HSR — Hybrid Supersonic Rotorcraft by Compagnie Apex | Stealth VTOL · Stop-and-Merge Coaxial Rotor · 60° Scissor Sweep · Mach 1.5

HSR — Baseline (preliminary)

Single source of truth for this document. Any figure quoted elsewhere refers back to this block; where a value is an assumption rather than a result it is marked TBD.

Masses	MTOW 10 800 kg · OEW 7 400 kg (target; +5 % / ≈ 370 kg growth margin carried outside the figures) · usable fuel 2 450 kg · payload 950 kg allowance — realisable strike load ≈ 400–650 kg on two trapeze stations (§ 8); the allowance is a structural and balance envelope, never quoted as a demonstrated load · W_mid 9 700 kg = OEW + full payload + 55 % of usable fuel (1 350 kg), i.e. the mass at the start of the supersonic leg
Propulsion	1 × adaptive-cycle engine (three-stream adaptive turbofan) · PTO 3 800 kW · 52 kN dry / 75 kN reheat · ISG 1.1 MW continuous / 1.4 MW peak
Geometry	Rotor Ø 12.8 m (2 stages × 2 blades) · length ≈ 14.2 m · height ≈ 4.1 m · merged box thickness 90 mm · merged chord 1.55 m (= MAC) · inter-stage spacing 0.62 m
Reference areas	S_wing 15.6 m² · S_body ≈ 12 m² (TBD by CFD) · S_eff ≈ 27.6 m² — W/S_wing and W/S_eff are quoted separately throughout, never combined. Aspect ratio is quoted with its sweep: AR ≈ 10.5 unswept, AR ≈ 2.6 at 60° (projected span 6.4 m); each half-panel is 5.0 m exposed × 1.55 m chord (panel slenderness ≈ 3.2)
Rotor system	Ø 12.8 m, 2 stages × 2 blades · tip speed 210 m/s · advancing-blade Mach ≤ 0.85 at 265 kt / 70 % N_r · disc loading ≈ 84 kg/m² · solidity σ ≈ 0.169 · required shaft power ≈ 3 150 kW against a 3 800 kW PTO (≈ 21 % margin) — build-up in § 4, expanded in § 6
Design point	Mach 1.5 at 12 km · supercruise M 1.15 dry at 45° sweep · disc loading ≈ 84 kg/m² · sweep schedule 30° / 45° / 60° · merge window 235–265 kt, sea level – 2.5 km
Electrical	± 270 VDC main bus for all high-power actuation (28 VDC retained for legacy avionics only) · 12 MJ / 280 kg supercapacitor buffer · ISG 1.1 MW continuous · isolated essential bus

How the Mach 1.5 target is treated in this document:
Mach 1.5 is not reachable with a turning rotor, and hover is not reachable with a fixed wing sized for Mach 1.5 — which is why this aircraft carries no permanent lateral fixed wings at all. Instead, its two coaxial two-blade rotor stages stop in flight aligned dead straight at 90° to the fuselage, merge vertically into a single wing box — and scissor-sweep its two halves symmetrically aft into a pure 60° delta (“Stop-and-Merge”). Three class choices are locked in from the start — lift: rotors in helicopter mode, the merged wing beyond · propulsion: one rear adaptive-cycle engine (ACE) doing both jobs, mast drive through a power take-off and pure jet thrust · survivability: stealth shaping in both natures. Individual bricks have precedent — the stopped rotor (X-50, RSRA/X-Wing), the rigid coaxial rotor (X2/S-97), 60°-class symmetric variable sweep (F-14, Tornado), single-mast pivot bearings (NASA AD-1), thrust-vector control (X-31), adaptive-cycle engines (XA100/XA102, ground-tested), Nitinol shape-memory actuation (NASA SMA). These are cited as design inspiration and existence proofs of the individual physics — not as evidence that this configuration works. None of them was flown in this combination, at this scale, or on one engine. What has never been done is merging two rotor stages into one wing in flight. That, not the speed, is the real risk — and the engineering below is concentrated on exactly that point.

Written versus demonstrated — the seven items that keep this document conditional

Successive reviews have closed descriptions and gates. That is not the same as closing the engineering. This block separates the two so that no reader mistakes a written action for a retired risk. Every row below is a statement about what does not yet exist.

Item	What this document now contains	What still does not exist	Retired at
ACE engine and its PTO	The trade study is specified, the partner agreement is a Phase 0 exit gate, the kill criterion at T0+24 is written, and three candidate PTO architectures are now compared (§ 4)	A contracted partner and a demonstrated PTO concept. Until one exists, every performance, mass and schedule figure downstream of the powerplant is conditional — this is the structural risk, not merely the first one	Phase 0 gate (§ 12)
Inter-stage spacing 0.62 m (0.048 D)	The 0.95 m fallback, its + 18 kg / + 3 s cost, and a quantitative freeze rule with pass/fail clearance numbers (§ 3)	The rig data the rule is applied to. The decision is defined, not taken	Phase 0 whirl rig, frozen at the Phase 1 gate
Roll authority by aeroelastic twist	Numerical requirements at both ends of the envelope and a first-order reversal / divergence check that converts them into a stiffness and elastic-axis constraint (§ 6)	A validated aeroelastic model, a GVT, and any measurement at all. The check narrows the design space; it does not clear it	Phase 1 structural freeze, Phase 5 flight
Supersonic L/D	A drag build-up with the semi-flush, iris-seam and chine-junction penalties actually debited (§ 6) — the objective end of the old 5.5–6.5 band is withdrawn as a result	CFD. The build-up is a hand calculation with declared assumptions, and the thrust it must be compared against is an ACE-partner input that does not exist	Phase 0–1 CFD, Phase 5 measurement
Signature	A ranked relative contributor budget (§ 17) and a preliminary RCS model as a Phase 0 gate item	Any number. No dBsm value, modelled or measured, exists anywhere in this programme	Preliminary model Phase 0, measurement Phase 6
Alternatives trade study	The mission set, the comparison axes and a first-order qualitative answer (§ 17)	The quantitative study. Until it exists, the M 1.5 dash is an assumed requirement, and it is the assumption the whole configuration rests on	Phase 0, before Phase 1 funding
CPFH and life-cycle cost	A driver decomposition and the reasons the answer is adverse (§ 17)	An estimate. An operator buys on this number and this programme cannot yet quote one	Phase 1

1. Project Objectives

Design a hybrid military aircraft able to operate with the flexibility and hover capability of a stealth helicopter, while possessing the ability to transform in flight to reach supersonic speed (Mach 1.5) with the performance of a fighter aircraft — without using permanent lateral fixed wings. The designation states the aircraft plainly: HSR — Hybrid Supersonic Rotorcraft — intercept and strike targets that a helicopter cannot catch, and insert or extract personnel where a fighter cannot land.

0–200 kt

Helicopter Mode

Stealth VTOL · hover · terrain masking

235–265 kt

Stop-and-Merge

Sea level – 2.5 km · rotors stop → merge into one wing

M 0.5–1.0

Acceleration & Climb

Halves scissor-sweeping 30° → 60°, full thrust

Mach 1.5

Design Point

Supersonic intercept at 12 km

What the aircraft must do

• Take off and land vertically from any 25 × 25 m pad — no runway in any nominal mission state
• Hover and manoeuvre nap-of-the-earth with a low acoustic and radar signature
• Transform in flight and dash at Mach 1.5 with fighter-like agility — an objective, not a demonstrated capability
• Carry two crew plus two extracted personnel or an internal weapons load
• Return, un-merge, and land vertically at the point of extraction

Declared exception to “no runway”: the degraded states of § 3 — asymmetric Nitinol pin set, jammed collar, failed un-merge — recover as a fixed-wing aeroplane at 170–190 kt and require a prepared strip of ≈ 1 200 m. The HSR is runway-independent in every nominal state and in most failure states, but it is not runway-independent in all of them, and basing plans must carry a diversion field.

What it deliberately gives up

• Permanent lateral fixed wings — excluded by requirement, replaced by the merged rotor-wing
• Heavy-lift capability: payload is measured in hundreds of kilograms
• Supersonic endurance: Mach 1.5 is an intercept dash, not a cruise regime
• A second engine: one ACE does everything — accepted single-engine risk, managed by autorotation
• A whole-airframe ballistic recovery parachute — traded out at § 15: no precedent exists above ≈ 2 t and the honest mass was 400–800 kg

2. Global Architecture & Airframe (Fuselage)

The airframe is a faceted, blended lifting body sized around three fixed volumes — cockpit, internal bay, engine — and one moving mechanism: the central mast. Its visual and stealth DNA is deliberately that of the RAH-66 Comanche: the same canted flat facets, retractable gear, serrated apertures, all-internal carriage and suppressed hub line — carried here into a shape that must also survive supersonic flow. Because the requirement excludes permanent lateral wings, the fuselage itself must carry a meaningful share of lift in jet mode, and every surface is angled for radar-signature control in both natures.

Area ruling — the wasp waist

The fuselage obeys the transonic area rule: the flanks pull in sharply at the wing root — a visible “wasp waist” — so the total cross-sectional area distribution of body plus merged wing stays smooth from the nose apex to the nozzle (there is no radome: see § 7). Where the swept wing adds area, the body gives it back, and the volume wave drag of the combination falls drastically below that of an un-ruled fuselage of the same volume. The waist is cut for the wing in its swept positions (30–60°), where wave drag actually exists.

Chine line — compression lift

A sharp chine line springs from the nose apex and runs aft along both flanks, dividing the upper facets from a flat, slightly dished lower body. Subsonically it sheds a stable vortex pair that feeds the lifting body; at Mach 1.5 the intent is to pin the bow shock along its own edge, trapping the high-pressure field under the flat belly so the forebody recovers compression lift the wing never has to generate. The magnitude of that contribution is TBD by CFD and tunnel test — the shock-riding behaviour is claimed here as a shaping intent, not as a validated lift budget. The chine doubles as the fuselage's principal RCS alignment edge.

Raked canopy — no bubble

The single-piece canopy is highly raked (windscreen ≈ 62° from vertical) and blended into the chine shoulder. A hemispherical “bubble” canopy would be a transonic drag generator and a broadband radar scatterer; the raked, gold-film-coated wedge keeps the canopy inside the forebody's shock system and the crew inside the faceted mould line — vision requirements are met by the raked geometry plus a distributed-aperture sensor belt.

Element	Architectural choice	Justification
Fuselage	Faceted blended body, ≈ 14.2 m long, area-ruled “wasp waist” at the wing root, sharp chine line from the nose apex down both flanks, S_body ≈ 12 m² of effective lifting planform	Comanche-heritage facets and chines are intended to control the radar return in helicopter and jet mode; the area ruling cuts supersonic wave drag and the chined forebody is shaped to ride its own bow shock for compression lift at Mach 1.5. The body's share of supersonic lift is assumed ≈ 40 % at preliminary level and is TBD by CFD — the merged wing is sized on that assumption and is re-sized if it does not hold.
Cockpit	2 crew side by side, single-piece highly raked low-IR canopy blended into the chine shoulder, zero-zero ejection seats	Pursuit and extraction are two-operator missions; the raked wedge avoids the transonic drag and radar return of a bubble canopy, and side-by-side keeps the cabin volume usable for the extracted personnel directly behind.
Internal bay	One ventral bay, 3.4 × 1.1 × 0.8 m: 2 fold-down seats or 950 kg of stores, released by pneumatic trapeze ejectors	External carriage is forbidden by the stealth requirement; the same volume serves extraction and strike. At Mach 1.5 a store cannot simply drop: the cavity shear layer and the bay's own shock system would trap it against the airframe, so the trapeze arms punch each munition through the supersonic boundary layer at 40 g, clearing the shear layer into free stream before separation dynamics can bounce it back.
Forward gun	Fixed forward-firing 25 mm rotary cannon (Gatling), fully internal in the lower nose (chin), behind a flush faceted stealth trapdoor	The trapdoor snaps open only for the burst and re-seals immediately, so the chin keeps its Mach 1.5 mould line and its low-observable aspect the rest of the time. The chin position is synergistic with the dorsal S-duct inlet: muzzle gas is generated below and ahead of an intake that breathes from above. The geometry is designed to preclude gun-gas ingestion; this must be verified by CFD and firing trials (§ 12, Phase 6) before any firing clearance is granted.
Central mast	Single load-path titanium mast on the centre of gravity carrying the smart articulated hub; retracts semi-flush into a shallow dorsal saddle sealed by spring-loaded carbon-composite iris doors — the cavity depth is constrained by the S-duct, the engine and the aft fuel cell (§ 3)	Hosts both two-blade stages; once merged, the mast becomes the scissor-sweep pivot, incidence actuator and carry-through spar — then sinks flush so no hub or pylon remains in the supersonic flow.
Empennage	Aft fuselage tapers into a flat diamond spine (YF-23 style) carrying a canted V-tail with seamless morphing ruddervators (compliant aeroelastic trailing edges, no hinge line) — no tail rotor, no fenestron, no rear propeller	The fins are passive stabilising facets in normal flight: in jet mode all active pitch and yaw moments come from the 2-D vectoring nozzle, and in helicopter mode those axes belong to rotor cyclic pitch and differential coaxial torque. The morphing trailing edges exist for one case — engine-out at Mach 1.5, when the nozzle dies with the engine — and being hingeless they add that authority without a single radar-scattering gap or actuator bulge.
Sensor apertures	Conformal AESA “smart skin”: T/R modules laminated into the composite cheeks, wing roots and spine — no nose radome	The chin cannon and the sharp chined forebody leave no volume for a conventional radome, and a dielectric nose cone would break both the chine line and the forebody shock. Distributing the array into the structure recovers the aperture area, gives 360° coverage and folds the EW function into the same skin (§ 7).
Landing gear	Retractable tricycle, high-sink-rate rated (4 m/s), rough-field tyres	Vertical landings on unprepared sites at mission weight; fully enclosed by serrated doors.
Signature control	Edge-aligned facets, serrated doors, S-duct inlet, shielded nozzle, RAM on leading edges; mast and hub retract flush behind iris doors in jet mode	The hub is the classic helicopter radar spike: here it is faceted and RAM-coated in helicopter mode, and in jet mode it retracts semi-flush into a shallow dorsal saddle — the intent being to minimise hub drag, hub return and blade-flash signature. A full-depth cavity is not available: the S-duct, the engine, the aft spine cell and the ventral bay already occupy the depth (§ 3). Residual root-step drag and seam return are TBD by test.
Reference areas	S_wing = 15.6 m² (merged wing, 2 panels) · S_body ≈ 12.0 m² (lifting body) · S_eff ≈ 27.6 m²	Quoted separately and used consistently throughout: wing loading is W/S_wing (≈ 692 kg/m² at MTOW, ≈ 622 kg/m² at W_mid), effective loading is W/S_eff (≈ 391 kg/m² at MTOW, ≈ 351 kg/m² at W_mid). S_body is the preliminary assumption of § 2 and is TBD by CFD.
Dimensions / masses	Length ≈ 14.2 m · rotor Ø 12.8 m · height ≈ 4.1 m · MTOW 10 800 kg · OEW ≈ 7 400 kg · W_mid = 9 700 kg (defined as OEW + full payload + 55 % of usable fuel — the mass at the start of the supersonic leg; used for all performance figures)	Sized so the effective loading W/S_eff closes at Mach 1.5 while disc loading stays in the VTOL class (≈ 84 kg/m²).

General arrangement — Orthographic views

Indicative dimensions for 10 800 kg MTOW · Rotor Ø 12.8 m · Length ≈ 14.2 m · One rear ACE engine · No permanent fixed wings

3. "Stop-and-Merge" Wing System (Rotors & Wings)

The heart of the innovation. The system rests on a single central titanium mast hosting two coaxial two-blade contra-rotating rotor stages — exactly two stages of two blades, four blades in all. In helicopter mode they are a classic rigid coaxial pair — torque cancelled, no tail rotor. For supersonic flight they perform the manoeuvre that gives the aircraft its name: both stages brake and stop dead straight at 90° to the fuselage (Stop Straight); the upper stage then executes a two-axis merge stroke — 0.62 m down the mast and 0.70 m aft in chord — so that its blades seat behind the lower stage's rather than on top of them, and Nitinol locking pins clamp and pre-load the two stages onto shear keys into a single rigid wing box (Vertical Merge & Chord Offset); then the two halves of that straight wing pivot symmetrically aft about the mast — the Scissor Sweep — into a pure 60° delta. Two blades per stage is not a styling choice: it is the only blade count for which a stopped stage forms one clean, straight lifting surface with no odd blade left pointing into the airstream — and no side-folding hinge anywhere in the load path.

The chord-offset degree of freedom — why the merge cannot be purely vertical. A blade is 0.85 m in chord and 90 mm thick. If the two stages merged by pure vertical descent onto the same straight line, they would land exactly on top of each other: the merged section would be 0.85 m in chord and 180 mm thick — t/c ≈ 21 %, which is not a supersonic wing by any reading. The 1.55 m merged chord and the 5.8 % thickness ratio that the rest of this document depends on require the upper stage to be displaced ≈ 0.70 m aft in chord relative to the lower stage before the pins fire (0.85 + 0.85 − 0.15 m of deliberate overlap = 1.55 m). Earlier issues of this specification quoted the 1.55 m chord without ever naming the mechanism that produces it; that omission is closed here.

• The mechanism: once the mast is stopped and indexed, the ball-screw merge collar is no longer a pure vertical slide. It is an L-stroke carriage: the upper hub descends 0.62 m on the mast, then translates 0.70 m aft along a pair of chordwise rails machined into the stopped mast head, becoming deliberately eccentric to the mast axis. Because the whole upper stage translates as one rigid body, both upper panels move aft together and left–right symmetry of the merged wing is preserved exactly.
• The profile consequence: the two stages are not identical. The lower stage carries the forward half of the biconvex-diamond section (morphing Nitinol leading edge, maximum thickness at its own trailing face) and the upper stage carries the aft half (blunt forward face, sharp trailing edge). In rotor mode each runs as a conventional 0.85 m / 90 mm blade with a spring-loaded compliant fairing closing its bluff mating face; the fairings are retracted by the collar at the start of the stroke.
• What it costs: ≈ 60 kg of rails, carriage and eccentric-load fittings (booked in § 9, which is why the mechanism line is now 740 kg and not 680 kg); ≈ 7 s added to the sequence; a cantilevered upper-blade root reaction that no longer passes straight down the mast axis but through the carriage; and a second indexing axis with its own ± 3 mm tolerance. The cantilever and the fairing-retraction reliability are new Phase 1 rig items.
• The alternative, stated honestly: delete the chord offset and the aircraft still merges — but at 0.85 m of chord and t/c ≈ 10.6 %, volume wave drag rises by a factor of ≈ 3.3, S_wing falls to ≈ 8.5 m², and the Mach 1.5 point does not close on 52 kN dry. There is no cheap version of this concept: either the chord-offset degree of freedom exists, or the design point moves down to ≈ M 1.1.

The five steps, in words

M0 — Partial Unload — level at 250 kt between sea level and 2.5 km, jet thrust carries the thrust line, collective is reduced to a residual 30–40 % of weight and rotor speed is brought down to ≈ 70 % N_r to keep advancing-tip Mach below 0.85 at this speed. The rotors are not zeroed here: at 250 kt at low level the body and stopped surfaces alone would need C_L ≈ 0.34, and any higher or slower entry pushes that beyond what a 5.8 % biconvex section at a 0°-datum fuselage can produce. This is why the merge window was cut from 200–250 kt / 2–7 km to 235–265 kt / sea level – 2.5 km: it is the only band in which the lift equation closes at every instant of the sequence.
M1 — Stop Straight — rotor brakes stop the upper stage, then the lower stage, both aligned dead straight at 90° to the fuselage to ± 0.2° — four blades forming one clean perpendicular line. In parallel the torsion actuators untwist the blades from their −10° hover washout to a flat 0° datum, so the stopped stages are already two clean, geometrically flat wings. The residual 30–40 % of rotor lift is bled to zero only once strain gauges confirm the stopped stages are carrying — not on a timer.
M2 — Merge Stroke (down + aft) — the L-stroke collar first lowers the upper stage 0.62 m down the mast, then translates it 0.70 m aft in chord on the rails of the stopped mast head, both axes at 100 mm/s — ≈ 13 s in all — under near-zero blade load. The chordwise leg is not cosmetic: it is what turns two 0.85 m blades into one 1.55 m chord instead of one 0.85 m chord of double thickness. The descent path is mechanically clear because there is no swashplate: pitch is commanded electrically inside the hub (see below), so no pitch link, scissor or slider crosses the travel. No side-folding hinge is involved at any point.
M3 — Nitinol Lock & Morph — Nitinol shape-memory locking pins contract electrically, drawing the two stages together and pre-loading the joint onto steel shear keys to close a rigid, 90 mm ultra-thin wing box of biconvex diamond section. The pins supply clamping pre-load only — the shear keys and the common shear web carry the flight loads; no heavy hydraulics. Resistively heated Nitinol leading edges transform from the blunt hover profile to the sharp supersonic profile.
M4 — Scissor Sweep & Semi-Flush · law "JET" — the central titanium mast becomes a pure pivot: the two halves of the merged straight wing pivot symmetrically aft — the Scissor Sweep — into a delta, 30° for the transition, scheduled to 60° with Mach, with fuel transferring forward-to-aft in step so the CG tracks the retreating centre of lift — and with the mechanism's own mass migration booked, not ignored (see the CG card below). The mast column then descends into its dorsal saddle and spring-loaded carbon-composite iris doors slide shut around the swept root: the wing box upper surface is intended to come level with the flat dorsal spine, minimising hub drag. The stroke is limited by available depth, not by choice — the retraction-clash callout below states why, and withdraws the earlier claim of a fully flush spine. PTO declutched, nozzle and aeroelastic twist live, envelope open to Mach 1.5. Total sequence ≈ 32 s.

Why this sequence is intended to be survivable

The historic killer of stopped-rotor aircraft (X-50) is the lift gap while the rotor slows through partial rpm. The design intent here is to close that gap before it opens: the aircraft flies on jet thrust, body lift and a deliberately retained 30–40 % of rotor lift until the stopped two-blade stages are measured to be carrying. This is an intent, not a demonstrated result. Five effects remain open and are precisely what Phase 2 exists to measure: the magnitude and duration of the residual lift gap through the deceleration; biplane interference between the two stacked stages before merge, which degrades both; whether abort authority genuinely remains at every step; whirl flutter of the stopped, mast-mounted stages at 235–265 kt — the mechanism that destroyed the XV-3, and the single most under-analysed dynamic in this concept; and the 2/rev vibration signature of two two-blade stages through the deceleration, including the rpm band where a rigid coaxial head can couple with the airframe modes. Until flown, the merge must be treated as a flight-critical manoeuvre, not as a performance option.

