Light Aircraft Design & Certification — Master Workflow
From initial desired parameters to a finished, certified, flying aircraft.
Microlight class, UK jurisdiction primary (BCAR Section S / CAP 482 Issue 8 two-seat; SSDR single-seat). US Part 103 and Experimental Amateur-Built as comparators.
Compiled 10 June 2026. All items marked [REG] against BCAR Section S were verified directly against the CAP 482 Issue 8 text (15 May 2023), confirmed current on the CAA publications page on the compilation date.
Tag legend — used throughout:
- [REG] regulatory requirement (paragraph cited; CAP 482 Issue 8 unless stated)
- [STD] industry standard practice (not legally required, but expected by inspectors/engineering offices)
- [JC] judgement call (defensible engineering opinion; reasonable people differ)
- Confidence: H (verified against primary text / settled physics), M (good secondary evidence or strong recall), L (estimate; verify before relying on it)
Standing caveats: BMAA/LAA procedural leaflets and fee schedules change without much notice — TIL 064 is Issue 4 (Feb 2022) and TIL 045 is Issue 4 (Jan 2018); check for later issues before acting. All cost and timeline figures are [JC, L] unless stated.
- PHASE 1 — REQUIREMENTS DEFINITION
- PHASE 2 — CONCEPTUAL DESIGN
- PHASE 3 — PRELIMINARY DESIGN
- PHASE 4 — STRUCTURAL DESIGN
- PHASE 5 — SYSTEMS DESIGN
- PHASE 6 — DETAIL DESIGN
- PHASE 7 — MANUFACTURING
- PHASE 8 — TESTING & CERTIFICATION
- PHASE 9 — CROSS-CUTTING DISCIPLINES (run from Phase 1, audited at every gate)
- DEPENDENCY MAP — what locks earliest and costs most to change
- READING & REFERENCE LIST BY PHASE
- UNCERTAINTY REGISTER — verify these yourself before relying on them
- SOURCES
PHASE 1 — REQUIREMENTS DEFINITION
Objective: A written requirements document in which every performance number is traceable either to your mission or to a regulatory limit, and the regulatory class is locked.
1.1 The class decision comes first, because it back-constrains everything
| Parameter | UK SSDR (single-seat) | UK two-seat microlight | US Part 103 | US E-AB |
|---|---|---|---|---|
| Max weight | 300 kg MTOM landplane; 315 kg with whole-aircraft parachute; 330 kg sea/amphibian [REG, ANO Art. 33 via BMAA TIL 045, H] | 600 kg landplane; 650 kg amphibian/floatplane [REG, S 2, H] | 254 lb (115 kg) empty weight (not MTOM), excl. floats & emergency safety devices; ≤5 US gal fuel [REG, 14 CFR 103.1, H] | None |
| Stall limit | VS0 ≤ 35 kt CAS at MTOM, landing config [REG, H] | VS0 ≤ 45 kt CAS at MTWA, landing config [REG, S 2, H] | ≤24 kt CAS power-off [REG, H] | None |
| Other perf. limits | — | — | ≤55 kt CAS max level speed at full power [REG, H] | None |
| Design code | None — airworthiness wholly deregulated; pilot carries responsibility [REG, H] | BCAR Section S, compliance demonstrated to BMAA/LAA [REG, H] | None | None — FAA inspects workmanship/documentation, not the design [REG, AC 20-27G, H] |
| Build oversight | None required | Stage inspections mandatory; ≥51% amateur effort; built start-to-finish under BMAA supervision [REG/procedural, TIL 064 §5.2, H] | None | 51% rule; DAR/FAA final inspection [REG, H] |
| Flight test | None required | Programme + check schedule under a Certificate of Clearance for test flying [REG/procedural, H] | None | Phase 1: 40 hr (25 hr with type-certificated engine/prop combo) or task-based plan per AC 90-89C ch. 2 (17 tasks + Aircraft Operating Handbook) [REG, H] |
| Pilot/ops | Microlight licence, registration, 3rd-party insurance, 3-point harness, logbooks; day-VMC class norms [REG, TIL 045, H] | Permit conditions: non-aerobatic (≤60° bank), day VMC, no icing [REG, TIL 064 §1.4, H] | No licence, no registration, no medical [REG, H] | Normal US licensing |
Notes that matter at this stage:
- The UK 600 kg / 45 kt definition came in via ANO amendment (2021) and Section S Issue 8 (2023); BMAA brands 450–600 kg types "Light Sport Microlights" [REG/STD, H]. SSDR limits did not rise with the 600 kg change [REG, H].
- A grandfather clause exists for pre-2003 390 kg amateur single-seaters — irrelevant to a new design [REG, TIL 045, H].
- US MOSAIC final rule: sport-pilot changes effective 22 Oct 2025; new LSA airworthiness provisions effective 24 July 2026 — if the US comparison matters, re-check; the new LSA category at VS0 ≤ 61 KCAS may become a more attractive US route than E-AB for a clean-sheet two-seater [REG, H on dates, M on detail].
