LLNL Develops Semi-Liquid Metal 3D Printing Method: Direct Metal Writing

Metal additive manufacturing is one of the oldest forms of 3D printing, but the technology is still constantly being developed. The most common forms of metal 3D printing still have flaws and limitations; for example, most metal powder-based forms of additive manufacturing are prone to gaps and defects. The issue of porosity in metal 3D prints is something that Lawrence Livermore National Laboratory has been working on for a long time, and now the researchers at the California laboratory have developed a new printing method that they say eliminates the problems of powder altogether – by eliminating powder altogether.

The technology, which they call direct metal writing, involves heating a metal ingot until it reaches a semisolid state, something like a metal paste, composed of a core of solid metal particles surrounded by liquid. The material is what is called a shear thinning material, meaning that it behaves like a solid when still, but like a liquid when force is applied – force such as, for example, being pushed through an extruder. It then solidifies again once it’s been extruded, and hardens as it cools, so that there’s less incorporated oxide and thus less residual stress.

“We’re in new territory,” said materials scientist Wen Chen. “We’ve advanced a new metal additive manufacturing technique that people aren’t aware of yet. I think a lot of people will be interested in continuing this work and expanding it into other alloys.”

Chen is the lead author of a study entitled “Direct metal writing: Controlling the rheology through microstructure,” which you can access here. Additional authors include Luke Thornley, Hannah G. Coe, Samuel S. Tonneslan, John J. Vericella, Cheng Zhu, Eric B. Duoss, Ryan M. Hunt, Michael J. Wight, Diran Apelian, Andrew J. Pascall, Joshua D. Kuntz, and Christopher M. Spadaccini.

The technology hasn’t been perfected yet; according to the researchers, a lot of work will still need to be done before they can create higher-resolution parts with more commonly used metals such as aluminum and titanium. For the study, they printed parts with a bismuth-tin mixture, which has a low melting point of below 300ºC. The process took several attempts, as bits of solid metal called dendrites would get stuck in the nozzle.

“The main issue was getting very tight control over the flow,” said Pascall. “You need precise control of the temperature. How you stir it, how fast you stir it, all makes a difference. If you can get the flow properties right, then you really have something. What we’ve done is really understand the way the material is flowing through the nozzle. Now we’ve gotten such good control that we can print self-supporting structures. That’s never been done before.”

Interestingly, Adrian Bowyer, the creator of RepRap, begs to differ. Today on Twitter, Bowyer pointed to a 2009 blog post that details a student’s successful efforts to 3D print circuits by directly extruding melted metal. The blog post doesn’t go into much detail, so it’s difficult to fully assess any differences there may be in the two processes (though the RepRap post doesn’t mention self-supporting structures), but it’s another example of how, in the 3D printing world, it’s almost impossible to say with confidence that a new process or machine or material is brand new – there’s always a chance that someone, somewhere, may have done the same thing, or something very similar, already.

The LLNL researchers are now adapting the technology to work with aluminum alloys, which are much more commonly used in industries such as aerospace and transportation. Aluminum is much more of a challenge, however, because of its higher melting point.

“Being able to print parts out of metal in this way is potentially important,” said Thornley, who helped engineer the bismuth-tin mixture. “So much of the work that goes into validation and analyzing for defects would be eliminated. We can use less material to make parts, meaning lighter parts, which would be big for aerospace.”

Luke Thornley

The study was funded by the Laboratory Directed Research and Development Program. Discuss in the LLNL forum at 3DPB.com.