Advanced Real Time Imaging: Additive Manufacturing
Sponsored by: TMS Functional Materials Division, TMS Structural Materials Division, TMS: Advanced Characterization, Testing, and Simulation Committee, TMS: Alloy Phases Committee, TMS: Biomaterials Committee
Program Organizers: Jinichiro Nakano, MatterGreen; David Alman, National Energy Technology Laboratory; Il Sohn, Yonsei University; Hiroyuki Shibata, Tohoku University; Antoine Allanore, Massachusetts Institute of Technology; Noritaka Saito, Kyushu University; Anna Nakano, US Department of Energy National Energy Technology Laboratory; Zuotai Zhang, Southern University of Science and Technology; Candan Tamerler, University of Kansas; Bryan Webler, Carnegie Mellon University; Wangzhong Mu, KTH Royal Institute of Technology; David Veysset, Stanford University

Monday 8:30 AM
March 15, 2021
Room: RM 14
Location: TMS2021 Virtual

Session Chair: David Veysset, Stanford University; Jinichiro Nakano, USDOE National Energy Technology Laboratory


8:30 AM  
An In Situ and Operando Additive Manufacturing Process Replicator for High Speed Optical, Infra-red and Synchrotron X-ray Imaging: Sebastian Marussi1; Chu Lun Alex Leung1; Samuel Clark1; Leigh Stranger2; Robert Atwood3; Veijo Honkimäki4; Alexander Rack4; Mike Besston5; Jon Willmott2; Peter Lee1; 1University College London; 2The University of Sheffield; 3Diamond Light Source Ltd; 4European Synchrotron Radiation Facility; 5Oxford Lasers Ltd
    With increasing demand in using additive manufacturing (AM) technologies to produce high-quality industrial products, a better understanding of the underlying phenomena is required. To achieve this, we developed an in situ and operando process replicator (ISOPR) to simulate industrial laser powder bed fusion in a synchrotron beamline. We determined the design based on the build process and experimental constraints, evaluating the trade-offs required. The key design criteria included the capacity to be integrated on a range of beamlines (5 to date), to enable high-speed X-ray radiography and diffraction to capture melt pool dynamics and microstructural feature evolution. We also incorporated a correlative high-speed infra-red and optical imaging, enabling the synchrotron imaging to be used for calibrating surface-based process responses. We conclude that in situ process replication of laser additive manufacturing provides a means to transform the understanding of AM processing and expedite the optimisation of strategies that improve product quality.

8:50 AM  
In Situ Characterization of the Balling Phenomenon in Additive Manufacturing: Debomita Basu1; Jack Beuth1; Bryan Webler1; 1Carnegie Mellon University
    A major challenge of adopting Laser Powder Bed Fusion into an industrial setting is the relatively slow build rates characteristic of this process. Although using higher laser powers and velocities increases the build rate, undesirable surface tension defects known as the balling phenomenon, or bead-up, may form along the length of the laser track. This creates uneven surfaces for subsequent layers, which can result in embedded pore-type flaws. Since these artifacts form very quickly during the manufacturing process, real time imaging is a necessary tool used to study the formation mechanisms of this phenomenon. In this work, 316 Stainless Steel was examined to determine differences in balling behavior using optical microscopy, high-speed radiography, and IR imaging. Possible processing strategies to change melt pool shape at high powers and high velocities and mitigate balling are also discussed.

9:10 AM  
In-Situ Machine Learning Enabled Spatter Detection in Laser Powder Bed Fusion Additive Manufacturing: Brandon Abranovic1; Jack Beuth1; Rishikesh Magar1; Lalit Ghule1; Amir Farimani1; 1Carnegie Mellon University
    This work focuses on the analysis and classification of fusion images from laser powder bed fusion (LPBF) additive manufacturing processes for the detection of laser spatter on the powder bed. In-situ laser spatter identification is a key component in assuring build quality since it can be correlated to severe failures later in the building process, making it of vital interest to the additive manufacturing community. The work relied upon the thousands of fusion images normally collected in the operation of the EOS M290 LPBF system. Various convolutional artificial neural network architectures including AlexNet, VGG16, and UNnet were tested for both identification of spatter behavior on the powder bed as well as localization of spatter flaws in the powder bed. Preliminary results have shown substantial promise for these approaches in the identification of spatter yielding identification accuracy of roughly 90% and patch-wise segmentation accuracy of roughly 87%.

9:30 AM  
High-speed Synchrotron X-ray Imaging of Metal Additive Manufacturing Processes: Tao Sun1; Kamel Fezzaa2; 1University of Virginia; 2Argonne National Laboratory
    The in situ/operando studies on many dynamic material behaviors have been limited by the characterization tools. Synchrotron x-ray techniques have been demonstrated to be very effective for studying real materials under real conditions in real time. In particular, the high-flux hard x-rays generated at the third-generation high-energy synchrotron facilitates allow imaging of dynamic structure evolution in bulk materials with extremely high spatial and temporal resolutions. At the Advanced Photon Source, we recently applied high-speed x-ray imaging to study multiple metal additive manufacturing processes. By taking advantage of the short pulse and diverse time structure of the source, high-resolution imaging experiments with min exposure time set by the natural width of the pulse (~100 ps) and max frame rate set by the bunch repetition frequency (6.5 MHz) can be performed, which enables quantitative measurements of many dynamic structure parameters during laser-based and binder jetting additive manufacturing.

9:50 AM  Invited
Characterizing Laser-driven Metal Ejecta Interactions: Alison Saunders1; Camelia Stan1; Kyle Mackay1; Suzanne Ali1; Hans Rinderknecht2; Hye-Sook Park1; Jon Eggert1; Fady Najjar1; Tomorr Haxhimali1; Brandon Morgan1; Marcho Echeverria3; Jeremy Horwitz1; Yuan Ping1; 1Lawrence Livermore National Laboratory; 2Laboratory for Laser Energetics; 3University of Connecticut
     The understanding of metal ejecta interactions presents a unique challenge in materials physics due to the high strain rates and fast time scales involved in collisions. There exist few examples of ejecta interaction studies. To that end, we present the first movies of ejecta-ejecta interactions from experiments performed on the OMEGA and EP lasers. Lasers drive shocks through two tin foils with planar trenches carved into their back sides. As the shocks break out, the trenches invert to form planar jets of micron-sized ejecta moving towards each other at speeds of several km/s. We use point-projection radiography to image the jets and discuss the observed interaction dynamics for tin releasing into solid and liquid phases.LLNL-ABS-811923. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 and supported by Laboratory Directed Research and Development (LDRD) Grant No. 18-ERD-060.

10:10 AM  
Quantifying Spatter in Powder Bed Fusion Processes with High-speed Video Observations and Machine Learning: Christian Gobert1; Evan Diewald1; Jack Beuth1; 1Carnegie Mellon University
    During laser powder bed fusion (L-PBF) additive manufacturing, spatter particles ejected from the melt pool region can be detrimental to material performance and powder recycling. Quantifying spatter generation with respect to processing conditions is a step towards mitigating spatter and better understanding the phenomenon. A high-speed camera was used to observe the L-PBF process at multiple laser power and velocities. A machine leaning network was trained to segment regions of spatter particles in the captured high-speed images. A separate machine learning network was used to generate affinity matrices between spatter particles of subsequent frames to aid object tracking. The detection and tracking tool were used to quantify spatter generation of multiple laser power and velocity settings for L-PBF.