Publications

Results 26–50 of 59

Search results

Jump to search filters

Deep Convolutional Neural Networks as a Rapid Screening Tool for Complex Additively Manufactured Structures

Additive Manufacturing

Garland, Anthony; White, Benjamin C.; Jared, Bradley H.; Heiden, Michael J.; Donahue, Emily; Boyce, Brad L.

Additively manufactured metamaterials such as lattices offer unique physical properties such as high specific strengths and stiffnesses. However, additively manufactured parts, including lattices, exhibit a higher variability in their mechanical properties than wrought materials, placing more stringent demands on inspection, part quality verification, and product qualification. Previous research on anomaly detection has primarily focused on using in-situ monitoring of the additive manufacturing process or post-process (ex-situ) x-ray computed tomography. In this work, we show that convolutional neural networks (CNN), a machine learning algorithm, can directly predict the energy required to compressively deform gyroid and octet truss metamaterials using only optical images. Using the tiled nature of engineered lattices, the relatively small data set (43 to 48 lattices) can be augmented by systematically subdividing the original image into many smaller sub-images. During testing of the CNN, the prediction from these sub-images can be combined using an ensemble-like technique to predict the deformation work of the entire lattice. This approach provides a fast and inexpensive screening tool for predicting properties of 3D printed lattices. Importantly, this artificial intelligence strategy goes beyond ‘inspection’, since it accurately estimates product performance metrics, not just the existence of defects.

More Details
TYPE Journal Article YEAR 2020
DOIOSTIScopus

High-throughput additive manufacturing and characterization of refractory high entropy alloys

Applied Materials Today

Melia, Michael A.; Whetten, Shaun R.; Puckett, Raymond V.; Jones, Morgan; Heiden, Michael J.; Argibay, Nicolas; Kustas, Andrew B.

Refractory High Entropy Alloys (RHEAs) and Refractory Complex Concentrated Alloys (RCCAs) are high-temperature structural alloys ideally suited for use in harsh environments. While these alloys have shown promising structural properties at high temperatures that exceed the practical limits of conventional alloys, such as Ni-based superalloys, exploration of the complex phase-space of these materials remains a significant challenge. We report on a high-throughput alloy processing and characterization methodology, leveraging laser-based metal additive manufacturing (AM) and mechanical testing techniques, to enable rapid exploration of RHEAs/RCCAs. We utilized in situ alloying and compositional grading, unique to AM processing, to rapidly-produce RHEAs/RCCAs using readily available and inexpensive commercial elemental powders. We demonstrate this approach with the MoNbTaW alloy system, as a model material known for having exceptionally high strength at elevated temperature when processed using conventional methods (e.g., casting). Microstructure analysis, chemical composition, and strain rate dependent hardness of AM-processed material are presented and discussed in the context of understanding the structure-properties relationships of RHEAs/RCCAs.

More Details
TYPE Journal Article YEAR 2020
DOIOSTIScopus

Evolution of 316L stainless steel feedstock due to laser powder bed fusion process

Additive Manufacturing

Heiden, Michael J.; Deibler, Lisa A.; Rodelas, Jeffrey; Koepke, Joshua R.; Tung, Daniel J.; Saiz, David J.; Jared, Bradley H.

Some of the primary barriers to widespread adoption of metal additive manufacturing (AM) are persistent defect formation in built components, high material costs, and lack of consistency in powder feedstock. To generate more reliable, complex-shaped metal parts, it is crucial to understand how feedstock properties change with reuse and how that affects build mechanical performance. Powder particles interacting with the energy source, yet not consolidated into an AM part can undergo a range of dynamic thermal interactions, resulting in variable particle behavior if reused. In this work, we present a systematic study of 316L powder properties from the virgin state through thirty powder reuses in the laser powder bed fusion process. Thirteen powder characteristics and the resulting AM build mechanical properties were investigated for both powder states. Results show greater variability in part ductility for the virgin state. The feedstock exhibited minor changes to size distribution, bulk composition, and hardness with reuse, but significant changes to particle morphology, microstructure, magnetic properties, surface composition, and oxide thickness. Additionally, sieved powder, along with resulting fume/condensate and recoil ejecta (spatter) properties were characterized. Formation mechanisms are proposed. It was discovered that spatter leads to formation of single crystal ferrite through large degrees of supercooling and massive solidification. Ferrite content and consequently magnetic susceptibility of the powder also increases with reuse, suggesting potential for magnetic separation as a refining technique for altered feedstock.

More Details
TYPE Journal Article YEAR 2019
DOIOSTIScopus
Results 26–50 of 59
Results 26–50 of 59