A multivariate adaptive regression splines (MARS) nonparametric algorithm is used to approximate the relationship between the fused deposition modeling (FDM) process parameters: infill density, infill pattern, layer thickness, print speed, and shell thickness and the FDM’s energy consumption, production time, and mechanical properties. The FDM is one of the most used additive manufacturing (AM) processes for production of parts across different industries due to its capabilities in building free-form and complex shapes that are otherwise challenging to manufacture by conventional methods. However, there is not yet comprehensive understanding of how the process parameters impact FDM’s energy consumption, mechanical, and physical properties of engineered FDM products. Various advanced computational learning tools such as neural networks have been increasingly used for modeling the highly nonlinear multivariate relationship between the inputs and outputs of engineering problems. Still, a challenge posed to the use of neural networks is lack of a priori optimal configuration for each presented case. Here, MARS is used to develop simple and easy to interpret models of the FDM process parameters with built part quality characteristics. In addition, the MARS will estimate the contributions of the FDM process parameters with their computational efficiency.

Estimation of fused deposition modelling (FDM) part quality characteristics based on multivariate adaptive regression splines

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.