Yes, it could happen thanks to research into the additive manufacturing of electrical steel, copper conductors, and motor dielectrics.
Leland Teschler, Executive Editor, Design World
Examine the inside of an industrial electric motor and you’ll see components that all operate in environments characterized by high temperatures and significant levels of electrical energy. These sorts of surroundings would seem to be incompatible with the parts coming out of 3D printers today.
But don’t tell that to researchers at United Technologies Corp.’s Research Center. Under a contract from the Dept. of Energy’s Arpa-E (Advanced Research Project Agency-Energy) program, a UTRC team is working toward producing a 30 kW induction motor using just additive manufacturing methods. Specifically, the team is trying to define an additively manufactured induction motor capable of delivering 50 kW peak, 30 kW continuously over a speed range of zero to 12k rpm, using motor technology that does not involve rare-earth magnets. The target application: powering electric vehicles.
“Our original proposal was to devise a single machine that made copper conductors, dielectric components, and steel laminations that were all co-located and adjacent to each other, with all having state-of-the-art properties,” said UTRC project leader Wayde Schmidt. “But there is a large technical gap to be bridged before we can fabricate parts of all three materials simultaneously on a single machine. So we are dividing the problem into smaller chunks that can each be addressed with a focused effort.”
UTRC has been working on additively manufactured motors for about a year and half with partners that include the Connecticut Center for Advanced Manufacturing, Penn State University, and the engineering consulting firm Ricardo PLC. As an interim step, the researchers aim to make parts from each of the three materials using fabrication methods that seem to best suit the task at hand.
“It is a challenge to fabricate multiple materials where each requires different processing steps that may be incompatible with that of the other two,” said Schmidt. “It may ultimately make more sense to have three separate machines.”
Schmidt said the group is exploring laser additive methods such as direct metal laser sintering or powder-bed fusion to create copper conductors. But fabricating 3D-printed copper conductors that work as well as ordinary copper wires is turning out to be extremely tough.
“The state-of-the-art for additive copper isn’t far enough along to give us what we want. So we are looking at alternative methods while we push the state-of-the-art with our laser technology,” said Schmidt. “We are examining simpler methods such as infiltrating molds for making finely featured copper castings that have the conductivity we need.”
Similarly, the group is studying surrogate dielectric materials such as alumina (aluminum oxide) that might do double duty, functioning as a dielectric within the motor itself and as a mold into which copper conductors can be cast.
“For example, we can take off-the-shelf alumina honeycomb structures used for catalytic combustors and turn them into molds into which we can cast copper. This lets us get good copper properties immediately adjacent to an insulating structure, giving us a testable prototype,” explained Schmidt. “We aren’t necessarily using additive methods for the ceramic, but we are looking at powder-bed methods and additively making castings that create the dielectric structures.”
In the same vein, the researchers admit that the most economical way of getting sheet-metal stampings for motor components is from sheet stock. But longer term, they are looking at laser deposition methods for creating electrical steel with the right grain structures. In particular, they think laser-engineered net shaping has promise for electrical steel. The technique is characterized by the use of a metal powder injected into a molten pool created by a focused laser beam and can create objects that have dimensions suitable for the industrial motor application.
There is more to the UTRC work than just 3D printing motor parts. It’s likely that conventional motor architectures won’t translate well into versions that are 3D printed. So researchers are also trying to figure out what the design of a 3D-printed induction motor should really look like.
“A major portion of the project is devoted to finding a motor design that maximizes power efficiency under the assumption that the structures can be made with additive manufacturing,” said motor design task leader and UTRC staff scientist Jagadeesh Tangudu. “We haven’t yet constrained the motor design. It could end up being quite non-traditional. But there are a lot of challenges. Traction applications, for example, need constant power over a wide range of speed. The usual approach is to use flux weakening as a way of hitting high-speed operation, but induction motor characteristics don’t lend themselves to this sort of field weakening. There are some opportunities to innovate here, but that’s all I can say at this point.”
Though researchers won’t say much about the motor design, they admit that the first prototype they hope to deliver at the end of 2015 will only have additively manufactured stator components. The rotor will be made with conventional methods.
Additionally, the team has yet to involve any 3D printer makers in the project. “We haven’t pursued any discussions with machine makers yet,” said Schmidt. “We are still focused on trying to demonstrate the structures we need and on unraveling technical challenges. As we get closer to the end of the program in 2015, we might identify the right technology and the machine that goes with it. Collaborating with a machine maker to develop something that incorporates what we learn could be a follow-on program opportunity.”
Magnets for motors
Of course, the induction motor design UTRC is working on doesn’t use magnets. Does that mean other motor technologies such as permanent-magnet motors can’t be made additively?
Not necessarily. Another research group that is part of the DoE”s Critical Materials Institute is trying to develop a super-strong magnet that sharply reduces use of rare-earth materials. Besides cutting rare-earth content by perhaps 75% or more, the effort looks as though it could result in a magnet that lends itself to construction via 3D printing.
The effort is led by a team at Lawrence Livermore National Laboratory (LLNL) that is working closely with a group at Brown University. The aim is to develop what’s called an exchange-spring magnet. An exchange spring magnet basically combines soft magnetic material that has a high magnetic remanence (the amount of magnetism remaining after the magnetizing field has been removed) with hard magnetic material that has a high coercivity (meaning the magnet cannot be easily demagnetized once initially magnetized). When the hard and soft magnetic materials are combined in a special way, the resulting magnets have both a high remanence and a high coercivity.
“The trick is that you need to have the hard material spaced between soft materials in such a way that the separation distance between them is only on the order of a magnetic domain wall. In many materials that turns out to be a handful of nanometers,” explained CMI’s Deputy Magnet Thrust Leader Dr. Scott K. McCall.
The making of an exchange-spring magnet requires layout a checkerboard pattern of hard and soft magnetic material on a spacing of 5 to 10 nm. “It is a simple idea but trying to control materials at that level is something people have been working on for 20 years,” said McCall. “We think with some of our advanced manufacturing methods we will be able to build up exchange-spring magnets brick-by-brick.”
One of the ideas CMI has for devising the hard-soft magnetic material is to start with a hard magnet and coat it somehow with a shell of a soft one in a manner analogous to that of M&M candy. For this method to be successful, the core materials must be smoothed – nanoparticles tend to have jagged edges when prepared by grinding bulk materials into ultrafine powders. Researchers at Brown University came up with a way to make hard magnets using chemistry that have smooth surfaces. The CMI team at LLNL is developing a method of coating these hard magnet cores that can precisely control the thickness of the soft layer and thus the size of the particles involved.
McCall thinks one intriguing thing about creating magnetic material with an M&M-like structure is that its deposition could be handled by 3D printing having enough control of material thickness. He also thinks the sintering process normally used to 3D-print metals could be managed in a way that wouldn’t destroy the grain structure of the magnet. “If you have a large surface area then the temperature you need to sinter the material is somewhat lower,” he said. “The advanced manufacturing methods we have in mind will use a process of pressure and heat to sinter the overall end product and will likely apply a magnetic field at the same time.”
But it will be awhile before CMI’s exchange-spring magnets are ready for prime time. McCall thinks the effort might approach industrial partners in perhaps three years. He says CMI hasn’t yet talked with 3D printer makers about the possibility of printing magnets. But he says the industrial magnet manufacturers who are aware of the idea find it “intriguing.”