Rotor dynamics not yet closed — stated rather than hidden:

• Whirl flutter of the stopped rotor (pylon pitch/yaw coupling with the mast-mounted stage mass at 235–265 kt). Requires a whirl-mode stability analysis and a mast-stiffness floor before Phase 2 — added to the risk register as HIGH.
• Ground and air resonance of a rigid coaxial head with no swashplate and no lag dampers in the classic sense: lag-mode damping is provided electrically by the IBC loop, which makes resonance stability a software failure case. Loss of the IBC damping law must be shown to be benign at every rpm.
• 2/rev forcing is inherent to two-blade stages; contra-rotation makes it 2/rev per stage plus a 4/rev beat. Airframe modes must be placed clear of both across the full N_r range, including the 70 % merge-entry condition.
• Inter-stage clearance: 0.62 m is 0.048 D, roughly half the 0.09–0.10 D of flown rigid coaxial practice. Clearance under the +3.5 g gust case is not demonstrated: mitigation is blade-tip proximity sensing, IBC flap suppression and a hard flapping placard, with a fallback to 0.95 m spacing (+18 kg of mast, +3 s of stroke) if the whirl-rig data does not close. A blade-to-blade contact is not a degraded state; it is a loss of aircraft. The decision is no longer left open — the rule that freezes it is stated now:
- – Measured on the Phase 0 whirl rig at 265 kt equivalent and 70 % N_r, with the +3.5 g gust case superimposed: worst-case tip approach = flapping excursion + coning + mast bending + build tolerance stack.
- – Pass: worst-case tip approach ≤ 0.31 m, i.e. half the installed 0.62 m retained as clearance, with IBC flap suppression inhibited during the measurement so the margin does not depend on a software function.
- – Fail: anything above it, or any dependence on the IBC loop to stay inside it, adopts the 0.95 m fallback immediately — the + 18 kg and + 3 s are booked and the mast drawing is re-issued.
- – Decision point: Phase 0 exit, frozen at the Phase 1 gate (§ 12). It is not permitted to travel further into the programme as an open item, because the mast, the collar stroke and the merge timeline all depend on it.

Deep-flush retraction — the geometry does not close yet. Every earlier issue described M4 as “the mast sinks into its dorsal cavity and the wing sits flush with the spine”, and never drew the section. Drawn, it clashes:

• There is no depth to sink into. Below the dorsal spine, on the same stations, sit the S-duct from the flush dorsal inlets, the engine, the aft spine fuel cell (1 000 kg) and, further forward, the 3.4 × 1.1 × 0.8 m ventral bay — inside an airframe only ≈ 4.1 m tall on its gear. A cavity deep enough to swallow the hub stack has to be taken out of one of those four volumes, and none of them is spare.
• The aperture is not a hole, it is a slot. The merged box is 1.55 m in chord and its root sweeps through 30° → 60°. The iris doors must therefore seal against a root that changes both its angle and its chordwise station — and, since M2 has already moved the upper stage 0.70 m aft, the root is no longer symmetric about the mast axis. A door set that seals at 60° does not seal at 30°.
• The swept panels arrive over the empennage. At 60° a tip sits 6.4 sin 60° ≈ 5.5 m aft of the mast axis at only 6.4 cos 60° ≈ 3.2 m of half-span — that is over the aft spine, in the neighbourhood of the canted V-tail roots. Lowering the whole wing by the retraction stroke on top of that is what turns a clearance question into a clash.

Resolution carried in this issue — the claim is reduced, not defended: “deep flush” is replaced by a semi-flush retraction. The stroke is limited to the pylon height, so the upper surface of the wing box comes level with the spine crown and the hub sits in a shallow saddle rather than below the mould line; the doors become a two-part sliding aperture that seals against the panel root at the three scheduled sweep detents, with the intermediate positions accepted as unsealed; and a fin-root station constraint is imposed so the 60° tip locus stays clear of the V-tail. The residual step drag and seam return are TBD by CFD and by RCS measurement in Phase 6, and the earlier claim of a fully flush spine is withdrawn. This is a geometry item, not a detail: it is the one place where the § 2 volume distribution and the § 3 mechanism are competing for the same cubic metres.

Failure solutions

A rotor fails to stop or index
Automatic abort to rotor mode in < 3 s: brakes release, PTO re-engages, both stages respin. The jet keeps the speed while the discs spool up.

Upper stage jams mid-stroke (either axis)
Spring-loaded return to the top datum on a pyrotechnic disconnect; the mast fairing is flyable at any intermediate hub position up to 265 kt. A jam on the chordwise axis is the worse case: the stage is retracted forward before the descent axis is reversed, because an eccentric hub cannot pass back up the mast.

Asymmetric Nitinol pin set (one side locked, one not)
Detected by differential pin strain; response is to re-raise the upper stage and respin to helicopter mode where possible. If the collar will not retract, the aircraft lands as a stage-M1 aeroplane at 170–190 kt — which requires a prepared strip of ≈ 1 200 m and is the declared exception to the no-runway requirement (§ 1). Bank limited to 30° until landing.

One half jams mid-scissor (asymmetric sweep)
The FBW instantly commands the healthy half back to the jammed angle — symmetry restored at the lower sweep — while the vectoring nozzle absorbs the transient yaw–roll couple. Envelope capped at the sweep achieved (M 0.9 at 30°, M 1.2 at 45°); the mission degrades, the wing still flies. Return, de-sweep on the electric or spring path, un-merge, land.

Mast fails to flush, or iris doors fail to seal
The hub column stays proud of the spine: a drag and RCS penalty, not a hazard — envelope capped at M 1.2. The iris petals are spring-loaded open, so the reverse transition is never blocked.

Nitinol edge fails to morph
The wing stays in hover profile: envelope capped at M 0.85 — the mission degrades to subsonic, the flight does not.

Total engine failure
In helicopter mode: autorotation. Merged above 3 000 ft: un-merge and autorotate, or glide — the lifting body returns ≈ 9:1 (a target, § 4). Below 3 000 ft in jet mode the lifting body flies as a high-speed glider to the nearest usable surface; if none exists, the crew ejects and the airframe is lost. The whole-airframe ballistic parachute of earlier issues has been deleted — no certified system exists above ≈ 2 t and its honest mass was 400–800 kg, none of it budgeted. Extracted personnel in the bay are covered by the downward-ejection provision of § 15.

What the merge actually buys

Numbers for one half-wing of exposed length 5.0 m (root cut-out 1.4 m, tip radius 6.4 m), blade chord 0.85 m, maximum thickness 90 mm, merge overlap 0.15 m, chord offset 0.70 m. Weights quoted at the mid-mission reference mass W_mid = 9 700 kg — defined as OEW + full payload + 55 % of usable fuel, i.e. the mass at the start of the supersonic leg (used consistently throughout this document).

Parameter	Stopped, unmerged (end of M1)	Merged (after M3)	Effect
Configuration	Two stacked thin wings, 0.62 m apart, chords aligned	One hyper-rigid 90 mm wing box — biconvex diamond section, upper stage offset 0.70 m aft	Biplane interference (≈ 15 % lift loss) eliminated
Chord	0.85 m	1.55 m	× 1.8 — and only because of the 0.70 m chord offset (0.85 + 0.85 − 0.15 of overlap). Without that degree of freedom the merged chord is 0.85 m and every line below changes
Thickness ratio t/c	10.6 %	5.8 %	Physical thickness never grows past 90 mm — thickening a Mach 1.5 wing would be aerodynamically fatal; only the chord grows
Volume wave drag ∝ (t/c)²	reference	≈ 0.30 ×	≈ 3.3× reduction — what makes Mach 1.5 affordable on one engine
Torsional stiffness GJ	reference	≈ 5 ×	Two open sections → one closed box; flutter speed × ≈ 2.2
Wing area (2 panels)	2 × 8.5 m² stacked	S_wing ≈ 15.6 m² clean	With S_body ≈ 12 m² → S_eff ≈ 27.6 m²: W/S_wing ≈ 622 kg/m² and W/S_eff ≈ 351 kg/m² at W_mid
Leading-edge sweep	0° — stopped dead straight, perpendicular to the fuselage	30–60° — halves scissor-swept symmetrically about the mast	Normal Mach at M1.5: 1.5 → ≈ 0.75 at 60°. The sweep does not “split” or “cut” the shock — it reduces the Mach component normal to the leading edge below 1, so the section itself never sees supersonic flow. 45° (M_⊥ ≈ 0.81) is what makes M 1.15 dry supercruise physically realistic
Aspect ratio — quoted with its sweep	AR ≈ 10.5 at 0° sweep (12.8 m span on 15.6 m²)	AR ≈ 2.6 at 60° (projected span 6.4 m) — an ordinary delta value	The “AR-10” figure quoted in earlier issues is only true unswept. The aeroelastic problem is not delta aspect ratio: it is the slenderness of each half-panel, 5.0 m exposed on a 1.55 m chord (≈ 3.2), cantilevered from a single pivot — that is the quantity the flutter statements in § 6, § 11 and the risk register now refer to
Leading edge	Blunt, erosion-protected (hover shape)	Sharp, r < 2 mm (Nitinol morphed)	Kept sharp for the transonic acceleration and for edge-aligned RCS, even though the 60° scissor sweep makes the normal flow subsonic at M1.5
Hub / mast drag	Faired pylon in the stream	Minimised — hub in a shallow dorsal saddle behind a sliding aperture (semi-flush)	The retraction deletes most of the parasitic drag and RCS of the rotor system. It is not zero and is no longer claimed to be: the airframe has no depth for a full-depth cavity (see the retraction-clash callout above), so a residual root step and seam remain, TBD by CFD and Phase 6 measurement
Radar signature	Rotating blades — blade flash	Static faceted surface	Jet mode removes the single biggest helicopter RCS contributor

Structure

The two stages are not glued: after the chord offset, the upper stage's forward spar and the lower stage's aft spar close onto a common shear web through a row of steel shear keys, drawn up and held by the Nitinol locking-pin row — the pins contract ≈ 4 % on electrical heating and supply clamping pre-load only, latched in the contracted position by an over-centre detent so that clamp load does not depend on holding the alloy hot. Flight loads are carried by the shear keys and the shear web, never by the pins themselves; a relaxed pin therefore loses joint stiffness, not load path. Bending, torsion and shear share one path; the hollow titanium spars double as fuel-sink cooling galleries (§ 5), and the mast — no longer turning — becomes the scissor pivot and carry-through fitting.

Aerodynamics

The merged section is a biconvex diamond profile: Nitinol-sharpened LE, maximum thickness at 45 % chord, no trailing-edge surfaces at all — roll is generated by twisting the whole box (§ Smart Hub). The stages stop dead straight at 0° of sweep; the scissor pivot then folds both halves symmetrically aft to the scheduled 30–60° — conceptually the variable-sweep stroke the F-14 flew past 60° for decades, cited as inspiration only: here it is collapsed onto a single central mast bearing, which is a materially different and unproven load case.

The honest cost

Two hubs, two brakes, two indexers, an L-stroke collar with its chordwise rails, a Nitinol pin row, twin scissor actuators, the incidence gimbal, the flush-retraction jack and the iris doors: 740 kg, ≈ 7 % of MTOW that a conventional helicopter or a fixed-wing fighter does not carry — the Nitinol pins delete the merge hydraulics, the swashplateless hubs delete the entire pitch-link and boost-actuator train, and the deleted tail rotor, ailerons and rudder actuators buy more of it back. The residue is paid out of payload and hover margin, permanently.

Rotor tilting — from the swashplate to IBC

Before explaining what the HSR does, it is worth stating plainly what a conventional helicopter does, because the two are routinely confused. On a classic helicopter nothing pivots mechanically to make the aircraft go forward. The mast stays where it is; what tilts is the rotor disc — the aerodynamic plane traced by the blade tips — and it tilts because the swashplate applies cyclic pitch: a once-per-revolution variation of each blade's incidence, so that a blade produces more lift on one side of the azimuth than on the other. The disc plane, and with it the rotor's lift vector, leans in the commanded direction; the horizontal component of that vector is the propulsive force. Collective pitch — the same incidence change applied to every blade at once — sets how much lift there is; cyclic pitch sets where it points. This is the mechanism the HSR cannot use.

Why the swashplate is excluded here

• A swashplate is a rotating-to-stationary mechanical bridge wrapped around the mast: a stationary ring, a rotating ring, pitch links, scissors and a slider. It is not a component that can be miniaturised out of the way — it occupies the mast.
• The upper stage must travel 0.62 m down that same mast during M2 (and then 0.70 m aft in chord). Every millimetre of that path is where a swashplate's linkage lives. A swashplated HSR would jam on its first merge.
• There is no partial solution: a shortened, relocated or split swashplate still has to cross the rotating interface somewhere on the column the upper hub descends. The mechanism has to be deleted, not moved.

What replaces it — swashplateless IBC

• Four electromagnetic actuators (EMA), one per blade, buried inside the two hub bodies and acting directly on each blade root spindle. Collective, cyclic and higher-harmonic inputs become different current commands to the same four units.
• Individual Blade Control up to 5/rev — per blade, per azimuth. The disc still tilts exactly as it does on any helicopter; what has changed is that the once-per-revolution pitch variation is generated electrically inside the hub rather than mechanically outside it.
• Only current and photons cross the rotating joint: a slip-ring / rotary-transformer stack at the hub base carries 270 VDC power and a dual-redundant fibre data ring. Nothing mechanical crosses it, so the mast between the two hubs is a clean cylinder.
• The cost is stated rather than hidden: no mechanical fallback exists (a swashplate would block the merge), which is why the IBC hub is one of the pacing items of § 5 and a HIGH risk in the register.

Terminology — three different things that are all called “tilting”:

• (a) Rotor-disc tilt (classic helicopter). The disc tilts under cyclic pitch; the mast, the hub and the nacelle do not move. This is how every conventional helicopter translates, and the HSR does exactly this in helicopter mode — just without a swashplate.
• (b) Tilting nacelles (tiltrotor). On a V-22 Osprey or an AW609 the whole nacelle and rotor assembly rotates through ≈ 90° on a wing-tip pivot, converting the rotor into a propeller in front of a permanent fixed wing. The HSR has no tilting nacelle, no wing-tip pivot and no permanent fixed wing, and is therefore not a tiltrotor — the comparison in § 17's alternatives trade study is between architectures, not a family resemblance.
• (c) The HSR choice — nothing tilts in flight. The rotor stops dead straight at 90° to the fuselage, merges into a single wing box, and then sweeps: the two halves scissor aft about the mast on the 30° / 45° / 60° schedule. The mast pivots the wing in sweep, in the horizontal plane, after the rotor has already ceased to be a rotor. At no point does a lifting system rotate from a vertical-thrust orientation to a horizontal-thrust orientation.

Architecture	What physically moves to change flight regime	Lift in cruise	Consequence for this specification
Classic helicopter	Nothing structural — the rotor disc tilts under cyclic pitch from the swashplate	The rotor, always	Advancing-blade compressibility caps forward speed. The HSR keeps this mechanism for hover and low speed, but implements it with IBC because the swashplate blocks the merge path
Tiltrotor (V-22, AW609 — general reference)	The nacelles rotate ≈ 90° on wing-tip pivots; the rotors become propellers	A permanent fixed wing	Excluded by requirement: the specification forbids permanent lateral fixed wings, and a propeller-driven configuration cannot reach the Mach 1.5 design point
HSR	The rotor stops (M1), merges into one wing box (M2–M3), then the two halves scissor-sweep aft 30° → 45° → 60° (M4)	The merged wing plus the lifting body (S_body ≈ 12 m², TBD by CFD)	Buys a supersonic wing out of a hover-sized rotor — and pays for it with the 740 kg mechanism of § 9 and the flight-critical merge of this section

Swashplateless Individual Blade Control — why the upper hub can descend at all

A conventional rotor head cannot perform this manoeuvre, and the reason is geometric rather than aerodynamic: a swashplate is a rotating-to-stationary mechanical bridge wrapped around the mast, and its pitch links, scissors and slider physically occupy exactly the 0.62 m of mast travel the upper stage must pass through. A swashplated HSR would jam on its first merge. The rotor system is therefore swashplateless throughout:

Individual Blade Control (IBC) by electromagnetic actuation

• Actuation inside the hub: each of the four blades is pitched by its own electromagnetic actuator (EMA) buried in the hub body, acting directly on the blade root spindle — collective, cyclic and higher-harmonic inputs are simply different current commands to the same four units.
• Power and data across the rotating joint: a slip-ring / rotary-transformer stack at the hub base carries 270 VDC power and a dual-redundant fibre data ring to the actuators. Nothing mechanical crosses the rotating interface — the only things that cross are current and photons.
• Zero linkage in the merge path: no swashplate, no pitch links, no rotating or stationary scissors, no boost actuators, no hydraulic swivel. The mast between the two hubs is a clean cylinder, and the upper stage's 0.62 m descent is mechanically unimpeded from first millimetre to last; the chordwise rails live in the stopped mast head, above the rotating interface, and cross nothing.
• Blade-by-blade authority: IBC gives per-blade, per-azimuth control at up to 5/rev — which is what makes the −10° active washout, the flat-untwist schedule and the ± 0.2° stop-straight indexing achievable at all, none of which a mechanically ganged head could deliver. The consolidated rotor figures — tip speed, rotor speed, solidity, advancing-blade Mach — are tabulated in § 6.

What deleting the swashplate buys elsewhere

• Hub RCS collapses: the classic helicopter radar signature is not the blades but the hub scatterer — a rotating forest of links, horns and bearings acting as a dense corner-reflector array. A smooth, closed, RAM-coated EMA hub is a single faceted body, and in jet mode it retracts into its saddle and largely disappears — largely, because the semi-flush retraction above leaves a root step and a door seam.
• Mass and maintenance: pitch links, scissors, sliders, boost actuators and their hydraulic runs are simply not built — booked as a credit in § 9 — and with them go the highest-maintenance, shortest-life items on any rotor head.
• No rotating hydraulics: the merge already deletes the locking hydraulics via the Nitinol pins; IBC deletes the last hydraulic circuit that would have had to cross a rotating joint. The rotor system is electrically actuated end to end.
• The electrical consequence: four EMA channels, the Nitinol pin row and the torsion actuators are an all-electric actuation suite with a violently transient load profile. Their honest peak demand is ≈ 300–500 kW, not a megawatt — the megawatt ISG of § 4 is sized by engine start, growth and high-power mission systems, not by actuation. What actuation drives is the power requirement (the supercapacitor bank), not the energy requirement (the generator).

The bill for an all-electric hub: electromagnetic interference — previously unaddressed. Deleting the swashplate does not delete the physics; it converts a mechanical problem into an electromagnetic one, and this document had not said so. What now sits at the geometric centre of the aircraft is four switching motor drives on a ± 270 VDC bus, rotating at ≈ 313 rpm, feeding through a sliding electrical contact — immediately beneath the dorsal spine, which is exactly where § 7 laminates conformal AESA modules, and immediately beside the routing of a quadruplex flight-control system. Three consequences follow, and all three are qualification items:

• Broadband conducted and radiated emission. Switching inverters produce steep dV/dt on long feeders; a brushed slip ring transferring hundreds of amperes adds micro-arcing, which is broadband by nature and worst exactly when the contact is worn — i.e. late in the maintenance interval, not on the qualification bench. Baseline is therefore a brushless rotary transformer for the power path, with the brushed ring retained only as a low-current backup; the mass and efficiency penalty of that choice is accepted rather than argued away.
• Victim side: the AESA and the FBW. A conformal array in the spine is a sensitive receiver a metre from the source. Mitigation is architectural, not filtration: the hub body is closed as a Faraday enclosure with a 360°-terminated shielded feeder; the IBC drive return is a dedicated conductor, never the composite structure (a composite airframe has no usable ground plane); common-mode chokes sit at the rotating interface; and all data — already — crosses on fibre.
• Signature side: unintentional RF emission. For a low-observable aircraft an arcing rotating contact under a RAM cover is not only an EMC nuisance, it is a beacon. Hub emission is therefore treated as part of the signature specification and measured in § 12 Phase 6 alongside RCS, not only in the EMC campaign.

Means of compliance: MIL-STD-461 CE/RE limits and MIL-STD-464 system-level margins (§ 14), demonstrated with the hub turning and the EMAs commanded, on the Phase 1 whirl rig and again on the Phase 3 aircraft. A static bench result is explicitly not acceptable evidence for this item, because the dominant emission mechanism only exists in rotation.

The Smart Articulated Hub — three active adjustments, one buried mechanism

The merged wing is never thickened and never grows a control surface — both would be fatal at Mach 1.5. Instead, everything the wing must do after the merge is done by the smart articulated hub at the top of the titanium mast, working invisibly beneath the closed iris doors:

A · Variable incidence

The mast head tilts the whole wing box a few degrees fore and aft, setting the incidence the wing needs to lift — while the fuselage stays at its 0° angle-of-attack datum, level. A level fuselage keeps every facet at its designed radar aspect and the airframe at its absolute-minimum-drag attitude: the wing works, the body never has to.

B · Continuous dynamic sweep

The scissor sweep is not a fixed setting but an actively scheduled variable: 30° out of the transition for maximum subsonic efficiency, 45° for the M 1.15 dry supercruise (normal Mach ≈ 0.81), 60° for the Mach 1.5 dash (normal Mach ≈ 0.75). The FBW moves the two halves symmetrically and continuously along that schedule as Mach changes — the wing is always at the sweep the flow wants. Sweeping aft also drags the aerodynamic centre aft, so the sweep schedule is slaved to the active CG fuel transfer below: geometry and mass move together, in one commanded law.

C · Active aeroelastic wing

There are no trailing-edge ailerons — heavy, gap-producing and flutter-prone on a 90 mm box. The same root torsion actuators serve two regimes. For roll in jet mode, the FBW twists the entire merged wing box elastically (≈ ±0.8° at the tip): differential lift by aeroelastic torsion, the X-53 Active Aeroelastic Wing technique — turning the slender wing's flexibility from a structural threat into the roll effector itself. For hover, the same actuators solve the blade-twist paradox described below.

Solving the blade-twist paradox — active washout

A rotor blade and a supersonic wing want opposite geometries, and no passive structure can be both:

• Hover wants twist. Local blade speed grows linearly with radius, so an untwisted blade over-loads its tip and starves its root — hover figure of merit collapses. Every helicopter blade therefore carries built-in washout.
• Mach 1.5 wants flat. A twisted supersonic panel presents a spanwise-varying incidence to a shocked flow: premature shock-induced separation at the tip, a wandering shock foot, and a torsional divergence risk on a 90 mm box.
• The resolution is active, not built-in. The aeroelastic torsion actuators hold a commanded −10° of washout root-to-tip throughout helicopter mode — full hover efficiency — then dynamically untwist the blades to a flat 0° datum during the Stop-and-Merge sequence, so the panels that the Nitinol pins clamp together are already geometrically flat when the box closes. Pure, unsheared supersonic flow, with no permanent structural compromise at either end.
• Failure state: loss of torsion authority freezes the blades at their last commanded twist — merge is inhibited above ±2° residual, and the aircraft returns as a helicopter.

