Session: 03-09-01: Data-Driven Innovation in Smart Product Design and Manufacturing

Paper Number: 173040

Real Time Monitoring of Anomalies During the Laser Powder Bed Fusion With Optical Tomography 

Introduction

Laser Powder Bed Fusion (LPBF) additive manufacturing is a 3D printing process that enables the fabrication of complex, lightweight, and porous components that are difficult or impossible to produce using conventional methods. However, LPBF can still result in various types of defects—such as porosity, lack of fusion, surface irregularities, and thermal inconsistencies—that compromise part quality and performance. Detecting these issues using traditional post-build characterization tools is often labor-intensive, time-consuming, and costly. In-situ monitoring offers a promising alternative by enabling real-time, layer-by-layer observation of the build process, reducing reliance on extensive post-processing and accelerating quality assurance in metal additive manufacturing.

Contribution

This project aims to advance in-situ, real-time monitoring techniques for identifying and evaluating a wide range of defects and anomalies during LPBF fabrication. By developing a scalable, non-destructive framework for quality assessment, the work seeks to provide a low-cost alternative to traditional inspection methods such as X-ray CT. Ultimately, the project contributes to the broader integration of real-time feedback systems and data-driven quality assurance strategies in metal additive manufacturing.

Methodology

In this process, the EOS M290 LPBF system is used for fabrication, and an integrated Optical Tomography (OT) system captures thermal emissions during each layer’s exposure. A range of defects—including lack of fusion, porosity, and geometric deviations—will be studied. Various post-process characterization tools, such as X-ray CT and metallographic analysis, will be used to verify the detected anomalies and assess the accuracy of the in-situ monitoring data. Image processing techniques will be used to classify defect types and support the development of automated detection algorithms.

Preliminary Status and Expectations

The monitoring system has been successfully installed and verified for basic defect detection capabilities. Initial data collection is planned using test builds with varying geometries and process conditions. The project will focus on evaluating the system’s ability to detect and characterize a range of build anomalies. Detected anomalies will be validated using established post-process inspection techniques, such as X-ray CT and metallographic analysis. It is expected that the captured data will offer sufficient resolution and reliability to identify macroscale defects.

Conclusion and Future Work

This work aims to develop a scalable, non-destructive quality assurance framework that minimizes reliance on expensive post-build inspection methods such as X-ray CT and laser scanning. By leveraging Optical Tomography (OT) data collected during the LPBF process, the project lays the groundwork for real-time detection and classification of a wide range of defects and anomalies.

Future work will focus on evaluating the resolution limits of the OT system for detecting various defect types and implementing machine learning algorithms to enable predictive, real-time analysis during the build process ultimately supporting more reliable and efficient production in industrial settings.

Presenting Author: Yusef Qazzaz University of Memphis

Presenting Author Biography: Yusef Qazzaz is a Ph.D. candidate in Mechanical Engineering at The University of Memphis, where his research focuses on additive manufacturing, specifically real-time monitoring and process control in Laser Powder Bed Fusion. His work integrates experimental techniques and data-driven approaches to improve part quality and reliability in metal 3D printing. Yusef has served as a teaching assistant and researcher in advanced manufacturing labs and has presented his work at multiple conferences. He is passionate about bridging the gap between manufacturing science and industrial application through innovation and collaboration.

Authors:

Yusef Qazzaz University of Memphis
Craig Bowlin University of Memphis