Building a sixth-generation fighter is, in many ways, an engine problem dressed up in a stealthy airframe. The F-47 — Boeing’s now-flying Next Generation Air Dominance demonstrator — needs an engine that does three things at once: it has to produce more thrust than the F-35’s F135, sip fuel like an airliner in cruise, and dissipate enough heat that the airframe’s directed-energy weapons and AI-driven sensors do not melt their own racks. Nothing in current operational service does any one of those things at the level a 2030s air-dominance fighter will demand. So the U.S. Air Force, quietly, has been spending the better part of a decade building the engine first.

On 12 May 2026, that programme — the Next Generation Adaptive Propulsion (NGAP) effort — hit a critical milestone. GE Aerospace announced it had completed the Assembly Readiness Review (ARR) for its XA102 engine on 11 May. Pratt & Whitney followed by completing the technical assessment for its XA103 on 8 May. Both companies are now cleared to begin physical assembly of full prototype engines. The era of computational models, digital twins and rig tests is over. The era of hardware on the test stand begins.

What “Adaptive” Actually Means

A conventional fighter turbofan has a fixed bypass ratio. The compressor sucks in a fixed amount of air, pushes a fixed fraction through the core, and exhausts the rest around it as cool thrust. The geometry is welded into place. It is optimised for a single flight regime — typically supersonic dash, in the case of a fighter engine — and it is suboptimal everywhere else.

An adaptive-cycle engine adds a third air stream that can be opened or closed by the engine’s flight control system. In high-thrust mode (climb, dash, supersonic) the third stream closes and air is routed through the core for maximum thrust. In cruise the third stream opens, increasing bypass, dramatically improving fuel economy. In ground attack or low-altitude penetration the airflow can be biased to reduce IR signature. The same engine becomes three different engines, switched by software.

That is not a new concept. GE built a working three-stream adaptive demonstrator — the XA100 — for the F-35 around 2020. It proved the technology in test stands but never made it onto a Lightning, partly for political reasons (the Pentagon downselected to a single F135 upgrade path) and partly because the F-35’s airframe was not optimised for an engine with significantly different airflow characteristics.

XA102 vs XA103: The Lessons Carried Forward

The XA102 (GE) and XA103 (P&W) are both descended from that earlier work, but redesigned around a clean-sheet sixth-generation airframe rather than the F-35’s existing nacelle. Both engines target roughly 45,000 lb of thrust — about 7-10% more than the F135. Both aim for around 25-30% improvement in fuel burn at cruise. Both promise dramatically increased onboard electrical-power generation, the figure the Pentagon cares about most for next-gen directed-energy weapons and sensor fusion compute loads.

Where they differ is in design philosophy. GE’s XA102 leans on the company’s experience with the LEAP commercial turbofan, applying mass-production manufacturing techniques to military hardware in ways the Pentagon has historically been allergic to. P&W’s XA103 leverages F135 production tooling and supply chains, betting that incremental refinement of a proven base is a lower-risk path to operational service. Both arguments are defensible. Only one will win the F-47 production contract.

Why the Pentagon Is Funding Both

The decision to keep two adaptive engine competitors in development simultaneously is itself a strategic statement. For most of the last 30 years, the Pentagon has selected a single propulsion supplier per fighter generation — Pratt for the F-22 and F-35, GE for legacy F-15s and F-16s. The result is a propulsion industrial base whose health depends on whichever side wins, with sometimes-disastrous consequences when that side stumbles.

NGAP is the first major fighter-engine programme to deliberately fund two competitors all the way to prototype assembly. The FY2027 budget request, $514 million for competitive prototyping, is a hedge: the Air Force wants two functioning engines in 2028, then a downselect based on actual test-stand performance, then full-rate production for the winner. The loser stays in the industrial base as a competitor for upgrade contracts.

Heat, Power, and the Sixth-Gen Problem

The F-47 is being designed around requirements no current fighter has to meet. Onboard directed-energy weapons — laser pods capable of dazzling or destroying incoming missiles — require electrical generation of several hundred kilowatts. AI-driven sensor fusion suites running multiple radar and EO-IR feeds simultaneously require compute hardware that produces correspondingly large thermal loads. The aircraft itself, flying long Pacific missions to maintain a credible deterrence over the South China Sea, requires range that conventional fighter engines simply cannot deliver.

Adaptive cycle engines solve all three problems with a single technology. The third air stream provides the heat-rejection capacity needed for high electrical loads. The variable-cycle behaviour produces the fuel-burn efficiency needed for range. The thrust-to-weight ratio matches or exceeds the F135. A sixth-generation airframe without an adaptive engine is, structurally, a fifth-generation aircraft with worse manners. That is why the Pentagon is willing to spend a decade and several billion dollars getting the engine right before the airframe enters production.

The Long Test Stand

Both companies will spend the rest of 2026 and most of 2027 assembling, instrumenting and bench-testing their engines. Initial test-stand runs are expected in the late 2020s. Flight testing on a flying test bed — likely a modified B-1B or a dedicated single-engine pylon configuration — would follow in 2028-2029. The F-47 itself is targeting first flight of a production-representative airframe around 2029-2030, with initial operational capability in the mid-2030s.

For now, the milestone is hardware. After fifteen years of digital design and rig testing, after the bruising downselects of the early 2020s, the United States Air Force is finally going to find out whether adaptive cycle engines work in practice. The XA102 and XA103 are about to leave their CAD environments and become metal on a stand in West Palm Beach and Cincinnati. By 2030, one of them will be powering America’s next fighter.

Sources: The Aviationist (Parth Satam, 13 May 2026); Aviation A2Z; Aerotime; Aero Magazine; Defense Post; eplane.ai; Aerospace Global News; USAF NGAP programme office statements.