Detecting Additive Manufacturing Defects in Real Time

Zhongshu Ren (left) and Tao Sun, University of Virginia, display the results of L-PBF research.

A research team led by University of Virginia Associate Professor Tao Sun has made significant strides in advancing AM, particularly in aerospace and industries requiring robust metal parts. Their study focuses on detecting keyhole pores, which are major defects in laser powder bed fusion (L-PBF).

Porosity in L-PBF parts is considered a defect and it poses challenges for applications that are sensitive to fatigue, such as aircraft wings. Keyhole pores form as deep, narrow vapor depressions during the manufacturing process and are influenced by laser power, scanning speed, and material properties. Stable keyhole walls enhance laser absorption, improving manufacturing efficiency, while unstable walls can trap air pockets, making the material brittle and susceptible to cracking.

Sun's team, along with experts from Carnegie Mellon University and the University of Wisconsin-Madison, developed a method to detect the exact moment a keyhole pore forms during printing. By combining operando synchrotron x-ray imaging, near-infrared imaging, and machine learning, they achieved sub-millisecond temporal resolution with a 100% prediction rate for keyhole pore generation.

This innovative approach not only improves the capabilities of synchrotron imaging but also identifies two modes of keyhole oscillation. According to Professor Anthony Rollett, Carnegie Mellon University, the findings could broaden the commercial application of L-PBF in metal part manufacturing. Sun emphasized that addressing keyhole porosity is crucial for the wider adoption of L-PBF and that their method offers a reliable solution for real-time detection in various AM scenarios.