Active CG management — fuel and the mechanism's own mass

Sweeping a wing aft moves its aerodynamic centre aft — violently, on a planform this slender. It also moves mass, which earlier issues denied:

• The problem: going from 30° to 60° of scissor sweep marches the aerodynamic centre roughly 0.9 m aft — on a 1.55 m MAC that is 58 % MAC. Left alone against a fixed CG, that is a large nose-down pitching moment that the nozzle would have to hold out permanently — paying continuous trim drag and eating vectoring authority the aircraft needs for manoeuvre.
• What fuel transfer can actually deliver: the ledger has to be shown, because earlier issues of this document promised ± 2 % MAC and “near-zero trim deflection” without ever quoting a lever arm. Tank centroids are 7.0 m apart (forward ventral at station 4.9 m, aft spine at 11.9 m). Moving 1 000 kg at W_mid = 9 700 kg shifts the CG by 1 000 × 7.0 / 9 700 = 0.72 m. The aft spine cell is therefore burned first on the outbound leg specifically to be empty and available when the sweep schedule starts.
• The mechanism moves mass too — and it helps. The four blades weigh ≈ 600 kg with their centroids ≈ 3.9 m from the mast axis (spanning 1.4–6.4 m). Sweeping both halves aft from 30° to 60° moves each blade centroid aft by 3.9 (sin 60° − sin 30°) ≈ 1.43 m, i.e. + 0.09 m of CG travel at W_mid, in the right direction and automatically in phase with the sweep. Earlier issues asserted that “sweeping does not move the mass, only the lift”; that was simply wrong, and the term is now credited.
• The residual, restated: 0.90 m of aerodynamic-centre travel against 0.72 m of fuel travel plus 0.09 m of blade-mass migration leaves ≈ 0.09 m ≈ 6 % MAC that nothing chases. That residual is trimmed by the nozzle: 95 kN of lift on a 0.09 m arm is 8.6 kN·m, which on a 6.5 m tail arm is a 1.3 kN download — ≈ 1.5° of vector deflection, ≈ 7 % of the ± 20° authority, and ≈ 0.2 kN of trim drag (≈ 1.2 % of dash drag). That is the honest price; it is paid, not hidden.
• The one that was genuinely forgotten — the M2 mass step. During the merge stroke the upper stage (two blades plus the upper hub, ≈ 520 kg) translates 0.70 m aft in chord. That is 520 × 0.70 / 9 700 = ≈ 0.038 m ≈ 2.4 % MAC of aft CG shift, arriving as a step in ≈ 7 s, in the middle of the most flight-critical manoeuvre the aircraft performs — and no earlier issue accounted for it anywhere. It is small in magnitude and badly placed in time: it lands while the rotors are unloading and the effectors are being handed over. The MERGE law therefore pre-biases the transfer pumps forward before the chordwise leg begins and feeds the residual as a scheduled nozzle input, so the step is cancelled by an open-loop command rather than caught by the pitch loop. The same stroke also lowers the CG by ≈ 0.033 m as the upper stage descends 0.62 m — negligible in pitch, but it is in the ledger now rather than absent from it.
• The payoff: static margin held inside ± 3 % MAC out to 45° of sweep — where the aircraft spends its supercruise — and a bounded, quantified 6 % MAC nose-down trim at the 60° dash condition, rather than the unbounded moment a fixed CG would produce. The fuselage stays at its 0° AoA datum and ≈ 93 % of vector authority stays available for manoeuvre.
• Interlock: sweep rate is limited by transfer-pump flow (20 kg/s → 50 s for the full stroke) — the wing is never allowed to out-run the fuel. A pump failure caps sweep at the last balanced angle (typically 45°, M 1.15), which is a degraded mission, not a departure.

Where the idea is weakest — and why the target is Mach 1.5, not more: hover sizes the rotor at Ø 12.8 m, and a 12.8 m rotor forces a 12.8 m span — an unswept aspect ratio of ≈ 10.5, three times a purpose-built supersonic delta. Left dead straight at 0° of sweep, that long thin wing would meet the shock head-on and generate a wave drag no single engine could push through: the 60° scissor sweep is what saves the equation, cutting the normal Mach at M1.5 from 1.5 to ≈ 0.75 and the projected aspect ratio to ≈ 2.6 — an ordinary delta — at a wave drag 52 kN of dry thrust can actually sustain. What the sweep cannot fix is aeroelasticity: each half-panel is still 5.0 m of exposed span on a 1.55 m chord, cantilevered from one pivot, and it is that slenderness — not the delta's aspect ratio — whose bending, torsion and pivot bearing loads cap the design at Mach 1.5. At Mach 2 they would not close, even with the twist law exploiting the flexibility for roll. The aerodynamic compromise is balanced for Mach 1.5, which is precisely what the specification asks and no more.

4. Powerplant (Engine)

The aircraft is propelled by a single engine located at the rear, handling both the rotation of the mast (through a power take-off) and the horizontal thrust. The engine is an adaptive-cycle engine (ACE): a three-stream turbofan that changes its own bypass ratio in flight — wide open and cool for hover and loiter, closed down to a low-bypass supersonic cycle for the Mach 1.5 dash. One engine, one fuel system, two jobs.

Rear jet thrust only — no propeller or rotor aft of the mast. All horizontal thrust comes from the rear adaptive-cycle engine, which is a three-stream adaptive turbofan (not a pure turbojet) exhausting through a single 2-D nozzle. There is no tail rotor, no fenestron, and no rear propeller anywhere on this aircraft. The aft fuselage tapers into a flat diamond spine (YF-23 style) carrying only two fixed canted stealth fins. Control authority is regime-dependent, and the distinction is absolute:

• Helicopter mode — pitch and roll come from rotor cyclic pitch on the coaxial stages; yaw comes from differential coaxial torque (splitting rpm/collective between the contra-rotating stages), which is also why no anti-torque device is needed at all. The nozzle only trims residual thrust.
• Jet mode — the rotors are stopped, merged and swept, so cyclic and differential torque no longer exist: pitch and yaw are provided exclusively by the 2-D low-observable vectoring nozzle of the rear ACE turbojet, with roll by aeroelastic wing twist. No moving tail surface, no hinge gap, no actuator bulge ever breaks the mould line.
• Engine-out in jet mode — a vectoring nozzle with no gas through it is a dead effector, so the aircraft must not rely on it alone: authority is designed to revert within 150 ms to the seamless morphing ruddervators on the canted V-tail (§ 2, § 7). This reversion is a design requirement, not a demonstrated capability — it is the object of the deliberate nozzle-kill trials of § 12 Phase 5, and until then the 9:1 dead-stick glide is a target figure, with crew ejection as the declared fallback — there is no whole-airframe parachute (§ 15).

The MERGE law owns the hand-over between the two effector sets — authority is blended, never dropped.

Mode	Power take-off (mast)	Nozzle thrust
Hover / VTOL	Clutch closed — up to 3 800 kW to the coaxial gearbox	Idle, trimmed by the 2-D nozzle
Helicopter cruise (0–200 kt)	2 000–2 800 kW	Low — supplements the tilted disc above 120 kt
Merge window (235–265 kt, SL – 2.5 km)	Unloading to a 30–40 % residual through M1, then to zero; clutch opens at M4	Rising to full dry — carries the aircraft through the merge
Supercruise (M 0.5–1.2)	0 — PTO declutched, mast locked	52 kN dry, third stream narrowing
Dash (M 1.5)	0	75 kN with reheat — ≈ 8 min within the fuel budget

Why an adaptive-cycle engine — and why only one

A fixed-cycle engine is wrong at one end of this mission or the other: a helicopter turboshaft cannot push an airframe to Mach 1.5, and a fighter turbofan wastes fuel and blows hot, loud exhaust in hover. The ACE's variable third stream resolves the conflict in one machine: wide open, it behaves like a high-bypass engine — cool, quiet, efficient at shaft work through the PTO; closed, it becomes a low-bypass supersonic engine. That duality is what makes the single-engine requirement viable at all — and its 2-D vectoring nozzle earns its mass three times over: it trims the abrupt centre-of-lift transfer during the merge, it is the sole pitch/yaw effector in jet mode, and its flat, shielded aperture is the IR-signature answer at the tail.

Sizing sanity check (M1.5 at 12 km) — preliminary

• a ≈ 295 m/s → V ≈ 443 m/s ≈ 1 590 km/h; ρ ≈ 0.31 kg/m³ → q ≈ 30 kPa.
• W_mid = 9 700 kg on S_eff ≈ 27.6 m² → C_L ≈ 0.115 — provided the S_body ≈ 12 m² assumption survives CFD.
• L/D at the design point is now debited, not targeted: the clean build-up gives ≈ 6.6, and the semi-flush saddle, the iris seams and the chine junctions bring it to 5.3–5.7 — see the drag build-up in § 6. The old 5.5–6.5 band survives only at its lower end, and the 6.5 objective is withdrawn.
• Reheat budget 900 kg of fuel → ≈ 8 min at M1.5, against 2 450 kg of total internal fuel.

Hover power build-up — required shaft power, not “ideal”

Earlier issues quoted “≈ 3 600 kW ideal”, which conflated two different quantities. The ideal induced power is far lower than that; the required shaft power is what the PTO must deliver. Both are shown. Condition: MTOW 10 800 kg, ISA+20 at 4 000 ft (ρ ≈ 1.015 kg/m³), Ø 12.8 m (A = 128.7 m²), tip speed 210 m/s, solidity σ ≈ 0.169.

• Ideal induced power, sea level ISA: P_i = T^1.5/√(2ρA) = 1 940 kW. At 4 000 ft ISA+20: 2 130 kW. This is the number “ideal” refers to — it is not a rating.
• × κ = 1.16 for non-uniform inflow and coaxial interference → 2 470 kW.
• + profile power ρA(ΩR)³σC_d0/8 at C_d0 ≈ 0.010 → + 256 kW = 2 730 kW.
• + 3 % transmission losses, + 4 % vertical drag / download → 2 910 kW.
• + 8 % for hot-gas re-ingestion on an unprepared pad and hover control margin → required shaft power ≈ 3 150 kW (equivalent figure of merit ≈ 0.73).
• PTO rated 3 800 kW → ≈ 21 % margin at the hot-and-high hover point. The rating is set by this build-up, not by a round number.

Integration problems specific to this airframe

• PTO gearbox — HIGH risk, unresolved: 3 800 kW through a 90° bevel set and a dog clutch that must also open cleanly in 2 s during the merge. The 2 s disengage is a requirement, not a demonstrated capability; this remains the single most stressed mechanical item on the aircraft and does not “close” at preliminary level.
• ISG integration — HIGH risk: embedding a megawatt machine in the HP spool alongside a 3 800 kW power take-off is an unprecedented combination of shaft offtakes for an engine of this size. Rotordynamics and thermal margins are open items for the ACE partner.
• Hover breathing: flush inlets have no ram feed at 0 kt; spring-loaded auxiliary doors around the duct shoulder do the feeding, MiG-21/Concorde style. Door area is TBD — the ≈ 0.25 m² carried in the current mass and drag model is a placeholder pending inlet CFD and hover rig test.
• Downwash ingestion: the dorsal lips sit above the debris fountain, but re-ingestion of hot exhaust in ground effect is expected to cap hover at ISA+25 — accepted, and to be measured.
• One engine = one failure case: in helicopter mode the intended answer is autorotation; merged, the priority order is glide first, un-merge-and-autorotate second (available only with altitude and time), crew ejection third. None of the three is yet a trained, demonstrated procedure.
• Engine availability is a programme risk, not a technical one: no engine in the world combines 52/75 kN with a 3 800 kW power take-off and a megawatt ISG. The ACE is a second mega-programme sitting inside this one — see § 17.

The power take-off — three candidate architectures, none demonstrated

An external review observed correctly that this document had specified a 3 800 kW PTO without ever describing how it is taken. Naming the architectures does not create the engine, but it converts an assumption into a bounded trade that an engine partner can actually be asked to price. All three are stated with their consequence, and none of them exists in hardware at this rating.

Architecture	How the 3 800 kW is taken	Consequence for this aircraft	Verdict at preliminary level
A · Front bevel offtake from the LP spool (current baseline)	A 90° bevel set on the LP shaft ahead of the fan, through a dog clutch, into the coaxial gearbox	Shortest, lightest shaft path and the one already costed in the 2 150 kg propulsion line. But it loads the LP spool with a torque comparable to its own fan work, and the 2 s clutch disengage during the merge is the most stressed mechanical event on the aircraft	Baseline — lowest mass, highest single-item risk. Requires LP rotordynamic clearance the partner has not given
B · Dedicated free power turbine in the third stream	A separate power turbine spun by third-stream flow, mechanically independent of the propulsive spools	Decouples shaft work from thrust — the clutch problem largely disappears and hover and jet modes stop competing for the same turbine. Costs a turbine stage, its casing and its own gearbox: an estimated + 150–250 kg, and it consumes the third-stream air the nozzle and the SMA cooling loop also want	The most credible route to a real product, and the one an engine house is most likely to quote. Must be traded in Phase 0, not assumed away
C · Electrical PTO — no mechanical shaft to the rotor	Generation on the spool, distribution on the ± 270 VDC bus, electric motors driving the coaxial gearbox	Deletes the bevel set, the dog clutch and the 90° drive entirely, and makes the merge disengage a software event. But 3 800 kW of continuous shaft work is 3.5× the 1.1 MW ISG already baselined: generator, cabling, converters and their cooling would add several hundred kilograms and the whole hover case would then depend on power electronics	Rejected at this rating, and stated so explicitly — it is the architecture people assume is available and it is not. Retained only as a growth path if machine specific power improves materially

The honest position: the baseline is architecture A because it is the lightest, not because it is the most likely to be delivered. Architecture B is the one this programme should expect to buy, and the Phase 0 trade must be run with the mass penalty of B carried in the mass statement as a declared contingency rather than discovered at contract signature. Until a partner commits to one of the three, the PTO is a requirement written against a machine that has not been designed.

Fuel for the engine that turns the rotor

The question is asked often enough about rotorcraft to be worth answering explicitly in a specification of this kind. A helicopter turbine engine burns kerosene, never petrol. The standard is Jet A / Jet A-1, whose military equivalents are the NATO F-34 (JP-8 class) and F-44 (JP-5 class, high flash point) grades — the same kerosene with a defined additive package (anti-icing, static dissipator, corrosion inhibitor and lubricity improver). Avgas 100LL appears only on the small piston helicopters of the Robinson R22 / R44 class, which are not turbine aircraft at all. These are general reference values for the class, quoted here to situate the HSR rather than derived from it.

Why kerosene and not petrol — general reference

• Energy density. Kerosene carries more energy per unit volume than avgas, which is what buys endurance out of a fixed internal tankage — and on this airframe the tankage is fixed by requirement, since no external tank is permitted.
• Flash point, therefore safety. Kerosene's high flash point is the reason it is the only acceptable fuel in a shipboard or confined-deck environment; F-44 exists precisely for that case (§ 16).
• Military logistics. A single-fuel-for-the-battlefield policy puts F-34 in the same bowser as the ground fleet. An aircraft that needed avgas would need its own supply chain to every unprepared pad it is designed to operate from.
• Alternate-fuel tolerance. Turbines will run on emergency alternates where a piston engine will not — at the cost of hot-section life. That is a dispatch option, not a planning assumption.
• Specific consumption, general reference: a turboshaft of this class sits at 0.28–0.35 kg/kW·h. Applied to the hover build-up of this section it is what makes hover, not the dash, the dominant fuel consumer of a VTOL sortie.

What that means on the HSR specifically

• One engine, two fuel-burning natures, one fuel. The ACE runs as a turboshaft in helicopter mode, delivering up to 3 800 kW through the PTO, and as an adaptive turbofan in jet mode delivering 52 kN dry / 75 kN reheat. One fuel system serves both; there is no second fluid anywhere on the aircraft.
• 2 450 kg internal, in two cells: forward ventral 1 450 kg and aft spine 1 000 kg, centroids 7.0 m apart. The split is not a packaging accident — it is the active CG transfer loop of § 3 and § 9, and the aft cell is burned first outbound so that it is empty and available when the sweep schedule starts.
• Fuel is the aircraft's only heat sink. The leading-edge purge of § 5, the ± 270 VDC converters and ISG stator of the electrical section below, and the transmission oil all reject into the feed fuel, which is then burned. No ram-air scoop ever breaks the mould line.
• Low-freeze-point specification for arctic operations (−40 °C ground start, § 16) — which is a fuel-specification requirement, not a system one, and is therefore stated here.
• No external tanks, ever (stealth requirement, § 1). That is exactly why the air-to-air refuelling probe of § 16 is baselined rather than deferred: with 2 450 kg internal and ≈ 560 km of combat radius, the fuel state is the mission radius.

Electrical architecture & power generation — a transient problem before a megawatt one

This airframe has no hydraulic muscle left to speak of: the swashplate is gone, the merge latches are shape-memory, the ailerons are aeroelastic and the tail is compliant. Everything that used to be pressurised oil is now current — and the loads are neither small nor smooth. Four EMA blade-pitch channels, a 40 kW Nitinol pin-and-edge heating spike held for 8 s, the torsion actuators, the scissor and flush drives, the CG transfer pumps and a quadruplex FBW together peak at ≈ 300–500 kW, in short, violent, non-coincident bursts. That is a power-quality problem, not a megawatt problem, and it is answered by the supercapacitor bank. The 1.1 MW ISG is sized by something else entirely: engine start and windmill relight, the mission-systems and EW load of § 13, thermal-management pumping, and the deliberate directed-energy growth provision of § 11. Earlier issues justified the megawatt machine by the actuation suite; that was over-sold and is corrected here.

Element	Rating / architecture	Engineering rationale
Integral Starter-Generator (ISG)	Megawatt-class, shaft-embedded in the ACE HP spool — rotor and stator integrated into the engine core rather than hung on an accessory gearbox; 1.1 MW continuous, 1.4 MW for 60 s	An accessory-gearbox generator of this rating would not fit inside a stealth tail, and its drive train would be another single-point failure beside the PTO. Embedding it in the spool also makes it the engine's own starter and windmill-relight motor — one machine, three jobs.
Primary distribution	± 270 VDC main bus, solid-state contactors, two independent channels routed on opposite sides of the keel	High-voltage DC is what makes megawatt-class distribution possible at acceptable feeder mass; physical separation means no single battle-damage event takes both channels.
Graphene supercapacitor bank	Buffer store, ≈ 12 MJ usable at ≈ 12 Wh/kg cell level → ≈ 280 kg installed, charge/discharge at > 500 kW, sited on the CG beside the bay; graphene-electrode cells for power density rather than energy density	The transient answer. The Nitinol phase draws ≈ 40 kW for 8 s and the four EMA channels draw hard, non-linear peaks; taking those straight off the generator would sag the bus. The bank absorbs and re-emits the spikes, so the ISG sees a smoothed load and the transients never reach the rest of the aircraft. Mass corrected: earlier issues implied ≈ 30 Wh/kg, which is double any commercial supercapacitor; at a realistic 10–15 Wh/kg the bank is a 280 kg item and is booked as such in § 9.
Flight-critical bus isolation	FBW, inceptors, IBC actuator drives and the morphing-ruddervator amplifiers sit on a galvanically isolated, supercapacitor-backed essential bus with its own converters	Non-negotiable: no Nitinol heating event, no gun-drum inrush and no pump start may ever be visible as a voltage excursion on a flight-control channel. Isolation is by architecture, not by filtering.
Emergency power	Windmilling ISG is the primary dead-stick source; the bank is the bridge. Split, explicitly: bank alone holds the ≈ 15 kW essential bus for ≈ 13 min; the windmilling ISG supplies ≥ 25 kW above 150 kt, which covers the whole glide; the bank alone covers the final approach below windmill threshold and any restart attempt. No ram-air turbine.	A deployed RAT would be an aperture and a drag/RCS event exactly when the aircraft is gliding for its life. Earlier issues quoted “≈ 4 min of essential bus” against a 15–20 min glide from 12 km — the two did not meet. With the corrected 12 MJ / 280 kg bank and the windmill contribution stated separately, the morphing ruddervators are powered continuously from flame-out to touchdown, which is the actual requirement.
Thermal rejection	Converters, ISG stator and bank all reject into the feed fuel through the existing fuel-sink loop	Same principle as the Nitinol purge (§ 5): the fuel is the heat sink, so megawatt power electronics never force a ram-air scoop through the mould line.

Why this closes the concept electrically: every exotic mechanism in this document — swashplateless IBC hubs, Nitinol locking pins and morphing edges, aeroelastic twist, morphing ruddervators, scissor and flush drives — was chosen because it deletes mechanical complexity, and each one converts that complexity into an electrical transient. The megawatt ISG supplies the energy; the graphene supercapacitor bank supplies the power. Separating those two jobs is what lets a 10 800 kg aircraft run a near-all-electric actuation suite without ever putting a ripple on the flight-control bus.

5. Materials & Thermal Management

At Mach 1.5 the airframe is warm, not incandescent: recovery temperature at 12 km is ≈ 45 °C, rising to ≈ 120–130 °C only in sustained low-altitude supersonic flight. That is what makes this aircraft buildable out of fighter-standard composites instead of exotic hot structure. It also sets a hard rule for the shape-memory alloys: every transformation temperature in this aircraft must sit above the maximum soak temperature it can ever see, with margin. Earlier issues of this document placed the leading-edge A_f at 85 °C, below the ≈ 104 °C recovery temperature of M 1.5 at 3 km — so “commanded, never accidental” was true at 12 km and false lower down. The alloys are re-specified below.

Zone	Material	Thermal condition	Why
Blade / wing box	IM carbon-epoxy torque box, hollow Ti-6Al-4V spars doubling as fuel-sink cooling galleries, Nitinol pin seats	≤ 130 °C worst case — epoxy limit 180 °C holds; spars fuel-purged before un-merge	Each blade is both a rotor blade (centrifugal + fatigue) and half a supersonic wing (bending + torsion); the titanium root carries both load cases into the mast, and its hollow bore is the thermodynamic exit route for the Nitinol's heat.
Leading edges	NiTiHf high-temperature shape-memory alloy (HTSMA) skin over a flexcore, resistively heated, fuel-sink cooled	A_s ≈ 150 °C / A_f ≈ 175 °C (sharp) · M_s ≈ 120 °C / M_f ≈ 95 °C (blunt)	Two trained shapes: blunt erosion-tolerant radius for rotor mode, sharp (r < 2 mm) profile for supersonic flow. A_s is set 20 °C above the 130 °C maximum soak case, so kinetic heating cannot trigger the morph anywhere in the envelope — including M 1.5 at 3 km, where recovery temperature reaches ≈ 104 °C. Morphing takes ≈ 8 s per edge during M3; reversion is driven by the fuel-sink loop, not by waiting on the airflow. Baseline NiTi cannot reach these temperatures: NiTiHf/NiTiPd class alloys (NASA HTSMA) are required, and that raises the Phase 0 alloy risk rather than lowering it.
Fuselage skins	Carbon-BMI facets over aluminium-lithium frames, RAM outer layer	≤ 120 °C at the nose in the dash	BMI's 230 °C margin keeps the RAM coating (the real thermal weak point, ≈ 170 °C) inside its envelope.
Mast, hubs, locking pins	Ti-6Al-4V mast, scissor-pivot bearing and collar, steel dog clutch, steel shear keys and Nitinol shape-memory locking pins in René 41 seats	Gearbox oil 140 °C, brake packs 400 °C transient	The rotor brakes absorb ≈ 3–7 MJ per stop — re-derived, not carried over. Rotor-system inertia is I ≈ 10 400 kg·m² (4 blades of ≈ 150 kg on a 1.4–6.4 m span, plus hubs, mast and driveline); at 100 % N_r (Ω ≈ 32.8 rad/s) that is ≈ 6.5 MJ, and at the 70 % N_r merge-entry condition ≈ 3.2 MJ. The 18 MJ figure of earlier issues would have required ≈ 550 kg per blade and is withdrawn. Packs are sized at 8 MJ with third-stream cooling; the pin seats use Nitinol pins whose A_s ≈ 170 °C sits clear of both soak and brake conduction. Open item: the conducted heat path into hub, spindles and blade roots is still not budgeted — a Phase 1 rig deliverable.
Engine bay / nozzle	Titanium firewall, C/C nozzle flaps washed by third-stream air, carbon-carbon (C/C) shield plating on the aft ventral spine	Reheat liner 900 °C, bay ≤ 200 °C, ventral shield 700 °C transient	Third-stream cooling is what lets a reheated engine live inside a composite stealth tail. The ventral C/C plate answers a purely VTOL problem: on an unprepared pad the plume bounces back off the ground and re-ingests along the belly, soaking the aft under-spine far beyond any composite's limit — C/C takes that thermal spike for the seconds it lasts, every landing, for the life of the airframe.

