Additive Manufacturing: Advanced Characterization with Synchrotron, Neutron, and In Situ Laboratory-scale Techniques: In Situ Monitoring and Diagnostics: Powder Bed
Sponsored by: TMS: Additive Manufacturing Committee
Program Organizers: Fan Zhang, National Institute of Standards and Technology; Tom Stockman, Los Alamos National Laboratory; Tao Sun, Northwestern University; Donald Brown, Los Alamos National Laboratory; Yan Gao, Ge Research; Amit Pandey, Lockheed Martin Space; Joy Gockel, Wright State University; Tim Horn, North Carolina State University; Sneha Prabha Narra, Carnegie Mellon University; Judy Schneider, University of Alabama at Huntsville
Thursday 2:00 PM
February 27, 2020
Room: 8
Location: San Diego Convention Ctr
Session Chair: Joy Gockel, Wright State University
2:00 PM Invited
A Machine-Agnostic Approach to Layer-wise Process Monitoring and Control of Powder Bed Additive Manufacturing Technologies: Luke Scime1; Derek Siddel1; Vincent Paquit1; 1Oak Ridge National Laboratory
It is well-accepted that robust in-situ process monitoring and control schema are required for the broad adoption of Additive Manufacturing technologies, particularly in industries producing safety-critical components. With new metal Additive machines entering the market every year, it is impractical to develop process monitoring solutions specific to each machine. Therefore, we propose a fully machine and camera agnostic algorithm for the layer-wise classification and localization of anomalies in powder bed systems. This is achieved through a novel Convolutional Neural Network (Machine Learning) architecture that automatically adapts to the input data while maintaining the ability to share learned knowledge across Additive platforms. This capability is demonstrated for multiple binder jet, laser fusion, and electron beam fusion machines. The dynamic architecture also enables fusion of different sensing modalities, including multispectral imaging and combining layer-wise images with melt pool-scale data. Finally, the utility of this algorithm in a process control scenario is demonstrated.
2:30 PM
Rapid Characterization of AM Components for Alloy Design and Process Optimization: Ryan Dehoff1; Alex Plotkowski1; Kevin Sisco2; Paul Brackman3; Pradeep Bhattad3; Curtis Frederick3; Andres Rossy1; Amit Shyam1; 1Oak Ridge National Laboratory; 2University of Tennessee; 3Zeiss AG
The alloy design cycle for additive manufacturing requires significant time for characterization, both to optimize process parameters for minimization of defects, and to understand the alloy microstructure and properties. Reducing time required for these characterization tasks is an opportunity to accelerate alloy development for AM, rapidly optimize process conditions, increase system utilization, and improve part quality. This presentation will outline an approach for efficiently and rapidly selecting and characterizing laser powder bed fusion process parameters for a new Al alloy using x-ray computed tomography, optical microscopy, hardness testing, and other techniques. The advantages and disadvantages of possible characterization approaches will be discussed, and recommended approaches for rapid parameter development will be presented. Finally, the outlook for alloy and process development utilizing new advanced characterization techniques and their impact on the AM community and broader industrial interests will be considered.
2:50 PM
Simultaneous High-speed Measurements of Laser Absorptance and Melt Pool Geometry in Metal Powder Bed Systems: Brian Simonds1; Jack Tanner1; Paul Williams1; Niranjan Parab2; Cang Zhao2; Tao Sun2; 1National Institute of Standards and Technology; 2Argonne National Laboratory
During the metal powder bed fusion additive manufacturing process, the intense laser-metal interaction dramatically and rapidly changes the built surface, which in turn alters the amount of laser energy absorbed locally. These transient changes in energy deposition can lead to the formation of defects including porosity and cracking. In this work, we measured the absolute absorbed laser energy using an integrating sphere while simultaneously imaging the melt pool and keyhole evolution using synchrotron high-speed x-ray imaging. These measurements were performed on both titanium and aluminum alloys with the experimental conditions incrementally increased in complexity from a stationary laser spot on solid metal to a scanning laser beam incident to a metal powder bed. This is the first experimental work that simultaneously correlates the effects of specific melt pool geometries, like vapor cavity shape and collapse, with the amount of laser energy absorbed.
3:10 PM
High Speed Video of the Influence of Preheating on Tungsten Microcracking During Laser Scanning: Bey Vrancken1; Rishi Ganeriwala1; Aiden Martin1; Manyalibo Matthews1; 1Lawrence Livermore National Laboratory
Microcracks in additively manufactured tungsten are currently limiting the range of applications to non-load bearing parts, such as waveguides and collimators. By combining high speed video and thermomechanical modeling, previous work has shown that the microcracks occur after a certain time delay behind the moving melt pool, corresponding to the time to cool down to ductile-to-brittle transition temperature. This study leverages the same procedure to visualize the effect of preheating temperatures up to 1000°C on the microcracking behavior of pure tungsten and tungsten alloyed with different rare earth oxides. The results are an indication of the steps required to produce crack free tungsten parts by additive manufacturing.Prepared by LLNL under Contract DE-AC52-07NA27344, funded by LDRD 18-ERD-057 and 18-SI-003.