- Part 103's 254 lb empty weight is the most brutal constraint in any of these regimes; it forces a sub-minimal aircraft. Don't choose it by default just because it's paperwork-free.
1.2 How the class limits back-constrain the design space
Stall speed caps wing loading. W/S_max = ½ρ₀VS²·CLmax. At sea level ISA (settled physics, H):
| CLmax (landing config) | 35 kt (SSDR/old µL) | 45 kt (600 kg µL) | 24 kt (Part 103) |
|---|---|---|---|
| 1.6 (no flaps, modest section) | 32 kg/m² | 54 kg/m² | 15 kg/m² |
| 2.0 (plain/slotted flaps, realistic) | 41 kg/m² | 67 kg/m² | 19 kg/m² |
| 2.2 (good slotted flaps, optimistic at this Re) | 45 kg/m² | 74 kg/m² | 21 kg/m² |
A 450 kg aircraft at 35 kt and CLmax 2.0 therefore needs ≥11.1 m² of wing; at 600 kg / 45 kt only ≥9.0 m². Design to ≤90% of the stall-limited W/S to leave margin for weight growth and CLmax shortfall [JC — flight-test stall busts are a classic late failure; CAS, not IAS, is what's measured].
MTOM minus realistic empty weight caps payload + fuel. Microlight empty-weight fractions cluster at We/W0 ≈ 0.55–0.65 for conventional two-seat 3-axis types (comparators: C42, EV-97 Eurostar, Skyranger ≈ 0.58–0.62 at 450 kg) [STD — comparator data, M]. Section S additionally forces a minimum credible payload: MTWA must be ≥ empty weight + minimum equipment + fuel for 60 min at max continuous power + occupant weights of 86 kg/seat (≤472.5 kg MTWA), 90 kg/seat (≤525 kg), 100 kg/seat two-seat / 110 kg single (≤650 kg) [REG, S 25 a)2), H]. You cannot paper over a heavy structure by placarding a featherweight crew.
The CG envelope is also regulated, not chosen freely: it must cover a 55 kg solo pilot up to max placarded occupants, with fuel from zero to full; ballast provision required if your assumed minimum pilot exceeds 55 kg [REG, S 23 / S 25 b), H]. Decide now what crew-weight range you honestly intend.
1.3 Mission specification contents
State: seats; payload (use the S 25 table values as floors); range with reserves (fuel mass at ~0.72 kg/L for mogas); cruise speed target; field length (S 51 / S 75 require determining distances over 15 m on dry short grass — there is no pass/fail distance, but your own strip sets one [REG distinction, H]); climb (regulatory floor: ≤4 min from brakes-off to 1000 ft AAL, corrected to SL — roughly ≥250 fpm average [REG, S 65, H]; prudent target ≥500 fpm at MTOM hot-day [JC]); cost ceiling and annual budget; build hours available; hangar/rigging constraint (Section S has specific de-rig/mis-rigging design requirements, S 612 [REG, H]).
GATE 1 — exit criteria
Requirements document signed off against this checklist: class chosen and its limits written as hard constraints; payload ≥ regulatory floor; W/S feasible per table above with margin; cost ceiling stated with a 30% contingency [JC]; comparator list (≥10 similar aircraft with published W0, We, S, P, VS0) compiled — this becomes your statistical base in Phase 2.
Deliverables: Requirements/TLAR document; regulatory constraints register; comparator database.
PHASE 2 — CONCEPTUAL DESIGN
Objective: One configuration, frozen, with a closed first-order sizing loop and a constraint diagram showing simultaneous satisfaction of stall, climb, takeoff, and cruise.
2.1 Configuration selection
Decide, with stated reasons (these lock in early and are near-impossible to change later — see dependency map):
- Wing position & bracing: strut-braced high wing dominates the class for reasons that survive scrutiny: strut bracing cuts spar bending mass substantially, gives benign stall visibility/ground access, simplifies fuel gravity feed (S 955 allows gravity feed at 150% takeoff flow — a pump system needs 125% + redundancy thinking) [REG fuel numbers, H; configuration reasoning STD/JC].
- Tail: conventional fixed fin/stabiliser. Note a hard regulatory hook: the simplified flutter compliance route (FAA Report No. 45 criteria) is only available to conventional designs with fixed fin and stabiliser, no T-/V-/boom-tail, no large outboard masses, no significant sweep [REG, S 629 c), H]. Choosing a T-tail or all-flying tail commits you to flight flutter testing with excitation (S 629 b)) and probably ground vibration work [REG + JC].
- Gear: nosewheel vs tailwheel — Section S ground loads cover both with prescribed conditions (S 479–S 499); nosewheel carries punitive supplementary loads (2.25× static vertical with drag/side components) that size the nose gear and its backup structure [REG, H].
- Engine placement: tractor vs pusher. Pusher behind/above the cockpit or tanks triggers the 15g ultimate engine-retention case (S 561 e)) [REG, H], plus mandatory protection from propeller debris within ±20° of the prop disc plane (S 905 + AMC) [REG, H]. Cooling and CG also argue tractor for a first design [JC].
