Archon Aerospace Inc. B.2.N Boomerang

B.2.N Boomerang Composite Airframe Fabrication and Airworthiness Strategy

In-House Foam Tooling Assessment • One-Piece Upper and Lower Shells • Carbon Fiber Ribs • nasa and F.A.A-Referenced Structural Substantiation • Building-Block Qualification • Bonded-Joint Safety • Damage Tolerance and N.D.E

Aircraft: B.2.N Boomerang — 50 ft span / 288 in (24 ft) root chord flying wing

Production site: Clark Freeport Zone, Pampanga, Philippines

July 30, 2026— Revision 4

Supersedes Revision 3, July 24, 2026

Classification: Internal — Engineering & Manufacturing Strategy

Strictly Confidential

Purpose of This Revision

How do we get wing shells built, who builds them, and what do the materials cost. Revision 4 answers a different and more consequential question, raised directly by the C.E.O: what has to be true for this airframe to be safe to put a pilot in, and can we get there while building more of the tooling ourselves?

Three specific proposals prompted this rewrite:

- Build the tooling in-house from cut and assembled rigid foam board rather than buying molds from an outside shop.

- Produce the entire upper surface in one mold and the entire lower surface in one mold, rather than in panels.

• Build the ribs in carbon fiber rather than 7075-T6 aluminum.

All three are examined here on their merits, with a direct verdict on each. Two of the three are endorsed with conditions. One of them — using the foam itself as the mold surface — does not survive scrutiny and is replaced with a corrected version that captures most of the cost saving without the safety exposure.

The larger change in this revision is structural. Rev 3 treated qualification and inspection as a section near the back. In Rev 4 they are the spine of the document, because the honest answer to "how do we make this airplane extremely safe" is not a material choice or a tooling choice. It is a test program. The tooling decision moves a few tens of thousands of dollars. The test program decision is what decides whether the airplane holds together.

Bottom line up front: cut the foam plug in-house, but do not fly parts off a bare foam tool. Build carbon ribs, but design and test the rib-to-skin joint as the primary failure mode. And build two wing sets — fly one, break the other. If only one budget line in this document survives, it should be the second wing set, not the mold.

1. Executive Summary

The B.2.N Boomerang is a 50 ft span, 288 in root chord flying wing intended to carry a pilot and to be flown hard enough to support a record claim. Two facts follow from that and drive everything in this document. First, there is no fallback if the structure fails — a flying wing has no separate empennage to trade against and a structural failure in the wing box is not survivable. Second, an airframe built by hand, in a tropical climate, by a team building its first aircraft, has a variability problem that must be bought down with testing rather than assumed away.

nasa has run this exact experiment. The Composite Crew Module program designed, built and destructively tested a full-scale carbon/epoxy primary structure built as one upper shell and one lower shell on two male tools, then spliced. Its two headline findings are directly applicable here:

• Finding F-1: many of the design, analysis, materials and manufacturing lessons were not evident in coupon, element or subcomponent testing and only became evident when building the full-scale assembly.

- Lesson L.L-3: engineering models predicted mass and structural response well, but did not predict production challenges.

Translated to Archon: the analysis package will not tell you where this airplane is going to give you trouble. Building it will. That argues for a full-scale manufacturing trial early, and for a dedicated structural test article — not for skipping steps to save tooling money.

1.1 Verdicts on the three proposals

: Table summary: A set of design proposals and their verdicts for an aircraft project. The proposal to build the entire upper and lower sections in single molds was adopted to reduce failure sites, provided the multi-cure sequence is properly costed and the closing bond is treated as safety-critical. Using carbon fiber ribs instead of 7075-T6 was adopted with a hard condition that rib-to-skin and rib-to-spar joints must be mechanically fastened and tested for out-of-plane pull-off. The proposal to cut and assemble rigid foam board for tooling was partially adopted, as foam is suitable for the male plug but rejected as a mold surface due to thermal growth, creep, and surface quality issues. Finally, using expanded or extruded polystyrene as structural material was rejected because these are tooling materials with poor shear properties and low service temperatures.

Table summary: Proposals for the flying article. The proposal to use core in the flying article is ruled out due to sensitivity and flammability, with the structural core remaining PVC or Nomex. The proposal to use a single wing set for the first article is rejected in favor of building two, allowing one to be tested to failure while the other is flown.

1.2 What this actually costs

The cost intuition behind the in-house tooling proposal is sound but aimed at the wrong line item. Tooling is not where the money goes on a first composite airframe. Labor, test articles and the qualification program are.

Section 13 lays this out in detail; the short version is that building the plug in-house plausibly saves $15,000 to $40,000, while skipping the structural test article "saves" $30,000 to $60,000 and buys a materially higher chance of losing the aircraft and the pilot. Those two decisions should not be traded against each other.

2. What "Airworthy" Means for This Airframe

2.1 The regulatory reality, stated plainly

Under an Experimental certificate — Research and Development under 14 C.F.R 21.191(a), which remains the correct category for B.2.N #1 — the F.A.A does not make a structural compliance finding. Nobody from the F.A.A is going to check the laminate schedule, review the allowables, or witness a static test. The certificate comes with operating limitations (restricted airspace, no persons carried other than required crew, a flight test area, a phase-one hour requirement) and the airworthiness burden sits with Archon.

This is a trap that has killed a lot of experimental aircraft builders. The absence of a regulator checking the work is frequently misread as the absence of a standard. The airplane does not know what category it is certificated in. The loads are the same either way.

Archon's position should be: we build to Part 23 structural substantiation rigor and document it as if we intended to certify, even though we are flying under an Experimental R&D certificate. That posture is also the only credible bridge to the type-certification and P.S.E-listing narrative in the investor materials — a company that can show a building-block data package is a fundamentally different proposition from one that can show a flying prototype and no substantiation.

Two clarifications on category, to be confirmed with a Designated Airworthiness Representative before any application is filed:

- Experimental Amateur-Built (the 51% rule) is not available to Archon. That category exists for individuals building for education and recreation; a corporate prototype does not qualify, and attempting to use it would be a misrepresentation.

- Exhibition is a poor fit for a flight test program. R&D is the right lane for #1, with Market Survey available later if aircraft are to be demonstrated to prospective customers.

2.2 The technical standards Archon should self-impose

The following documents form the working standard for this program. They are all publicly available at no cost except C.M.H-17, which is purchased from S.A.E International.

Table summary: Key regulatory and technical documents for composite aircraft design. The most critical is FAA AC 20-107B, Change 1, which governs material and process control, environmental design, proof of structure, damage tolerance, and bonded joints, with Paragraph 8 specifically driving the joint strategy. CMH-17 Rev G provides the building-block test pyramid and methods for design allowables in Volume 3, Chapter 4, while Volume 6 governs sandwich panel design in skins. Additionally, DOT/FAA/AR-03/19 defines the lamina-level qualification test matrices.

Table summary: Reference documents and their application to the Archon project, organized by source, number, and governing purpose. The list includes FAA and NASA publications that provide technical guidance for composite material systems. Key examples include Equivalency for Polymer Matrix Composite Material Systems, which helps size coupon programs, and the Composite Crew Module report, which serves as a primary analogue for structure, testing, and damage-tolerance criteria. Other documents provide specific guidance for shop-level manufacturing, NDE protocols for CFRP, electrical bonding and lightning protection, and structural element testing templates.

