GA-EMS industrializing SiC/SiC and other CMC via MAITrX facility

In February 2026, General Atomics Electromagnetic Systems (GA-EMS, San Diego, Calif., U.S.) announced an MOU with Oak Ridge National Laboratory (ORNL, Oak Ridge, Tenn., U.S.) to advance industrial manufacturing of advanced ceramic matrix composites (CMC). Under the agreement, GA-EMS will examine its advanced manufacturing processes for ceramic precursors, fibers and composites using resources from the U.S. Dept. of Energy’s (DOE) Manufacturing Demonstration Facility (MDF) at ORNL.

GA-EMS president, Scott Forney, explained the goal is to accelerate innovation, strengthen critical U.S. supply chains and deliver advanced materials essential for national security and energy security. “The initiative complements our growing advanced materials and technology capabilities and launch of our Materials Acceleration, Innovation and Transition Exchange [MAITrX] lab.” Featuring capabilities built on more than 70 years of nuclear expertise, MAITrX was established to drive commercial implementation of customized advanced materials, with CMC front and center.

Advanced materials development hub for General Atomics

General Atomics comprises four main business segments and several affiliate companies. The most well-known affiliate is perhaps General Atomics Aeronautical Systems Inc. (GA-ASI), which produces unmanned aerial systems (UAS) including the MQ-9A Reaper, MQ-9B SkyGuardian/SeaGuardian, MQ-1C Gray Eagle and the MQ-20 Avenger and Gambit series.

GA-EMS develops and manufactures advanced electromagnetic systems, including the Electromagnetic Aircraft Launch System (EMALS) for the U.S. Navy, rail gun technologies and high-power energy lasers, satellite and space sensing systems, hypervelocity projectiles and specialized pulsed-power and energy conversion systems that also support next-generation fission energy applications. “We tend to invest in nuclear technologies,” says Dr. Christina A. Back, vice president of nuclear technologies and materials for GA-EMS, “but also in materials that support all of General Atomics and its affiliates, including areas like hypersonics. Composites are a common investment for the company, and we develop those technologies and specialize them for specific end-use products.”

Range of CMC materials development

CW has reported previously on GA-EMS’ development of SiGA high-temperature cladding comprising silicon carbide (SiC) composites for nuclear fuel rods. In this material, both the fiber and matrix comprise crystalline beta-phase SiC (β-SiC), explains Back, because it resists embrittlement from neutrons in the nuclear reactor. This environment currently requires replacing Zircaloy metal fuel rods in less than 5 years. “In the advanced gas-cooled reactors we’re working with, the SiC/SiC clad fuel rods, which have exceptional resistance to neutron damage, could have a 30-year lifetime in helium coolants,” says Back.

“You could also use carbon fiber in an SiC matrix,” she continues, “and we’re also working on a zirconium carbide fiber (ZrC, melting point ~3540°C) with a different matrix that can go to even higher temperatures.” Carbon/carbon (carbon fiber-reinforced carbon matrix, C/C) is also used heavily in high-temperature applications, “but its drawback is that it erodes in the presence of oxygen at those high temperatures, which influences system design and makes C/C inadequate for some applications.”

“SiC is a stronger matrix, depending on the application,” notes Back, “while C/SiC tries to take advantage of using carbon fiber, which is less expensive and more readily available than SiC fiber.” She notes that these non-oxide CMC typically require an interphase or coating between the fiber and the matrix that allows the fibers to slip, “which gives the pseudo ductile mechanical response required for durability.” This interphase is typically a 0.1- to 1.0-micron-thick layer and may include boron nitride (BN) or pyrolytic carbon (PyC), SiC and multilayer combinations, deposited on SiC fibers before matrix infiltration.

“Many of the processes for these different CMC are similar,” says Back, “but you have different precursors or infiltration materials. We’re pursuing a wide range of technologies, trying to provide a kind of incubator for solutions. We choose materials with an end use in mind and then develop the technology to commercialize them.”

MAITrX as a collaboration center to accelerate CMC

Acceleration, Innovation and Exchange are key pillars of MAITrX. “We want to bring together the ideas, people and organizations of importance to help accelerate commercialization of key materials like CMC,” says Back. “We don’t want to recreate the wheel but instead enable collaboration.”