3:30 PM
In-situ Measurement of the Kinetics of Homogenization and Aging Treatments in A205 Alloy Produced Through Additive Manufacturing: Guilherme Faria1; Antonio Ramirez1; 1Ohio State University
The aerospace industry is an enticing field for the deployment of additive manufactured (AM) parts. This brings the need for the development of AM techniques in materials relevant for this industry, such as Ni, Al and Ti based alloys. A205, a Cu alloyed Al alloy with desirable mechanical properties and a refined grain structure. Particularly, when produced through selective laser melting, A205 presents micron sized equiaxed grains. Before service, these alloys must go through a careful set a heat treatments to produce an intricate precipitate structure. The schedule for the heat treatment in cast alloys is known, but may be optimized for the unique SLM as-built structure. In this work we present a set of in-situ XRD experiments to track phase transformations during homogenization and aging and in A205 alloys produced through SLM. The kinetics of solidification second phase dissolution and precipitate formation are tracked for several temperatures.
3:50 PM Break
4:10 PM Invited
Quantifying Defect Signatures in Metal Additive Manufacturing Using In-situ Diagnostics: Manyalibo Matthews1; Bradley Jared2; John Carpenter3; Elena Garlea4; Benjamin Brown5; 1Lawrence Livermore National Laboratory; 2Sandia National Laboratories; 3Los Alamos National Laboratory; 4Y-12 National Security Complex; 5Kansas City National Security Campus
Prediction of process-structure-property relationships is a significant challenge that must be overcome to facilitate the rapid, widescale adoption of metal additive manufacturing (AM) for operation-critical applications. Specifically, on-line defect characterization and control remains elusive. To address these issues, in situ techniques that probe and clarify the physics of defect formation are required, along with diagnostics suitable for on-line process monitoring. Here we present studies of metal AM process signatures that have been monitored in-situ using pyrometry, imaging and optical coherence tomography (OCT). These signatures are correlated to as-built material structures captured using X-ray computed tomography (CT), radiography and metallography. Probability relations for defect detection were constructed based on pyrometer signals. Machine learning algorithms were also used to analyze thermal emission data and predict surface roughness. Resulting process-structure-property correlations will be discussed, along with their relevance to part qualification and production acceptance. Prepared by LLNL under Contract DE-AC52-07NA27344.
4:40 PM
Unsupervised Learning Applied to Powder Metals for Additive Manufacturing: Ryan Cohn1; Andrew Kitahara1; Srujana Rao Yarasi1; Elizabeth Holm1; 1Carnegie Mellon University
Current methods of characterizing powders used for additive manufacturing do not provide a complete understanding of powder flow (particle size distribution, hall flow) or are expensive and time consuming (rheology measurements.) In an attempt to provide more reliable high throughput measurements of powder characteristics a computer vision pipeline is proposed. Powder images are segmented so each particle may be analyzed individually. A pre-trained convolutional neural network is used to extract quantitative visual information for each particle. Unsupervised learning is used to cluster the powder particles into distinct groups. The distribution of particles in each group then provides a fingerprint which can be compared between different powders, including powder that is recycled during the additive manufacturing process.
5:00 PM
Coherent Scanning Interferometry for Characterization of Recycled Metal Powder and Reusability Assessment in Additive Manufacturing: Susana Castillo1; Anna Hayes1; Rongguang Liang1; Gregory Colvin1; Krishna Muralidharan1; Douglas Loy1; Barrett Potter1; Christopher Shanor1; 1University of Arizona
Direct Metal Laser Sintering (DMLS) is a widely used additive manufacturing (AM) method. DMLS produces unused metal powder that can undergo significant change in particle size, particle chemistry and morphology. These changes determine the extent of reusability of the powder as a function of cycle-number (the number of cycles over which the powder has been re-used). Using Coherent Scanning Interferometry (CSI), the evolution in particle size distribution and morphology as a function of number of cycles is studied for Inconel 718. It is seen that CSI provides an effective, accurate and a much faster method as compared to traditional electron microscopy and optical techniques; more importantly it provides the ability to be integrated with existing DMLS systems to enable ‘in situ’ monitoring of the powder sintering process. The efficacy of CSI for AM monitoring will be discussed and recommendations for powder reuse will be provided.
5:20 PM
In-line Powder Packing Density Analysis During Selective Laser Melting: Tan-Phuc Le1; Karl Davidson1; Bernard Gaskey1; Po-Ju Chang1; Matteo Seita1; 1Nanyang Technological University
The powder packing density in powder bed additive manufacturing (PBAM) is known to affect the quality and consistency of the consolidated parts. Investigating and monitoring variations in powder packing density in different PBAM processes is challenging owing to the small particle size and the large powder bed area. In this work, we use a “powder bed scanner” to capture high-resolution and large field of view images of the entire powder bed during the powder re-coating process. We quantify powder particle density via numerical analysis of the acquired images and analyze it as a function of powder layer thickness, re-coating speed, re-coater blade type, and powder particle size distribution. Our preliminary results show that the particle density may vary up to 20% from start to end of an individual powder spread. We use this information to study the effect of powder packing density on site-specific mechanical properties in PBAM-produced parts.