Thermodynamics of the reverse transition — fuel-sink cooling

Turning a jet back into a helicopter is a thermodynamic problem before it is a mechanical one, and it is asymmetric. Heating the HTSMA leading edge past A_f ≈ 175 °C to snap it sharp is trivial — 40 kW of commanded resistive power for 8 s. Cooling it back below M_f ≈ 95 °C to recover the blunt hover profile is not: after a supersonic leg the edge arrives heat-soaked, and at the Mach ≈ 1 recovery temperatures of the deceleration the boundary layer holds the skin above M_f rather than pulling it decisively below — aerodynamic cooling alone would hold the aircraft hostage in jet mode for minutes it may not have.

Fuel-sink cooling closes the loop. Before un-merge is enabled, the FBW opens the purge circuit: cold jet fuel from the ventral tanks is pumped through the hollow titanium spars of the wing box, sweeping the span behind the Nitinol edge. The fuel absorbs the edge's stored heat almost instantly — ≈ 8 MJ of soaked edge-and-spar heat carried away by ≈ 180 kg of fuel (ΔT ≈ +22 °C at 6 kg/s) in under 30 seconds — and is then fed straight into the ACE and burned: a regenerative cycle that wastes no energy, opens no ram scoop and leaves no thermal trace in the mould line. The Nitinol drops below M_f, the flexcore springs the edge blunt, and the aircraft is cleared to un-merge — a rapid, safe return to hover instead of a multi-minute subsonic cool-down orbit.

Nitinol used twice — the morphing edge and the locking pins

• Morphing edge — two-way trained HTSMA: the skin remembers both profiles; heating past A_f ≈ 175 °C drives it sharp, and the fuel-sink purge pulls it below M_f ≈ 95 °C so the flexcore springs it blunt on demand.
• Locking pins — structural SMA, separately specified: pins are a different alloy from the edge, with A_s ≈ 170 °C / A_f ≈ 195 °C, and are held contracted by an over-centre mechanical detent so clamp pre-load never depends on keeping the alloy hot. On heating the pins contract ≈ 4 %, pre-loading the joint onto steel shear keys. The pins clamp; the shear keys carry. A relaxed pin degrades joint stiffness and fatigue life — it does not release the load path.
• Resistive heating, not kinetic: fed from the ± 270 VDC main bus (§ 4) — ≈ 40 kW peak for 8 s per morph or pin set, buffered by the supercapacitor bank; the legacy 28 VDC rail serves conventional avionics loads only and is never used for shape-memory actuation. Commanded, never accidental — now true everywhere in the envelope, because every A_s sits ≥ 20 °C above the 130 °C maximum soak, not merely above the 45 °C recovery temperature at 12 km.
• Conflict with ice protection — resolved by temperature separation: an electrothermal leading-edge de-icer pulsing to 150–200 °C would have fired an 85 °C edge straight into its supersonic shape in the middle of a hover. The anti-ice mats are therefore surface-temperature limited to 110 °C and hard-interlocked with the SMA controller, which sits 40 °C below A_s. See the ice-protection block below.
• The honest limit: sustained M1.5 below 3 km soaks edge and pins toward 130 °C. The placard survives — not because the alloy would transform, but to protect the RAM (≈ 170 °C) and the 20 °C margin under A_s.
• Failure state = safe: unheated pins stay relaxed (abort to rotor); an unheated edge stays blunt (envelope caps at M 0.85). The aircraft comes home either way.

Heat sinks and signature

• Fuel is the heat sink: gearbox oil, hydraulics, avionics, the Nitinol power electronics — and, through the hollow-spar purge loop, the Nitinol edges themselves — all reject into the feed fuel — no ram-air scoops to break the mould line.
• Brake energy management: at ≈ 3–7 MJ per stop the pack runs far cooler than earlier issues assumed — four merge cycles per sortie are allowed, in two pairs separated by a 4-min cooling hold, enforced by the FBW rather than by the pilot's memory.
• IR discipline: third-stream air films the nozzle flaps and mixes the plume; the merged (stopped) rotor removes blade flash, the biggest radar spike of any helicopter.
• No hot structure anywhere: every primary material is fighter-industry standard — the programme's materials risk is concentrated in one place only: the Nitinol edge.

Ice protection — the strategy the shape-memory edges force on the aircraft

A rotorcraft that must hover in cloud needs blade ice protection; a shape-memory leading edge means that ice protection cannot simply be a hot strip. The two systems are designed together rather than sequentially.

Surface	Method	Limit / interlock
Blade / wing leading edges	Electrothermal parting-strip and shed-zone mats laminated beneath the HTSMA skin, cyclically sequenced blade by blade over the ± 270 VDC bus	Surface temperature hard-limited to 110 °C (40 °C below A_s), dual-channel over-temperature cut-out voting with the SMA controller; the morph command and the anti-ice command are mutually exclusive by design, not by procedure
Inlet lips, S-duct throat	Engine bleed, continuous in icing	No SMA in the path; sized for FAA/EASA App. C and App. O (freezing drizzle)
Canopy, sensor windows, probes	Conventional electrothermal	Independent of the SMA bus
Merged wing in jet mode	None required — kinetic heating above M 0.9	Icing is a helicopter-mode and transition-mode problem only
Mast cavity and iris doors	Warm-air purge from the third stream, doors cycled on the ground before departure in icing conditions	Open item: ice accretion inside the cavity could inhibit the flush retraction — a Phase 7 icing-tunnel item, not yet analysed

the materials themselves are conventional, but the programme's pacing items are four, not one: (1) the Nitinol — a full-span, load-bearing, two-way morphing edge and a structural locking-pin row with a 10 000-cycle fatigue life, beyond any flown example; (2) the in-flight merge itself, which has no precedent at all; (3) the PTO / ACE integration — 3 800 kW of shaft offtake plus a megawatt ISG and a 2 s dog-clutch disengage; and (4) the swashplateless EMA/IBC hub, for which no mechanical fallback exists. NASA has flight-tested SMA-actuated surfaces and Nitinol is a mature alloy, but that maturity does not transfer to this application. The specification is right to call for sustained development in Nitinol alloys — provided it is understood that Nitinol is the first of four pacing items, not the only one. The fallbacks (a mechanically hinged droop edge held in the baseline risk plan, and conventional hydraulic latches — both heavier, both forfeiting their § 9 mass credit) are carried in parallel until SMA fatigue data closes.

6. Expected Performance

Rotor specification — consolidated

The rotor is quoted in one place because almost every other figure in this document is downstream of it: the span, the merged wing area, the aspect ratio, the PTO rating and the merge window all follow from the disc. Column three states the general reference band for the class so that the HSR values can be read against normal practice rather than in isolation.

Rotor parameter	HSR value	General reference band	What it drives
Diameter / configuration	Ø 12.8 m, 2 coaxial contra-rotating stages × 2 blades (4 blades)	—	Sized by the 25 × 25 m pad requirement; it then forces the 12.8 m span, the 5.0 m × 1.55 m half-panels and, with them, the Mach 1.5 cap (§ 11)
Tip speed	210 m/s	200–220 m/s — bounded above by advancing-blade compressibility and by noise, below by autorotation energy	Sets rotor speed, profile power and the advancing-blade Mach limit. Chosen mid-band, not at either edge
Rotor speed	Ω ≈ 32.8 rad/s (≈ 313 rpm) at 100 % N_r; ≈ 219 rpm at the 70 % N_r merge-entry condition	Light helicopters typically turn at 300–400 rpm; heavy helicopters at ≈ 200 rpm or below — rotor speed falls as diameter grows, because tip speed is the invariant	Places the HSR at the bottom of the light-helicopter band, which is what a 12.8 m disc at a 210 m/s tip speed produces
Advancing-blade Mach	≤ 0.85 at 265 kt with N_r reduced to 70 %	Compressibility onset is the universal forward-speed limit of any helicopter rotor	The slowed-rotor condition is the only reason a 235–265 kt merge entry exists (§ 3, M0). At 100 % N_r the aircraft is limited to 200 kt rotor-borne
Disc loading	≈ 84 kg/m² at MTOW 10 800 kg on 128.7 m²	True VTOL hover class, an order below a jet-lift configuration	Hover induced power, downwash velocity on the pad, and the C/C ventral plate case of § 5
Solidity	σ ≈ 0.169 (4 blades × 0.85 m chord on R = 6.4 m)	High for a helicopter rotor — a direct consequence of a blade chord chosen to become half of a supersonic wing	Profile power (≈ 256 kW in the § 4 build-up) and blade loading. It is a cost of the merge, paid in hover efficiency
Required shaft power	≈ 3 150 kW at ISA+20 / 4 000 ft, OGE, MTOW	Turboshaft SFC for the class is 0.28–0.35 kg/kW·h (§ 4), which is what makes hover the dominant fuel consumer of a VTOL sortie	Against the 3 800 kW PTO rating → ≈ 21 % margin. Full build-up, including κ = 1.16 for coaxial interference, in § 4 — the figure is a required shaft power, never an “ideal” power
Blade pitch control	Swashplateless IBC, 4 EMAs, up to 5/rev, ± 0.2° stop-straight indexing, −10° commanded washout in hover	Conventional practice is a swashplate with mechanically ganged pitch links	Mandatory here, not preferred: a swashplate occupies the 0.62 m of mast travel the merge needs (§ 3)
Inter-stage spacing	0.62 m = 0.048 D	Flown rigid coaxial practice is 0.09–0.10 D	Roughly half of practice, and therefore a declared open item: tip-clearance sensing, IBC flap suppression and a 0.95 m fallback (§ 3, risk register)

How to read this block: the diameter is a hover requirement, the tip speed is a rotor requirement, and the chord is a supersonic wing requirement — three constraints that do not normally meet on the same blade. The high solidity and the tight inter-stage spacing are where they collide, and both are carried as costs rather than presented as features.

Parameter	Expected value	What sets it
Crew / extraction seats	2 side by side / 2 in the bay	Pursuit & extraction is a two-operator mission with a human cargo case
MTOW / OEW / W_mid	10 800 kg / ≈ 7 400 kg / 9 700 kg	Merge + flush mechanism 740 kg ≈ 7 % MTOW. W_mid is defined as OEW + full payload + 55 % of usable fuel (1 350 kg) — the mass at the start of the supersonic leg; it is not “MTOW minus half fuel”
Internal payload	950 kg allowance · realisable load ≈ 400–650 kg (or 2 extracted personnel)	Bay volume, two trapeze stations and the stealth requirement — nothing hangs outside. The allowance is what the structure and the CG envelope are sized for; it is not the load two stations can carry (§ 8, § 17)
Rotor Ø / disc loading	12.8 m / ≈ 84 kg/m²	True hover class; sized by the 25 × 25 m pad requirement
Hover OGE	ISA+20 at 4 000 ft	Required shaft power ≈ 3 150 kW by the build-up of § 4 against the 3 800 kW PTO rating — ≈ 21 % margin
Max helicopter speed	200 kt rotor-borne at 100 % N_r · 265 kt at 70 % N_r under partial unload	Advancing-tip Mach ≤ 0.85 on the two coaxial 2-blade stages; the slowed-rotor condition is what makes a 235–265 kt merge entry possible at all
Merge window	235–265 kt, sea level – 2.5 km, ≈ 32 s	Set by lift, not by mechanism: below 235 kt or above 2.5 km the body plus stopped stages cannot reach the C_L the sequence demands (§ 3, M0)
Jet → helicopter reversion	≈ 30 s fuel-sink purge, then un-merge below M 0.8	Cold ventral-tank fuel through the hollow spars drops the HTSMA edges below M_f ≈ 95 °C — no cool-down orbit
Sweep schedule	30° transition · 45° supercruise · 60° dash	Continuous dynamic scissor sweep, FBW-scheduled with Mach — the wing is always at the angle the flow wants
Fuselage attitude in jet mode	0° AoA datum — level	The smart hub's variable incidence tilts the wing, never the body: stealth aspect and minimum drag preserved
Supercruise	M 1.15 dry at 11 km · stretch target M 1.20–1.25	Halves scissor-swept to 45° — normal Mach ≈ 0.81 — with the ACE third stream closed; without the sweep this line would not close. Dry supercruise, not the reheated dash, is the real employment regime, and it is also the cheapest headroom in the design: M 1.20–1.25 is bought by third-stream scheduling and residual-drag reduction, with no additional hardware (§ 11)
Design point	Mach 1.5 at 12 km (1 590 km/h)	Reheat, ≈ 8 min budget, sweep at 60° (normal Mach ≈ 0.75, projected AR ≈ 2.6); scissor-pivot loads and torsional flutter of the 5.0 m × 1.55 m half-panels cap it here
Reference areas / loadings	S_wing 15.6 m² · S_body ≈ 12 m² · S_eff ≈ 27.6 m²	W/S_wing ≈ 692 kg/m² and W/S_eff ≈ 391 kg/m² at MTOW — quoted separately, never combined; S_body is TBD by CFD
Internal usable fuel	2 450 kg	Forward ventral 1 450 kg / aft spine 1 000 kg, centroids 7.0 m apart — the two ends of the CG transfer loop (§ 9)
Combat radius	≈ 560 km with one supersonic leg · unlimited with air-to-air refuelling (§ 16)	2 450 kg internal fuel at the corrected 10 800 kg MTOW; subsonic transit at M 0.85 — preliminary, no reserves policy fixed yet
Ferry range	≈ 1 650 km unrefuelled	Helicopter take-off, merged subsonic cruise, no external tanks possible — which is exactly why the refuelling probe of § 16 is not optional for a strategic-geography operator
Load factor	+6 / −2 merged · +3.5 / −1 helicopter	Fighter agility only once merged — the wing box takes it, the free rotors cannot
Roll performance — requirement, not a claim	≥ 90 °/s at M 0.9 / 30° sweep; ≥ 50 °/s at M 1.5 / 60° sweep — both unverified	“Fighter-like agility” was never a number in earlier issues. Roll comes solely from active aeroelastic twist (± 0.8° at the tip) on a 90 mm box: whether that produces these rates, and how much it degrades with dynamic pressure through aeroelastic reversal, is a Phase 5 measurement and a live risk
Peak structure temperature	≈ 120–130 °C (leading edges, dash)	M1.5 kinetic heating — composites survive, no hot structure needed
Signature	Objective: fighter-like LO merged, low-observable helicopter unmerged	Shaping objective only — no RCS has been measured or modelled to a validated standard. Facets, S-duct, internal carriage and the stopped, flush-retracted hub are the design means; the target is confirmed or rejected in § 12 Phase 6.
Propulsion	1 × ACE: 3 800 kW PTO + 52/75 kN	One engine, both natures — the specification's central bet

CONSISTENT

The physics

The first-order budgets are self-consistent at preliminary level: disc loading, the two loadings, thrust, fuel and temperatures. They are order-of-magnitude sizing figures, not closed calculations — every one of them is conditional on assumptions (S_body, L/D, PTO rating) that only CFD, rig and flight test can settle.

HARD

The mechanism

Two rotor brakes, an L-stroke collar with a chordwise offset axis, a Nitinol locking-pin row, twin scissor pivots on one titanium mast, a flush-retraction cavity with iris doors, and an HTSMA edge that must survive 10 000 morph cycles. This is where the programme is won or lost.

DELIBERATE

The trade

A 12.8 m rotor forces a 12.8 m span — AR ≈ 10.5 unswept, and half-panels 5.0 m long on a 1.55 m chord. That slenderness, not the swept delta's AR ≈ 2.6, is the exact reason the target is Mach 1.5 and not Mach 2. The compromise is balanced for the specification, not against it.

Aerodynamics that are not closed — declared, not omitted

Open item	Why it matters here specifically	Where it is retired
Whirl flutter of the stopped rotor	Two heavy stages cantilevered on one mast at 235–265 kt is the XV-3 failure mode almost exactly. A mast-stiffness floor and a pylon-mode stability boundary must exist before any airborne merge	Analysis + whirl rig, Phase 1; scaled confirmation Phase 2. Now carried as a HIGH risk
Deep stall / pitch-up of the merged delta	A 60°-swept delta with an aft-mounted V-tail and a body that carries ~40 % of lift is a classic deep-stall geometry. There are no elevons; recovery authority is a vectoring nozzle that needs thrust to work	Low-speed tunnel and spin-tunnel work in Phase 2; AoA limiter authority proven before Phase 4
Dorsal-inlet distortion at manoeuvre AoA	Dorsal inlets are shadowed by the forebody at positive incidence — the opposite problem to hover breathing. At +6 g subsonic the distortion case is unquantified	Inlet CFD and rig testing, Phase 0–1; engine surge margin agreed with the ACE partner
Ground and air resonance	A rigid coaxial head with electrically damped lag modes makes resonance stability a software failure case, not a hardware one	Phase 1 rig, Phase 3 ground-resonance test at MTOW
2/rev and 4/rev vibration	Two two-blade stages, contra-rotating: 2/rev per stage plus a 4/rev beat, across a rotor-speed range that now includes 70 % N_r	Airframe mode placement, Phase 1; measured in Phase 3
Roll rate by aeroelastic twist	The only roll effector. Aeroelastic reversal or divergence at high q could reduce, invert or dangerously amplify its authority — and the binding case is not the dash	First-order check below; Phase 1 structural freeze; Phase 5 envelope expansion against the requirement stated in the table above

Supersonic drag build-up — the penalties debited, not deferred

Earlier issues quoted L/D 5.5–6.5 as a target and separately admitted that the semi-flush saddle, the iris doors and the chine junctions would cost something. A review observed correctly that the cost was never taken out of the budget. It is taken out here. Condition: M 1.5 at 12 km, q ≈ 30 kPa, W_mid = 9 700 kg (L ≈ 95 kN), 60° sweep, S_eff ≈ 27.6 m². Counts are 10⁻⁴ of C_D referred to S_eff. This is a hand build-up with declared assumptions, not CFD — its purpose is to stop the penalties being invisible.

Drag term	Counts	At q = 30 kPa	Basis / assumption
Skin friction	76	6.3 kN	≈ 95 m² wetted at C_f ≈ 0.0022 (M 1.5, 12 km, largely turbulent)
Wave drag — volume and lift-independent	55	4.6 kN	Area-ruled waist, 60° leading-edge sweep, sharpened HTSMA edge; assumes the area distribution actually achieved matches the intent of § 2
Lift-dependent drag	40	3.4 kN	C_L ≈ 0.115 with K ≈ 0.31 (AR ≈ 2.6 projected, e ≈ 0.4)
Nozzle trim at ≈ 1.5°	2	0.2 kN	The ≈ 6 % MAC nose-down residual of § 9, already quantified
Clean sub-total	173	14.3 kN	⇒ L/D ≈ 6.6 — which is where the old 6.5 objective came from, and it was a clean number
Semi-flush saddle step + hub cavity	12–20	1.0–1.7 kN	The debit that was missing. "Deep flush" was withdrawn (§ 3): a forward-facing step and an open cavity mouth at the highest-q station on the airframe
Iris-door edges and three-detent sealing gaps	5–9	0.4–0.7 kN	The apertures seal at 30° / 45° / 60° only; at the dash detent the gap is minimum but not zero
Chine-to-wing-root junction interference	6–10	0.5–0.8 kN	A swept root meeting a compression-lift chine at three different sweep angles is three different junctions
Excrescence, leakage and roughness allowance	5	0.4 kN	Standard preliminary allowance; deleted only by a build standard, never by analysis
Debited total	201–217	16.6–18.0 kN	⇒ L/D 5.3–5.7
Sensitivity: S_body does not deliver	—	17.7–19.0 kN	If body lift is half the ≈ 12 m² assumption, S_eff falls to ≈ 21.6 m², C_L rises to ≈ 0.147 and lift-dependent drag rises with its square ⇒ L/D 5.0–5.4

What the build-up actually says. Three conclusions, none of them comfortable:

• The 6.5 end of the old 5.5–6.5 band is withdrawn. It was a clean-configuration number and this configuration is not clean. The credible band at the design point is 5.3–5.7, falling to 5.0–5.4 if the compression-lift assumption disappoints — i.e. at or just below the bottom of the band this document has been quoting.
• The dash is closer to being reheat-dependent than the earlier text implied. That is a range-and-tactics consequence against an 8-minute reheat budget, not merely a performance one.
• The comparison that matters has never been made. A drag of 16.6–19.0 kN must be set against installed net thrust at M 1.5 and 12 km — not against the 52 kN sea-level static dry rating, which is the figure this document has been comparing to. Installed thrust at altitude is an ACE-partner input that does not exist, so the thrust–drag closure at the design point is formally open, and it is the single most consequential unclosed line in § 6.

Aeroelastic reversal and divergence — the first-order check behind the roll requirement

Roll rates of ≥ 90 °/s at M 0.9 / 30° and ≥ 50 °/s at M 1.5 / 60° were stated as requirements without any analysis behind them, on an aircraft whose only roll effector is commanded aeroelastic twist of the wing box. The check below does not clear those numbers — it converts them into structural constraints that can be frozen before metal is cut.

The binding condition is not the dash

• M 1.5 at 12 km: q ≈ 30 kPa. High Mach, thin air.
• M 0.9 at low level: q ≈ 57 kPa — nearly double. The transonic roll requirement, not the supersonic one, sets the aeroelastic case, and every earlier discussion of "reversal at the dash condition" was looking at the wrong point of the envelope.
• Margin requirement adopted: divergence / reversal dynamic pressure q_div ≥ 1.5 × q_max ≈ 85 kPa, consistent with the 25 % / 50 % margin practice used elsewhere in this document.

Why this box is geometrically unfavourable

• At 60° sweep the normal Mach is ≈ 0.75, so the section aerodynamic centre sits near 0.25 c.
• The merged box is two half-profiles pinned together on a mid-chord row of shear keys (§ 3). That joint drives the shear centre aft, toward ≈ 0.45 c.
• The resulting offset e ≈ 0.20 c is nose-up — amplifying. The coupling is therefore of the divergence type, not the classical reversal type: commanded twist is magnified by dynamic pressure rather than washed out.
• Amplification is not a bonus. Effectiveness η = 1/(1 − q/q_div) reaches ≈ 2.0 at the low-level transonic point for q_div = 85 kPa: the same actuator command produces twice the twist. That is a gain-scheduling requirement — twist command must be scheduled on measured q, and loss of that schedule is a control-authority failure case in both directions.

