Optimizing the impact toughness of PLA materials using machine learning algorithms

This study investigates the influence of various printing parameters on the impact toughness of PLA products, utilizing small sample data and machine learning algorithms to expedite the identification of optimal toughness. We propose a 3D printing tree algorithm model that focuses on the temporal distribution of impact toughness. The relationships between printing parameters and impact toughness are explored and predicted using three algorithms: Random Forest(RF),Light Gradient Boosting Machine(LightGBM) and Adaptive Boosting(AdaBoost). Experimental results indicate that the LightGBM algorithm yields the highest predictive accuracy, with the optimal parameter combination being a layer height of 0.3 mm, an infill density of 100 %, a printing speed of 60 mm/s, and a nozzle temperature of 200-210 degrees C.