Failure Detection and Prevention for Additive Manufacturing

Abstract

Additive manufacturing, more commonly referred to as 3d printing, is currently an open-loop process. Even for a perfectly-prepared print, random variations in input conditions can cause failure modes to occur hours or days into a print job. Common failure modes can include separation of the part from the print bed, a clogged print nozzle, or the print head crashing into a malformed piece of the model. Currently, the only way to prevent failures is for a human to close the loop by visually inspecting the print periodically throughout the job. Without human intervention (either in person or through remote monitoring), a failed print can waste hours or days of time and cost the user up to $50 per print job depending on the amount and type of filament wasted. The current work describes a system that automates the detection of print failures and, upon detection of a failure state, automatically pauses the print job and notifies the user. This system has been implemented and tested at the United States Air Force Academy Aeronautics Laboratory and is demonstrated to reduce waste, save time, and improve overall efficiency in the fabrication process. The system is comprised of commercial-of-the-shelf components, making it simple to construct and implement a similar system in other facilities.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 10, 2024
Accession Number
AD1227904

Entities

People

  • Christopher J. Coley
  • Judson T. Babcock

Tags

Readers

  • Educational Psychology
  • Manufacturing Engineering.
  • Robotics and Automation.