The number this produces — and it is a demanding one. Taking q_div ≈ GJ·(π/2L)² / (a·e·c²) with L = 5.0 m exposed half-panel, c = 1.55 m, a ≈ 4.0 /rad and e ≈ 0.20 c, the requirement q_div ≥ 85 kPa demands GJ ≥ ≈ 1.7 MN·m² per half-panel. A plain single-cell box of 90 mm × 1.55 m with 8 mm walls at G ≈ 6 GPa delivers only ≈ 0.54 MN·m² — a factor of about three short. Three consequences, now design constraints rather than aspirations:

• The "closed-box GJ × 5" line that has sat in the risk register as an assertion is now quantified: it is what buys this margin, and roughly three of those five are needed, not spare.
• |e| ≤ 0.20 c with a controlled sign becomes a frozen structural constraint, set by spar-cap and shear-web placement. Pulling the shear centre forward to e ≈ 0.10 c halves the required stiffness — it is the cheapest lever available and it must be exercised before the Phase 1 rig, because the mid-chord shear-key row is what pushes it aft in the first place.
• The torsion actuators must be sized against the amplified hinge moment at 0.9 q_div, not the rigid-wing value, and η must be shown bounded within 0.8–2.2 across the whole envelope.

Status: this is a strip-theory estimate with no GVT, no validated model and no test behind it. It narrows the design space and it raises the roll-authority risk rather than retiring it — the merged-wing flutter and twist-authority items stay HIGH.

7. Avionics (Flight Controls)

An aircraft that is a helicopter, then a wing, then a helicopter again cannot be flown by memory and muscle: the pilot commands intent, the system flies the machine. The flight controls are a quadruplex digital fly-by-wire with four mode laws and full-authority management of the one manoeuvre that matters — the merge.

Law	Active when	What the stick commands	Envelope protection
HELO	Rotors turning, 0–250 kt	Attitude / translational rate: pitch and roll by rotor cyclic pitch, yaw by differential coaxial torque between the contra-rotating stages; collective on the left hand; blades held at −10° active washout	Tip-Mach governor, torque limit, vortex-ring-state guard, hover-hold and pad-precision modes
MERGE	Sequence M0 → M4, 235–265 kt, sea level – 2.5 km	Flight-path only — the system owns collective, brakes, indexing, blade untwist, the two-axis collar, Nitinol pins, the scissor pivots, the CG fuel transfer and the flush-and-iris mechanism, and blends control authority from cyclic/differential-torque onto the vectoring nozzle. One button: ABORT TO ROTOR	Speed clamp 235–265 kt, altitude clamp ≤ 2.5 km, load factor clamp +1.5/+0.5 g, residual-rotor-lift floor of 30 % until the stopped stages are measured to carry, open-loop CG pre-bias against the 2.4 % MAC M2 mass step, effector-authority overlap monitored continuously, automatic abort on any out-of-window sensor in < 200 ms
JET	Merged wing, 250 kt → M 1.5	Classic fighter C-law with an unconventional effector suite: pitch and yaw exclusively on the 2-D vectoring nozzle* (no moving tail surfaces), roll by commanded aeroelastic twist of the wing box; scissor sweep scheduled automatically with Mach, 30° → 45° → 60° with CG fuel transfer slaved to it, incidence servo holding the fuselage at exactly 0° AoA	+6/−2 g, AoA limiter, static-margin monitor (± 3 % MAC to 45° sweep; scheduled nose-down trim beyond it) gating sweep rate against transfer-pump flow, twist-authority and flutter-margin monitor on the swept half-panels, thermal-soak placard below 3 km, gun-firing placard at M 1.2 until the Phase 6 firing trials clear the full envelope
ABORT / EMERGENCY	On demand or automatic	Reverses the merge, or configures for autorotation / lifting-body glide and steers to the nearest survivable field. On flame-out in jet mode, pitch and yaw revert automatically from the dead nozzle to the seamless morphing ruddervators in < 150 ms	Autorotation entry automated below 200 kt; glide director when merged; morphing-tail authority monitor and supercapacitor-backed essential bus guaranteeing surface power through the descent; ejection-envelope director (see below)

Merge management — the avionics' real job

• 200 ms decision loop: rotor azimuth, Nitinol pin strain, collar position, both scissor-pivot angles, iris-door state and Nitinol temperature fused at 5 kHz; any red parameter triggers the reverse sequence automatically.
• Index sensing: optical azimuth targets plus inductive fine sensors — ± 0.2° on the 90° stop-straight datum confirmed before the collar is allowed to move.
• Thrust-line trim: the vectoring nozzle compensates the abrupt transfer of the centre of lift as the rotors stop lifting and the lifting body takes over — the pilot sees nothing but a green sequence bar.
• Active CG management: the sweep law and the transfer pumps run as one closed loop — forward ventral tanks to aft spine tanks, up to 1 000 kg over a 7.0 m arm on the 30° → 60° stroke — and the law also books the mechanism's own mass migration: + 0.09 m from the blades sweeping aft, and a 2.4 % MAC step during the M2 chord offset that is pre-biased out before the stroke begins. Static margin ± 3 % MAC to 45°, residual ≈ 6 % MAC at 60° trimmed on the nozzle (§ 3). Sweep rate is gated by pump flow, never the reverse.
• Reverse sequencing: un-merge is armed only after the fuel-sink purge reports the leading edges below M_f — a thermodynamic interlock, not a timer.
• Four merges per sortie enforced by the brake thermal model, in two pairs separated by a cooling hold; the fifth attempt is locked out, not advised against.

Crew survivability logic — with the escape envelope stated

• Rotors turning: ejection is preceded by pyrotechnic blade severance — blade-first, pilot-second, never simultaneous. Below 200 kt autorotation remains the trained recovery.
• Rotors stopped and merged — severance is not used: the panels are locked laterally at 90–150° of azimuth, well clear of the seat trajectory, so the jet-mode ejection sequence fires straight through the canopy with no rotor event at all. Earlier issues carried the severance logic into jet mode; that was redundant complexity in the most time-critical failure case and has been removed.
• Declared ejection envelope: zero-zero to 600 kt EAS / 15 km. The design point is inside it — M 1.5 at 12 km is ≈ 430 kt EAS — but not comfortably: it sits at the upper end of what flown seats have demonstrated, and the 62°-raked canopy is a harder MDC penetration case than any vertical windscreen. Canopy severance and seat sequencing at 430 kt EAS are a qualification item, not an assumption.
• Recovery priority when merged: glide first, un-merge-and-autorotate second (requires altitude, time and a successful reverse sequence), ejection third.
• The dead zone is admitted, not engineered away: engine failure in jet mode below ≈ 1 500 ft with no usable surface within gliding range is a loss of aircraft. The whole-airframe supersonic ballistic parachute that earlier issues used to close this case has been deleted: no certified system exists above ≈ 2 t, and 400–800 kg of canopy, mortar and staging were never in the mass statement. The honest recovery matrix has a hole in it, and § 15 states what is done about the bay occupants instead.
• Extracted personnel: bay seats are crash-rated in the baseline; the extraction module can be fitted with two downward-firing ejection seats (B-52 lower-deck precedent) usable above 1 500 ft — see § 15.

Design rule carried through every law: the pilot is never asked to sequence hardware. The stick means “go there”; the buttons mean “change nature” and “stop changing nature”. Everything else — rpm, brakes, collar, Nitinol pins, scissor pivots, incidence, twist, iris doors, Nitinol heating and fuel-sink purge, nozzle angle — belongs to the machine, with the abort path always one action deep.

Dead-stick control — the seamless morphing ruddervators

Thrust vectoring is a superb primary effector and a catastrophic only effector. At Mach 1.5 a flame-out kills the nozzle in the same instant it kills the thrust: a purely passive V-tail would leave the aircraft with static stability but no pitch or yaw authority at all, and a slender 60°-swept configuration with no control power is expected to depart and tumble — there is no glide to be had from a tumbling airframe. The tail is therefore mostly passive, but not entirely.

How the morphing surface works

• No hinge line: the aft ≈ 25 % of each canted fin is a compliant aeroelastic trailing edge — a graded-stiffness composite skin over a flexcore, deflected by embedded actuators. It bends; it does not pivot.
• No radar-scattering gap: conventional ruddervators pay an RCS penalty at the hinge gap and the actuator fairing. A seamless surface has neither, so the fins keep a continuous edge-aligned mould line at every deflection — the stealth case is unchanged whether the surface is at 0° or at full travel.
• Authority sized for one case: ≈ ± 12° of effective camber change, intended to trim and steer the dead-stick descent from M 1.5 down to the flare. Compliant-surface authority at supersonic hinge moments is unproven and is the object of the Phase 5 nozzle-kill trials.
• Powered to the ground: the amplifiers sit on the isolated essential bus, held by the graphene supercapacitor bank and the windmilling ISG (§ 4) — surface authority does not expire before the airframe lands.

Reversion logic

• Detection: nozzle gas-path pressure and spool speed are monitored at 5 kHz; loss of vectoring authority is declared on physics, not on an engine fault code.
• Hand-over in < 150 ms: the C*-law re-allocates pitch and yaw to the morphing surfaces and re-weights roll onto wing twist, blending gains rather than switching them — the pilot feels a gradient change, not a transient.
• Configuration for glide: sweep is commanded back toward 45° as speed decays and the CG fuel law follows it, so the aircraft arrives at the best-glide configuration already balanced.
• And then the seat: with the whole-airframe parachute deleted, the morphing tail is what stands between a flame-out and an ejection — which is precisely why its authority is a Phase 5 gate rather than a footnote.

Sensors — Conformal AESA smart skin

There is nowhere to put a radome. The chin volume is occupied by the rotary cannon and its drum, and the nose itself is a sharp chined forebody apex whose whole job is to hold the bow shock — a dielectric cone bolted onto it would blunt the chine, disturb the shock the forebody is designed to ride, and hand back a specular return at exactly the aircraft's most exposed aspect. The radar is therefore not in a nose; it is in the airframe.

Aperture	Installation	Function
Cheek arrays (port & starboard)	T/R modules laminated into the composite forebody flanks below the chine, as a structural layer of the skin	Primary fire-control sectors: ± 60° each side of the nose with no radome loss, no gimbal, no bulkhead. The two arrays overlap ahead of the aircraft.
Wing-root arrays	Embedded in the fixed root fairing structure either side of the mast cavity — not in the moving wing	Beam and rear coverage, and the long-baseline interferometry pair for passive emitter location. Fixed installation means the scissor sweep never disturbs array alignment or the RF feed.
Spine & fin arrays	Thin conformal panels along the diamond spine and the inboard faces of the canted fins	Upper-hemisphere and rear-quadrant coverage, closing the sphere: 360° situational awareness with no mechanically scanned aperture anywhere on the aircraft.
Shared EW function	The same T/R modules time-share between radar, ESM and jamming through the mission computers	One skin, three roles: detect, listen and radiate. Distributed aperture also means graceful degradation — battle damage costs a sector, never the sensor.
Structural integration	Arrays are load-carrying laminate plies with RAM-compatible frequency-selective outer surfaces; thermally bonded to the fuel-sink loop	The array is part of the structure rather than a payload inside it: no radome mass, no aperture cut-out, no separate cooling scoop — and at Mach 1.5 the skin the modules live in stays inside its 120 °C limit.

Why smart skin is the enabling choice, not a luxury: it is what allows the nose to remain a pure aerodynamic and stealth object — sharp, chined, aerodynamically clean and full of gun — while still targeting fighter-like fire control (unvalidated) and full spherical awareness. The radome was not miniaturised or relocated. It was deleted, and the aperture was distributed into the airframe that replaced it.

8. Lethality & Payload Integration

Everything this aircraft carries, it carries inside. The stealth requirement forbids a single external pylon, and the Mach 1.5 design point forbids a single permanent aperture — so both the gun and the bay are treated as momentarily openable volumes rather than as installations. Two hard problems follow: getting muzzle gas away from an air-breathing engine, and getting a store out of a cavity into supersonic free stream. Both are solved by geometry, not by placarding the envelope.

Installation	Configuration	Engineering rationale
Chin rotary cannon	Fixed forward-firing 25 mm rotary (Gatling) cannon (GAU-12/U class), fully internal in the lower nose; 350 rounds in a linkless drum on the bay's forward bulkhead; boresighted to the DAS/HMD cue	A fixed gun on the aircraft's own thrust axis is the only gun installation compatible with a faceted, chined forebody — no turret, no barrel proud of the mould line, no cavity ahead of the shock. Magazine raised from 220 to 350 rounds: at 3 600–4 200 rd/min, 220 rounds is 3–4 seconds of fire — one or two firing opportunities — which is not an interceptor's magazine. 350 rounds gives ≈ 5 s and costs ≈ 190 kg of payload.
Snap-action stealth trapdoor	Flush, edge-aligned faceted door over the muzzle port; opens in ≈ 120 ms, closes in ≈ 150 ms, commanded by the trigger and re-sealed automatically at burst end	The port exists only while rounds are leaving it. Between bursts the chin is a continuous RAM-coated facet: no permanent aperture, no cavity resonance, no RCS penalty and no Mach 1.5 drag rise.
Gun-gas isolation — in forward flight	Muzzle plane below and forward of the nose datum; engine fed by the dorsal S-duct inlets on the upper spine	The classic strafing failure mode — ingesting oxygen-depleted propellant gas and stalling the compressor — is addressed by geometry: the gas is produced under the belly, the free stream carries it aft, and the engine breathes over the back. This argument is valid only where a free stream exists, and earlier issues did not say so. See the hover case in the row below and in the card set that follows.
Gun-gas ingestion in the hover — the case the geometry does not cover	Firing inhibited below 40 kt in the baseline, by FBW interlock, until the Phase 6 firing trials produce a measured surge margin at zero airspeed	At 0 kt there is no free stream to carry the efflux away. There is instead a rotor downwash column pressing air down over the whole aircraft and, in ground effect, a recirculation fountain climbing the fuselage sides — and § 4 states that at zero airspeed the engine is fed by spring-loaded auxiliary doors around the duct shoulder, i.e. by exactly that recirculating air. Firing in a hover therefore injects an oxygen-depleted plug into the intake path at the one condition where the engine is at maximum shaft power (3 800 kW to the PTO), has the least surge margin and cannot be relit by windmilling. This is not a strafing nuisance; in a hover it is a loss-of-lift event. Mitigations to be traded in Phase 0–1: a muzzle gas deflector, closing the auxiliary doors for the burst duration (not possible at 0 kt — the engine would be starved instead), or accepting the placard. The baseline accepts the placard, which costs the aircraft its hovering fire-support case and must be stated as a capability limitation rather than buried.
Modular ventral bay	One bay, 3.4 × 1.1 × 0.8 m, serrated doors: 2 crash-rated extraction seats or 950 kg of internal ordnance, swapped as a palletised module in ≈ 40 min on the pad. Maximum store length 3.4 m — which excludes AIM-120 AMRAAM (3.65 m) and every weapon of that class	Pursuit and extraction from one airframe means the same volume must sell itself twice. The module carries its own floor, restraints and umbilicals, so neither role compromises the other and no structure is flown that the sortie does not need. The bay length is a hard weapons-integration constraint, not a detail: the compatible set is MICA / IRIS-T / AIM-9X / ASRAAM for air-to-air, and SPEAR 3, GBU-53/B SDB II, Brimstone or a 500 lb-class PGM for strike — see the weapons matrix in § 17.
Pneumatic trapeze ejectors	Two scissor-trapeze arms, high-pressure pneumatic (stored gas, no cartridges): stroke ≈ 0.55 m, peak 40 g push-out as the design target, store clear of the shear layer in ≈ 90 ms	A gravity release at Mach 1.5 does not release: the bay's own shear layer and shock system push the store back onto the airframe. The trapeze is intended to punch the munition bodily through the supersonic boundary layer into clean free stream. Clearance is progressive — subsonic separation first, then transonic, then M 1.5 (§ 12 Phase 6); the 40 g figure is the sizing objective, to be confirmed by CFD and separation trials. Two arms means two stores — which is the constraint that makes the payload arithmetic below not close.
Door timing	Bay doors open, store away and doors closed inside a ≈ 1.6 s window, FBW-sequenced with the sweep and CG laws	Cavity-open time is the RCS and buffet exposure; keeping it under two seconds keeps the supersonic release a non-event for the airframe and for the store's separation dynamics.

Chin gun + dorsal inlet — compatible in forward flight, placarded in the hover

The two decisions were taken for different reasons — the gun went to the chin for mould-line and boresight reasons, the inlets went dorsal to stay above the hover debris fountain and hide the fan face — and in forward flight they are mutually protective: muzzle efflux expands downward and aft along the belly while the intake lips look upward and aft along the spine. The geometry is designed to preclude gun-gas ingestion — to be verified by CFD and by firing trials (§ 12 Phase 6) before any firing clearance is granted. Until those trials pass, firing is placarded to M 1.2 by an FBW interlock (§ 7). If verification holds, that placard is lifted and the cannon is cleared across the entire envelope including a Mach 1.5 gun intercept. Earlier issues of this document asserted both “interlock above M 1.5” and “no firing restriction whatsoever”; the sequence above is what was meant, and is now stated once. What none of them addressed is the opposite end of the envelope: with no free stream, the argument inverts, and firing is inhibited below 40 kt (table above).

Modular ventral bay — dual use, one volume, two very different masses

In the extraction fit: two crash-rated seats (not ejection-rated), 5-point restraints, oxygen and intercom, rated to the 4 m/s vertical-landing sink case. In the strike fit: internal stores on the trapeze rails. The bay is on the CG and inside the area-ruled waist, so the aircraft balances the same either way — the balance ledger of § 9 is written once, for both roles, and the pilot never re-learns the aircraft. It does not weigh the same either way, and earlier issues blurred the two: an extraction fit is a few hundred kilograms, a full strike fit is the whole payload line. Same CG, different gross weight, different hover margin.

40 g — the number that makes supersonic release safe

Inside an open bay at Mach 1.5 the cavity flow is separated, unsteady and at lower total pressure than the free stream: a released store sees an upward pressure gradient pushing it back in. The trapeze answers with momentum — 40 g over 0.55 m is sized to put the store's centre of gravity below the shear layer inside 90 ms, past the bounce-back region and into flow its own control surfaces can bite. Pneumatic, not cartridge: no residue in the bay, and it re-arms from the onboard bottle for the next release. Release clearance is built up subsonic → transonic → M 1.5.

The payload line does not close — the arithmetic, in full. § 9 books 950 kg of payload. What can actually be carried is smaller, and the two figures had never been reconciled:

• Ammunition first. 350 rounds of 25 mm is ≈ 190 kg, charged to the payload line. That leaves ≈ 760 kg for everything else before a single store is loaded.
• Two arms means two stores. The bay has two scissor-trapeze ejectors. Against the § 17 compatible set — SPEAR 3 and Brimstone at ≈ 1.80 m, GBU-53/B SDB II at ≈ 1.76 m, a 500 lb-class PGM — two stations carry roughly 200–460 kg, not 760 kg and certainly not 950 kg.
• The extraction fit is lighter still: two occupants, the module floor, restraints and umbilicals, plus the optional downward-firing seats of § 15 — a few hundred kilograms against a 950 kg allowance.
• Therefore: 950 kg is a structural and balance allowance, not a demonstrated store load. The realisable strike fit today is ammunition + two stores ≈ 400–650 kg, and the aircraft flies the strike mission 300–500 kg below MTOW — which is useful for hover margin and useless as a magazine.

Closing the gap requires a decision, not a rewrite: either add stations inside the same 3.4 × 1.1 × 0.8 m volume (which competes with store diameter and with the trapeze stroke), or accept that this is a two-shot aircraft plus a gun whose magazine depth comes from the manned-unmanned teaming of § 13 rather than from its own bay. The baseline states the second position and carries the first as an open trade for Phase 1. Until it is taken, § 9's payload line must be read as an allowance.

Lethality summary: one fixed 25 mm rotary cannon on the thrust axis with 350 rounds (≈ 5 s of fire), permanently loaded and permanently hidden — but inhibited below 40 kt until the hover ingestion case is measured; one modular ventral volume that is either a rescue cabin or a two-station weapons bay, 3.4 m long and therefore closed to AMRAAM-class stores; and a release mechanism sized for the worst case in the envelope rather than for the easy one. Nothing hangs outside the mould line at any point in any mission — and nothing much hangs inside it either.

9. Mass Breakdown & Balance (Target MTOW: 10 800 kg)

A hybrid airframe lives or dies on its mass statement, and the reviewer's first question is always the same: what does the transformation actually cost, and where is it paid from? The answer is isolated below as a single explicit line rather than smeared across the structure — 740 kg, ≈ 7 % of MTOW — and set against the conventional hardware the concept deletes outright. The mass statement below is the frozen preliminary-design baseline; the OEW figure is contractual.

Why OEW moved from 6 400 kg to 7 400 kg, and MTOW from 9 800 kg to 10 800 kg. The previous statement was under-budgeted in four identifiable places, and re-closing them is not optional:

• Propulsion 1 520 → 2 150 kg (+630). A 52/75 kN core is 1 150–1 250 kg dry before a 2-D vectoring nozzle (+150–250 kg), the S-ducts, a 3 800 kW bevel PTO with 90° drive, dog clutch and coaxial gearbox (+450–600 kg) and the ISG.
• Armament 120 → 320 kg (+200). A GAU-12/U-class 25 mm rotary cannon weighs ≈ 122 kg on its own; the drum, feed, snap-action door, bay structure and trapeze ejectors are the rest.
• Supercapacitor bank (+170 within Systems). 12 MJ at the implied ≈ 30 Wh/kg does not exist; at a real 10–15 Wh/kg the bank is ≈ 280 kg.
• Mechanism 680 → 740 kg (+60). The chord-offset carriage and rails of § 3 are new hardware and are booked, not absorbed.
• Offset by − 90 kg: the whole-airframe ballistic parachute is deleted (it was never properly budgeted either — see § 15).

Fuel and payload are held at 2 450 kg and 950 kg, so the growth lands on MTOW. Disc loading rises to ≈ 84 kg/m² and the hover power build-up of § 4 still closes on the 3 800 kW PTO with ≈ 21 % margin — which is the reason this rebalance was possible at all rather than a re-sizing of the rotor.