Rev 3's reference to Rutan's moldless manual and the E.A.A technical resources remains valid and useful for workmanship, and those references carry over. They are craft guidance, not substantiation standards, and should not be cited in place of the documents above when the question is whether a load path is proven.

3. Assessment of the In-House Foam Tooling Proposal

3.1 What the proposal gets right

The instinct is correct and should be preserved. Rigid foam board is cheap, locally available in Luzon, cuts cleanly with a hot wire or a router, bonds to itself with polyurethane or epoxy adhesive, and can be driven straight from the B.2.N C-A-D loft. For establishing a large, complex, single-curvature-dominant aerodynamic shape, it is by a wide margin the least expensive way to get from a C-A-D file to a physical master. Scaled Composites, Burt Rutan's shops, and essentially every low-rate composite aircraft program in the world have used exactly this to build masters.

It also puts the shape under Archon's own control, at Clark, on Archon's schedule — which removes a real dependency and a real lead-time risk from the D.M8 relationship. That is worth something independent of the dollar saving.

3.2 Where "foam board as the mold" breaks down

The proposal as stated is to cut and assemble to am boards into the entire upper and the entire lower surface, and lay carbon directly into or onto that. Six independent mechanisms defeat this on a 50 ft span flight structure. Each one alone is disqualifying; together they are not close.

(a) Thermal growth over a 50 ft span

Polystyrene foams have a coefficient of thermal expansion on the order of 60 to 70 times 10 superscript minus 6 per °C. Cured carbon/epoxy laminate is between roughly 0 and 3 times 10 superscript minus 6 per °C along the fibers. Over a 15.24 m (50 ft) span, a 15 °C swing between a cool night and a hot Clark afternoon moves a foam tool by approximately 15 millimeters — about 0.6 in. The part laid on it moves essentially not at all. There is no way to hold twist, dihedral, incidence or spar-station location on a tool that changes length by half an inch depending on the time of day. This is the reason aerospace production tooling is Invar, steel, or carbon/epoxy tooling laminate, and never bare foam at this scale.

(b) Creep under vacuum

Full vacuum applies about 14.7 psi (101 kilopascal) uniformly over the tool surface. A.S.T.M C578 Type 4 X.P.S — the common 25 psi board — has a compressive strength around 25 psi at 10% deformation, but sustained-load design values are typically taken at roughly one third of that. A full-vacuum bag held for a multi-hour cure will creep and dish a Type 4 foam tool, and the dishing prints straight into the part. Working around this means either high-density tooling foam (40 psi and up), a fully supported eggcrate substructure under the foam, or reduced bag pressure of roughly 8 to 10 inHg (4 to 5 psi) — and reduced bag pressure means higher void content, which is the opposite of what a flight laminate needs.

(c) Exotherm and service temperature

E.P.S and X.P.S have a service ceiling of roughly 75 to 80 °C before shrinkage and warping begin. A thick epoxy laminate curing against an insulating foam substrate — and foam is an excellent insulator, which is precisely the problem — traps its own exotherm. Thick sections of room-temperature-cure epoxy routinely exceed 100 °C internally. The tool softens under the part while the part is curing. Worse, any post-cure above ambient (which Section 7 shows is mandatory for a tropical-service aircraft) is flatly impossible on a polystyrene tool.

(d) Resin chemistry

Styrene-containing resins — polyester and vinylester, which are exactly what a marine yard like D.M8 uses by default — dissolve polystyrene foam on contact. This must be stated explicitly in any work package that goes to a boatbuilding partner, because it is not an obvious constraint to someone whose entire career has been in polyester hull work. Epoxy is compatible with E.P.S and X.P.S, but the compatibility only holds if the resin is genuinely epoxy and the shop has not substituted.

(e) Surface quality

A bare foam surface is cellular, not smooth. Assembled from boards, it also has a bond line every 600 to 1200 millimeters. Both print through into the part. For a general-aviation airframe that would be a cosmetic annoyance; for an aircraft whose entire commercial claim rests on a approximately 25:1 lift-to-drag ratio, laminar-flow-relevant surface waviness on the upper surface is a performance defect, not a finish defect. Any surface quality recovered afterward by filling and sanding the part is weight added to a wing that has already been sized.

(f) Tool life

A foam tool is effectively single-use. That is acceptable for a plug. It is a poor trade for a mold, because the whole argument for going molded rather than moldless (Rev 3, Section 1) was that the tool amortizes across aircraft #1, the structural test article, and aircraft #2 and #3.

3.3 The corrected architecture

The saving the C.E.O is reaching for is real, and it survives intact if the foam is moved one step back in the process chain. The recommended sequence:

Table summary: A five-step workflow for aircraft component production. The critical-path begins with Archon (Clark) performing the loft release to lock the root-chord OML in native CAD and export surfaces. This is followed by Archon (Clark) handling plug fabrication using CNC-cut EPS or XPS board on a frame for cost savings, and plug surfacing using barrier coats and tooling paste to achieve the required aerodynamic waviness. The process then moves to DM8 Composites or F.R.P. Philippines for the female tool pull, where tooling-grade epoxy laminate is used to create female shells. Finally, Archon at Clark performs the part layup, infusing carbon skins into the tools under full vacuum.

Table summary: Post-cure procedures for demolded shells. The process involves an elevated-temperature post-cure performed by Archon (Clark) using a purpose-built oven or heat blanket, as detailed in Section 7, because this step is impossible on foam but straightforward on a tooling-epoxy shell or demolded part. An earlier step involves or DM8 using female tools under full vacuum with the core inserted per Rev 3 Section 3.1, as the tool can withstand the process.

This split gives Archon the in-house control and most of the cost avoidance, moves the technically demanding step to the partner best equipped for it, and produces a tool that survives to build the structural test article and aircraft #2 and #3.

3.4 Plug material and process specification

Table summary: Specifications and safety requirements for plug fabrication. Foam must be at least 40 kg/m3 EPS or XPS ASTM C578 Type VI, with tooling-grade PU board recommended for leading edges to ensure durability. Adhesives must be epoxy or polyurethane, as polyester will dissolve the material. Board joints should be staggered and gap-filled to prevent print-through, and the plug must be supported by a rigid steel or timber substructure rather than relying on the foam itself. Surfacing requires 2 to 3 plies of light E-glass in epoxy followed by high-build tooling surfacer, with surface fairness verified by a 1 meter straightedge. Fabrication must occur in a temperature-logged, shaded environment to protect the foam from heat. Safety protocols are critical, specifically the use of forced local extraction and organic vapor respirators during hot-wire cutting, as polystyrene releases decomposition products and is highly flammable.

3.5 If the female tool genuinely cannot be funded

The moldless (Rutan-method) fallback retained in Rev 3 Section 2.3 stays valid and is the correct contingency: foam shaped directly to the O.M.L, skinned in place, foam remaining as permanent structural core. Two things must change if it becomes the primary plan rather than the fallback.

- The core material changes. The permanent core in a moldless flying structure is not construction E.P.S or X.P.S — it is the same P.V.C or P.U structural foam specified in Rev 3 Section 3.1. Section 3.6 below is not negotiable on this point.

- The surface finish penalty must be carried in the performance model. Fill-and-sand to an aerodynamically acceptable surface on 1,080 sq ft of wetted area is a large amount of added weight and a large amount of labor. Estimate both explicitly before choosing this path, because the L/D claim and the range claim in the April 2026 white paper both depend on the surface actually being achieved.

