AMPulse

April–July 2026 signal analysis

The Welding-Scale Front

Institutional investment in metal AM is concentrating at welding scale: naval sustainment orders, a nuclear reference case, and welding incumbents entering wire DED. Four months of AMPulse signal data, read with their limits.

AMPULSE DATA REPORT · 2026 · UPDATED SEPTEMBER 2026Data snapshot: Aug 7, 2026

During Exercise Rim of the Pacific 2026 in July, the US Navy ran a containerized metal manufacturing system aboard the USS Theodore Roosevelt. The ADDiTEC HYBRiD-X, a laser-wire deposition unit that builds metal parts and machines them in the same envelope, was evaluated for its ability to make and repair components at sea. The Navy called it the first deployment of such a platform aboard a US aircraft carrier; it did not identify a repaired part returned to service. A sailor quoted in the coverage described the goal plainly: diagnose the problem, come up with the solution on board, and cut the lead time.

A trial is not an adoption. But the trial is the right image for what this report defends: accelerating institutional investment in wire-fed directed energy deposition for large-part manufacturing and sustainment, led by the US naval industrial base, reinforced by a nuclear reference case, and answered by welding incumbents on the supply side. That claim is about where money, certification frameworks, and new entrants concentrated. It is not a claim that DED overtook powder bed fusion in coverage volume, installed base, or revenue; it did not.

Of the 490 industry stories AMPulse tracked in July 2026, DED accounted for 34, from 29 companies across 14 countries. (The DED tag counts wire and powder-fed systems together; the event ledger labels each event's feedstock.)

Two regimes, one axis

The simplest way to see the industrial logic is to put metal AM on a single size axis. Laser powder bed fusion owns the left of it: microns of resolution, sub-meter build plates, flight-qualified brackets. To the right of roughly one meter, propellers, impellers, pressure sections, the commercially mature additive routes are deposition processes: wire and powder DED to build the part, cold spray to repair it, with sand binder jetting reaching the same scale through cast patterns rather than direct metal. Systems built like welding cells are what now ship into that end.

The loudest channel is not the densest

Volume and substance rank the industry's processes in opposite orders. Product launches were July's largest news category and its least substantive. Ranked by the share of stories that carried real signal, metal DED came first, ahead of laser powder bed fusion, with every polymer segment far behind. The ranking metric is high-significance share; on mean significance the two were level in July (1.35 against 1.37), because PBF's larger base spreads its scores more evenly. Powder bed's much larger base still produced the most high-significance stories in absolute terms; DED's lead is in density, not volume.

Four consistent months

July was not a one-off. DED's high-significance share was above PBF's in all four months scored, clearly in June and July and within a single story's swing in April and May. Two other metrics moved in DED's favour over the window: DED's mean novelty rose from 1.13 to 1.32, and PBF's mean significance fell from 1.58 to 1.37 while its novelty stayed flat. Mean significance alone does not rank DED ahead; it was level or slightly below PBF in April and July. Four months is a short window; the monthly numbers:

What the DED stream was made of, story by story:

Three threads carry the high-scored tail of that stream: naval demand, a nuclear reference case, and a supply-side response from the welding industry.

Naval sustainment is the most coherent demand cluster

Six of the 16 high-significance July DED stories are tied to US Navy programs. The naval evidence is concentrated in one institutional ecosystem, the Navy's industrial-base and training programs plus one Korean classification approval, and within that scope it is the clearest cluster in the data: visible in purchase orders, backlog, and capacity plans, not just memoranda of understanding.

AML3D, the Australian wire-arc DED company, entered its FY2027 carrying A$16.8 million in orders after booking A$20 million of new orders during the year, with an order book that peaked at A$29 million (company-reported figures). The company attributes the run directly to the US Navy Maritime Industrial Base program converting into orders: two ARCEMY X systems commissioned at Newport News Shipbuilding with four more on order for early 2027 delivery, an order for five non-safety-critical replacement submarine components, and AML3D's own A$12 million investment to double capacity at its Ohio Technology Center. For a company of AML3D's size, that is not a pilot program. That is a customer.

The Navy is also building the workforce to match the hardware. In July, Phillips Corporation completed delivery of 24 manufacturing systems to the Navy's training schoolhouse in Danville, Virginia: twelve hybrid machines pairing Haas CNC platforms with Meltio wire-DED heads, and twelve composite printers. The stated logic is to train sailors on the same systems they will encounter aboard ship. Read together with the Theodore Roosevelt trial, the sequence from schoolhouse to shore facility to carrier trial is consistent with a multi-year sustainment program rather than a one-off demonstration. It is still preparation, not production adoption.

The pattern is not confined to the US. In late July, the Korean Register awarded AM Solutions its first manufacturing-method approval for additive repair of metal ship structures, based on a laser-wire DED process. KR described it as the first time it had verified an AM process for restoring damaged structures rather than fabricating new parts, and it built a new certification framework, covering materials, process, and product, to do it. A method approval is not a license to repair anything on a classed vessel; specific applications still qualify one by one. What it is, is scaffolding for the applications that come next.