Mission mass statement

Group	Mass	% MTOW	Notes
Operating empty weight (OEW)	7 400 kg	68.5 %	Includes crew (2 × 105 kg equipped), unusable fluids, oils and the fixed 25 mm gun installation
Internal usable fuel	2 450 kg	22.7 %	Forward ventral 1 450 kg / aft spine 1 000 kg, centroids 7.0 m apart — the two ends of the active CG transfer loop; Jet A-1 / F-34 class, low-freeze-point specification for arctic operations (§ 4, § 16); no external tanks, ever
Payload	950 kg	8.8 %	An allowance, not a demonstrated load. 350 rounds of 25 mm (≈ 190 kg) plus two trapeze stations (≈ 200–460 kg against the § 17 compatible set) realise ≈ 400–650 kg; the extraction module is lighter still. The balance case is written at 950 kg because the structure and the CG envelope are sized for it — see the payload arithmetic in § 8
MAXIMUM TAKE-OFF WEIGHT	10 800 kg	100 %	Disc loading ≈ 84 kg/m² · W/S_wing ≈ 692 kg/m² · W/S_eff ≈ 391 kg/m² (S_body TBD by CFD)

OEW build-up — the transformation penalty isolated

OEW group	Mass	% OEW	Content
Airframe structure	1 800 kg	24.3 %	Faceted carbon-BMI fuselage, area-ruled waist, diamond spine and canted fins, dorsal mast cavity, landing gear, all doors
Rotor-wing system	1 040 kg	14.1 %	4 blades (2 stages × 2), both swashplateless EMA/IBC hubs with their slip-ring and rotary-transformer stacks, hollow Ti spars, HTSMA morphing edges and locking-pin rows
Stop-and-Merge / scissor mechanism penalty	740 kg	10.0 %	The isolated cost of the concept: 2 rotor brakes and indexers, L-stroke merge collar with its chordwise offset rails and carriage (≈ 60 kg of the total), twin scissor-sweep pivots and actuators, incidence gimbal, aeroelastic torsion actuators, semi-flush retraction jack, iris doors and their drive — ≈ 7 % of MTOW
Propulsion installed	2 150 kg	29.1 %	ACE dry (≈ 1 200 kg for a 52/75 kN core), S-ducts, 2-D vectoring nozzle (≈ 200 kg), shaft-embedded megawatt ISG, PTO bevel gearbox with 90° drive, dog clutch and coaxial gearbox (≈ 520 kg). Corrected upward by 630 kg — the previous 1 520 kg line was not survivable against any real engine of this class
Systems	600 kg	8.1 %	Fuel system incl. CG transfer pumps and the fuel-sink purge loop, ± 270 VDC distribution and converters, graphene supercapacitor bank — 280 kg for 12 MJ at a realistic 12 Wh/kg, OBIGGS, environmental and thermal management
Avionics & mission systems	470 kg	6.4 %	Quadruplex FBW, active inceptors, conformal AESA smart-skin modules and their feeds, DAS/EO belt, EW and self-protection suite, datalinks and GPS-denied navigation (§ 13), mission computers, morphing-ruddervator amplifiers
Crew & escape	280 kg	3.8 %	Cockpit fit, 2 zero-zero ejection seats, OBOGS and pressurisation pack. No whole-airframe ballistic parachute — deleted (§ 15); the two downward-firing seats of the extraction module are carried on the payload line, not here
Armament installation	320 kg	4.3 %	25 mm rotary cannon (≈ 122 kg alone), 350-round drum and linkless feed, snap-action trapdoor, bay structure and trapeze ejectors (ammunition counts as payload). Corrected upward by 200 kg
TOTAL OEW	7 400 kg	100 %	Design target, monthly mass reviews. A standard preliminary-design growth margin of + 5 % (≈ 370 kg) is carried against it and is not included in the figures above; releasing it consumes fuel or payload one for one.

What the concept deletes — the mass credits

The 740 kg is a gross figure, and quoting it alone is dishonest. A conventional compound helicopter or a conventional fighter of this class carries hardware that this architecture does not build at all:

Deleted by architecture	Credit	Why it does not exist here
Tail rotor / fenestron, tail drive shaft, intermediate and tail gearboxes	− 210 kg	Contra-rotating coaxial stages cancel torque; yaw in hover is differential coaxial torque
Ailerons / elevons, their hinges, actuators and hydraulic runs	− 95 kg	Roll is aeroelastic twist of the wing box — there is no trailing-edge surface to build
All-moving ruddervator actuation and empennage hydraulics	− 70 kg	Pitch and yaw in jet mode are the 2-D vectoring nozzle; the fins are fixed facets whose morphing trailing edges have no hinges, bearings or hydraulic runs to build
Merge / locking hydraulics, accumulators and plumbing	− 85 kg	Nitinol shape-memory pins do the clamping electrically, on the ± 270 VDC bus
Swashplates, pitch links, rotating and stationary scissors, sliders and boost actuators (× 2 stages)	− 105 kg	Swashplateless IBC: pitch is commanded by EMAs inside each hub over a slip ring — the entire mechanical pitch train, and the rotating hydraulics that drove it, are never built
Nose radome, its bulkhead, gimbal ring and dedicated liquid-cooling loop	− 65 kg	Conformal AESA smart skin makes the array a structural laminate in the cheeks, roots and spine — there is no radome to carry and no aperture cut-out to reinforce
TOTAL CREDIT	− 630 kg	—
NET TRANSFORMATION PENALTY	+ 110 kg (≈ 1.0 % MTOW)	740 kg built, 630 kg never built — the honest arithmetic the concept is actually judged on

Balance ledger

Static balance

• Mast on the CG — with one correction: the rotor-wing system, the heaviest concentrated mass after the engine, sits on the balance point in every configuration. Earlier issues concluded from that that “sweeping it does not move the mass, only the lift”. That is false: both halves rotate aft, so ≈ 600 kg of blade mass migrates ≈ 1.43 m rearward between 30° and 60° (+ 0.09 m of CG), and the M2 chord offset translates ≈ 520 kg by 0.70 m (+ 0.038 m, as a step). Both terms are now in the dynamic ledger opposite.
• Bay on the CG: extraction module and the strike fit occupy the same volume at the same station: no re-balancing between roles, no role-dependent trim charts. The two fits do not weigh the same — see § 8 — so the hover margin differs even though the balance does not.
• CG envelope: 22–38 % MAC placarded on a MAC of 1.55 m; loaded range across all fits and fuel states 26–34 % MAC — four points of margin at each end.
• Gun and drum forward: deliberately balanced against the aft engine mass. Corrected figure: 350 rounds is ≈ 190 kg on a ≈ 2.5 m arm, so firing the magazine out shifts the CG by ≈ 0.049 m — ≈ 3.2 % MAC, not the “< 0.4 %” claimed in earlier issues, which was wrong by roughly a factor of eight. It is inside the placarded band, but it is a real number that the fuel law must account for.

Dynamic balance — the sweep loop

• Transfer capacity and arm, both stated: up to 1 000 kg movable forward-ventral → aft-spine over a 7.0 m centroid separation, at 20 kg/s (50 s for the full stroke), which is what sets the maximum sweep rate. The aft cell is burned first outbound so that it is empty and available.
• Tracking law and its residual: the aerodynamic centre marches ≈ 0.9 m aft between 30° and 60° of scissor sweep; the fuel law delivers ≈ 0.72 m of CG travel at W_mid and the blade-mass migration a further ≈ 0.09 m. Static margin is held inside ± 3 % MAC to 45°; the remaining ≈ 6 % MAC at 60° is trimmed by the nozzle at ≈ 1.5° of vector deflection, costing ≈ 0.2 kN of trim drag. The “± 2 % MAC across the whole schedule” of earlier issues was not achievable with any tank arm this airframe can offer and has been withdrawn.
• M2 mass step — new line: the chord-offset leg moves ≈ 520 kg by 0.70 m aft (≈ 0.038 m ≈ 2.4 % MAC) in ≈ 7 s, and the 0.62 m descent lowers the CG by ≈ 0.033 m. Handled by an open-loop forward pre-bias of the transfer pumps before the stroke, plus a scheduled nozzle input — not by the pitch loop reacting to it (§ 3, § 7).
• Reserve interlock: the transfer is inhibited below 450 kg total fuel — at that point sweep is capped at 45° and the aircraft is already committed to recovery.
• Fuel-sink coupling: the ≈ 180 kg purged through the wing spars during the reverse transition is drawn from, and returned to, this same ledger — one fuel system, three jobs: propulsion, balance and cooling.

Mass verdict: the concept costs 740 kg gross / 110 kg net, and it buys VTOL and Mach 1.5 in one airframe. That is the entire trade, stated in one line. The risk is not that the number is wrong today — it is that it grows: hence the mechanism mass is frozen contractually at 740 kg, reviewed monthly, with payload the declared victim of any breach. Note that the credit column is only collectable if the exotic technologies hold: an aircraft that has to retreat to a swashplate, a radome or hydraulic latches loses the offset, not just the feature. Note also that the 1 000 kg OEW correction above was found by review, not by the design team — which is the strongest argument in this document for the independent mass audit called for at the Phase 0 gate.

10. Human-Machine Interface (HMI) & Cockpit

No pilot has ever been asked to fly two aerodynamic regimes this far apart inside a single sortie. A helicopter pilot's left hand pulls lift; a fighter pilot's left hand pushes thrust. The two habits are muscle-memory opposites, and swapping between them at 250 kt in a hostile environment is exactly the kind of workload that kills aircraft. The HSR does not ask the pilot to swap: the cockpit itself changes nature, physically, in step with the airframe.

Regime	Left inceptor	Right inceptor	Force gradients & feel	Primary display state
HELO	Collective — vertical axis live, up/down commands lift; twist grip inert	Cyclic — short throw, light breakout, near-zero centring spring	Soft, low-gradient, high-sensitivity: ≈ 8 N/cm, breakout 3 N — helicopter feel, position-dominant	Hover page: velocity vector, pad cue, torque and disc state, DAS ground picture
TRANSITION (M0 → M4)	Morphing — motorised force-feedback progressively locks the vertical axis and hands the axis over to the fore/aft throttle rail	Hardening — centring spring and damping ramp up continuously with the sweep schedule	Gradients ramp linearly with sequence progress; a distinct haptic detent at each of M0–M4 tells the hand where the aircraft is without a glance	Sequence bar M0–M4, effector-authority blend, one guarded button: ABORT TO ROTOR
JET	Fighter throttle — linear fore/aft rail, detented idle / mil / reheat gate; vertical axis mechanically dead	Fighter stick — stiff, force-dominant, small displacement	Hard, high-gradient: ≈ 34 N/cm, breakout 12 N — Mach 1.5 fighter stiffness, force-dominant C*-law	Tactical page: HUD/HMD symbology, sweep and CG-transfer state, gun and bay status
EMERGENCY	Reverts to whichever function the current configuration can actually use, announced by a triple haptic pulse	Soft-stops the pilot out of any input the degraded configuration cannot accept	Tactile shaker on both inceptors; stick-shaker semantics preserved across both regimes	Recovery director: autorotation entry, glide profile, or ejection-envelope cue

The FBW Unified Active Inceptor

A single smart HOTAS pair with fully motorised, software-defined force-feedback on every axis. There is no mechanical mode switch, no second set of controls and no folding hardware: the inceptors are actuators as well as sensors, so stiffness, breakout, damping, detents and even axis availability are commanded parameters. The aircraft changes what the pilot's hands can physically do, at the moment the aerodynamics change what they should be doing.

The left-hand hand-over, in detail

Entering M0 the collective is still live in the vertical axis. As the rotors unload, force-feedback progressively raises the vertical-axis gradient until that axis is physically locked at the M2 detent — the hand simply cannot pull lift that no longer exists. Simultaneously the fore/aft throttle rail unlocks, its gate hardening into the idle / mil / reheat detents. By M4 the same grip is a conventional fighter throttle. The reverse sequence runs the choreography backwards, and the collective's vertical axis is restored only after rotor respin is confirmed.

Right-hand gradient scheduling

The stick's feel is scheduled against dynamic pressure and sweep angle, not against mode flags: from helicopter-cyclic sensitivity (light, long-throw, position-dominant — correct when the aircraft answers with disc tilt) to Mach 1.5 fighter stiffness (heavy, short-throw, force-dominant — correct when a centimetre of stick is 6 g). The ramp is continuous, so there is never a step change in feel at a mode boundary — the classic source of pilot-induced oscillation in multi-regime aircraft.

Cockpit fit

• 2 crew side by side under the raked canopy: pilot left, mission operator right, both on zero-zero seats.
• Two 20-inch large-area displays plus a central standby panel; no discrete instruments, no round dials.
• Helmet-mounted display fused with the distributed-aperture sensor belt — the raked canopy costs geometric vision, the DAS gives it back through the airframe.
• Nature buttons only: “MERGE” and “ABORT TO ROTOR” are the only two controls that change what the aircraft is; everything else changes only where it goes.

Why force-feedback, not a mode annunciator

• Mode confusion is a haptic problem, not a labelling problem: a caption saying “JET” does not stop a hand from pulling collective. A locked axis does.
• Eyes stay out: the M0–M4 detents are read through the hand, so the pilot flies the transition looking at the threat, not at a sequence bar.
• Reversibility is muscular: ABORT TO ROTOR is a guarded button under the left thumb throughout the transition — always in the same place, in both natures.
• Training payoff: the aircraft teaches its own regime through feel, which is what makes a dual-qualified crew a realistic proposition rather than a programme risk.

HMI design rule: the pilot flies intent, and the cockpit physically forbids intent the current configuration cannot honour. Every regime change is felt before it is read — the inceptors become the primary annunciator, and the displays merely confirm what the hands already know.

11. Theoretical Maximum Capability Evaluation

Every figure in this document so far is a design point — a number the aircraft is contracted to deliver with margin. This section asks the opposite question, the one a board asks before it funds anything: where is the physical wall? For each parameter below, Compagnie Apex has identified the single binding constraint, the value the airframe would reach if that one constraint were relieved, and what relieving it would cost elsewhere. Nothing here is promised. It is a map of the headroom.

Parameter	Design point (contracted)	Theoretical maximum	The binding constraint	Cost of relieving it
Maximum Mach	M 1.5 at 12 km	Sensitivity only — no sellable headroom	Torsional flutter of the 5.0 m × 1.55 m half-panels and the scissor-pivot bearing load line — not thrust, not heating. The flutter boundary is currently predicted at, not above, M ≈ 1.5, on unvalidated models	A sensitivity study indicates ≈ M 1.8 if ≈ 90 kg of box stiffening and a second active-flutter channel were added and the flutter prediction were confirmed by GVT and flight test. Until then M 1.5 is a cap, not a design margin, and no figure above it may be quoted as capability.
Dry supercruise	M 1.15 at 11 km, 45° sweep	M 1.28	Third-stream schedule optimisation and ≈ 2 % of residual airframe drag (iris-door steps, chine junction)	Pure development effort — CFD, tunnel hours and an ACE control-law block. The cheapest headroom on this table.
Service ceiling	12 km design point	14.5 km sustained / 17.2 km zoom apogee	Thrust lapse against wing loading at 60° sweep; above 15 km the merged wing simply runs out of dynamic pressure	Nothing structural — it is an engine-rating question, and the intercept mission never asks for it
Load factor	+6 / −2 g merged	+7.4 g ultimate-limited	Blade-root titanium fitting at full 60° sweep, where bending and pivot shear superimpose	Root fitting redesign, ≈ 25 kg; the placard also protects the extracted personnel, who are crash-rated rather than g-rated
Hover ceiling OGE	ISA+20 at 4 000 ft	ISA+20 at 7 500 ft	PTO shaft torque limit (3 800 kW) and hot-gas re-ingestion on the pad, at the corrected 10 800 kg MTOW	PTO uprate to ≈ 4 200 kW and a heavier bevel set: the single most stressed part of the aircraft becomes more stressed still
Combat radius	≈ 560 km, one supersonic leg	≈ 740 km unrefuelled · theatre-limited with AAR	Internal volume: 2 450 kg of fuel is all the area-ruled body can hold without breaking the waist	≈ 350 kg of conformal spine tankage — which competes directly with the aft cell the CG transfer law depends on. The cheaper answer is the refuelling probe of § 16, which changes the radius question entirely
Internal payload	950 kg	1 250 kg	Bay volume, not structure — the trapeze rails and the area-ruled waist set the envelope	Fuel traded 1:1; a 1 250 kg sortie is a short-radius strike profile, not the design mission
Merge window	235–265 kt, SL – 2.5 km	225–300 kt, SL – 4 km	Lift, not brake energy. The re-derived 3–7 MJ per stop is no longer the binding constraint; what binds is the C_L the body plus stopped stages can produce at the entry condition	Either accept residual rotor lift deeper into the sequence, or add fixed lifting area — which the no-fixed-wing requirement forbids
Electrical growth	1.1 MW continuous / 1.4 MW peak	≈ 300 kW-class directed-energy provision	Not the ISG — the supercapacitor bank's 12 MJ store and its thermal rejection into the fuel	Doubling the bank (≈ 280 kg at a realistic 12 Wh/kg, not the 110 kg quoted in earlier issues) and a dedicated fuel-cooled loop; the architecture of § 4 was deliberately sized to allow it, but the mass is real
Mach 2 and beyond	—	Not reachable on this airframe	The hover requirement itself. A 25 × 25 m pad at 10 800 kg forces Ø 12.8 m of disc, a 12.8 m rotor forces a 12.8 m span, and a 12.8 m span forces half-panels 5.0 m long on a 1.55 m chord — an unswept AR ≈ 10.5 that no amount of sweep makes structurally slender enough for M 2	A smaller rotor. Which is to say: a different aircraft, one that no longer hovers. This is the wall, and it is not moveable by engineering.

Where the headroom actually is

• Cheapest: dry supercruise. M 1.28 costs analysis and control laws, not hardware — and supercruise, not the reheated dash, is what an intercept profile actually lives on.
• Most strategically valuable: the electrical growth line. The megawatt ISG and the isolated bus were sized once, for actuation; a directed-energy or high-power-EW fit is a bank-and-cooling change rather than a re-architecture.
• Most tempting and most dangerous: the Mach sensitivity line. It is a study result on unvalidated flutter models, and pursuing it consumes the very margin that makes the variable-sweep concept flyable. Compagnie Apex's engineering position is that this margin is not for sale and that M 1.5 is to be treated as a hard cap until flight test says otherwise.

What the table does not mean

• These maxima are not additive. The headroom lines cannot coexist: they draw on the same mass, the same volume and the same thermal budget.
• They are not a growth plan. Publishing headroom is how programmes acquire requirements they never agreed to; every line above is explicitly out of scope for the baseline aircraft, and the Mach line is a sensitivity, not an offer.
• The design point is where the aircraft is balanced. M 1.5 is not a limitation reluctantly accepted — it is the equilibrium at which one engine, one merged wing, fighter-standard materials and a true hover capability all close simultaneously.

Chief Engineer's position on the maxima: the honest reading of this table is that the HSR may have modest aerodynamic and useful electrical headroom above its contracted numbers, no confirmed Mach headroom whatever, and none at all against the constraint that defines it. The rotor that lets it hover is the wing that caps it at Mach 1.5. Any programme conversation that begins “could it reach Mach 2” is, whether it knows it or not, a conversation about deleting the hover.

12. Development Roadmap

The programme is sequenced around a single conviction stated in the risk register: the speed is not the risk, the merge is. Every phase below therefore exists to retire the transformation before the envelope, and each is gated on a demonstrated capability rather than on elapsed time. Durations are measured from programme go-ahead (T0) and overlap deliberately — the rigs never stop running once the aircraft starts flying.

Phase	Window	Focus	Exit gate — pass/fail, not opinion
0 · Technology maturation	T0 → T0+24	The four pacing items in parallel: HTSMA alloy and two-way fatigue programme; swashplateless EMA/IBC hub on a whirl rig; graphene supercapacitor abuse and qualification testing; and the ACE partnership itself — no engine of 52/75 kN with a 3 800 kW PTO and a megawatt ISG exists today (§ 17), which makes the adaptive-new-core versus proven-core-plus-PTO-module trade a Phase 0 deliverable in its own right. In parallel: independent mass audit, whirl-flutter stability analysis of the stopped rotor, the S_body and semi-flush step-drag CFD campaign, a preliminary RCS model, the alternatives trade study (compound helicopter vs tiltrotor vs HSR on one common mission set), arctic and shipboard design requirements written into the specification rather than deferred to Phase 7, and the systems-safety framework of § 14 (FHA and preliminary SSA)	TRL 5 on all three technology items, plus 10 000 SMA morph cycles on a full-span coupon, a 500-hour rotating slip-ring endurance run, a signed ACE development agreement with a demonstrated PTO concept, and a whirl-mode stability boundary carrying 25 % margin on the 265 kt merge ceiling (≥ 330 kt) with the mast/pylon stiffness floor declared as a number, not an intention. Also gating: an independent mass audit closed, a modelled RCS value on record for the semi-flush saddle, its seams and the iris doors, and CFD verdicts on S_body ≈ 12 m² and on the semi-flush step drag folded into the § 6 lift and L/D budgets
1 · Iron bird & 1:1 mechanical rig	T0+18 → T0+42	Full-scale ground rig of mast, both hubs, the L-stroke merge collar with its chordwise offset axis, scissor pivots, semi-flush retraction and iris doors; a full-scale dorsal-section mock-up to close the retraction clash against the real S-duct, engine and aft-cell envelopes; separate megawatt electrical iron bird with the real ISG and bank	2 000 ground merge/un-merge cycles without a pin, bearing, rail or door failure; eccentric-carriage load case substantiated with the ± 3 mm second-axis indexing tolerance demonstrated, not assumed; retraction stroke, door sealing at all three sweep detents and V-tail clearance demonstrated on the mock-up; essential-bus isolation demonstrated under a live 40 kW SMA transient; MIL-STD-461 emission limits met with the hub rotating and the EMAs commanded; inter-stage spacing decision frozen — 0.62 m confirmed on rig and proximity-sensing data, or the 0.95 m fallback adopted with its + 18 kg and + 3 s; and the first CPFH and life-cycle-cost estimate issued against the closed alternatives trade study of § 17
2 · Sub-scale autonomous demonstrator	T0+30 → T0+54	1:4 unmanned vehicle, jet-powered, flying the complete M0 → M4 sequence and its reverse. No crew, no weapons, no stealth qualification — one job only	25 airborne merges and 25 un-merges, including 5 deliberate mid-sequence aborts; scaled flutter clearance to the equivalent of M 1.2
3 · Full-scale prototype — helicopter envelope	T0+48 → T0+66	Aircraft 001 flies as a pure coaxial helicopter: mast locked, stages pinned, merge mechanism installed but inhibited. IBC control laws, hover performance, autorotation	Hover-to-200 kt envelope cleared, autorotation demonstrated at MTOW, IBC washout schedule validated against predicted figure of merit
4 · First manned in-flight merge	T0+66 → T0+78	The programme's decisive phase. Aircraft 001 with a spin-recovery chute and telemetry (a test installation only — there is no operational airframe parachute), over water, chase-escorted, expanding from stop-straight only, to merge-without-sweep, to the full sequence	Repeatable merge, sweep, un-merge and vertical landing in one sortie, plus ABORT TO ROTOR demonstrated from every step M1 through M3
5 · Supersonic envelope expansion	T0+78 → T0+96	Aircraft 002 joins: progressive sweep-and-Mach expansion 30° → 45° → 60°, CG transfer law validation, thermal soak survey, fuel-sink reversion timing, deliberate in-flight nozzle-kill and dead-stick trials on the morphing tail	M 1.5 at 12 km achieved and repeated; M 1.15 dry supercruise confirmed; dead-stick glide ratio measured against the 9:1 target
6 · Mission systems & lethality	T0+90 → T0+114	Conformal AESA smart-skin integration and calibration; datalink, IFF, EO/IR and self-protection integration (§ 13); chin cannon firing trials through the trapdoor, including the hover ingestion case; supersonic store separation from the trapeze ejectors; extraction-module fit and rescue trials	Gun placard lifted from M 1.2 to the full envelope with no compressor event, and the 40 kt lower placard resolved by measured hover firing over a pad; clean store separation demonstrated at M 1.5; RCS and hub unintentional emission measured against the signature specification; datalink and MAWS/DIRCM performance demonstrated against the § 13 requirement
7 · Qualification & initial production	T0+108 → T0+138	Fatigue and static test articles to completion; icing, EMC/HIRF/lightning, environmental (MIL-STD-810), arctic and shipboard trials; air-to-air refuelling clearance; two-aircraft operational evaluation; LRIP tooling	Type qualification and first delivery; mechanism life substantiated to 10 000 merge cycles by test article, not by analysis; arctic and shipboard limits published as numbers, not intentions

Sequencing principle

Nothing supersonic is attempted until the merge is boringly repeatable, and nothing is attempted on a crewed aircraft that has not first been flown unmanned at 1:4 scale. Phase 3 deliberately flies the prototype as a helicopter with the transformation inhibited — proving the aircraft can always come home before it is ever allowed to change nature.