3.6 What must never be foam board

E.P.S and X.P.S are tooling materials. They are not aerospace structural core and must not appear in the flying article under any circumstance. Service temperature around 75 to 80 °C, low and poorly characterised shear strength, solvent sensitivity, no aerospace qualification data, and high flammability with toxic combustion products all disqualify them. Structural core in the B.2.N remains P.V.C (Divinycell / Airex class) per Rev 3 Section 3.1, with Nomex honeycomb reserved for weight-critical panels once the edge-seal process is proven.

This warrants a written material substitution prohibition in the Archon process specification, because polystyrene board is cheap, locally abundant, and superficially similar in appearance to P.V.C foam. It is the single most likely uncontrolled substitution on a Philippine shop floor.

4. One-Piece Upper and Lower Shells

Adopted. The reasoning is sound and it is what nasa did on the Composite Crew Module: that structure was built as an upper shell and a lower shell on two separate male tools, then joined with a double-lap-shear splice using internal and external doublers. Fewer parts means fewer joints, and joints — not acreage laminate — are where composite airframes fail.

4.1 What this buys

- A continuous fiber path from tip to tip on both surfaces, with no spanwise splice carrying bending load.

- A single controlled surface per side, which is what the L/D claim depends on.

- A dramatically smaller inspection burden than a panelised wing, because there are no acreage splices to scan.

- Parallel work: internal systems, wiring and plumbing can be installed in an open half-shell before closure, as C.C.M did deliberately.

4.2 What it costs, and what to plan for now

The Composite Crew Module's largest unplanned cost growth came from precisely this architecture, and the causes are worth quoting into Archon's own planning because they will recur:

- The multi-cure process was not accounted for in the original estimate — inner skin cure, inspection, core installation and cure, then outer skin cure, are four separate operations on each shell, not one.

- Core application proved more complex than expected, including core forming, impression testing and final bonding.

- Support tooling and shop aids — drill fixtures, router templates, trim jigs — were omitted from the original estimate entirely.

• Post-cure assembly was a large and unbudgeted effort (on C.C.M, roughly 500 precision reamed holes).

Archon should build the manufacturing plan around a per-shell cure sequence document before quoting the build, and should budget explicitly for trim and drill fixtures, a bonding/assembly jig, and a shell handling and rotation fixture. A 50 ft cured shell is a large, floppy, expensive object; the fixture that lets four people turn it over without inducing damage is not optional and is not free.

4.3 The consequence that matters most

Building the wing as two continuous shells concentrates essentially all of the assembly risk into one feature: the leading-edge and trailing-edge closing bond that joins upper shell to lower shell around the spar and rib skeleton. That bond runs the full 50 ft span on both edges, is loaded in peel at the trailing edge, and — once closed — is largely inaccessible to inspection from inside. It is the single most safety-critical feature on the aircraft and Section 6 is devoted to it.

5. Carbon Fiber Ribs

5.1 Why this is the right call

• Weight. Ribs on a 288 in chord are large parts; the mass saving over 7075-T6 is material at this scale.

- Thermal compatibility. Aluminum runs about 23 times 10 superscript minus 6 /°C; carbon/epoxy laminate runs near zero. Every aluminum rib bonded or bolted into a carbon shell is a built-in thermal stress riser that cycles every flight and every day on the ramp. nasa hit this exact problem on C.C.M and noted that a flight design would have used titanium rather than aluminum at the composite interface purely to manage the expansion mismatch. Carbon ribs make the problem disappear rather than managing it.

• Galvanic corrosion. Carbon is strongly cathodic to aluminum. A carbon skin bonded to a 7075-T6 rib in a humid tropical environment is a corrosion cell, and it corrodes the aluminum. It can be managed — a glass isolation ply and wet-installed sealant at every interface — but managing it forever, on an aircraft based in Clark, is worse than designing it out.

5.2 The rib is not the design problem — the joint is

A carbon rib web is easy. Carbon is extremely strong in the fiber direction and a rib web is a straightforward shear panel. The difficulty is entirely in the orthogonal joint: rib-to-skin and rib-to-spar. Composite laminates are strong in-plane and weak through the thickness, and an orthogonal joint loads the laminate in exactly the direction it is weakest — interlaminar tension and peel. This is the classic composite structural failure and it is not intuitive to anyone whose structural intuition was formed on metal.

nasa's C.C.M team hit this and solved it with three-dimensional woven preforms, with results worth knowing:

Table summary: Joint concepts for B2N ribs are evaluated by their reported capability and applicability, with performance increasing significantly from the baseline bonded composite L-clip to 3D woven options. The bonded composite L-clip serves as the baseline but is the weakest, as it loads the bond in peel at the radius and is not recommended for primary structure. The 3D woven Pi-preform offers approximately two times the capability of the L-clip with roughly 2,000 lb/in pull-off, making it a strong commercial candidate that supports out-of-autoclave curing. The 3D woven cruciform provides the highest performance at greater than three times the Pi-preform, or roughly 7,000 lb/in bi-directional pull-off, though it is considered overkill for most stations and reserved for the highest-load areas. Regardless of the chosen concept, bonded joints in critical locations must include mechanical fasteners to arrest disbond growth and carry limit loads.

5.3 Conditions attached to adopting carbon ribs

1. Rib-to-skin pull-off is tested at element level before design freeze, not after. This is a small, cheap test (flatwise tension per A.S.T.M C297 on representative joint coupons, plus a representative rib-foot pull-off element) and it decides the joint concept. nasa's finding on C.C.M element testing was that detailed pre- test analysis was crucial to the success of the element tests — model it first, then test it, and reconcile the two.

2. Every rib-to-skin and rib-to-spar joint on primary structure is bonded and fastened. The bond carries the design load; the fasteners exist so that a defective or degraded bond cannot progress to a catastrophic failure. This is standard practice in certified composite aircraft for precisely this reason.

3. Metallic fittings do not disappear. Splice fittings, hardpoints, control surface hinges and landing gear attachments remain metal, and the carbon-to-metal interface at those points still needs glass isolation, sealant and a bearing/bypass analysis. Carbon ribs reduce the number of these interfaces; they do not eliminate them.

4. Bearing allowables at every bolted joint in carbon must be established by test, not assumed. Bolted joints in laminate are bearing-critical and hole-quality-sensitive. On C.C.M, bearing margins at fittings were among the lowest in the entire structure — several at or below +0.06, one at -0.03 — which is a useful signal that this is where the design gets tight.

5. Drilling procedure is a controlled process. Carbon delaminates on exit if drilled like aluminum. Backing plates, correct geometry drills, controlled feed, and no dry drilling. This belongs in the process specification, not in tribal knowledge.

6. Bonded Joints — The Hardest Safety Problem in This Airframe

6.1 Why this section exists

The architecture chosen — two continuous shells closed around a bonded skeleton — means that the aircraft's primary load path passes through adhesive bonds that cannot be seen after closure. There is a specific, well-documented failure mode that makes this dangerous: the weak bond.

A weak bond is a bond line that is geometrically perfect — full adhesive coverage, no voids, no disbonds, and it passes every ultrasonic and thermographic inspection — but has essentially no strength, because chemical adhesion to the substrate never occurred. Contamination, an out-of-date surface preparation, mould release transfer, moisture on the surface at bond-up, or an incorrect mix ratio will produce it. The F.A.A's position, restated in A.C 20-107B, is that no non-destructive inspection method has been demonstrated capable of detecting it. Transport Canada's guidance material is blunter still: the weak bond phenomenon has defied attempts at detection by N.D.I, which is why the F.A.A adopted a fail-safe approach to bonded joints and why mechanical fasteners became common practice on bonded primary structure.

