Session: ASME Undergraduate Student Design Expo

Paper Number: 176082

Defect Detection in Laser Powder Bed Fusion Using Optimal and Computed Tomography 

Metal Additive Manufacturing (MAM) has become a vital part of modern manufacturing, particularly in the aerospace, medical, and energy sectors. It has allowed the production of complex parts that would have otherwise not existed or would have needed to be constructed out of separate components, rapid prototyping, designing novel alloys that would be difficult or impossible to process using traditional methods, and reducing the waste of material during the build process. In our lab, we utilize the EOS M290 printer, a Laser Powder Bed Fusion (LPBF) system, to study and optimize metal 3D printing. It is capable of printing with a variety of materials, and our lab has mainly focused on Titanium. Printers that utilize Laser Powder Bed Fusion (LPBF) are highly precise and can create incredibly small structures that are impossible to make using any other manufacturing method. One of the main challenges in LPBF is ensuring consistent print quality, as minor variations in print conditions can lead to defects such as porosity, lack of fusion, surface irregularities, warping, etc... To combat this, the machine has an optimal tomography (OT) system that uses a near-infrared camera to capture a temperature map of the part layer by layer. This powerful in situ monitoring system allows us to determine whether the print process has issues such as overheating, porosity, layer misalignment, or recoater failures. After the print is complete, the part is scanned using Computed Tomography (CT) to create a 3D model of the part layer by layer, which is then compared to the (OT) data. Through analyzing the data, our lab can determine the efficiency of the printing process, the internal structure of the part and whether there are defects such as porosity, cracks, warping, and uneven surface finish. The lab’s goal for the future is to create a robust in situ monitoring, closed-loop, real-time data collecting system that uses machine learning to predict defects before they occur and alter the printing process to prevent them. It would collect data such as melt pool and OT data and through this system correct errors before they propagate. This system helps eliminate the waste of material and time through reducing the need to reprint defective parts. It also has potential to produce better printing techniques and help researchers uncover the complex relationships between parameters such as material, laser power, scan speed, and grain size. In conclusion, the integration of advanced in situ monitoring techniques like optimal tomography with post-process CT scanning provides a comprehensive approach to understanding and optimizing the Laser Powder Bed Fusion process in Metal Additive Manufacturing. By leveraging these technologies, alongside machine learning, our lab aims to reduce defects and improve print quality for the many industries it impacts.

Presenting Author: Adeel Khawaja University of Memphis

Presenting Author Biography: An undergraduate senior mechanical engineering student concentrating in advanced manufacturing. Working in the University of Memphis Metal Additive Manufacturing Lab.

Authors:

Adeel Khawaja University of Memphis
Madison Cole University of Memphis