A robotic 3D printer for UV-curable thermosets: dimensionality prediction using a data-driven approach

ABSTRACT

This paper presents a robotic 3D printer specifically designed for ultraviolet (UV)-curable thermosets, whose printing parameters can be selected by using a predictive modeling strategy. A specialized extruder head was designed and integrated with a UR5e robotic arm. Software packages were developed to enable the communication between the extruder head and the robotic arm, and control systems were implemented to regulate the printing process. A predictive approach using either a feedforward neural network (FNN) or convolution neural network (CNN) is proposed for estimating the dimensions of future prints based on the process parameters. This enables selection of the appropriate parameters for high-quality prints. This strategy aims to decrease expensive trial-and-error campaigns for material and printing parameter tuning. Experimental results demonstrate the capabilities of the robotic 3D printer and the accuracy of the predictive approach.

KEYWORDS:

• UV-curable thermosets
• robotic system
• additive manufacturing
• machine learning

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Louisiana Board of Regents [LEQSF-EPS(2022)-LAMDASeed-Track1B-11]; Louisiana Board of Regents [LEQSF-EPS(2021)-LAMDASeed-Track1B-01]; Office of Integrative Activities [OIA1946231].