Ultrasonic and thermographic inspection tell you the adhesive is present and continuous. They do not tell you it is stuck. Any process built on the assumption that a clean N.D.I scan proves a good bond is built on a false premise, and this is the mechanism most likely to produce a fatal outcome on a hand-built composite airframe.

6.2 The three permitted answers

A.C 20-107B, paragraph 8, sets out three ways to substantiate a bonded joint whose failure would be catastrophic. Every safety-critical bond on the B.2.N must comply with one of them:

Table summary: Design features are the feasible and recommended approach for Archon, using mechanical fasteners to prevent disbond propagation because it is cheap, robust, and independent of inspection sensitivity. Proof testing each article by applying critical limit design loads is feasible for Aircraft #1 using a sandbag or whiffletree rig at Clark, though it becomes burdensome as production rates increase. Repeatable and reliable NDI is not available, as no currently accepted method exists for adhesion quality, and should not be used to build the safety case.

6.3 Archon's bonded joint policy

6. Every bond whose failure is catastrophic uses Option (i) — bonded plus fastened, with fastener pitch set by disbond arrest analysis and confirmed by element test.

7. Aircraft #1 additionally receives a full-airframe proof load to limit under Option (2) before first flight, instrumented with strain gauges at the locations predicted to be critical.

8. Option (3) is used only as supporting evidence, never as primary substantiation.

9. A written bonding process specification governs surface preparation, and it is treated as a controlled process. A.C 20-107B calls for a "process control mentality" combining in-process inspections and tests, with the bonding environment and cleanliness controlled to a validated level. In a Clark shop that means: a dedicated bonding bay, controlled humidity, defined maximum time between surface preparation and bond-up, peel ply removed immediately before bonding, no silicone-containing products anywhere in the building, and a documented operator.

10. Every bond-up cure produces a witness coupon from the same adhesive mix, cured alongside the part, and destructively tested. Adhesion failure — adhesive separating cleanly from the substrate rather than tearing cohesively within itself — is an automatic reject of the entire bond, not a data point to average. A.C 20-107B treats adhesion failure as an unacceptable failure mode in all test types.

11. The trailing-edge closing bond is peel-loaded and gets specific attention: a peel-type test in the witness coupon program, and a design that turns peel into shear wherever geometry permits.

There is a design-level implication here that should be reflected back into the structural package. Wherever the design can be arranged so that a bond failure is not catastrophic — redundant load paths, fail-safe rib spacing, a secondary shear path through the spar caps — that arrangement is worth real weight, because it moves the joint out of the catastrophic category entirely and takes it out of the scope of paragraph 8.

7. Environmental Design Point — Clark, Bali, and Hot/Wet

This section exists because the B.2.N is being built and operated in one of the most hostile environments in the world for a hand-laid epoxy structure, and because the failure mode is silent. A composite airframe that is 25% weaker than the analysis assumed does not look any different from one that is not.

7.1 The mechanism

Epoxy is a thermoset with a glass transition temperature (Tg). Above Tg the resin softens and matrix-dominated properties — compression strength, shear strength, bond strength, core-to-skin adhesion — fall off sharply. Two things degrade the margin between Tg and the temperature the structure actually sees:

- Absorbed moisture depresses Tg. A saturated laminate typically loses on the order of 15 to 25 °C of Tg relative to dry. In Clark and Bali, the laminate will approach saturation over its life; it is not an edge case.

- Solar soak raises the structure temperature well above ambient. A dark upper surface parked on a tropical ramp reaches 70 to 80 °C. Archon should measure this directly at Clark with a surface thermocouple on a painted witness panel rather than assume a number.

The accepted design rule is that the maximum operating temperature must sit meaningfully below the wet glass transition temperature — commonly Tg(wet) minus approximately 28 °C (50 °F). Run that backwards from a 75 °C solar soak and the required Tg(wet) is above 103 °C, which implies a dry Tg comfortably above 120 °C.

7.2 Why this rules out an ambient-cure-only process

A typical room-temperature-cure laminating epoxy reaches a dry Tg of only about 55 to 70 °C with no post-cure. Wet, that is roughly 40 to 50 °C — below the temperature the upper skin will reach sitting on the ramp at Clark on a clear afternoon. An airframe laid up and left to cure at ambient in a tropical shop is not airworthy for tropical service, and this is not a marginal call.

Elevated-temperature post-cure of every structural laminate and every structural bond is mandatory for this aircraft, and the required post-cure schedule must come from the resin manufacturer's data sheet and be verified by D.S.C or D.M.A on witness coupons — not assumed from the cure time on the container. This requirement alone is sufficient to disqualify a polystyrene foam mould (Section 3.2c), because the tool cannot survive the post-cure.

7.3 Environmental controls at Clark

Table summary: Quality control requirements and recording procedures for material handling. Layup bay conditions must be dehumidified with stable temperature and relative humidity, tracked via continuous logged timestamps for every cure and bond. Material storage requires prepreg to be kept in a freezer with out-time tracking, while resins, hardeners, and adhesives must remain sealed within their shelf life and with desiccant, recording batch numbers, receipt dates, and expiry. Dry carbon fabric is noted as hygroscopic and must remain sealed until use, with a drying cycle considered before layup.

Table summary: Quality control requirements and records for material curing. The process requires materials to be stored in a humid bay. Cure verification is performed by producing a witness panel from identical materials and cycles, with records including fiber volume fraction, void content, mechanical checks, and Tg measured by DSC or DMA. Post-cure must follow the resin data sheet using thermocouples at the thickest and thinnest sections, with the resulting thermocouple trace retained in the build record. Finally, oven surveys are conducted before first use and periodically to ensure no cold zones, documented via a survey report.

7.4 Moisture in service

Beyond cure quality, water intrusion into sandwich structure is the long-term degradation mechanism for a tropical-based composite aircraft. The Rev 3 recommendation of closed-cell P.V.C foam over Nomex honeycomb for aircraft #1 is correct on these grounds and stands. Add to it: every core-to-solid-laminate transition is edge-closed and sealed, every fastener through a sandwich panel is sealed or potted, and the aircraft gets a drainage design — low-point drains that actually drain in the parked attitude.

7.5 Lightning and electrical bonding

A carbon airframe conducts, but poorly and anisotropically, and it will not carry a strike the way an aluminium skin does. Without protection, a strike attaches locally, vaporises resin, and can blow a hole through the laminate; joints and fastener heads are the typical attachment and damage points. nasa C.R-4784, with the supporting fault-current and lightning-effects reports C.R-4774 and C.R-4783, is the reference. At minimum the design needs a conductive layer (expanded copper foil or bronze mesh) in the strike zones, a defined electrical bonding and grounding scheme so that no metallic fitting is isolated, and protection at the fuel system — which on the B.2.N means the four conformal fuel bays and every fitting and vent in them.

Flag: on an aircraft with 200 gallons of fuel in conformal bays inside a carbon wing, lightning protection is a fuel-system safety item, not an avionics nuisance item. It should be reviewed as part of the fuel system design, not deferred to a later avionics phase.

8. Structural Substantiation — Loads, Factors and Allowables

8.1 Load set

Before any of the qualification work below is meaningful, the design load set has to be frozen and documented: a V-n diagram with the selected limit load factors, gust cases, the flight and ground load conditions, control surface hinge moments, and the mass and C.G envelope from which they are derived. For a Part 23-track aircraft the load derivation follows the relevant Part 23 subpart C requirements, and Archon should follow that structure even under an Experimental certificate. Ultimate load is 1.5 times limit unless a specific case justifies otherwise.

