emerging body of work has attempted to generate parametric models through statistical learning to predict and compensate for shape deviations in AM. However, generating such models for 3D freeform shapes currently requires extensive human intervention. This work takes a completely different path by establishing a nonparametric, random forest model through learning from a small training set. One novelty of this approach is to extract features from training shapes/products represented by triangular meshes, as opposed to point-cloud forms. This facilitates fast generation of predictive models for 3D freeform shapes with little human intervention in model specification. A real case study for a fused deposition modeling (FDM) process is conducted to validate model predictions. A practical compensation procedure based on the learned random forest model is also tested for a new part. The overall shape deviation is reduced by 44%, which shows a promising prospect for improving AM print accuracy.

Potential of 3D Printing for Heat Exchanger Heat Transfer Optimization—Sustainability Perspective

“…Additive technology allows for a seamless transition from a digital model to its physical form [34]. One of the main obstacles to the widespread adoption of additive manufacturing is the observed larger-than-expected shape difference between the final print and the intended design [35]. Every 3D-printed object begins as a virtual spatial model (Figure 2).…”

Application of artificial intelligence in 3D printing physical organ models

“…However, the computational complexity as well as the variability of various scenarios, and the cumbersome modeling process during the simulation have limited its widespread application. Decker [ 86 ] et al proposed a geometric accuracy prediction method based on triangular mesh shape data to easily build complex 3D models, and the researchers extracted key geometric features and parameters in the 3D printing process and combined it with an ML algorithm for prediction and optimization. The method not only predicts the possible geometric accuracy problems in the 3D printing process but also optimizes and improves based on the prediction results, thus improving the geometric accuracy and quality of printed products [ [87] , [88] , [89] , [90] , [91] ].summarize some ML-based compensation for material properties and geometry prediction.…”

Predicting geometric errors and failures in additive manufacturing

“…Quality assurance can be achieved by using supervised learning methods to predict the quality of a manufactured item. Decker et al (2020) analytically present research on shape deformation through predictive modeling and compensation approaches, dividing them into two main categories of predictive modeling approaches: physics-based approaches that use finite element modeling (Meier et al , 2021) and data-driven approaches based on statistical analysis and machine learning (Zhu et al , 2018). In the scope of this study, we focus on statistical analysis and machine learning approaches owing to their more generalized outlook.…”

“…This model can then be used to predict the dimensional deviations of other models, even for free-form design. Decker et al (2020) improved their previous method to achieve more accurate shape deviation prediction. This is an interesting first approach to error prediction that works well on small subsets of low-complexity objects.…”