Our Sites

What metal additive manufacturing means for the metal fabricator

For those in some niches, the technology will be hard to ignore

This conceptual photo shows the laser metal fusion (LMF) process in action. In an actual process, a layer of powder covers the part through the build cycle. Photo courtesy of TRUMPF Inc.

Metal additive manufacturing (AM) crosses markets. Both the machining world and specialty welding and fabrication arenas have been paying attention. A high-end, previously machined part could be totally redesigned to take advantage of all that AM has to offer, complete with impossible-to-drill internal contours and impossible-to-mill contours. In this sense, metal additive complements machining.

On the other hand, metal additive technology fuses metal particles together; it is essentially welding and, therefore, potentially in the wheelhouse of a high-end specialty fabricator.

If a specialty fabricator—like those serving industries such as aerospace, medical, and defense—is eyeing metal additive technology, it should start with the basics: what these technologies are, what they can and can’t do, and (not least) what’s hype and what’s not.

The Technology of Fused Metal

To that end, The FABRICATOR spoke with Plymouth, Mich.-based Franziska Maschowski, business development manager for TruPrint products, and Tobias Noack, applications engineer, additive manufacturing, at TRUMPF Inc., a company that of course has roots in sheet metal fabrication but also has delved into the additive arena for years.

The company offers two metal additive processes: a directed energy deposition process it calls laser metal deposition, or LMD, and a powder bed fusion process it calls laser metal fusion, or LMF. These and myriad other trademarked processes, most of which fall into the directed energy deposition and powder bed fusion process families, have dominated metal additive manufacturing for years. Although each takes a very different approach, both build up components metal layer by metal layer.

Noack added that the company uses “melt” in LMF and LMD to describe what’s really going on when the laser beam deposits metal powder. The laser technically doesn’t perform a sintering process, which uses a combination of heat and pressure. “This is different. We have heat energy from the laser beam, and we use no pressure at all.”

He added that the act of completely melting metal particles has been a key advancement. Years ago some metal additive technologies would melt just the outside surface of the metal particles. “This resulted in densities of around 90 percent, if you were to measure a part cross section. Today, however, all the powder particles are being melted and transformed into a liquid phase completely and then solidified again. This gives a dense part of 99.9 percent, which makes a major difference.”

Directed Energy Deposition

Conventional directed energy deposition, LMD included, doesn’t involve building parts from scratch. In many respects LMD is an extension of laser cladding. A three-axis laser cladding system builds up cladding layer by layer. Add another two axes and start building up actual shapes, and you have LMD.

“In terms of hardware, for laser cladding you can have a 3-axis system do most of the work, and for LMD you may need a 5-axis system,” Noack said. “But regarding the processing head and the powder hopper, it’s basically the same,” adding that in some cases, laser cladding and LMD can be performed with the same system. “Where it gets more complicated [in LMD] is the process parameters, including heat management, simply because you’re now creating a part rather than going back and forth over a surface.”

Noack described the differentiation by describing a laser depositing metal powder to create a pyramid on top of an existing part. For the first few layers, the area is so large that by the time the processing head makes it back to its origin point to start the next layer, even if the head travels at high speed, the previous metal layer will have cooled.

Laser metal deposition (LMD) builds up a shape on a fixtured component. Photo courtesy of TRUMPF Inc.

“But the closer you get to the tip [of the pyramid], the shorter the weld path will be and the more likely the part will overheat,” Noack said. “So you need to slow the process to ensure you have the same mechanical properties throughout the part. Otherwise, you will perform some heat treatment while building the part and ultimately change material properties in ways that were not intended. And if it’s a certified manufacturing process, you’ve just created some very expensive scrap.”

Tuned correctly, though, directed energy deposition processes offer flexibility in hardware options. The head can be integrated with a 5-axis system or with an articulating robot arm. The work itself can be placed on another automated axis. This in turn allows the laser to build geometries out of fused metal powder that would be impossible to produce with other metal additive processes. For instance, Maschowski described an application in which a workpiece is rotated on its side, after which the laser from above begins depositing a long, vertical column, bead after solidified bead, layer after layer. The laser finishes, and the workpiece is rotated 90 degrees, after which the laser starts building up another element perpendicular to the first. When it’s finished, the two structures together look like an upper-case “L.”

Powder Bed Fusion

Try building this “L” piece with powder bed fusion, including LMF, and you’d run into several insurmountable roadblocks, at least without a few changes. Powder bed fusion is inherently a 2-axis process; the laser head moves in X and Y. A “re-coating” device sweeps a layer of metal powder over the build area; the laser melts the powder that will become the first layer in the part’s geometry, after which a re-coater places another powder layer on the bed. The laser melts the second layer, and the process continues.

At first glance, the process would look like it’s perfectly suited to build any complex shape you could possibly imagine, and in many cases that’s just what it can do. After all, the powder bed supports the workpiece throughout the process.

But as sources described, trying to build that “L” shape alone in the powder to an accurate, consistent dimension would be next to impossible without some changes to the build—and this is thanks to something that those in the welding business are all too familiar with: distortion. As the layers solidify, the part shifts, usually upward, especially so-called “overhangs,” or elements of a structure that protrude outward with nothing underneath. Hence, the part needs support structures to keep it from moving. And at this point, those support structures need to be removed manually.