The MAITrX lab has 65,000 square feet of space in the GA-EMS Torrey Pines facility, north of La Jolla in the San Diego area. “We span the whole process for CMC including a fiber lab and composites lab,” says Back. “Because we’re making SiC/SiC parts, we’re onshoring SiC fiber [see sidebar below]. The SiC fiber produced by GE Aviation is well-suited for its turbine engine parts, but for nuclear applications, we need the crystalline β-SiC, which isn’t currently produced in adequate quantities in the U.S.”

“We also have braiders, winders and extruders as well as high-temperature furnaces and other processing equipment. The manufacturing approach we use depends on what we’re making.” For example, the SiGA cladding for nuclear fuel rods uses braided SiC/SiC tape to create sleeves which are then infiltrated multiple times with high-purity β-SiC. “We’re doing that in two ways and have proved out the component properties, strengths and material quality that we need,” explains Back. “Our focus now is to improve the manufacturing. The current infiltration processes are too time-consuming, and we know there are ways to improve.”

“ORNL is helping us to answer that scale-up part,” she continues. “The whole point of our MAITrX lab is to really bring together different technological pieces and skill sets. So, we have no problem working with different companies.” 

Why ORNL?

They are a partner that GA-EMS has worked with for a long time, says Back. “They’ve prioritized advanced ceramics and CMC more than other national laboratories and have equipment that we’re not going to invest in until we fully prove out a process.”

“The MAITrX lab tends to focus and be driven by specific end-use applications,” she explains. “We'll look at what the needs are for a hypersonic vehicle or nuclear thermal rocket, for example. But we don’t have the time for exploration the way that a national lab can — they are not making an end product like we are. Although we have a substantial lab facility and manufacturing center, ORNL has done more exploration into scaling up some of the processes. For example, how you wind and unwind ceramic prepreg tape matters, and what temperatures you use in processing. All of those details are important in moving to industrial scale and this is where we are pairing with ORNL.”

Advancing SiC/SiC for next-gen nuclear power

Several examples of GA-EMS advancements in both nuclear fission and fusion applications are detailed in a March 2026 Journal of Nuclear Engineering article that includes Dr. Back and the head of the MAITrX lab, Dr. Hesham Khalifa, among many co-authors. This article explains that for irradiation stability, nuclear-grade SiC/SiC requires using PyC for the interphase because BN is considered a neutron poison while chemical vapor infiltration (CVI) or deposition (CVD) is favored to produce the β-SiC matrix required. Other matrix infiltration methods reportedly have issues — polymer infiltration pyrolysis (PIP) with lower crystallinity and reactive melt infiltration (RMI) with unreacted free silicon — that tend to produce a SiC matrix not as well suited for irradiation environments.

CVI uses a high-temperature vacuum furnace to decompose a precursor gas and deposit β-SiC between the filaments of a SiC fiber preform. This is then repeated to densify the CMC. However, for the meter-long scales required for nuclear fuel cladding and fusion reactor components, controlling the densification uniformity is challenging, and production quantities of nuclear-grade SiC/SiC using CVI have yet to be demonstrated.

The March article explains that the MAITrX lab is capable of fabricating full-length, 12-foot-long, high aspect ratio SiC/SiC parts in hundreds of parts per batch. These comprise the SiGA cladding for fuel rods to retrofit nuclear fission reactors, including current light water reactors and high-temperature, gas-cooled reactors. MAITrX has also prototyped flow channel inserts (FCI) for fusion reactors, discussed further below.

MAITrX has installed a 35-foot-tall CVI/CVD processing furnace, designed to produce up to 300 SiGA fuel rods per batch. It’s being used to prove out the path for production scale-up. The facility also has machining capability for final finishing and control of surface roughness.

“We’ve made a huge amount of progress,” says Back, “and we’re now starting to test samples with utilities. Our goal is to demonstrate the properties of the rods and show their performance, but also to look at implementation.” She notes that to retrofit a nuclear reactor, one-third of the full 50,000 fuel rods are replaced at a time. “We’ve scaled from single digits to hundreds of rods and now we’re scaling to tens of thousands.”

