3D-Printed Ceramics Can Increase Gas Turbine Efficiency

Ed Brown

Tech Briefs: How did you get started on this project?

Professor Stephen Lynch: A colleague of mine in material science, Professor Mike Hickner, who has since moved to Michigan State University but at the time was at Penn State in the Material Science Department, was talking with me about his work in 3D-printed ceramic materials. He was developing resins for that kind of process when a proposal call came out from the Department of Energy looking at advanced manufacturing topics in ceramic materials. So, we put together an idea along with partners at the University of Wyoming (Professor Carl Frick, now at Colorado School of Mines, and Professor Ray Fertig) who could do the printing part. Mike is an expert in resin development, my expertise is in the turbine-cooling part and the design of the pieces, and the University of Wyoming's expertise was in the actual manufacturing process to make the parts. So, we wrote a proposal. Looking back, it was a fortunate confluence of things that came together at the right time — my discussions with Mike, and the proposal call, and the fact that Carl and Ray had just published a paper on 3D-printed ceramics.

Tech Briefs: Has this combination of resin and ceramics been done before?

Lynch: In some sense, yes. The idea for using a polymer as a precursor to a ceramic material is not new, it’s actually been done for a long time. The uniqueness of this particular project was twofold. One is that we integrated the resin into a standard 3D-printing process, using a laser to build it rather than a form or a mold for the resin. And the second is to apply it to an application in high-temperature materials, in this case the turbine environment. So, coupling the capabilities of the manufacturing process with the needs of the application.

Tech Briefs: I understand that the ceramic can take really high temperatures but how about the polymer?

Lynch: The polymer cannot — that's the interesting part. The polymer is what is known as a ceramic precursor polymer. It has elements of a polymer, and in the form we start with, it's a liquid, with a viscosity similar to honey, that gets mixed with the elements required to solidify it in the 3D-printing process.

Once it's solidified you have what's called a green body, which is the part that is in the shape you want but is not purely ceramic yet. Then the final step of the process, which is probably one of the more important ones, is you have to pyrolyze that part. You put it in an inert environment at high temperature, essentially burn out everything that is not a ceramic material and you're left with the ceramic.

Tech Briefs: How do you feed the ceramic into the printer?

Lynch: The ceramic starts out as a liquid that has a silicon carbide backbone but is in liquid form. There is a liquid bath, which is the resin, and an ultraviolet laser that shoots into the resin bath. Wherever the laser shines, it solidifies the part. There's a reaction within the polymer liquid due to special additives, which solidifies the resin into a solid material. So, you trace out the shape you want, move the resin vat down (or move the part up out of the resin), trace out the next shape you want, and repeat this process over and over until you generate a three-dimensional part.

Tech Briefs: So, you’re making a part that goes inside a gas turbine?

Lynch: Yes, sort of. Right now, our technology is still at what we call a low-technology-readiness level (TRL), it's not yet ready to plug directly into a gas turbine. What we ended up creating from the results of this research is a turbine vane part, which is the component immediately downstream of the combustor in a gas turbine. It experiences the highest temperatures in the engine and therefore ceramics are a great option if they can work. They will be an excellent substitute material for the nickel superalloys that are currently used in that part of the engine. We want extreme temperatures in a gas turbine because it is directly coupled to fuel efficiency — the higher the temperature, the more fuel efficient the turbine is.

Tech Briefs: I read that you modified the shape — you changed it from the traditional shape.

Lynch: Yes, although there were some constraints. We wanted to test a realistically shaped part, so we kept the outer shape the same as it would be if it were a metal component — essentially the aerodynamic shape that defines the flow that goes around it. Keeping that the same, was a big constraint, an interesting one, which we didn't want to violate. However, we were able to do whatever we wanted to the interior of the part.

In modern gas turbine engines made from metal components, the interior is hollow so that you can pass cooling air through it to keep it cold. This is really difficult to do with conventional ceramic materials, because a hollow ceramic is challenging to manufacture. But for a 3D-printed part, you can create, within reason, any shape you want. So, we were able to create a hollow ceramic material and cool it in the same way that we do for nickel superalloy materials. It was kind of a merger between the best things we know about metal cooling in gas turbines, with a new high-temperature material that maybe doesn't even need that much cooling but now has the capability of doing it.

Tech Briefs: Did you do mathematical modeling before you built it?

Lynch: Yes. Ray Fertig, one of the principal investigators on the project at the University of Wyoming, was responsible for that. His group did finite element analysis of the turbine blade parts to figure out what the stresses would be in the baseline component we manufactured, as well as in what direction to drive the manufacturing to get better performance. In our case, better performance was defined as being able to put it in my test facility at Penn State, measure its surface temperature with an infrared camera, and determine that for a given amount of cooling flow, the temperature of the ceramic was colder. That would mean that it's being cooled inside more effectively. He did modeling of that process, and we compared the model results to the experiment.