This connects to an open item from the aerodynamic work: the static margin from the AeroSandbox vortex lattice analysis was +0.7% to +1.9% with a stability augmentation system required. Any structural load set has to reflect the actual achievable control authority and the S.A.S failure case, because a S.A.S failure on a marginally stable flying wing is a structural load case, not just a handling case.

8.2 Why composites do not inherit the metallic factor

A common and dangerous simplification is to apply the same 1.5 ultimate factor used on metal and consider the job done. nasa C.R-186010 exists specifically to clarify this: composites carry additional knockdown factors on top of the factor of safety, to account for environmental degradation, damage tolerance, scatter in the material, and process variability. The factor of safety accounts for uncertainty in the load; the knockdown account for uncertainty in the strength. They multiply.

The Composite Crew Module's final margin summary shows how nasa actually applied this in practice on a comparable structure, and the pattern is instructive: different factors were applied to different failure modes, with the higher factors reserved for the failure modes least well understood and least inspectable.

: Table summary: Factors of safety applied to various failure modes based on NASA CCM margin summary and their read-across to B2N. The highest factors of safety are 2.00, applied to core shear in sandwich panels and critical cases of bonded or woven-preform joint pull-off and shear. Laminate strength uses a factor of 1.40 in most locations and 2.00 in some, while bearing at bolted joints is set at 1.40. The lowest factor of safety is 1.25, applied to metallic bolt yield for fasteners and metallic fittings.

The pattern to carry across is not the specific numbers — those were set for a spacecraft load environment — but the principle: bonded joints and core shear get a higher factor than laminate acreage, because they are the modes where process variability dominates and inspection is weakest. Archon should adopt the same asymmetry and document the rationale.

8.3 Allowables — the hand-layup penalty

Rev 3's process comparison table is correct that hand wet layup runs 2 to 5%+ void content against under 1 to 2% for prepreg and autoclave, at lower fiber volume fraction. The structural consequence needs to be stated more directly than Rev 3 stated it: this is not a weight penalty that can be trimmed later, it is a reduction in the design allowables that must be carried through the entire sizing analysis.

- Do not use published prepreg/autoclave allowables from a supplier data sheet or from C.M.H-17 Volume 2 to size a wet-laid or infused structure. They will overstate strength, and matrix-dominated properties (compression, shear, interlaminar) will be overstated the most.

- Generate allowables from Archon's own process, on Archon's own materials, cured in Archon's own facility, at the conditioning states that matter (room temperature dry, and elevated temperature wet). DOT/F.A.A/AR-03/19 gives the test matrices — batches, panels and specimens per property — needed for a defensible B-basis value.

- If a supplier or partner's data is used, an equivalency program per DOT/F.A.A/AR-03/19 is the accepted route to claiming it. It is cheaper than a full qualification and is the pragmatic option for aircraft #1.

- Vacuum infusion is strongly preferred over hand wet layup for the structural shells, and this is worth paying D.M8 for. Infusion cuts void content and raises fiber volume fraction relative to hand wet layup, and — because it is a controlled process rather than an operator-dependent one — it cuts the scatter, which is what actually drives a B-basis number down.

9. The Building Block Program

This is the core of the airworthiness case and the part of the program that cannot be compressed without transferring the risk directly onto the pilot. The building block approach is the accepted method in C.M.H-17 Volume 3 Chapter 4 and it was what nasa used on C.C.M. Its logic is that a large number of cheap tests at the bottom controls material and process variability, which reduces the number of expensive tests needed at the top.

9.1 The pyramid, sized for the B.2.N

Table summary: The structural testing hierarchy for the Archon aircraft, organized into four progressive levels of component complexity. Level 1 covers coupons, including flat panels, sandwich coupons, and bond specimens, which provide essential design allowables and confirm bond integrity. Level 2 focuses on elements, testing small structural features like joints and inserts to validate joint concepts. Level 3 involves subcomponents, such as a multi-bay wing box section, to validate analysis methods on real structures. Level 4 is the full-scale test of a complete wing set, which serves as the final determination of aircraft safety. Each level builds upon the previous, moving from material-specific data to full structural validation.

Table summary: A flight article is evaluated using representative tests consisting of an intended inspection interval followed by a load to failure.

9.2 The second wing set

This is the recommendation most likely to be argued with on cost grounds, so the case is worth making explicitly.

- The tooling is already paid for. The marginal cost of a second shell set is materials plus labor — using the Rev 3 planning estimate, roughly $9,000 to $15,000 in raw material plus the layup and assembly labor. It is not a second tooling program.

- It is the only way to test to ultimate and beyond. A flight article cannot be loaded to failure, and loading a flight article even to ultimate leaves an aircraft with an unknown amount of accumulated damage that will then be flown.

- It is where the manufacturing lessons get learned on a part that does not matter. nasa's finding F-1 was that the lessons only showed up at full scale. Better that they show up on the test article. nasa's recommendation R-1 was explicitly to implement full-scale manufacturing development early so that lessons can be incorporated as the design evolves.

- It builds the investor and regulator narrative. A photograph of a wing loaded to destruction in a test rig at Clark is a more credible technical asset than a photograph of a finished aircraft, and it is the kind of evidence that changes an institutional conversation.

Sequencing recommendation: build the structural test article First, not second. The first shell set off a new tool is the one most likely to contain manufacturing defects, and it is the one you want to break rather than fly. Build it, inspect it, break it, feed the lessons into the process, then build the flight article.

9.3 Where the testing happens

Coupon and element testing is well within the capability of a university materials laboratory, and the De La Salle University relationship established for the plasma actuator work is the natural home for it.

Subcomponent and full-scale testing needs a load frame or a reaction floor with hydraulic actuators and instrumentation — that is a bigger conversation, and the options are Philippine University or another Philippine university with a structures lab, a rented commercial test house, or a purpose-built whiffletree and sandbag rig at Clark. A sandbag or water-bag rig is genuinely viable for a static test to ultimate and is how a great deal of light aircraft structural testing has historically been done; it is slow and labor-intensive but it does not require capital equipment.

Whatever the venue, the test plan is written and independently reviewed before the test, with predicted strains at every gauge location, so that the test either confirms or refutes the analysis rather than just producing a number.

10. Damage Tolerance and the Impact Threat

10.1 The problem specific to composites

An aluminium skin that gets hit leaves a dent. A carbon skin that gets hit can look essentially undamaged on the surface while carrying substantial internal delamination, and can have lost a large fraction of its compression strength. This is barely visible impact damage (B.V.I.D) and it is the reason composite airframes are designed to a damage tolerance requirement rather than to a pristine-strength requirement.

The design principle: the structure must carry ultimate load with damage present up to the threshold that is reliably detectable by the inspection method in use, and must carry limit load with larger damage that is obviously detectable. That means the design allowables in Section 8 are damage-tolerant allowables, not pristine coupon values.

10.2 Threat assessment for the B.2.N at Clark

The threat set has to be specific to this aircraft in this operating environment, not generic:

Table summary: Design responses to various physical threats for the aircraft. The most frequent threat is dropped tools during build and maintenance on the upper wing surface, which dictates minimum skin gauge and the need for walkway zones or no-step markings. Other critical considerations include designing wingtips and winglets for repairability due to ground handling and towing strikes, and establishing lower-surface minimum gauge to protect against runway debris and stone impact. Additionally, the design accounts for bird strikes on the long leading edge, tropical storm debris on the upper surface, and lightning strikes at the extremities and fuel bay region.

