Additive Manufacturing: Equipment, Instrumentation and In-Situ Process Monitoring: Equipment, Instrumentation and In-Situ Process Monitoring I
Sponsored by: TMS Additive Manufacturing Committee
Program Organizers: Ulf Ackelid, Freemelt AB; Joy Gockel, Colorado School Of Mines; Sneha Prabha Narra, Carnegie Mellon University; Ola Harrysson, North Carolina State University

Tuesday 8:00 AM
October 11, 2022
Room: 304
Location: David L. Lawrence Convention Center

Session Chair: Sneha Prabha Narra, Carnegie Mellon University


8:00 AM  Invited
Wire DED Process Monitoring and Controls: Andrzej Nycz1; Chris Masuo1; William Carter1; Derek Vaughan1; 1Oak Ridge National Laboratory
    Wire based direct energy deposition processes are new AM techniques with potential to revolutionize high deposition AM market. Due to part size and complexity it is not always possible to apply nondestructive evaluation techniques. The alternative is to perform in situ process monitoring leading to the concept of building the digital history of part also known as the digital twin. This work will present the system architecture, data acquisition techniques and controls for wire-arc and laser wire additive manufacturing.

8:40 AM  
Design of a Glovebox for In Situ Monitoring of a Directed Energy Deposition Process: Marwan Haddad1; Hui Wang1; Ethan Weinberg1; Ronald Sellers1; Karan Kankaria1; Mahsa Valizadeh1; Sarah Wolff1; 1Texas A&M University
    The objective of this work is to present a laser-based, metal powder blown directed energy deposition experimental setup. The operando glovebox setup is used for conducting and monitoring experiments to study certain phenomena in the process, such as powder-melt pool-laser interactions and defects formation. The main advantage of a customized open architecture design is to allow more flexible process control and monitoring. The glovebox setup includes a chamber, a purge antechamber, a fixed laser welding head with four nozzles for powder feeding, three powder feeders, an argon gas regulator, and a substrate holder with a three-axis motor system. The chamber has an opening in the back to perform X-ray imaging and X-ray diffraction on the printed sample and two glove ports in the laser safety glass in the front to handle reactive powders. Off-axis in situ monitoring is performed from the outside using an ultra-high-speed camera installed on a tripod.

9:00 AM  Cancelled
Identifying Melt Pool Dimensions and Melt Pool Variability in Laser Powder Feed Directed Energy Deposition Additive Manufacturing: Ryan Utz1; Jose Loli1; Brandon Abranovic1; Jack Beuth1; Sneha Narra1; 1Carnegie Mellon University
    Defect detection is desired during directed energy deposition additive manufacturing processes to optimize process variables and improve material properties. This work focuses on analysis of in-situ melt pool monitoring data acquired on a TRUMPF TruLaser Cell 3000 using their in-process monitoring system. Computer vision techniques are applied to determine melt pool dimensions during printing, including the experimental determination of conversion factors between measured dimensions in pixels and physical dimensions. The developed algorithms are fast enough to be applied in real time as a part of a process monitoring and control platform. In addition, machine learning techniques are used to track melt pool variability for post-processing evaluation of process consistency. Together these methods are allowing for the maintenance of key melt pool dimensions during builds and the identification of locations within builds where flaws may exist, as an aid for non-destructive and destructive evaluation.

9:20 AM  
Melt Pool-Scale Process Monitoring of Laser Powder Bed Fusion: Christian Gobert1; Guadalupe Quirarte1; Syed Zia Uddin1; David Guirguis1; Jonathan Malen1; Conrad Tucker1; Jack Beuth1; 1Carnegie Mellon University
    This talk gives an overview of a variety of approaches used at Carnegie Mellon to track melt pool characteristics using high speed cameras, including the analysis of high speed video frames using machine learning techniques. These include the tracking of melt pool shape as viewed from above using moderate camera speeds of 6500 fps, and its correlation to the generation of keyhole-induced flaws. Higher camera speeds near 22,000fps are used to track the emission of melt pool spatter under a variety of conditions. Ultra high speed imaging approaching 200,000 fps is being used to track variability in melt pool shape and width. Finally, color camera imaging at high speeds is being used to identify temperature fields outside of the melt pool and to some distance within it. Each of these techniques is being linked to process manipulation, modeling or other process monitoring research.