Nuclear provides the reference case

The period's largest capacity investment in DED came from an industry that measures components in tons. (The largest DED-relevant transaction by value was a change of control: the RMB 1.02 billion sale of a controlling stake in Chaozhuo Hangke, a Chinese aviation-repair 3D printing company, which is in the story ledger and outside this report's three threads.) Framatome opened a EUR 25 million, 6,000 square meter additive manufacturing centre in Romans-sur-Isère, France, in May, serving nuclear and defence programs; the trade coverage that carries it here arrived in July. The facility runs both laser powder bed fusion and wire-arc additive manufacturing, with MX3D supplying the WAAM capability, and its stated production envelope is the interesting part: parts from a few kilograms to several tons, up to five meters in diameter, with one-ton impellers cited as representative work.

One facility does not mean the nuclear industry has qualified WAAM for safety-critical service, and Framatome has not disclosed the safety classification of what the centre will produce. What the investment does show is a major nuclear OEM moving WAAM out of the lab and into an industrial production-and-qualification environment, which Framatome calls a supply-chain sovereignty measure. Whether it becomes the reference case that pulls European energy and defence primes in behind it is one of this report's watch items, not a conclusion.

The new entrants are welders, not printers

The third thread is who showed up on the supply side. DAIHEN, the century-old Japanese heavy-electrical and welding-equipment group, announced its first additive manufacturing product in May and drew wide international coverage in July: the ArcBuilder 3D, a wire-arc DED system priced at 75 million yen before tax, roughly $450,000 to $500,000. Rather than license a printer maker's technology, it productized its own AC Synchro-Feed welding process, which cuts spatter by up to 98 percent by DAIHEN's own measurement, and showed propellers and rocket nozzles as demonstration parts. The ambition is stated in numbers: 20 systems in the first year, and 10 billion yen in AM revenue by 2030.

Germany's GEFERTEC, meanwhile, launched the arcTitan, a WAAM system with a sealed inert-gas chamber built for titanium, a high-value material for aerospace and selected naval applications that arc processes have struggled to handle in open atmosphere.

The DAIHEN entry matters for two reasons. First, the price puts a credible wire-DED system within reach of shipyards, heavy-equipment repair shops, and machine shops that were never going to buy a seven-figure laser machine, though installed cost with shielding, post-processing, and inspection will land higher than the list price. Second, the channel: welding incumbents bring distribution and service organizations in exactly the industries, shipbuilding and energy above all, where DED demand is forming. Lincoln Electric made this move in the US in 2019. If ArcBuilder systems move in volume through welding-equipment channels, other welding OEMs have every reason to follow; that is a scenario to watch, not a forecast.

Why this scale fits these buyers

Powder bed generally wins on feature resolution, geometric complexity, and the breadth of its serial-production qualification base. Wire DED wins on deposition rate, build volume, feedstock cost, and the ability to add metal to an existing component. The buyers in this period's high-scored events are the ones for whom the second list matters more: their parts are measured in meters and tons, and the sustainment enterprise behind them, the Navy's depots and shore facilities, spends on keeping old components in service, including parts whose original supplier no longer exists. A process that builds onto a worn part, which is what a welding-derived process naturally does, fits that economics; it is telling that KR's first AM method approval was for repair rather than fabrication.

There is also an institutional fit. Shipyards and energy companies have employed certified welders for a century; their codes, inspectors, and workforce already speak the language of arc and wire, which lowers the organizational barrier that powder bed's cleanroom discipline raises. The extension has a cost: WAAM brings its own procedure set for build-path planning, interlayer temperature control, residual stress, and inspection of layered deposits. But combined with point-of-need deployment, ADDiTEC's HYBRiD-X ships as a containerized unit, and AML3D delivered its first containerized ARCEMY to a US Navy additive manufacturing facility in July, the machine can go to the ship or the forward site rather than parts traveling from a central factory.

The July event ledger

The events behind the three threads, plus the two aerospace qualification rows, one row per underlying event with syndicated and follow-on coverage folded in. This is a curated subset: the full 34-story July table, including the DEEP subsea habitat, BMW's serial WAAM plan, Hyundai's AM centre and the Chaozhuo Hangke sale, ships in the report's data file.

How far up the ladder this is

Every claim in this report sits somewhere between "a product was announced" and "a qualified part went back into service." Placing July's events on that ladder shows both the momentum and its limit. The rungs just below the empty top one are aerospace, not naval: GKN Aerospace and Pratt & Whitney are targeting 2028 certification for a laser-wire F135 engine casing, BMT Aerospace joined Safran Aero Boosters and Pratt & Whitney in July to qualify large titanium F135 parts, and GKN's $8.4 million TITAN-AM programme with the US Air Force Research Laboratory, launched in April, is industrializing laser-wire deposition for titanium aerostructures. If a qualified wire-DED part enters service first, it may well be an engine part.

The counter-read

What to watch

Methodology

Lee, S. (2026). The Welding-Scale Front. AMPulse Data Reports. https://www.ampulse.online/reports/wire-ded-welding-scale-2026