Tech Briefs: What sort of testing did you do? Was it just temperature?

Lynch: Yes, it was primarily temperature testing. What we do in my facility here at Penn State is we test turbine blade parts at realistic scale. So, a lot of the parts we test are what you would find in an engine. However, we don't run them up to super-high temperatures yet. My facility is at a little bit lower technology level, we don't get to thousands of degrees, but we do get to a few hundred degrees. What we can do is supply the cooling air that you would see in a real turbine part and look at the difference between the cooling temperature and the hot gas temperature and how it affects the temperature of the solid part. Essentially, we're trying to keep the temperature of the part within a certain range with the amount of cooling flow we provide. In our testing, we were looking at different cooling flow rates to see how cold we could get the ceramic material.

Tech Briefs: What are the next steps? How long do you think it will be before you actually try this out, at least in a prototype engine?

Lynch: That's a good question. There are some interesting mechanical challenges to solve. A gas turbine engine operates at really high temperatures, and we haven't tested it at those temperatures yet, so we need to do that.

The other challenge is that some of these components, in particular the component that we tested, has a lot of stress in the engine. Besides thermal stress, it has mechanical stresses that happen when the interfaces to other components meet up with each other and grow thermally. So, there are some mechanical stresses that we would need to address in a full-scale part. I would say we next need to focus on increasing the mechanical properties of the material.

Right now, it has relatively low fiber-loading content. It's more like a pure ceramic rather than a ceramic matrix composite material. Because of that, its fracture toughness is relatively low. So, if you hit it with, say, a pebble, it would shatter like a piece of glass. That's not desirable in a gas turbine. So, we have to work on increasing the fracture toughness of the material. There was some work toward that at the end of the project, where we were trying to introduce fibers within the printing process.

Our next big challenge is to increase the fracture toughness of the material, increase its mechanical strength so that it can handle real gas turbine temperatures, and then also see how it performs at those high temperatures. It will probably be a few years before it can become a part in an engine. But certainly, we've demonstrated that it is possible, and we can create novel features.

Tech Briefs: Have you thought about whether this process could have other applications?

Lynch: Yes, absolutely, there's a lot of interest in hypersonics now, where there are very high heat-flux requirements because of the high speed of the air for hypersonic vehicles. There are usually ceramic materials at the leading edges of the aircraft because they get so hot. We need something high-temperature-tolerant and if we can embed cooling into the ceramic leading edges of hypersonic vehicles, that would be game changing, it would enable them to last longer and be more durable.

Tech Briefs: Are you referring to hypersonic airplanes?

Lynch: Yes, or maybe missiles or interceptors, or any sort of any application where you're trying to fly above Mach 5. The air temperature gets very hot when you go that fast.

Also, gas turbines not only provide propulsion for aircraft engines, but are also used in electric-power generation in many of our power plants. We certainly want power generation to be fuel efficient, which can mean running hot or with less parasitic cooling air. One other application would be in nuclear reactors. There are many locations where you would want ceramic materials to be high-temperature-tolerant, but also have some cooling capability.

It could be used any place where you have high temperatures or high heat fluxes on a surface that needs to be cooled and where temperatures are so high that you need to cool the ceramic material. Ceramics are difficult to cool because you can't just drill cooling holes into them, you have to sort of manufacture them after the fact or try to cast them in.

Tech Briefs: What excites you about this project?

Lynch: I'd say, of the research we've done so far, the capability of the 3D-printing process to produce novel designs and the resolution of the process to produce high-quality parts, is exciting. I think that a neat aspect of this project was that we were able to create quite small parts that have very fine, intricate features that were only possible because we 3D printed them. When you 3D print these parts, you have to generate a green body that’s about 30 percent oversized relative to the final dimensions. When you bake it in an inert environment, it shrinks down so that you get the desired feature resolution. This means we can get very good resolution during printing, since it’s larger than the final part will be.

Also, this uses a standard digital light processing (DLP)-style printer. Our collaborators at the University of Wyoming adapted a standard stereolithography (SLA) printer to use this novel resin. So, it wasn't a very expensive process because they were using parts that already exist. All we had to do was develop and refine resins and tailor them to the manufacturing process.

Another interesting aspect of this is that creating these 3D-printed ceramic geometries was more cost effective than the conventional manufacturing for ceramic vane parts because it's essentially just printing a plastic.

An exciting part for me is that my background is in turbine cooling technologies — how do we make these parts survive extreme temperatures in a gas turbine. I didn't know much about the material science of ceramics, so this was a really great learning experience for me.

Ultimately, our hope is that because this technology already exists for standard resins, we just have to change a few things about the printing process to make it work for these new resins, and ultimately that could be scaled pretty effectively.

---