Parallel de-risking

The Phase 0 fallbacks are carried, funded, to the end of Phase 4: a mechanically hinged droop leading edge, hydraulic merge latches, and a subsonic-cool-down reversion. Each is heavier and each forfeits its mass credit in § 9 — they exist so that a Nitinol setback delays a feature, never the programme.

Cost centre of gravity

Roughly 55 % of non-recurring effort sits in Phases 0–2, before a crewed airframe ever leaves the ground. That distribution is intentional and is the programme's principal financial signature: this is a concept that is cheap to fly and expensive to prove.

Programme kill criteria — stated in advance, deliberately: if the sub-scale demonstrator cannot complete 25 airborne merges in Phase 2, or if the 1:1 rig cannot reach 2 000 cycles in Phase 1, the concept is not viable and the programme stops there — before a crewed prototype exists and before the sunk cost makes the decision for us. Compagnie Apex commits to those two gates as written, and to publishing the result of each either way.

13. Mission Systems, Connectivity & Self-Protection

Earlier issues of this specification described an airframe and a transformation in great detail and were almost silent on what makes it a combat aircraft. A stealth VTOL interceptor with no datalink, no designator and no missile-warning system is an airshow item. This section closes that gap. Nothing below is exotic; all of it is mass, volume, power and cooling that the mass statement of § 9 now carries in the 470 kg avionics line.

Capability	Baseline fit	Why it is not optional here
Datalinks	Link 16 for coalition interoperability, a low-probability-of-intercept directional link (MADL/TTNT class) for stealth-to-stealth work, and beyond-line-of-sight SATCOM on a conformal spine aperture	The aircraft's whole value is arriving somewhere nobody expected. Without BLOS it cannot be re-tasked once it is there, and without an LPI link it must broadcast to be useful — which deletes the stealth it paid 740 kg for
Navigation in GPS-denied conditions	Laser-gyro INS with celestial and vision-aided drift correction, terrain-referenced navigation against the DAS, M-code GNSS with CRPA anti-jam	A nap-of-the-earth insertion profile into a defended area is by definition a jammed environment; navigation accuracy is what the extraction point depends on
Identification	Mode 5 / Mode S IFF with conformal interrogator and transponder apertures	A supersonic interceptor that cannot identify is a supersonic fratricide risk; also the condition of entry into any civil or coalition airspace
Targeting	Internal retractable EO/IR turret with laser designator and spot-tracker in a ventral bay-adjacent well, behind its own snap-action door	The strike role in § 1 is unexecutable without organic designation. External pods are forbidden by the same requirement that forbids external stores, so the well is designed in from the start rather than retro-fitted into the weapons volume
Missile warning and countermeasures	Distributed-aperture MAWS (shared with the DAS belt), laser warning, expendable chaff/flare/decoy dispensers in the tail boom, and provision for a DIRCM turret in the extraction fit	The nap-of-the-earth regime this aircraft is built for is MANPADS country. A rotorcraft without MAWS/DIRCM in that regime is not survivable at any RCS
Electronic warfare	Radar warning and jamming functions hosted in the same conformal AESA skin (§ 7), sectorised so battle damage degrades coverage rather than function	The array already exists; sharing it is what makes an EW fit affordable in mass terms
Sensor fusion & crew interface	Open-architecture mission computing (FACE/OMS-class), single fused track picture on the HMD of § 10	Two crew flying a machine that changes nature mid-sortie cannot also correlate raw sensor feeds
Manned-unmanned teaming	Level-4 control of two collaborative air vehicles from the right-hand seat over the LPI link	The single-airframe magazine is 350 rounds and one bay. Reach and magazine depth come from teaming, not from growth of this airframe

Honest note on the sensor budget: the 470 kg avionics line is a target, and it now has to cover a conformal AESA, a DAS belt, an EO/IR turret, MAWS, a laser warner, dispensers, three datalinks and an anti-jam navigation suite. On a conventional fighter that set is heavier. This line is the most likely source of the next mass growth, and it is flagged here rather than discovered later.

14. Certification, Qualification & Airworthiness Framework

A specification that never names a standard cannot be costed, scheduled or audited. The framework below is stated so that the Phase 12 gates have something to be measured against, and so that the novel items — a flight-critical transformation, a shape-memory structural joint, a megawatt bus — are pulled into an existing safety process rather than invented alongside one.

Domain	Standard	Programme-specific consequence
Airworthiness basis	Military type certificate under a national military airworthiness authority, with MIL-HDBK-516 as the certification criteria set; civil FAR/CS-29 and CS-25 used as reference where no military criterion exists	There is no existing criteria set for an in-flight rotor-to-wing transformation. A special condition must be negotiated and agreed before Phase 4, defining the merge as a flight-critical function with a quantified failure budget
Safety assessment	SAE ARP4754A / ARP4761 — FHA, PSSA, SSA, common-cause analysis	The merge, the single mast, the single engine and the single pivot all attract catastrophic-classification failure conditions. The target of < 10⁻⁷ per flight hour for loss of the aircraft due to a merge failure is the number the whole mechanism design must be shown against
Software / complex hardware	DO-178C DAL A for the FBW and merge laws; DO-254 DAL A for the flight-control and IBC drive hardware	The IBC damping law is flight-critical (§ 6, ground resonance) and cannot be treated as a performance feature
Environmental	MIL-STD-810 (temperature, altitude, vibration, sand, icing, salt fog)	Arctic and shipboard limits of § 16 are declared against these methods, not asserted
EMC / EMI	MIL-STD-461 equipment level, MIL-STD-464 system level	A ± 270 VDC megawatt bus beside a quadruplex FBW and a conformal AESA is an EMC problem of unusual severity; the galvanic isolation of § 4 is verified against 464, by test. The dominant source is the rotating IBC hub (§ 3): four switching drives and a sliding contact turning at ≈ 313 rpm a metre from the spine array. Compliance must be demonstrated with the hub turning and the EMAs commanded — a static bench result is not acceptable evidence, and hub emission is additionally assessed as an unintentional-radiation signature item
Lightning / HIRF	MIL-STD-464 and DO-160 sections 22/23	A composite airframe with no metallic skin needs a designed lightning current path around the mast, the SMA rows and the bay
Structural substantiation	Damage-tolerant / safe-life hybrid; full-scale static and fatigue articles	The SMA locking pins and shear keys have no accepted certification precedent and require an agreed means of compliance covering fatigue, inspection interval and failure detection (§ 17, NDT)
Weapons	MIL-STD-1760/1553 and national store-certification process	Store separation clearance is built subsonic → transonic → M 1.5, per Phase 6

15. Survivability, Safety & Escape

Escape — what replaces the airframe parachute

• The parachute is deleted. The heaviest certified whole-aircraft ballistic recovery system is in the ≈ 1.5–2 t class. A canopy for a 10.8 t airframe is 1 000–1 900 m² of fabric plus mortar and staging: 400–800 kg, none of which appeared in the 280 kg “crew & escape” line that also contained two ejection seats. It was not a conservative feature; it was an unbudgeted one.
• Crew: two zero-zero ejection seats, envelope 0/0 to 600 kt EAS / 15 km, MDC canopy severance qualified for the 62° rake. Blade severance is used only with rotors turning (§ 7).
• Bay occupants: the extraction module can be fitted with two downward-firing ejection seats (B-52 lower-deck precedent), usable above 1 500 ft; below that, and in the extraction fit without them, their case is the glide and the un-merge-and-autorotate path.
• Alternative traded and rejected: a two-place jettisonable crew capsule (F-111, B-1A precedent) at ≈ +300 kg. Rejected on mass and on the fact that it protects only the crew, not the bay — the same limitation as ejection seats, at three times the weight.
• The residual, stated: there is a low-altitude jet-mode dead zone with no escape solution. That is a declared limitation of this configuration, not an oversight.

Battle damage and single-point failures

• The mast is the aircraft. One titanium column carries lift, the pivot and the merged wing. It is designed with a dual load path (concentric inner and outer members sharing bending), damage-tolerant to a 12.7 mm through-shot in either member, with strain instrumentation feeding the HUMS.
• Each scissor pivot carries one half-wing. Redundant bearing races and a secondary retention collar are required so that a bearing failure is a jam, not a separation.
• Fuel: OBIGGS inerting of every tank and the vapour space; self-sealing bladders in the ventral cells; crashworthy break-away couplings and frangible fittings to the 4 m/s sink case; the fuel-sink purge loop isolatable in one command.
• Fire: dual-loop engine-bay fire detection with two-shot extinguishing; separate detection for the supercapacitor bay, whose failure mode is thermal runaway rather than fire and which vents overboard through a dedicated frangible panel.
• Electrical: lightning and HIRF paths designed into the composite; the ± 270 VDC bus segmented so a single arc-fault event cannot propagate to the essential bus; EMP hardening of the flight-control channels as a stated requirement.
• Bird strike: canopy, both rotor stages and the dorsal inlet lips certified to the applicable bird-strike case — noting that a stopped, merged blade is a fixed leading edge and must meet the fixed-wing criterion, which is more severe than the rotorcraft one.

16. Operations, Basing & Environment

Topic	Requirement / provision	Open issue
Air-to-air refuelling	Retractable probe behind a flush door on the starboard forebody, drogue-compatible, baselined rather than deferred; boom receptacle offered as a customer option in the spine	With no external tanks and ≈ 560 km of radius, AAR is the only way this aircraft reaches strategic distances — the fuel state is the mission radius (§ 4). Probe-and-drogue contact at the merged configuration's approach speeds, and the wake of a 60°-swept delta behind a tanker, are Phase 7 clearances
Arctic operations	− 40 °C ground start declared as a requirement: supercapacitor bank heated and pre-conditioned from ground power, SMA rows interlocked out below − 25 °C skin temperature, low-freeze-point kerosene specification (§ 4), snow-ingestion and whiteout-hover procedures	Cold is the SMA's second problem after fatigue: a −40 °C soak widens the heating energy required for every morph and every pin set. The 40 kW / 8 s figure of § 5 is a temperate-day number
Shipboard	Rotor Ø 12.8 m does not fold — the spot requirement is therefore a full 12.8 m circle plus clearance. Deck-lock harpoon, tie-down rings, folding-free stowage, marinised coatings	Galvanic corrosion is a design issue, not a coating issue: Nitinol, titanium and carbon in a salt-fog environment is an aggressive couple. Phase 7 currently promises shipboard trials without a naval design underneath them — that is a gap, and it is named here
Transportability	12.8 m rotor exceeds the cargo box of a CC-177 / C-17. Blade removal is therefore a designed operation: 4 blades off at the root fittings, mast unshipped, aircraft ≈ 14.2 × 4.2 m	Blade removal and reinstallation must not require re-indexing to ± 0.2° by shop equipment — a field-rig requirement that does not yet exist
Pad and deck limits	Slope limits 8° nose-up / 6° nose-down / 6° lateral; ship-motion limits to be established; brownout mitigation by the DAS-driven hover-hold and pad-precision modes of § 7	The C/C ventral plate answers plume bounce-back thermally; it does not answer recirculating debris into a dorsal inlet at low hover height
Cabin environment	Cockpit pressurised to a 2.4 km cabin altitude at the 12 km design point; OBOGS with backup oxygen for four occupants; bay conditioned and pressurised in the extraction fit	Never stated in earlier issues despite a 12 km design point and two passengers who are not aircrew
Embarking personnel	Baseline is landed pick-up only — ramp-less side door, 40 s target for two occupants. A rescue hoist is not in the baseline: it needs an external boom that the mould line forbids	If hover extraction is a requirement, it must be stated as one and paid for in signature and mass

17. Programme, Cost, Support & Requirements Traceability

The structural risk this programme has not been treating as one: the engine. No engine in existence, in development or in any published programme combines ≈ 52 kN dry / 75 kN reheat with a 3 800 kW power take-off and a megawatt shaft-embedded starter-generator. The XA100/XA102 adaptive-cycle demonstrators are F135-class in thrust and carry no comparable shaft offtake. What this document calls "the ACE" is therefore not a supplier selection — it is a second development programme of comparable ambition to the airframe, and it belongs in the risk register (where it now is) and in the cost model (where it now is) rather than in an assumption.

• Sovereignty: a GE-class solution places the aircraft under ITAR, which constrains export and, for a Canada/UK/France industrial axis, constrains sovereign upgrade rights. That is a programme-shaping consequence, not a paperwork one.
• The trade study that should exist: an M88-class derivative (Safran) sits at almost exactly the 50/75 kN of this specification and would be the natural European route; a Rolls-Royce partnership is the natural UK route. Neither has an adaptive third stream or a 3 800 kW PTO today. Trade: adaptive cycle with a new core, versus a proven core with a new PTO and a variable-bypass module bolted to it. This trade is a Phase 0 deliverable and has not been done.
• Kill criterion: if no engine partner is under contract with a demonstrated PTO concept by T0+24, the programme stops — added to the § 12 gates.

Cost and support — the two chapters that were entirely absent

Cost model (order of magnitude only)

• Non-recurring: airframe development is a fighter-class programme with a rotorcraft-class programme inside it; ≈ 55 % of non-recurring effort sits in Phases 0–2 (§ 12). The ACE is a separate non-recurring line of the same order and must be shown separately, never absorbed.
• Unit recurring flyaway: to be estimated against a 5th-generation single-engine fighter baseline plus the mechanism, the SMA rows and the IBC hubs, minus the deleted hardware of § 9. No credible number can be quoted at this stage; quoting one would be worse than the omission.
• Cost per flight hour and life-cycle cost are the figures an operator actually buys on, and they are adverse for this configuration: two dynamic systems, a life-limited transformation mechanism and an SMA joint with a 10 000-cycle life. The programme must produce a CPFH estimate at Phase 1, not at Phase 7.
• Alternatives trade study — missing and required: a compound helicopter and a tiltrotor both reach a large part of this mission at a fraction of the risk. The HSR is only justified if the supersonic dash is a genuine requirement rather than an aspiration, and that comparison must be documented before Phase 1.

Support, maintenance and training

• HUMS/PHM as a certification enabler, not a feature: mast strain, pivot bearing health, pin pre-load, collar rail wear and brake energy accumulation are recorded every flight. The merge-cycle counter is the aircraft's primary life-usage parameter.
• NDT of the SMA pin rows: there is no accepted field inspection method for a shape-memory structural pin. Developing one — eddy-current or ultrasonic, correlated to pre-load — is a Phase 0/1 deliverable, and its absence would force a conservative safe-life replacement interval instead.
• Gearbox TBO and engine change: the coaxial and PTO gearboxes are new designs with no service history; a 1 500-hour initial TBO is assumed and must be substantiated on the bench. Engine change on an unprepared pad requires the PTO to disconnect without a hoist — a designed-in requirement.
• Training and simulation: the merge cannot be trained in the air first. A full-motion simulator with a validated merge model is a Phase 3 deliverable, ahead of the first manned merge, and dual-qualification (rotary and fast-jet) is an aircrew-selection constraint with recruiting implications.

Weapons compatibility — set by the 3.4 m bay

Class	Compatible (≤ 3.4 m)	Excluded
Air-to-air	MICA IR/EM (3.10 m), IRIS-T (2.94 m), AIM-9X (3.02 m), ASRAAM (2.90 m), Meteor only if a 3.65 m bay were adopted	AIM-120 AMRAAM (3.65 m) — the single most consequential exclusion for an interceptor
Air-to-ground	SPEAR 3 (1.80 m), GBU-53/B SDB II (1.76 m), Brimstone (1.80 m), 250–500 lb-class PGM	2 000 lb-class weapons; any cruise missile of stand-off length
Programme choice required	Either the interceptor role accepts a short-range-only missile fit, or the bay grows to 3.65 m — which lengthens the area-ruled waist and re-opens the § 2 volume distribution. A second, independent decision sits beside it: two trapeze stations cap the strike fit at ≈ 200–460 kg regardless of the 950 kg payload allowance (§ 8). Neither decision has been taken; both should be taken before Phase 1.

Signature — the contributor budget that exists in place of a number

No RCS value, modelled or measured, exists for this aircraft. Quoting one would be inventing it. What can honestly be produced before the Phase 0 model is a ranked contributor budget: which features dominate the return, in which sector, and what the design lever against each one is. It is a list of where to look, not an answer — and it is deliberately expressed in relative rank rather than in dBsm.

Contributor	Dominant sector	Mechanism	Design lever — and what it costs
1 · Semi-flush saddle step and hub cavity mouth	Upper hemisphere, broad in azimuth	A forward-facing step, an open cavity and a discontinuous seal line on the spine — cavity resonance plus travelling-wave return	The lever is depth, and there is none (§ 3). Remaining levers: edge treatment, RAM in the cavity, and sealing at the three sweep detents. This is the direct signature price of withdrawing "deep flush", and it is expected to dominate
2 · Blade edges and hub, unmerged	All sectors, helicopter mode	Rotating blades are the largest radar spike of any helicopter; here they are also two stacked stages and an exposed hub	Merging deletes blade flash entirely — which is the single largest signature argument for the whole concept, and it only applies in jet mode
3 · Iris-door edges and sweep-detent seams	Upper and side	Every seam is an edge; three detents mean three different seam geometries, and only the detents are sealed	Edge alignment to the planform angles, serration, conductive gaskets. Cheap in mass, expensive in mechanism reliability
4 · Dorsal inlet and S-duct termination	Front and upper front	Duct cavity return; the auxiliary hover doors are additional apertures on the same shoulder	S-duct shielding is already baselined; the auxiliary doors are an unresolved addition and their area is still TBD (§ 4)
5 · Chine-to-wing-root junction	Front and side	A moving junction: the root meets the chine at a different angle at every sweep setting	Only two of the three sweep angles can be optimised. A variable-geometry aircraft cannot be edge-aligned in every configuration — the dash detent is the one to optimise
6 · Canopy and cockpit cavity	Front	Standard fighter contributor	Conductive coating and the 62° rake, both already in the configuration
7 · Hub unintentional RF emission	Not an RCS term at all	Four switching drives and a sliding contact at ≈ 313 rpm under a RAM cover (§ 14)	Faraday hub, rotary transformer, MIL-STD-461/464 with the hub turning. A stealth aircraft that broadcasts is not stealthy at any RCS

What the budget implies before anyone models it: the aircraft's front sector is conventional and probably achievable; its upper hemisphere is where the concept pays for its mechanism, and the upper hemisphere is exactly the sector a look-down radar or an airborne interceptor occupies. That is a mission-level consequence of a mechanical decision, and it is the first thing the Phase 0 model must be pointed at.

Alternatives trade study — the framework, and the first-order answer

The study itself is a Phase 0 deliverable and has not been done. What can be fixed now is the thing that usually makes such comparisons useless: the mission set and the axes. Four missions, one common set of assumptions, three configurations. The entries below are first-order and qualitative; the Phase 0 study replaces every one of them with a number.

Axis	Compound helicopter	Tiltrotor	HSR (this configuration)
M1 · VTOL insertion / extraction, 400 km radius, unprepared pad	Met, at lower risk and lower cost	Met, with a larger pad footprint	Met
M2 · Nap-of-the-earth penetration under a defended sky	Met; signature poorer, blade flash permanent	Met; prop-rotor signature and acoustics poor	Best — the merged, stopped rotor deletes blade flash
M3 · Time-critical response, 560 km, dash on arrival	Not met above ≈ 250 kt	Not met above ≈ 300 kt	Met — and this is the only mission the configuration uniquely serves
M4 · Supersonic intercept from a dispersed pad	Not met	Not met	Met
Technical risk	Low — flown (X2/S-97, Raider)	Low — in service (V-22, AW609)	Very high — the merge has never been flown
Non-recurring cost	Reference	≈ reference	Reference × several, plus a second engine programme
CPFH	Reference	Above reference	Well above — see below

The first-order answer, stated before the study is run so that the study cannot be written to a conclusion: M1 and M2 do not justify this aircraft — a compound helicopter reaches them at a fraction of the risk and cost. Only M3 and M4 do. Therefore: if the supersonic dash is not a hard, funded requirement, the correct answer is a compound helicopter and this programme should not start. That sentence is the actual purpose of the trade study, and the programme is committed to publishing its result either way.

Cost per flight hour — the drivers, ahead of the estimate

The estimate is a Phase 1 deliverable and no number is quoted here. The drivers can be ranked now, and they are ranked deliberately because they are what an operator will discover in service if the programme does not surface them first.

Driver	Why it is adverse here	What could contain it
Two dynamic systems on one airframe	A rotorcraft transmission and a supersonic propulsion installation, each with its own inspection regime, on an airframe that flies both duty cycles in the same sortie	Nothing deletes it — it is the concept. HUMS-driven on-condition maintenance instead of fixed intervals is the only real lever
The 10 000-cycle merge mechanism	Life-limited rails, collar, pivots, brakes and pins whose usage parameter is the merge counter, not flight hours. Four merges per sortie consumes life fast	Substantiating life above 10 000 cycles on the Phase 7 test article; FBW enforcement of the four-merge limit already exists
HTSMA pin rows with no field NDT method	Without an accepted inspection, the certification fallback is scheduled safe-life replacement — the single most expensive outcome available	The eddy-current / ultrasonic method correlated to pre-load, a Phase 0/1 deliverable. Its absence is a direct CPFH penalty, not just a paperwork gap
New-design gearboxes with no service history	The 1 500-hour initial TBO is an assumption; early TBOs on new rotorcraft gearboxes are historically optimistic	Bench life-cycle rig ahead of first flight; fleet-leader programme
Low-observable surface maintenance	RAM, seams and — uniquely here — a moving saddle and door set that must keep sealing after thousands of cycles	Restoration standards written at design time, not after the first depot visit
Aircrew currency	Dual rotary and fast-jet qualification, plus merge currency that cannot be maintained cheaply in the air	The full-motion simulator with a validated merge model — already a Phase 3 deliverable, and it is a cost driver as much as a safety one

Key performance parameters — the traceability table

Every KPP below now carries the section that specifies it, so that a reader can walk from a requirement to the engineering that claims to meet it and back again. Rows added in this issue are the ones an external review found to be asserted in prose but absent from the parameter set: merge-window lift closure, whirl-flutter margin, the dash-condition roll rate, the realisable strike load as distinct from the payload allowance, magazine depth, and the mechanism mass ceiling.

