The purpose of this study is to investigate the relationship between 3D printing input parameters and various output (target) variables such as wear, print time, printing cost and dimensional accuracy in both print and scan directions to show the order of parameter importance on each target variable and the overall order of feature importance across all target variables. Furthermore, to quantify the parameter-based sustainability and calculate the sustainability index (SI) (considering cost and CO2 emissions) by correlating the process, wear performance and sustainability.
This study uses machine learning (ML) to optimize process parameters in additive manufacturing (AM), focusing on wear resistance, accuracy, cost and printing time. A gradient boosting regressor is trained on experimental data to find relationships between parameters and performance. The wear performance was assessed by conducting sliding wear tests using a Pin-on-Disc tribometer, while sustainability was evaluated based on material cost and CO2 emissions. The connection between process parameters, wear and sustainability provides useful insights to make 3D printing more efficient and environmentally friendly.
The layer thickness was the overall most influencing process parameter affecting all the output variables studied. The solid infill pattern with 0.100 mm layer thickness consumed the maximum amount of material, took maximum printing time, consumed maximum electricity and was responsible for the highest amount of CO2 emissions and, hence, the least SI. The wear performance and the SI follow an opposite trend. The solid infill pattern with 0.200 mm layer thickness was found to be the most optimum case possessing the best wear performance in the most sustainable manner.
This study will help in decision-making for improved wear performance and setting the trade with sustainability of 3D-printed parts in various industries such as aerospace and healthcare.
This study supports sustainable manufacturing, reduces environmental impact, enhances product reliability and promotes artificial intelligence-driven innovation in industries and education.
This study uniquely correlates process parameters with multiple objectives, namely, dimensional accuracy, printing time, material consumed and associated cost and wear performance, both individually and cumulatively, using ML. The process–wear–sustainability correlation is addressed for the first time.
