Material Extrusion Filament Width and Height Prediction via Design of Experiment and Machine Learning.

Area of Science:

• Additive Manufacturing
• Materials Science
• Computational Engineering

Background:

• Filament dimensions in material extrusion 3D printing critically affect processing resolution and efficiency.
• Key process parameters influencing these dimensions include nozzle diameter, nondimensional nozzle height, extrusion pressure, and printing speed.

Purpose of the Study:

• To investigate the impact of four key process parameters on the geometric dimensions (width and height) of 3D printed filaments.
• To develop and evaluate machine learning models for predicting filament dimensions.
• To analyze the influence of process parameters on filament morphology.

Main Methods:

• Design of Experiments (DOE) and variance analysis were employed to assess parameter effects and interactions.
• Five machine learning models (Support Vector Regression, Backpropagation Neural Network, Decision Tree, Random Forest, K-Nearest Neighbor) were trained to predict filament dimensions.
• The Backpropagation Neural Network model achieved high predictive accuracy, with R-squared values of 0.9025 for width and 0.9604 for height.

Main Results:

• Nozzle diameter was identified as the most influential parameter on filament width and height, followed by printing speed, extrusion pressure, and nondimensional nozzle height.
• Nondimensional nozzle height affects filament dimensions through stretching (leading to thin filaments with less parameter regularity) or squeezing.
• Nozzle diameter significantly impacts dimensions when material is in a squeezed state.

Conclusions:

• The developed prediction models can accurately forecast the size of 3D printed filament structures.
• Findings provide guidance for selecting optimal printing parameters to control filament dimensions.
• This research contributes to determining the precise size of 3D printing layers for improved manufacturing outcomes.