“But we’re also looking at some of our formulations,” she notes. “We’re looking at going from something like SiC/SiC using CVI of a preform to a more manufacturing-friendly process for hypersonics. And we see the real importance of shifting from C/C composites to processes that are more efficient, starting from the formulation all the way through production and postprocessing. So, the goal is not only to move parts to more robust and higher temperature-resistant materials for each application, but also to reduce the production time and scale part volumes.”

This fabrication capability is also being used to develop fusion reactor applications, where SiC/SiC is critical to enhance efficiency, decrease maintenance and increase plant longevity. In dual-cooled lead lithium blanket (DCLL) designs, SiC/SiC enables components that withstand high temperatures, radiation and plasma interactions while enabling efficient tritium breeding. For example, the General Atomics Modular Blanket (GAMBL) design uses SiC/SiC structural supports that can use radiative cooling and don’t require a hermetic seal, which enables a range of cost and efficiency advantages.

Scaling production of SiC foam

In addition to monolithic SiC/SiC, low thermal conductivity is offered by SiC foam for applications like FCI, used to guide the flow of liquid metal coolants. However, as explained in the March 2026 article, the hardness of SiC foam makes it difficult and expensive to machine into target geometries, while large sizes of SiC foam are also expensive. Meanwhile, commercially available SiC foam offers limited porosity options, which may not meet the thermal conductivity required for fusion applications. Thus, GA-EMS has developed a scalable, low-cost SiC foam-based fabrication technology for FCI. The process does not begin with SiC but instead with carbon foam — which is easier and cheaper to machine. This is then converted into SiC foam by reacting silicon monoxide (SiO) gas with the carbon to form SiC.

GA-EMS has developed a patent-pending method to generate the SiO gas where a particle containing both Si and SiO₂ is fabricated and then infiltrated into the carbon foam. The reaction to release SiO gas begins at temperatures as low as 1200°C, is highly uniform and results in a highly crystalline β-SiC foam that is expected to be irradiation stable. The converted SiC foam also retains the same volume and geometry as the original carbon foam with <0.5% shrinkage.

GA-EMS has successfully manufactured 6 × 6 × 0.25-inch SiC foam plates and 3 × 3 × 3-inch SiC foam channels. It notes there isn’t a fundamental size limit to the conversion step as long as the Si/SiO₂ particles can be infiltrated into the carbon foam. While large blocks of carbon foam are commercially available, finding the desired pore size and porosity can be a challenge. The March article cites work that appears to provide a means for producing carbon foam to meet nuclear requirements and GA-EMS is continuing to develop and scale this SiC foam technology.

Path to manufacturing, need to revolutionize production

Back returns to the progress GA-EMS has made in SiC/SiC cladding for fuel rods and the path forward to parts production. “We’ve moved from formulating the materials and proving them out in test coupons to evaluating the tube components in a nuclear reactor — and each of those steps required different people.” She notes GA-EMS embeds experts from its manufacturing divisions into these teams. “Once we’ve demonstrated proof of concept, we can look at the costs and the equipment and the way to lay out a plant. We then progress toward a pilot plant and finally full-scale production in one of our manufacturing facilities, which are located near our end-use customers.”

“For CMC, we need to really look at revolutionizing how to produce them,” she continues, “and co-location is key, bringing the right people together at the right time in the life cycle — including innovators who are developing formulations that can be standardized with the reproducible properties you need, for example, how to make prepreg tape — to the teams engineering production and implementation. And we work similarly with other companies. That’s the spirit of MAITrX.”

“We’re also developing physics-based modeling and simulation to help reduce the empirical work required and more efficiently guide development but also qualification. Ideally, we will get there faster and integrate the technology in a way that’s already manufacturing friendly so that we can move more quickly to meet the needs.”

“The point is to transition from proof of principle to production at the speed of change today,” says Back. “AI and machine learning are enabling new development cycles. And we need to bring those capabilities together and work together where it makes sense, so that we don’t have these siloed technologies that are nice but not actually advancing our nuclear or hypersonic capabilities, for example. The developments required are difficult, but this is what MAITrX is trying to do — bring together the necessary partners and skills to really move industrialization of CMC and other advanced materials forward in an efficient way.”