Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Data-Driven Extrusion Force Control Tuning for 3D Printing

Xavier Guidetti, Ankita Mukne, Marvin Rueppel, Yannick Nagel, Efe C. Balta, John Lygeros

TL;DR

This work provides a novel framework for closed-loop 3D printing and proposes an automated calibration routine that produces high-quality prints for a desired combination of print settings, material, and shape.

Abstract

The quality of 3D prints often varies due to different conditions inherent to each print, such as filament type, print speed, and nozzle size. Closed-loop process control methods improve the accuracy and repeatability of 3D prints. However, optimal tuning of controllers for given process parameters and design geometry is often a challenge with manually tuned controllers resulting in inconsistent and suboptimal results. This work employs Bayesian optimization to identify the optimal controller parameters. Additionally, we explore transfer learning in the context of 3D printing by leveraging prior information from past trials. By integrating optimized extrusion force control and transfer learning, we provide a novel framework for closed-loop 3D printing and propose an automated calibration routine that produces high-quality prints for a desired combination of print settings, material, and shape.

Data-Driven Extrusion Force Control Tuning for 3D Printing

TL;DR

This work provides a novel framework for closed-loop 3D printing and proposes an automated calibration routine that produces high-quality prints for a desired combination of print settings, material, and shape.

Abstract

The quality of 3D prints often varies due to different conditions inherent to each print, such as filament type, print speed, and nozzle size. Closed-loop process control methods improve the accuracy and repeatability of 3D prints. However, optimal tuning of controllers for given process parameters and design geometry is often a challenge with manually tuned controllers resulting in inconsistent and suboptimal results. This work employs Bayesian optimization to identify the optimal controller parameters. Additionally, we explore transfer learning in the context of 3D printing by leveraging prior information from past trials. By integrating optimized extrusion force control and transfer learning, we provide a novel framework for closed-loop 3D printing and propose an automated calibration routine that produces high-quality prints for a desired combination of print settings, material, and shape.
Paper Structure (16 sections, 6 equations, 5 figures, 4 tables, 1 algorithm)

This paper contains 16 sections, 6 equations, 5 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Background
  3. Force Controlled Fused Filament Fabrication
  4. Bayesian Optimization
  5. Transfer Learning
  6. Proposed Framework
  7. Controller Structure
  8. Continuous Bayesian Optimization
  9. Transfer Learning with Bayesian Optimization
  10. Experimental Study
  11. Hardware Setup and Printing Conditions
  12. Results
  13. Single-Task Bayesian Optimization
  14. Transfer Learning
  15. Discussion
  16. ...and 1 more sections

Figures (5)

  • Figure 1: Simplified diagram of the extrusion process and feedback loop for fcp for FFF
  • Figure 2: Flowchart of the continuous Bayesian optimization method
  • Figure 3: Printed tower shells on the experimental setup with force feedback extrusion control. Both parts are continuously printed until user interruption, with force reference 0.3N.
  • Figure 4: Comparison of convergence plots from controller optimization for: (a) force reference of 0.3N with TL using data from the optimization of 0.4N, (b) force reference of 0.3N with TL using data from the optimization of 0.2N, (c) force reference of 0.2N with TL using data from the optimization of both 0.3N and 0.4N.
  • Figure 5: Comparison of convergence plots from controller optimization for a force reference of 0.3N