Elmet Can Print C103 Because It Owns the Atomizer, Not Because It Bought a Printer

Two weeks ago 3D Systems issued a release about a machine sale. The substance of it belongs to the buyer. Elmet Technologies will run a triple-laser DMP Flex 350 to print monolithic C103 niobium heat exchangers for hypersonic vehicles, and the printer is the least differentiated item in the package: Elmet atomizes its own C103, holds a granted US patent on additive fabrication of metallic parts, and is carrying a certification file that runs from powder conformance through to NASA-STD-6030 part classification. Only one of those is available on a purchase order.

C103's Real Gate Is Oxygen, Not Laser Count

C103 is a niobium alloy defined as much by what it must exclude as by what it contains. Elmet's own powder datasheet, on the Elmet Additive product page, specifies niobium at 88 percent minimum with hafnium at 9 to 11 percent, and caps oxygen at 0.05 percent. That ceiling carries the engineering. Niobium is oxygen-hungry, and interstitial pickup during atomization, handling or printing embrittles a part rather than merely dirtying it. The alloy earns its place because, on the same datasheet, its strength remains effective at temperatures up to 2700 degrees F (1482 C), which is why it turns up in thruster nozzles and hypersonic thermal hardware rather than in a nickel superalloy.

The conventional heat exchanger in this material is a brazed multi-piece assembly, and its joints are the leak and fatigue path under thermal cycling. Printing the part monolithically removes them. The comparison worth making is therefore not additive cost against machining cost, but a jointed assembly against a part with no joints.

A chamber that cannot hold near-zero oxygen converts qualified powder into scrap and destroys the reuse rate. Laser count buys throughput. The atmosphere buys the part.

Why Owning the EIGA Line Puts Elmet on Its Own Cost Curve

Elmet, successor to H.C. Starck Solutions, atomizes its C103 by electrode induction gas atomization and then prints it. No other named C103 printer in the public record holds both ends of that chain: ADDMAN's Castheon subsidiary buys qualified powder from Amaero under a standing purchase obligation, and Velo3D-based routes qualify someone else's material. 3D Printing Industry's coverage of the installation puts refractory feedstock in the thousands of dollars per kilogram, which is the difference between a pass-through cost and an internal transfer, and the same finding we reached on alloy rights and atomization capacity in the superalloy market.

Conformance here is verified rather than asserted. Auburn University's National Center for Additive Manufacturing Excellence confirmed the powder to ASTM F3635 Class B following heat treatment. The Nasdaq listing of parent company The Elmet Group this spring, and the AM process patent granted shortly before it, mean the additive position was underwritten into the offering rather than bolted on afterwards.

Our own index classifies Elmet Technologies under Materials in the value chain rather than equipment, a field populated for roughly 1,000 of 6,500 companies. Our patent linkage, which is name-matched and review-gated and so undercounts subsidiaries, associates 4 active patent records with the company. Our news pipeline logged 4 items on Elmet in the last 180 days against 0 in the 180 days before that, matched on company name rather than sector.

Elmet's Own Post-HIP Data Is the Open File

Nothing here is qualified. 3D Systems' release says Elmet expects machine, material and process qualification within a couple of months and NASA-STD-6030 certification a few months after that. There is no named customer, no named hypersonic program and no disclosed contract value attached to the heat exchangers.

It is also not a category first. NASA published crack-free laser powder bed fusion work on C103 for propulsion in 2020. Castheon has printed Niobium C103 and Super C103 out of a dedicated hypersonics facility since 2023, and Amaero qualified C103 powder into that supply chain. Elmet itself has a published Nb-C103 case study printed on a single-laser Renishaw AM400. The new machine buys throughput and atmosphere control, not a new capability, and coverage framing it as the latter overstates the technical jump.

The sharpest counter-evidence is Elmet's own. Its published C103 work shows post-HIP and post-coating specimens with decreased strength versus the as-printed condition, with some specimens failing ASTM F3635 strength requirements. The unresolved physics sits after the build, not during it. And the announcement is a vendor release from 3D Systems built around a single machine placement, which has repeatedly proven a weak proxy for segment demand.

Integration Versus Substitution: Amaero Takes the Other Side

Five days before the Elmet release, Amaero disclosed a US Department of Defense contract to develop alternative refractory alloy powders, explicitly aimed at reducing the cost of powders such as C103 and cutting dependence on foreign raw material sources. Two companies in the same alloy chain, in the same fortnight, named the same binding constraint and answered it in opposite directions. The merchant atomizer is being paid to engineer the feedstock cost away. The integrated producer is absorbing it internally and turning reuse into margin.

On balance the integration bet is the more durable of the two, but for a narrower reason than the coverage suggests. Elmet's advantage is an input-cost and documentation advantage, not a process moat. Metal AM has produced several process-level moats that read as permanent in their first year and eroded once competitors matched the parameter set. What does not erode as fast is atomization capacity, powder conformance history and a certification file, which is the same reason acceptance criteria rather than machine specs now decide who supplies qualified parts.

Three Things That Would Void the Vertical Integration Case

First, schedule. If machine, material and process qualification does not land in the stated window and NASA-STD-6030 classification slips with it, the stated production date becomes marketing and the file reads as capability demonstration again.

Second, post-processing. If hot isostatic pressing and coating continue to pull printed C103 below ASTM F3635 strength requirements, monolithic heat exchangers stay a demonstrator regardless of how clean the powder is. This is the one open item Elmet has published against itself, and it is the item a program office will read first.

Third, the alloy. If the Department of Defense-funded substitution work yields a cheaper refractory that primes are willing to respecify, C103 stops being the reference material for hypersonic thermal hardware. Elmet's position transfers only to the extent that its atomization and qualification capability is alloy-portable, which for a producer already running tungsten, molybdenum, tantalum and FS-85 it largely is.

What would settle the argument in the other direction is short and specific: a named hypersonic program, a disclosed contract value, and a second machine.