10.3 The demonstration program

nasa's C.C.M damage tolerance demonstration is a directly usable template and its numbers are a reasonable starting point for the B.2.N, to be adjusted once the threat assessment above is quantified:

- Impact the full-scale test article at multiple discrete locations at the defined allowable threat level (C.C.M used 6 ft-lb at 18 locations), covering acreage, joints, core ramps, and fitting regions.

- Impact a smaller number of specific design details at the critical threat level — the energy at which damage becomes reliably detectable (C.C.M used 26 ft-lb at five design details).

- Inspect all impact sites by the production N.D.E methods and record the damage extent.

• Load the damaged article through the design ultimate load cases.

- Cycle it through the intended service life — C.C.M cycled through four lifetimes — and re-inspect.

- The acceptance criterion is no detrimental damage growth. On C.C.M, no detrimental growth was observed in any case, which is the result Archon should be aiming to reproduce.

One further C.C.M result deserves attention because it is a warning rather than a reassurance: the full-scale impact testing highlighted variability in the paste bond at the docking ring — a bonded metallic interface — even though it met requirements. The bonded metal-to-composite interface was the noisiest feature in the structure. The B.2.N has several such interfaces at its splice fittings and hardpoints.

10.4 Repair

Plan for repair before the first damage occurs. C.C.M required repairs numerous times during manufacturing development — tooling shift during cure, improper cure of splice doublers, excess void in doublers, disbond from over-constrained tooling, and impact damage from a dropped tool. This is normal, not a sign of a bad program.

What is needed is a written, pre-qualified repair procedure with its own coupon data, so that repairs are made to a known standard rather than improvised, and a rule that any repair on primary structure is engineering-dispositioned rather than shop-dispositioned. F.A.A A.C 43 to 214 is the reference for repair process control.

11. Non-Destructive Evaluation Plan

11.1 Methods

nasa approved four N.D.E methods for C.C.M and the same four are appropriate for the B.Z.N, selected by flaw type, part geometry, thickness and access. The nasa Advanced Composites Project N.D.E Methods and Capabilities Handbook is the technique-selection reference and includes an applicability matrix mapping techniques to flaw types.

: Table summary: Inspection methods for composite materials and their roles at Clark. Ultrasonic testing is the primary workhorse, specifically portable phased-array or pulse-echo units, used for delaminations, porosity, and thickness, though it is slow and cannot detect weak bonds. Visual inspection serves as the continuous first line of defense for surface damage and gross defects, but misses subsurface issues. Infrared thermography provides fast wide-area screening for near-surface delaminations and water ingress, acting as a complement to ultrasound. Radiography and X-ray CT are reserved for anomaly investigations, such as core crush or foreign objects, due to costs and safety requirements.

11.2 Defect acceptance criteria

The C.C.M criteria, from specification C.C.M-SPEC-006, provide a defensible starting point that has been used on a full-scale flight-representative carbon/epoxy structure. Archon should adopt these as the initial acceptance limits and revise them only against its own element test data — never loosen them to accept a part that has already been built.

: Table summary: NASA CCM criteria for maximum permissible defects across different configurations. For facesheets, flat laminates, doublers, and laminate caps, the defect width or diameter must be less than 0.250 inches. Sandwich facesheet-to-core bonds allow a larger width or diameter of less than 0.50 inches, with a minimum of two times the cell size. Preform-to-laminate bonds, specifically for Section 5 rib joints, must have a bondline defect width or diameter less than 0.125 inches. For core-to-core bonds, foaming adhesive gaps cannot exceed 0.06 inches in length parallel to the splice. Finally, for core-to-insert bonds, no defect or adhesive void is permitted to span the entire panel thickness from the inner mold line skin to the outer mold line skin.

Note how tight the preform-to-laminate bondline criterion is — 0.125 in — relative to the acreage laminate criterion of 0.250 in. That asymmetry is deliberate and matches the factor-of-safety asymmetry in Section 8.2. Joints get held to a higher standard than acreage because joints are where the structure fails.

11.3 Practice

12. Every structural part gets a baseline scan after cure and before assembly, archived. Post-assembly and in-service scans are compared against this baseline; without a baseline, a later scan cannot distinguish a manufacturing defect from service damage.

13. Reference standards with intentionally embedded defects of known size and depth are fabricated from the same laminate and used to set up and calibrate every inspection. Inspection sensitivity is not assumed, it is demonstrated.

14. Multiple methods are used on the first article. nasa's N.D.E team recommended exactly this for the first article production unit and subsequent test articles.

15. Anything found is dispositioned by engineering, recorded with location, depth and size, and retained. Defects are not sanded out and forgotten.

16. Personnel are trained and their qualification is documented. An ultrasonic scan is only as good as the person running it; N.A.S 410 or equivalent is the reference for N.D.T personnel qualification.

12. Process Control and the Build Record

A.C 20 to 10/B's central theme, repeated across its material, fabrication and bonding sections, is that composite quality is a process control problem rather than an inspection problem. You cannot inspect quality into a composite part after the fact; the part is only as good as the process that made it. This section is what that means concretely at Clark.

Table 12.1 summary: The required documentation to be completed before the first structural cure, organized by document type and its specific contents. This includes material specifications covering manufacturer and storage details, and process specifications for laminating, bonding, and machining and drilling, which detail everything from layup sequences and resin mix ratios to drill geometry and surface preparation. Additionally, the set requires an NDE procedure for scanning and reporting, a repair procedure for damage classification and schemes, and personnel qualification records to verify the training and authorization of staff performing work on primary structures.

12.2 The traveler

Every structural part carries a traveler that follows it from raw material to installation. This is the document that, years from now, answers the question of why a part failed, and it is the document that turns a hand-built aircraft into a substantiated one.

• Part number, serial number, drawing revision.

• Every material lot number used, with certificate of conformance on file.

- Layup date, operator name, ply-by-ply sign-off against the layup drawing.

• Ambient temperature and R.H at layup and at bond-up.

• Vacuum level achieved and held, with the gauge trace.

• Cure and post-cure thermocouple traces.

• Witness panel identification and its test results — Tg, fiber volume, void content, mechanical checks.

- N.D.E scan results and any disposition.

• Any non-conformance, its disposition, and who signed it.

nasa's manufacturing quality plan for C.C.M used basic A.S9100 processes and systems, with defined exceptions to support a rapid prototype project. That is the right model: adopt the A.S9100 structure, take documented exceptions where the pace of a prototype program requires them, and record the exceptions rather than letting them happen informally.

12.3 The concurrent engineering point

One of the clearest findings from C.C.M was organisational rather than technical: success in making real-time changes to design and planning was made possible by the concurrent, on-site presence of key design, analysis, materials, manufacturing and quality personnel. The team ran a daily manufacturing tag-up between design and manufacturing throughout the build.

The read-across for Archon is uncomfortable but worth stating: if the structural shells are laid up at D.M8's yard in Danao City, Cebu, while the engineering team is at Clark or remote, that daily loop does not exist. Either the fabrication moves to Clark, or an Archon engineer is resident at D.M8 for the duration of the shell build with authority to stop work. Remote oversight of a first-of-type composite primary structure build is not oversight.