KPP	Threshold	Objective	Specified in	Verified by	Status
VTOL from a 25 × 25 m pad at MTOW	ISA+20, 4 000 ft	ISA+20, 7 500 ft	§ 4, § 6	Phase 3 flight test	Sized, unproven
In-flight merge and un-merge	Repeatable, with abort from every step	4 cycles per sortie	§ 3, § 6	Phase 2 (unmanned), Phase 4 (manned)	The programme's binding gate
Merge-window lift closure	C_L closed at every instant of M0 → M4 across 235–265 kt, sea level – 2.5 km	Closed with a + 3.5 g gust superimposed at the worst instant	§ 3, § 6	CFD Phase 0–1, sub-scale Phase 2, manned Phase 4	Body + stopped stages + 30–40 % residual rotor lift not yet summed instant by instant
Whirl-flutter stability of the stopped rotor	Boundary 25 % above the 265 kt merge ceiling (≥ 330 kt), mast/pylon stiffness floor declared	≥ 40 % margin	§ 3, § 6, risk register	Analysis + full-scale whirl rig Phase 0–1, scaled confirmation Phase 2	Stiffness floor not yet a number — XV-3 failure mode, no flown analogue
Inter-stage clearance	0.62 m (0.048 D) held under the + 3.5 g gust case with proximity sensing and IBC flap suppression	Decision frozen: confirm 0.62 m or adopt the 0.95 m fallback	§ 3, § 6	Phase 0 whirl rig, Phase 1 gate	Half of flown rigid-coaxial practice (0.09–0.10 D)
Dash speed	M 1.4 at 12 km	M 1.5 at 12 km	§ 6, § 11	Phase 5	Flutter-limited, predicted at the objective
Dry supercruise	M 1.10	M 1.15	§ 6, § 11	Phase 5	Sized. Stretch to M 1.20–1.25 by third-stream scheduling and residual-drag reduction — the cheapest headroom in the design, no new hardware
Supersonic lift-to-drag ratio	L/D ≥ 5.0 at M 1.5 / 60°	L/D 5.5–6.0 — the old 6.5 objective is withdrawn	§ 4, § 6	CFD Phase 0–1, measured Phase 5	Penalties now debited in the § 6 build-up: 5.3–5.7 clean-assumption, 5.0–5.4 if S_body disappoints. Thrust–drag closure remains formally open — installed thrust at 12 km is an ACE-partner input that does not exist
Combat radius, unrefuelled	500 km	560 km	§ 6, § 16	Phase 5/7	Sized at the corrected MTOW; strategic reach depends on the AAR clearance below
Roll rate, merged — transonic	≥ 70 °/s at M 0.9 / 30° sweep	≥ 90 °/s at M 0.9 / 30° sweep	§ 6, § 7	Phase 5	Unverified — aeroelastic twist is the only roll effector, and this is the binding aeroelastic case at q ≈ 57 kPa (§ 6)
Roll rate, merged — dash	≥ 40 °/s at M 1.5 / 60° sweep	≥ 50 °/s at M 1.5 / 60° sweep	§ 6, § 7	Phase 5, with dedicated aeroelastic-reversal trials	Unverified; q ≈ 30 kPa here, so the dash is not the critical case — the low-level transonic point is
Wing-box torsional stiffness (roll-authority enabler)	q_div ≥ 1.5 × q_max ≈ 85 kPa; twist effectiveness η bounded 0.8–2.2	GJ ≥ ≈ 1.7 MN·m² per half-panel with \|e\| ≤ 0.20 c, sign controlled	§ 3, § 6	Structural freeze Phase 1; GVT and flutter model Phase 2; flight Phase 5	First-order check shows a plain single-cell box ≈ 3× short — the "GJ × 5" mitigation is now quantified, not asserted
Dead-stick control authority	Morphing ruddervators take pitch and yaw in < 150 ms at realistic hinge moments, glide ratio ≥ 8:1	9:1 measured	§ 7, § 15	Rig demonstration at representative hinge moments before the Phase 5 in-flight flame-out trials	Critical since the airframe parachute was deleted
Realisable strike load	≥ 400 kg on two trapeze stations	≥ 650 kg, or a station-count increase inside the same 3.4 m bay	§ 8, § 9, § 17	Phase 1 store-integration study, Phase 6 separation trials	Distinct from the 950 kg allowance, which is a structural and balance envelope
Magazine depth — now a KPP	Level-4 control of 2 collaborative air vehicles	4 vehicles, with organic re-tasking over the LPI link	§ 13	Phase 6	Two stations and 350 rounds make teaming the magazine, not an accessory
Transformation mass penalty	≤ 7 % MTOW (756 kg) with payload the declared victim of any breach	≤ 6.5 % by targeted reduction on the offset rails and carriage, the iris doors and the incidence gimbal	§ 9	Monthly mass review; independent audit at the Phase 0 gate	Frozen contractually at 740 kg gross / + 110 kg net
Loss-of-aircraft rate from merge failure	< 10⁻⁶/FH	< 10⁻⁷/FH	§ 14, § 15	ARP4761 SSA + Phases 1–4	Not yet assessed
Signature	LO objective, unquantified in this document	A modelled RCS value on record before Phase 6 — semi-flush saddle, seams, iris doors and hub unintentional emission	§ 2, § 6, § 15	Preliminary RCS model at the Phase 0 gate; measurement Phase 6	No modelled or measured value exists
Air-to-air refuelling	Probe-and-drogue contact cleared in the merged configuration	Boom receptacle option cleared	§ 16	Phase 7	Probe baselined; the 560 km radius makes this the enabler, not an accessory
Extraction method	Landed pick-up, two occupants in 40 s	Hover extraction only if it is stated as a requirement and paid for in mass and signature	§ 16	Phase 6/7	Decision open — an external hoist boom is incompatible with the mould line, so this is a configuration decision, not a fit
Arctic and shipboard limits	− 40 °C ground start; galvanic compatibility of Nitinol / titanium / carbon in salt fog	Limits published as numbers	§ 14, § 16	Design requirements at Phase 0, trials Phase 7	Under-analysed — named as a gap rather than discovered in Phase 7

Placards and declared limitations — collected in one place

These limits are stated across §§ 3–16 in their engineering context. They are repeated here as a single list because an operator, a reviewer and a test pilot all need to read them together, and because a limitation that only exists inside a paragraph is a limitation that gets forgotten.

Placard / limitation	Value	Why it exists	How it lifts
Gun firing, upper limit	M 1.2	Dorsal S-duct ingestion of muzzle gas is designed to be precluded, not yet verified	Phase 6 firing trials with no compressor event
Gun firing, lower limit	Inhibited below 40 kt	At 0 kt the recirculation fountain feeds the auxiliary doors at maximum PTO power and minimum surge margin. Accepted capability loss: no hovering fire support in the baseline	Phase 6 measured hover firing over a pad; a muzzle deflector and burst-duration auxiliary-door isolation are carried as trades — door closure is not available at 0 kt, where it would starve the engine
Supersonic flight below 3 km	Prohibited	Thermal soak toward 130 °C against RAM (≈ 170 °C) and the 20 °C margin under HTSMA A_s (leading edge 150 °C, pins 170 °C)	Not lifted — a design limit, not a test placard
Merge window	235–265 kt, sea level – 2.5 km, ≈ 32 s	Set by lift closure, not by the mechanism	Widened only if CFD and flight test show C_L closes outside it
Merges per sortie	4, FBW-enforced	Rotor-brake energy 3–7 MJ per stop and the conducted heat path into hub and blade roots	Phase 1 brake-pack thermal substantiation
Load factor	+ 6 / − 2 g merged · + 3.5 / − 1 g helicopter	The wing box takes fighter loads; the free rotors cannot	Not lifted
CG and fuel transfer	22–38 % MAC; transfer inhibited below 450 kg of fuel, sweep then capped at 45°	Below that state the aircraft is committed to recovery	Not lifted
Static margin	± 3 % MAC to 45° sweep; ≈ 6 % MAC nose-down at 60°, nozzle-trimmed at ≈ 1.5° for ≈ 0.2 kN of trim drag	The aerodynamic centre marches further than any tank arm this airframe can offer	Not lifted — priced instead
Sweep detents and door sealing	30° / 45° / 60° — the semi-flush apertures seal at these three detents only	A 4.1 m airframe has no spare depth for a continuously sealed saddle	Residual step drag and seam return quantified by Phase 0–1 CFD and Phase 6 RCS
Retraction aperture jammed (hub proud)	Envelope capped at M 1.2	Drag and signature penalty only — panels are spring-loaded open so the reverse path is never blocked	Degraded state; no lift
SMA rows in cold soak	Interlocked out below − 25 °C skin temperature; anti-ice capped at 110 °C and interlocked	A − 40 °C soak widens the heating energy for every morph and every pin set — the 40 kW / 8 s figure is a temperate-day number	Phase 7 arctic trials against Phase 0 design requirements
Degraded-state recovery	Fixed-wing landing at 170–190 kt on a ≈ 1 200 m prepared strip	Asymmetric pin set, jammed collar or failed un-merge. The aircraft is runway-independent in every nominal state, not in all failure states	Not lifted — basing plans must carry a diversion field
Bay stores	Maximum length 3.4 m — AIM-120 AMRAAM (3.65 m) excluded; two trapeze stations	Bay volume inside the area-ruled waist	Only by a configuration decision to grow the bay to 3.65 m (§ 17)
Escape	Low-altitude jet-mode dead zone with no escape solution	The whole-airframe ballistic parachute was deleted as unbudgeted and unprecedented (§ 15)	Not lifted — a declared limitation of the configuration

Programme Risks & Mitigations

Risk	Level	Mitigation
Merge failure in flight test	HIGH	1:1 ground rig, sub-scale drone campaign, 235–265 kt window, automatic 200 ms abort, spin-recovery chute on the demonstrator
Whirl flutter of the stopped rotor at 235–265 kt	HIGH	New entry — previously absent from this register. Two heavy stages cantilevered on one mast is the mechanism that destroyed the XV-3. Mitigation: mast/pylon stiffness floor derived before Phase 1, whirl-mode stability analysis with a 25 % speed margin, whirl rig at full scale, scaled confirmation in Phase 2, and a hard FBW speed clamp. Not retired by analogy with any flown aircraft — there is none.
ACE dependency — the engine does not exist	STRUCTURAL	New entry. No engine anywhere combines 52/75 kN with a 3 800 kW power take-off and a megawatt shaft-embedded ISG. This is a second mega-programme inside the first one, with its own schedule, cost and sovereignty exposure (§ 17). Mitigation: signed development agreement as a Phase 0 exit gate, a sovereignty trade study against an M88-class derivative, and an explicit programme kill criterion if no partner commits by T0+24. The power take-off itself is now bounded to three candidate architectures (§ 4) — front bevel offtake (baseline, lightest), a dedicated free power turbine in the third stream (+ 150–250 kg, and the likeliest thing an engine house will actually quote), or an electrical PTO (rejected at 3 800 kW). The mass statement must carry the free-turbine option as a declared contingency rather than discover it at contract signature.
Chord-offset carriage — eccentric load path and fairing retraction	MEDIUM–HIGH	New entry, created by the corrected kinematics of § 3. The upper stage's root reaction no longer passes down the mast axis. Mitigation: eccentric-carriage load case substantiated on the Phase 1 rig; second indexing axis with ± 3 mm tolerance; spring-loaded fairings cycled 20 000 × on the bench.
Inter-rotor clearance (0.62 m = 0.048 D) under the +3.5 g gust case	MEDIUM–HIGH	Half of flown rigid-coaxial practice. Blade-tip proximity sensing, IBC flap suppression, hard flapping placard; fallback to 0.95 m spacing (+18 kg mast, +3 s stroke) if the whirl-rig data does not close. A blade-to-blade contact is a loss of aircraft, not a degraded state.
Ground / air resonance with electrically damped lag modes	MEDIUM	A swashplateless rigid coaxial head makes lag damping a software function. Loss of the IBC damping law must be shown benign at every rpm; ground-resonance test at MTOW in Phase 3.
Deep stall / pitch-up of the merged delta with no elevons	MEDIUM	Recovery authority is a vectoring nozzle that needs thrust. Spin-tunnel and free-flight model work in Phase 2; AoA limiter authority proven before the first manned merge.
Nitinol fatigue — morphing edge and locking pins (10 000 cycles)	HIGH	The specification's named development axis: alloy programme, coupon-to-full-span pyramid, hinged droop-edge and hydraulic-latch fallbacks carried in parallel
PTO gearbox + dog clutch (3 800 kW, 2 s disengage) and megawatt ISG on the same spool	HIGH	Named as a pacing item alongside the Nitinol, the merge and the IBC hub; bench life-cycle rig before first flight; torque-fuse protection. The 2 s disengage and the combined shaft-offtake rotordynamics do not close at preliminary level.
Asymmetric Nitinol pin set (one panel locked, one not)	MEDIUM–HIGH	Differential pin-strain detection, reversible collar, land as stage-M1 aeroplane; bank clamp 30°
Merged-wing flutter and scissor-pivot loads (5.0 m × 1.55 m half-panels, swept 60°)	HIGH	Single-mast pivot at this panel slenderness and Mach has no precedent; the flutter boundary is currently predicted at M 1.5 on unvalidated models. Closed-box GJ × 5, active-twist flutter suppression, wind-tunnel + GVT, incremental envelope expansion. This risk is what caps the design at M 1.5 and it is not retired by analysis.
Wing-box torsional stiffness — the enabler of the only roll effector	HIGH	New entry, and now quantified rather than asserted. The binding aeroelastic case is M 0.9 at low level (q ≈ 57 kPa), not the M 1.5 dash (q ≈ 30 kPa). Holding q_div ≥ 85 kPa on a panel whose mid-chord shear-key row drives the shear centre aft to e ≈ 0.20 c demands GJ ≈ 1.7 MN·m² per half-panel against ≈ 0.54 MN·m² for a plain single-cell box — a factor of about three (§ 6). Mitigation: elastic-axis position frozen as a structural constraint before the Phase 1 rig, ±45° shear-ply build-up, twist command gain-scheduled on measured q with effectiveness bounded 0.8–2.2, and loss of that schedule treated as a control-authority failure case.
Thrust–drag closure at the M 1.5 design point	STRUCTURAL	New entry. With the semi-flush, iris-seam and chine-junction penalties debited, drag at the design point is 16.6–19.0 kN (§ 6). This document has been comparing that against a 52 kN sea-level static dry rating, which is not the relevant quantity: installed net thrust at M 1.5 and 12 km is. That figure is an ACE-partner input that does not exist, so the closure is formally open and the dash may prove reheat-dependent. Mitigation: installed-thrust deck required as a deliverable of the Phase 0 engine agreement, before any performance figure in § 6 is treated as sized.
Asymmetric scissor sweep (one half jams)	MEDIUM	FBW re-symmetrises to the jammed angle in < 300 ms, nozzle absorbs the transient couple; envelope capped at the sweep achieved
Dorsal retraction clash — cavity volume, door sealing and V-tail clearance	HIGH	New entry. A 4.1 m-tall airframe has no spare depth under the spine: the S-duct, the engine, the 1 000 kg aft cell and the ventral bay already own it, and the 1.55 m root sweeps 30° → 60° over stations the V-tail occupies. Mitigation: “deep flush” withdrawn in favour of a semi-flush saddle, a two-part sliding aperture sealing at the three sweep detents only, and a fin-root station constraint. Residual step drag and seam return are TBD by CFD and Phase 6 RCS. This is a configuration-layout risk, and it is the one most likely to force an airframe change.
Retraction aperture jams or fails to seal	MEDIUM	Panels spring-loaded open (reverse path never blocked); proud-hub case is drag/RCS penalty only, envelope capped M 1.2
M2 chord-offset mass step during the merge	MEDIUM	New entry. ≈ 520 kg translating 0.70 m aft is ≈ 2.4 % MAC arriving as a step in ≈ 7 s, at the moment the effectors are being handed over. Small in magnitude, badly placed in time. Mitigation: open-loop forward pre-bias of the transfer pumps plus a scheduled nozzle input, validated on the Phase 2 demonstrator before any manned merge.
Gun-gas ingestion in the hover	MEDIUM–HIGH	New entry. The forward-flight geometry argument fails at 0 kt: the auxiliary doors breathe the recirculation fountain that the muzzle efflux enters, at maximum PTO power and minimum surge margin. Mitigation: firing inhibited below 40 kt by FBW interlock until measured in Phase 6; muzzle deflector carried as a trade. Accepted capability loss: no hovering fire support in the baseline.
Hub electromagnetic interference on the AESA and the FBW	MEDIUM–HIGH	New entry. Four switching drives on ± 270 VDC rotating at ≈ 313 rpm through a sliding contact, a metre from a conformal array and beside a quadruplex FBW — and a source of unintentional RF emission under a RAM cover. Mitigation: brushless rotary transformer for the power path, Faraday hub enclosure with 360°-terminated shielded feeder, dedicated drive return (never the composite structure), fibre for all data. MIL-STD-461/464 compliance must be shown with the hub turning, not on a static bench.
Payload allowance not realisable on two stations	MEDIUM	New entry. 950 kg is booked; ammunition plus two trapeze stations realise ≈ 400–650 kg (§ 8). Mitigation: declare the aircraft a two-shot platform whose magazine depth comes from MUM-T (§ 13), or open the station-count trade in Phase 1. Either way the allowance and the load must stop being quoted as the same number.
Fuel-sink cooling loop failure (Nitinol stuck sharp)	MEDIUM–LOW	Fallback is the legacy subsonic cool-down orbit (≈ 3 min below M 0.8) — slower reversion, never a lost aircraft; dual purge pumps, tank-to-spar plumbing bench-cycled 20 000 ×
Swashplateless IBC hub — EMA authority and slip-ring life in the rotating frame	HIGH	No mechanical fallback exists (a swashplate would block the merge), so redundancy is depth not diversity: dual-wound EMAs per blade, dual-redundant fibre ring, brush-and-rotary-transformer hybrid transfer, 10 000-hour rotating-rig campaign before first flight
Megawatt electrical architecture — transient management and supercapacitor safety	MEDIUM–HIGH	Iron-bird power rig ahead of the airframe; galvanic isolation of the essential bus verified by test, not analysis; graphene cell abuse-testing (thermal runaway, ballistic penetration) as a qualification gate; bank sited on the CG away from the bay and crew
Morphing ruddervator authority in the dead-stick case	MEDIUM–HIGH	Compliant-surface authority demonstrated on the demonstrator before any supersonic flame-out testing; deliberate in-flight nozzle-kill trials at increasing Mach; supercapacitor-backed amplifiers proven to hold the descent. Raised from MEDIUM: with the airframe parachute deleted, this surface is what stands between a flame-out and an ejection.
Conformal AESA smart skin — structural array under supersonic and rotor-wake loads	MEDIUM	Modules qualified as structural plies (fatigue and impact) rather than as avionics boxes; sectorised architecture degrades gracefully; fuel-sink thermal bonding validated at the 120 °C skin case
Single-engine loss cases	MEDIUM–HIGH	Recovery priority when merged is glide, then un-merge-and-autorotate, then ejection; the 9:1 glide ratio is a target and the morphing-tail reversion is undemonstrated. With no airframe parachute there is an admitted low-altitude dead zone. Accepted per specification, retired only in Phase 5.
Rotor-brake thermal management	MEDIUM–LOW	Downgraded: brake energy re-derived at 3–7 MJ per stop, not 18 MJ. 8 MJ carbon packs, third-stream cooling, FBW-enforced four-merge limit per sortie. The conducted heat path into hub and blade roots remains an open Phase 1 item.
Mast and scissor pivot as a single point of total failure	HIGH	New entry. One mast carries every flight load and one pivot per side carries the wing. Mitigation: dual load path through concentric mast members, damage-tolerant design to a 12.7 mm threat, redundant bearing races, and HUMS instrumentation on both pivots (§ 15, § 17).
Unrefuelled radius against strategic geography	MEDIUM	New entry. ≈ 560 km of combat radius with no external tanks caps the aircraft to a tactical instrument. Mitigation: the retractable refuelling probe of § 16 is baselined rather than deferred, and boom/drogue compatibility is a Phase 7 clearance.
Mass growth spiral	MEDIUM	Merge + flush mechanism frozen early at ≤ 7 % MTOW, monthly mass reviews, contractual OEW target
Competition from mature compounds / tiltrotors	STRUCTURAL	Differentiate on the only thing they cannot do: stealth VTOL and supersonic intercept in one airframe

Concept-killer risk: chasing Mach before the merge is boringly repeatable, loading the prototype with weapons, navalisation and stealth qualification at the same time — or assuming the engine. The first aircraft must prove one thing: stop, merge, un-merge and land vertically — repeatably. Speed comes second, everything else comes third, and none of it happens without an ACE partner under contract.

Chief Engineer's Validation

This specification describes a concept for which no physical show-stopper has been identified at preliminary level. It is not a validated design: it is conditional on five pacing items — high-temperature shape-memory alloy development, the in-flight merge itself, including the chord-offset degree of freedom, PTO / ACE and megawatt-ISG integration, the swashplateless IBC hub, and the existence of the engine at all — each of which must be retired at a declared gate before the next phase is funded.

What changed in this issue. An independent review found four design holes rather than four detail errors, and they are closed above rather than argued with: the merged 1.55 m chord required a chordwise offset that no earlier issue described (§ 3); the merge window did not close in lift and has been cut to 235–265 kt between sea level and 2.5 km with residual rotor lift retained (§ 3, § 6); the CG transfer could never have chased the aerodynamic centre, and the residual is now trimmed and priced rather than promised away (§ 3, § 9); and the whole-airframe supersonic parachute had no precedent and no mass budget and has been deleted (§ 15). Re-closing the mass statement moved OEW to 7 400 kg and MTOW to 10 800 kg.

The aerodynamic compromise — a hover-sized rotor becoming a slender supersonic wing, its halves scissor-swept symmetrically to a pure 60° delta about the central titanium mast and buried flush behind iris doors — is balanced for a Mach 1.5 objective: fast enough to intercept, slow enough that fighter-standard materials, one engine, one vectoring nozzle and one merged variable-sweep wing may close the design. Every figure in this document is a preliminary sizing value; the flutter boundary of the pivoted wing and the whirl-flutter stability of the stopped rotor remain the binding open risks.

Compagnie Apex · Programme HSR · Preliminary engineering — recommended for Phase 0, subject to the gates of § 12 and to a signed ACE development agreement

Document limits: Requirements specification and conceptual feasibility presentation. Not a certification dossier, not guaranteed performance, not a flight manual. Values are engineering order-of-magnitude figures to guide design. Images are technical illustrations, not manufacturing drawings.