“No one likes them, but [powder bed fusion] would not work without them,” Noack said. “Actually, I don’t know why they’re called support structures, because they’re not supporting anything. Their main purpose is heat dissipation out of the melt pool and process zone. It helps prevent uncontrolled changes due to the residual stresses that occur during the process.

“People hear ‘support structures’ and they think that the part would drop down into the powder bed without them. But, in fact, the opposite is the case. If you don’t have support structures where needed, the part will rise up out of the powder bed due to shrinkage when the fused particles solidify.”

The small rise not only creates an out-of-tolerance part but also could cause a part to rise up slightly out of the powder bed and crash with the re-coating system. “The re-coater deposits very thin layers, 20 to 60 microns,” Noack said, “so there’s not much space between the surface and re-coater”–hence, again, the need for support structures to keep the part in place.

How extensive do those support structures have to be? “As little as possible but as much as needed,” Maschowski said, adding that software can recommend how much support is needed. As a rule of thumb, if an overhang protrudes at an angle between 45 degrees and 0 degrees (that is, parallel to the bed), you need to support the structure.

Having too little support can scrap an expensive part and crash a build process. Still, those experienced with the process can tweak the support structures’ shapes to achieve that “little as possible but as much as needed” ideal.

Powders

“You need three factors to produce a good part,” Maschowski said. “You need a good machine, you need good process parameters, and you need good powders.”

Modern powders are specifically designed for metal AM. They exhibit properties, including “feasibility” traits (analogous to weldability), that make them perfectly suited for a certain process or additive application. This includes materials with low levels of carbon—very different from, say, tool steel alloys used for decades in the mold and die industry. A material chemistry that’s very machinable may not be feasible for a metal additive process.“Feasibility has really defined the quest for materials for LMF,” Noack said.

Lasers in Additive

Laser beam advancements have pushed metal AM forward, too. “We now have smaller focal diameters that result in better surface roughness and finer details, and you can build parts with higher dimensional accuracy,” Maschowski said.

Laser power in directed energy deposition hinges on the process and build quality requirements. As with laser welding and cladding, LMD’s solid-state lasers go into the multikilowatt range.

Powder bed fusion applications usually use considerably less laser power; 1 kW is typical. The laser beam wavelength of choice for LMF has been in the near-infrared spectrum, with a wavelength in the 1-micron range. For laser cutting, 1 micron has been a suitable beam wavelength for cutting through a variety of reflective materials—but for metal additive processes, not so much.

“Years ago, everybody used near-infrared-wavelength lasers,” Noack said, “knowing it’s not the ideal solution for reflective materials, because most of the energy is being reflected and not absorbed into the material.”

Recently, however, shorter wavelengths are aiding the process for these challenging materials. For instance, some additive systems now use green lasers, offering wavelengths between 300 and 400 nanometers. “You can broaden your material choice in terms of reflective materials, like copper alloys or precious metal like gold, silver, or platinum alloys,” Noack said.

What If?

These wavelengths can help melt material more effectively and completely. But to melt metal faster, shouldn’t there be another way? In the immortal words of Tim Allen, couldn’t you just give it more power?

“A common misconception is that you can just speed up the whole process by using more powerful lasers,” Noack said. “This is not true. There is a physical limitation to the whole process. You cannot take a 10-kW laser and increase the scan speed and build rate just because you have more power.”

The power question arises when people ask about how a process could be scaled up. Related to this is another question: Could a build bed be made bigger? Typical bed sizes are a few feet wide, not large enough to build large parts.

“But you have to be careful when talking about 3-D printing large metal parts,” Noack said. “When the part size grows, eventually there’s a tipping point, and eventually it would be much cheaper to redesign the part so it could be produced with conventional processes.”

Yes, the economics may not make sense for building a massive metal part—considering the costs of the powder and the very lengthy build cycle such a part would require—but what about a large bed with numerous small parts? This, Maschowski said, probably wouldn’t be a good approach either. Using a single laser, printing a large bed of tiny parts still would take a long time. A multilaser system is plausible, but again, instead of powering one giant system, why not run many smaller, right-sized machines that can produce small parts on demand?

The cost of the powder also needs to enter the equation. A large bed means a large amount of powder. And if a part is being produced with a certified manufacturing process, there may be limits as to how that powder can be reused.

Regardless, the reason companies turn to metal AM isn’t that they need to produce many parts quickly. After all, legacy manufacturing processes already excel at this.

Companies usually turn to metal AM for the design freedoms it allows, as well as the value a new design or build method brings to the entire manufacturing process chain. The process could eliminate numerous parts in an assembly, greatly simplifying manufacturing or perhaps shortening a product’s time to market. Or it could provide a huge customer benefit. GE’s 3-D-printed fuel nozzle, which saves customers billions in fuel costs over a life of an engine, did just this.

Metal AM isn’t a new technology; directed energy deposition and powder bed fusion have been around for years. But it is a technology on the cusp of future growth. Machine tool vendors are ramping up their offerings, and investors are paying attention. The what-if questions abound, perhaps because of where the market is right now. It’s a maturing market that has been around for decades, but it’s certainly not standing still.

About the Author
The Fabricator

Tim Heston

Senior Editor

2135 Point Blvd

Elgin, IL 60123

815-381-1314

Tim Heston, The Fabricator's senior editor, has covered the metal fabrication industry since 1998, starting his career at the American Welding Society's Welding Journal. Since then he has covered the full range of metal fabrication processes, from stamping, bending, and cutting to grinding and polishing. He joined The Fabricator's staff in October 2007.