13. Cost Reality — Where the Money Actually Is

The proposal to build tooling in-house is a cost-reduction move, so it deserves a cost answer rather than only a technical one. The table below is a planning-level allocation, not a quote, and should be replaced with real numbers as the D.M8, Hexcel and Teijin R.F.Q's land. Its purpose is to show the relative size of the line items, which is the part that drives the decision.

Table summary: The in-house foam proposal provides real savings specifically for the upper and lower master pattern plugs, which cost between 8,000 and 20,000 dollars in-house. However, the proposal is not recommended as a substitute for the female tooling shells, which are estimated between 25,000 and 70,000 dollars. Other cost drivers include the raw material for a second structural test shell set, noted as the single highest-value line item at 8,750 to 14,800 dollars, and layup, assembly, and finishing labor, which is the largest single line item overall. The table suggests further in-house savings by building the subcomponent and full-scale static test rig at Clark, rather than renting a commercial facility, which would cost between 15,000 and 50,000 dollars.

Three conclusions follow from the shape of this table:

17. Tooling is a minority of the cost. Making the tooling worse to save part of a minority line item, at the price of a compromised aerodynamic surface and a compromised cure, is a bad trade — especially on an airframe whose commercial claim is an aerodynamic efficiency number.

18. The in-house instinct is right but aimed one item too far up the table. Build the plug in-house. Build the assembly jig in-house. Build the static test rig in-house. Do not build the mould surface or the female tool in-house.

19. The second shell set is cheap relative to what it protects. It costs roughly what the coupon program costs and roughly a quarter of what the tooling costs, and it is the difference between an airplane that has been proven and an airplane that has been hoped for.

14. Recommended Baseline Plan

20. Freeze the corrected 288 in root-chord outer mold line in native C-A-D. Nothing downstream starts until this is done. This has been the critical path item since Rev G and remains it.

21. Write the material specification, laminating process specification and bonding process specification before any structural material is ordered. Use F.A.A A.C 23 to 20 as the template.

22. Stand up the coupon program with Philippine University. Target: allowables at room-temperature-dry and elevated-temperature-wet for the actual Archon layup, plus bond coupons proving cohesive failure, plus sandwich flatwise tension and core shear.

23. Run the element tests that decide the design. Specifically: rib-to-skin pull-off (which decides the Section 5 joint concept), the L.E/T.E closing bond specimen, and a spar-cap splice. Model each one before testing it.

24. Cut the plug in-house at Clark. High-density foam over a steel eggcrate, glass barrier, epoxy tooling surfacer, block-sanded and measured. Record the surface waviness.

25. Have D.M8 pull the two female tooling shells off the plug in tooling-grade epoxy laminate on a steel backing frame. Get this quoted against the finished plug, not against a C-A-D file.

26. Build the assembly jig, trim fixtures, drill fixtures and shell handling fixture at Clark, in parallel with the tooling.

27. Build the structural test wing set first. Inspect it fully. Feed every manufacturing lesson back into the process specification before starting the flight article.

28. Test the structural article: proof to limit with a strain survey, damage tolerance impacts, ultimate, repeated loading, then failure. Independently review the test plan before the test.

29. Build the flight wing set to the revised process specification, with full N.D.E and a complete traveler on every part.

30. Proof load the flight airframe to limit before first flight, instrumented.

31. Certificate and flight test: Experimental R&D certificate, operating limitations, and a phase-one flight test program with an incremental envelope expansion plan.

14.1 What is deliberately deferred

- Automated fiber placement. Rev 3's position stands: no A.F.P capital before the airframe is flight-proven with a funded order book. Exploratory R.F.Q's to Mikrosam, Coriolis and Ingersoll remain useful for lead-time intelligence and investor materials only.

- Thermoplastic tape for spar caps. Held for a later phase per Rev 3 Section 3.3.

- Nomex honeycomb. Held until the edge-seal and bonding process is proven, per Rev 3 Section 3.1.

- Type certification. The building-block data package generated here is the foundation for it, but the certification basis conversation is a separate program.

15. Risk Register

Table 15 summary: A risk management matrix for aircraft structural fabrication, mapping specific technical and operational risks to their consequences and mitigations. Key structural risks include undetected weak bonds, which are mitigated by using both bonding and fastening on catastrophic joints, and the use of incorrect design allowables, addressed by a coupon program based on the actual process. Manufacturing risks include inadequate post-cure leading to property collapse, mitigated by mandatory post-cure with thermocouples, and tool distortion, addressed by using tooling-epoxy female shells on steel frames. Operational and oversight risks include material substitution on the shop floor, remote fabrication oversight at DM8 in Cebu, and schedule pressure compressing the test program. Environmental and safety risks include undetected impact damage, lightning strikes to carbon wings, and galvanic corrosion at carbon-to-aluminum interfaces, with mitigations ranging from baseline NDE scans and conductive mesh to the use of glass isolation plies.

16. Action Items

Table summary: A 13-item action plan for aircraft composite development and certification. The key priority is a CEO decision to add a second structural test wing set to the plan of record and budget. Other critical tasks include Engineering freezing the corrected 288 in root-chord OML to unblock tooling actions, Archon building an internal reference library from SAE and NASA documents, and Archon drafting material and process specifications. Operational requirements include re-scoping the DM8 request for female tooling shells and vacuum infusion, pricing an in-house plug, and specifying post-cure oven systems and portable NDE equipment. Finally, the plan requires contracting a composite aircraft structures engineer for design review and confirming the Experimental R&D registration path with FAA and CAAP airworthiness representatives.

Table summary: A list of pending actions, including a review for the four conformal bays and a task assigned to Archon to update the Rev 3 Section 3.7 material cost estimate. The latter depends on the OML being locked, quotes from Hexcel and Teijin arriving, and a spar-cap material takeoff being pulled from structural CAD, with the addition of a second shell set.

Summary: vacuum infusion — strongly preferred over hand wet layup for the structural shells

A — Hand wet layup: dry fabric, resin brushed/rolled in by hand, vacuum bagged

B — Vacuum infusion: fabric laid in dry, bagged, resin drawn through by vacuum from this is what most people mean by "dry layup"

C — Prepreg, out-of-autoclave: fabric arrives pre-impregnated, freezer storage, oven cure

Table summary: The estimated dry fiber weight for a single airframe totals between 210 and 270 kilograms. The largest contributions come from the upper and lower skins at 80 kilograms, followed by spar caps at 50 to 80 kilograms and carbon ribs at 40 to 55 kilograms. Other components include spar and shear webs at 25 to 35 kilograms, and elevons, winglets, fairings, and panels at 12 to 18 kilograms.

Table summary: Material costs for one airframe across three different manufacturing methods. Prepreg OOA is the most expensive option, with total costs ranging from 46,600 to 79,750, driven largely by high costs for prepreg woven and prepreg UD carbon. Wet layup and Infusion have similar total cost ranges, between approximately 20,100 and 37,670. For carbon fiber, wet layup and infusion both use woven carbon costing 7,000 to 12,250 and UD carbon costing 3,900 to 6,500. Other cost variations include resin, which is more expensive for wet layup at 3,670 to 6,120 compared to 2,235 to 3,725 for infusion, and consumables, which are highest for infusion at 3,750 to 7,500 due to flow media, tubing, and traps.

Resin demand is where the methods split: wet layup runs ~0.85 kg resin/kg fiber, infusion ~0.62, prepreg ~0.70 (built in).

A and B are a dead heat on materials. Infusion saves $ \sim $55 kg of resin and spends it back on flow media and tubing.