9:40 AM  Invited
In-situ Monitoring of the EBM Process: From Powder Bed Homogeneity to Thermal Signatures: Marco Grasso1; Bianca Colosimo1; 1Politecnico di Milano
    The availability of in-situ monitoring solutions for Electron Beam Melting (EBM) is still quite limited compared to the wide number of studies and industrial developments devoted to laser-based additive manufacturing. Nevertheless, different quantities can be measured and used to determine the stability of the melting process and detect the onset of defects. This study presents different in-situ monitoring opportunities in EBM, ranging from powder bed homogeneity analysis to in-line characterization of the thermal history of the process. The presented real case studies refer to the production of Ti6Al4V parts using an industrial EBM system equipped with machine vision in the visible and in the infrared range. The results show that, depending on the sensing method and the data analysis techniques, various kinds of anomalies can be detected, enabling the use of in-situ monitoring methodologies to anticipate the detection of defects and to make process and product qualification procedures more efficient.

10:20 AM Break

10:40 AM  
Process Monitoring of Melt Pool Spatter at Melt Pool, Layer and Part Scales: Christian Gobert1; Syed Zia Uddin1; Brandon Abranovic1; Jack Beuth1; 1Carnegie Mellon University
    Emission of melt pool spatter is a significant concern for laser powder bed fusion AM. When spatter particles land on unfused material, large lack of fusion-type defects can result. This research involves in-process monitoring of spatter generation by machine learning analyses of high speed videos at the melt pool scale, infrared videos of entire builds, and individual images of fused layers. The combination of these is allowing determination of spatter counts and trajectories as spatter is emitted from melt pools, tracking of spatter particles as they are captured by the flow of argon across the build, and identification of spatter after it has landed on powder. This talk will present results showing changes in spatter counts and trajectories as a function of process parameters, the ability of argon flows to direct spatter away from unfused regions, and the correlation of spatter seen in fused layers with build conditions.

11:00 AM  
Automated Detection and Quantification of Spatter Generated During Laser Powder Bed Fusion Using Infrared Imaging and Computer Vision: Syed Zia Uddin1; Nicholas O'Brien1; Satbir Singh1; Jack Beuth1; 1Carnegie Mellon University
    Generation of melt pool spatter is commonly seen in laser powder bed fusion (LPBF) processing. Spatter can be a source of porosity in a subsequent layer if not remelted properly. Therefore, the larger the spatter particles, the greater the threat they pose to the quality of the build. Detection and quantification of spatter, and particularly large spatter particles can be critical in LPBF quality assurance. In this research, we have developed an infrared (IR) imaging setup and associated computer vision software to identify spatter particles generated during LPBF builds. Images cover the entire build area and attention is focused on the detection of large spatter particles. Spatter counts as a function of particle size are presented. Spatter detection and quantification is successfully performed on multiple builds and spatter transport across the build area is compared to CFD simulations of argon flow and associated spatter pickup and transport.

11:20 AM  
Several Ways Ultrasound Can Be Used during Powder Bed Fusion: Christopher Kube1; Nathan Kizer1; Corey Dickman1; Edward Reutzel1; 1Penn State University
     Ultrasonic or elastic waves in the MHz range are capable of propagating deep into metallic parts where they can interact and scatter from heterogeneities. Scattering from heterogeneities leads to sensitivity to defects, microstructure, and has even been used to monitor melt pool boundaries, dynamics, and solidification. Alternatively, ultrasound can be applied as a processing step to alter the microstructure. <br><br>This presentation describes research toward transitioning these capabilities into the powder bed fusion process. This is achieved by integrating an array of ultrasonic transducers into an EOS M280 build substrate. 9Cr-1Mo parts with a range of process parameters are manufactured directly above each transducer allowing for the assessment of the aforementioned capabilities. Particularly exciting is the possible simultaneous in situ monitoring and influence on microstructure formation, which could enable potential on the fly corrections along with understanding the mechanisms of how ultrasonic treatments could tailor AM microstructure.