5 Key Advances for Metal Additive Manufacturing

How metal 3D printing is leaping the “valley of death” between research and production.

Metal 3D printing has already seen a number of advancements in 2016. From the nanoscale to a $23.98M manufacturing facility, metal additive manufacturing (AM) is getting better all the time.

The last five years have seen explosive growth in the industry. As indicated by this graph from a Wohlers Associates report, sales of metal 3D printers increased by 75 percent in 2013 alone:

Graph showing metal 3d printer sales (Y-axis) over year (X-axis). (Image courtesy of Wohlers Associates.)

Graph showing metal 3d printer sales (Y-axis) over year (X-axis). (Image courtesy of Wohlers Associates.)

More recently, a UK metal powders company received a £20m investment to commercialize its powders for aerospace, automotive and medical industries. 

Metal 3D printing is here to stay and according to Jack Beuth, director of the NextManufacturing Center at Carnegie Mellon, this is only the beginning.

“I’ve never seen any research area that’s been like this,” said Beuth. “Additive manufacturing has been around for 20 years, but it wasn’t clear that you could automate the fabrication of 3D parts until about four years ago. Between that time and now, we’ve gone from a few aerospace companies experimenting to making real parts with metal AM that are getting FAA-certified and going into engines.”

What will the next five years look like?

Beuth recently predicted five key advances in metal AM to expect over the next five years.

 

1. Process Design

Users will be able to design the AM process concurrently with designing the geometry of a part.

“Current direct metal AM processes operate with a very narrow range of process variables, like beam power and travel speed,” said Beuth. “There’s also a very narrow range of powders.”

According to Beuth, this is because the techniques for metal additive manufacturing were developed experimentally. Once a working combination of beam powers and travel speeds was found, there was little interest in finding out what other combinations might work.

In addition to the powder, there are five major process variables in metal AM:

  • Beam power
  • Beam travel speed
  • Layer thickness/Material feed rate
  • Local geometry
  • Part temperature at fusing point

“In the future, you’ll have a wide variety of powders at your disposal and you’ll understand how to adjust the process to make those powders work,” Beuth predicted. “You’ll also have a wide range of process variables that you can manipulate so you can control your outcomes.”

 

2. Monitoring and Control

Users will be able to monitor and control the additive manufacturing process.

“In current processes, the most they have is a thermocouple or two which monitors the overall temperature in the build chamber. Some processes don’t even use that information in a control loop,” said Beuth.

Metal part made on a self-monitoring EOS 3D printer. (Image courtesy of Tom Lansford.)

Metal part made on a self-monitoring EOS 3D printer. (Image courtesy of Tom Lansford.)

This isn’t just a matter of improving dimensional and temperature sensors but adding new ones as well. For example, the laser powder bed machine that Beuth and his colleagues use incorporates an optical system which captures images of the powder bed before and after the powder is fused.

“It’s important to improve monitoring over the next five years. If you section a part and find an area where there’s a lack of fusion, you should be able to go back through the machine’s files and see why that happened,” said Beuth. “The next step would be feedback control; changing the process and fixing things as you go.”

 

3. Manipulating Microstructure

Users will be able to vary the material microstructures and properties of a part in different locations by manipulating process variables as the part is being built.

One of the main aims of the NextManufacturing Center is to demonstrate the problems with the current feed-forward approach to control in metal AM. Beuth and his colleagues believe that engineers should be able to manipulate variables in the 3D printing process to obtain specific variations in a part’s microstructure.

Metal 3D printing at the nanoscale. (Image courtesy of Alain Reiser.)

Metal 3D printing at the nanoscale. (Image courtesy of Alain Reiser/ETH Zurich.)

“We’ve shown how to do this with the ARCAN process, which is an electron beam powder bed process, for titanium alloy TI-64; we’re able to manipulate at least part of the microstructure for that system,” said Beuth.

“No one else is able to do that right now except us, but it’s coming,” Beuth continued. “In five years, it will be commonplace for some microstructural features to be controlled by choosing process variables in advance.”

 

4. More Powders

Users will be able to use a wide variety of metal powders.

Direct metal laser sintering. (Image courtesy of Tom Lansford.)

Direct metal laser sintering. (Image courtesy of Tom Lansford.)

In the next five years, that primarily means different diameters of existing metal powders. However, looking farther ahead to the five- to ten-year range, Beuth also expects to see developments in new alloys that exploit the high cooling rates of metal AM processes.

“For certain materials, if you know what the conditions are then you can use those alloys to get better material properties,” said Beuth. Until now, there haven’t been processes with such high cooling rates. My colleagues in the mechanical engineering and materials science departments are interested in exploring new alloys, but it’s still in the early stages.”

 

5. Manipulating Porosity

Users will be able to eliminate or design for internal porosity, which will have a significant effect on fatigue resistance and build rate.

Currently, some level of porosity in metal 3D printing is unavoidable, at least using the printer’s default settings. If your part is subjected to cyclic fatigue loading, that porosity can become a starter crack, which makes it a pressing problem for metal AM.

Porosity comes from three major sources:

  • Porosity of metal powders themselves
  • Lack of fusion (not all the powder is fused)
  • Keyholing (metal is locally vaporized and surrounded by molten metal, which cools and creates a pore)

“The first kind can be eliminated by working with the powder companies,” said Beuth. “The other two can be dealt with by changing the additive process. We currently know how to do that pretty well and no one else does, but in five years it will be fairly common.”

In addition to eliminating porosity, Beuth also believes that design engineers will soon be able to manipulate it.

“In the traditional way of thinking about parts, the first goal is to have a uniform microstructure and uniform properties,” said Beuth. “That makes it easier to design and understand.”

“But if you can make it so that the material has smaller grains, then you can make the part higher strength where you have stress concentrations,” Beuth continued. “It’s analogous to case-hardening in steels. You now have designers who aren’t just designing a part, but also the process to make the part faster or give it a nicer surface finish, or even designing the microstructure on a point by point basis.”

 

Metal 3D Printing and the Valley of Death

The role of additive manufacturing in the future of industry is still a controversial subject, but metal 3D printing may have already hit a critical turning point.

The

The “valley of death” between research and commercialization.

“I still get some people who are skeptical about the impact it will have in industry,” said Beuth. “They call it the Valley of Death: making the leap from research to commercial applications. That might have been true three or four years ago, but we’re well past that now.”

For more information, visit the NextManufacturing Center website.

Written by

Ian Wright

Ian is a senior editor at engineering.com, covering additive manufacturing and 3D printing, artificial intelligence, and advanced manufacturing. Ian holds bachelors and masters degrees in philosophy from McMaster University and spent six years pursuing a doctoral degree at York University before withdrawing in good standing.