Caught in the act

In a form of additive manufacturing, complex metal parts are built one ultra-thin layer at a time — similar to frosting a cake, but with far greater precision and intricacy. This technique allows for the printing of 3D parts that are difficult or impossible to make using traditional methods. It also offers a path forward to ease supply chain disruptions and modernize domestic manufacturing.

Additive manufacturing is already used to produce metal components for critical sectors like aerospace, healthcare and defense. But a major challenge remains: achieving consistent quality and repeatability from part to part.

Now, in a major scientific advance, researchers at the U.S. Department of Energy’s (DOE) Argonne National Laboratory, DOE’s Oak Ridge National Laboratory and universities have observed how the microstructure of metals changes in real-time during 3D printing. This breakthrough was made possible by Argonne’s Advanced Photon Source (APS), a DOE Office of Science user facility.

“Our analysis shows how powerful the APS is for studying atomic-scale defects that were previously only seen through after-the-fact analysis.” — Andrew Chuang, Argonne physicist

The team investigated an additive manufacturing technique in which a laser rapidly melts a thin metal wire. As the metal melts, it is deposited on the previous layer after the first, followed almost instantly by cooling and solidification. This process is repeated layer by layer to create intricate components. Previously, scientists could only analyze the microstructures of these components after the printing process was finished.

“Metals are made of atoms arranged in ordered crystal structures,” said Tao Sun, the project’s lead investigator and a professor at Northwestern University who also holds a joint appointment at Argonne. ​“But under rapid heating and cooling, some atoms fall out of alignment. These defects — called dislocations — can strengthen or weaken the final part.”

Using beamline 1-ID-E at the APS, the team conducted 3D printing of 316L stainless steel, a commonly used structural alloy. They tracked the printing process with real-time X-ray diffraction, directly measuring how and when dislocations form and spread.

“Our analysis shows how powerful the APS is for studying defects that were previously only seen through after-the-fact analysis,” said Andrew Chuang, a physicist at APS. ​“This is the first time this real-time technique has been applied to this laser-based method to study the dislocation evolution in a metal wire.”

The data revealed that dislocations form early, just as the metal changes from liquid to solid. It was previously thought that they form later as stresses build up during cooling and solidification. A key factor was a specific reaction in which two solid phases form at the same time from the liquid, creating a high density of dislocations.

This deeper understanding could help engineers improve the strength and reliability of 3D-printed parts. By adjusting printing variables, developers would be able to precisely control the formation of dislocations at the microscopic level. By this means, they could take full advantage of the dislocations’ beneficial attributes while minimizing the detrimental ones.

The insights gained could also spur the development of new alloys. Adjusting the chemical makeup of stainless steels — for example, by tweaking the ratios of chromium or nickel, or by adding elements like aluminum — can influence how dislocations form and how stress is distributed.

“This type of 3D printing could create customized metal parts that are reliable and extra-strong and would survive extreme conditions,” said Lin Gao, a postdoctoral researcher in in the Nuclear Science and Engineering division at Argonne. ​“It may be key to building advanced metal components for next-generation nuclear reactors now being designed at Argonne and other labs.”

The National Science Foundation provided partial funding for this project. The findings were first published in Nature Communications.

The research was carried out by scientists from Northwestern University, Argonne, Oak Ridge and the University of Virginia. In addition to Sun, Gao and Chuang, authors include Yan Chen, Xuan Zhang and Sean Agnew.