Two airframes (structural test article + flight article)

Including attrition — 25% on A/B, 35% on C for prepreg out-time and shelf-life losses.

Table summary: Production costs for airframes vary significantly by manufacturing method, with Prepreg being the most expensive option compared to Wet layup and Infusion. For materials across two airframes, Wet layup costs range from 50,000 to 89,000 dollars, Infusion ranges from 52,000 to 94,000 dollars, and Prepreg is substantially higher at 126,000 to 216,000 dollars. Capital costs for equipment such as pumps, freezers, ovens, and thermocouples follow a similar trend, costing 27,000 to 66,000 dollars for Wet layup, 29,000 to 70,000 dollars for Infusion, and 57,000 to 138,000 dollars for Prepreg.

Table summary: Labor and cost estimates for two airframes across three different scenarios. The highest total program cost is in the third scenario at 208,000 to 444,000 dollars, while the other two scenarios are lower and similar, ranging from roughly 104,000 to 254,000 dollars. These totals are driven by composite labor hours, which range from a low of 3,800 to 5,500 hours in the second scenario to a high of 6,600 hours in the first. Corresponding labor costs, calculated at 6 to 15 dollars per hour, range from 23,000 to 99,000 dollars across the scenarios.

The post-cure oven dominates capital in all three columns — and you need it regardless, because of the wet-Tg requirement in Rev 4 §7. It isn't a prepreg-only cost, which narrows the C gap somewhat. Labor is the widest band and the least certain line; it swings on whether Archon or DM8 does the touch labor.

Why infusion wins, beyond cost being a wash

121 pounds of dead resin. Wet layup carries approximately 55 kilograms more resin than infusion across the airframe. On an aircraft whose entire pitch is a 25:1 L/D and 2,348 nm range, that's not a rounding error.

Pot life makes wet layup genuinely unsafe at this scale. A laminating epoxy with 40 min working time at 25 superscript circle C gives you maybe 15 to 20 min in a 33 superscript circle C Clark shop. You cannot hand-wet 25 m ^2 of 4-ply skin in 20 minutes with a realistic crew. So you batch it — and every batch boundary is a wet-on-wet join against partially gelled resin, buried inside the laminate where no inspection reaches it. That's the 6 problem again, only now it's distributed through the acreage instead of concentrated at a joint you designed for.

Infusion supports the traveler. Dry fabric can be laid over days with ply-by-ply sign-off against the layup drawing, exactly as Rev 4 §12.2 requires. Wet layup can't be inspected ply-by-ply — it's covered in resin and on a clock.

Lower scatter. Infusion is a controlled process; wet layup is operator-dependent. Scatter is what drives a B-basis allowable down, so infusion buys back strength on paper as well as in the part.

One real risk it carries: a failed infusion loses the entire 50 ft shell in one event. Mitigate with a full-scale trial panel first, multiple valved feed lines, and — again — building the test article shells first, where a scrapped part costs materials and pride rather than schedule.

On prepreg

Not wrong, just wrong for #1. It roughly doubles program cost, needs a tool that survives full cure temperature under vacuum with matched C.T.E, and punishes a first-time team hard on out-time losses. Revisit it at the Phase 3 production step alongside the A.F.P decision — same category of capital, same timing.

Recommendation

Infusion for all structural laminate: both wing shells, spars, ribs. Reserve hand wet layup for secondary parts, fairings, closeouts, and repairs, where the areas are small enough that pot life isn't a factor.

Want me to fold this in as a new Section 3.8 and update the cost table in Section 13? It'd replace the Rev 3 §3.7 estimate, which was wing-shells-only and used an infusion resin ratio against a wet-layup process.

17. References

17.1 nasa

- nasa/TM-2011 to 217185, N.E.S.C-R.P-06 to 019, Composite Crew Module: Primary Structure, nasa Engineering and Safety Center, November 2011. Companion discipline reports: Design; Analysis; Materials and Processes; Manufacturing; Test; Non-Destructive Evaluation.

- Structural Element Testing in Support of the Design of the nasa Composite Crew Module, N.T.R.S 20120009213.

- nasa/TM-2013 to 217990, N.E.S.C-R.P-10 to 00685, Composite Crew Module Permeability Characterization.

• nasa S.P-8108, Advanced Composite Structures (structural design criteria monograph), December 1974.

• nasa C.R-186010, A Guide to Structural Factors for Advanced Composites Used on Spacecraft, 1989.

- nasa M.S.F.C G.D-E.D-2205, Design and Manufacturing Guideline for Aerospace Composites (nasa Preferred Practices).

- N.D.E Methods and Capabilities Handbook, nasa Advanced Composites Project, N.T.R.S 20200005253 and 20205007851.

- nasa C.R-4784, Design Guidelines for Shielding Effectiveness, Current Carrying Capability, and the Enhancement of Conductivity of Composite Materials; with C.R-4774 (Fault Current Through Graphite Filament Reinforced Plastic) and C.R-4783 (Lightning Effects on Composite Materials).

- Assessment of the State-of-the-Art in the Design and Manufacturing of Large Composite Structures, N.T.R.S 20010067276.

• Simplified Design Procedures for Fiber Composite Structural Components and Joints, Chamis and Murthy, N.T.R.S 19900015068.

17.2 F.A.A

- A.C 20-107B, Change 1, Composite Aircraft Structure. Paragraph 8 (bonded joints) is directly load-bearing on the B.2.N design.

- A.C 23 to 20, Acceptance Guidance on Material Procurement and Process Specifications for Polymer Matrix Composite Systems.

• A.C 43 to 214. Repairs and Alterations to Composite and Bonded Aircraft Structure.

• DOT/F.A.A/AR-03/19, Material Qualification and Equivalency for Polymer Matrix Composite Material Systems: Updated Procedure, Tomblin, Ng and Raju, N.I.A.R / F.A.A Office of Aviation Research, September 2003 (superseding DOT/F.A.A/AR-00/47).

- Transport Airplane Metallic and Composite Structures Working Group, Recommendation Report — Structural Bonding, F.A.A.

17.3 Industry standards

- C.M.H-17 (Composite Materials Handbook 17), Rev G. Volume 1 (characterization guidelines), Volume 2 (material properties), Volume 3 Chapter 4 (materials usage, design and analysis; building block approach), Volume 6 (structural sandwich composites). S.A.E International.

- A.S.T.M D3039, D6641, D695, D3518, D5379, D2344, D5528, D5961, D7136, D7137, D5868, D1002, D5573, C273, C297, C364, C393, D7249, C578.

- A.S9100 (quality management for aviation, space and defence organisations) — adopted in structure with documented exceptions per Section 12.2.

• N.A.S 410, N.D.T personnel certification.

17.4 Archon internal

- Archon B.2.N Composite Fabrication Strategy, Revision 3, July 24, 2026 — superseded by this document for the tooling architecture and airworthiness sections; still current for supplier identification (Section 3.6), core selection rationale (Section 3.1), tape-versus-fabric split (Section 3.3), material cost estimate basis (Section 3.7), and the Philippine partner assessment (Section 11).

- B.2.N Boomerang White Paper, April 2026 — authoritative for aircraft performance claims. Note that the L/D and range claims in that document depend on achieving the aerodynamic surface quality discussed in Sections 3.2(e) and 3.5 here.

- B.2.N Rev G structural C-A-D package — pending the 288 in root-chord O.M.L correction.

You have reached the end of the document.