2.2 First-order sizing — and where the textbooks break down
- Weight loop: W0 = Wcrew + Wfuel + We(W0), iterate. Do not use Raymer's empty-weight regressions below ~600 kg: his statistical base is certified GA and military aircraft (lightest data points around C-150/C-172 class) plus 1980s homebuilts; the regression exponents extrapolate badly into tube-and-fabric construction, and his composite adjustment factors are unvalidated at this scale [JC on the boundary, M; the underlying caution is standard]. Gudmundsson (General Aviation Aircraft Design, 2nd ed.) includes LSA-scale data and is the better single text here, but microlight-specific data is still thin. The honest empirical basis at this scale is your Phase 1 comparator table: regress We/W0 against your own class [STD among microlight designers, M].
- Component build-up beats regression at this scale: engine (a Rotax 912UL is ~57 kg dry, realistically 70–80 kg installed with exhaust, radiators, mount, prop [STD, M]), then structure by area-density comparison with comparators. At 300–600 kg the engine is 12–25% of MTOM, so the engine decision (Phase 5 logically, but make it now) drives the whole mass statement [STD, H].
- Carry a mass growth allowance of 10–15% on empty weight from day one [JC — but the single most common cause of dead projects is its omission].
- Constraint diagram: plot W/S vs W/P with: stall line (vertical, from §1.2 table); climb line (from the 4-min/1000 ft floor and your own target, using rate-of-climb = (ηP/W) − (D·V/W)); takeoff over 15 m for your strip; cruise. Method per Gudmundsson ch. 3 / Raymer ch. 5 — the method is scale-independent even where their statistics aren't [STD, H]. Typical class outcomes: W/S 25–45 kg/m² (450 kg / 35 kt era) or up to ~65 (600 kg / 45 kt), power loading 5–8 kg/hp [STD, M].
2.3 Propulsion candidates (decide the engine class now, exact model by Gate 3)
Section S does not require a certificated engine [REG, S 901–903 + BMAA TIL 064 §7, H]. Compliance is by installed compatibility: the engine/exhaust/propeller combination must function satisfactorily and safely within its limitations (S 903), with AMC acceptance evidence being a 3-hour structured ground test (1 hr at 75% MCP, then a prescribed start/idle/MCP cycling sequence, run twice — AMC S 901 b)1)) and 25 hours of flight without significant problems (AMC S 903) [REG/AMC, H — verified text]. Realistic candidates: Rotax 912UL/ULS (80/100 hp, the class default), Rotax 582 (2-stroke, lighter and cheaper, shorter TBO), Jabiru 2200, and for SSDR the paramotor-derived singles/twins (Polini, Vittorazi) or small V-twins (Verner) [STD, M]. Electric is technically open under SSDR (no code applies) but range/mass closes the case for most missions [JC].
GATE 2 — exit criteria
Weight loop closes with MGA intact; constraint diagram shows a feasible design point with stall margin ≥10%; configuration 3-view drawn; mass statement v0 itemised to ~20 line items; engine class chosen. If the loop only closes by deleting the growth allowance or assuming CLmax > 2.2, it does not close.
Deliverables: Configuration description + 3-view; mass & CG statement v0; constraint diagram; performance estimate v0; engine shortlist with installed-mass budget.
PHASE 3 — PRELIMINARY DESIGN
Objective: Geometry frozen; loads basis established per Section S Sub-Section C; CG envelope defined; no analysis showstoppers.
3.1 Aerodynamics
- Aerofoil: you need honest CLmax at Re ≈ 1–3×10⁶ (a 1.2 m chord at a 35 kt stall is Re ≈ 1.5×10⁶) and a gentle stall, because Section S demands docile behaviour: roll controllable to the stall, ≤20° roll in recovery, no spin tendency, and stall warning (aerodynamic buffet acceptable) if the bare stall isn't benign [REG, S 201–S 207, H]. Proven choices: NACA 4412/4415, GA(W)-1/LS(1)-0417, Riblett GA series; treat high-camber high-lift sections (e.g. FX 63-137) with caution — their large nose-down Cm raises tail loads and wing torsion, and Section S forces you to use at least |Cm0| = 0.025 in loads even if you claim less [REG, S 331 d)2), H]. XFoil/XFLR5 are adequate for trends but do not trust absolute CLmax closer than ±0.1–0.15 at these Re [JC, M]; Abbott & von Doenhoff wind-tunnel data starts at Re 3×10⁶ — note the mismatch at the bottom end [STD, H].
- Planform: AR 5.5–8; rectangular or taper ≥0.5 with 2–3° washout for root-first stall [STD, M]. Spanwise load distribution for loads: chord-proportional for cantilever wings, Schrenk for braced wings [REG/AMC, AMC S 331 d)3), H].
- Drag build-up: component method (Hoerner/Gudmundsson). At microlight scale with struts, wires, open gear and cooling drag, the textbook build-up systematically underpredicts; add 10–25% and calibrate against comparator measured cruise [JC, M].
- Stability & control: static margin 10–20% MAC for a first design [JC]; tail volumes from class comparators rather than Raymer's GA tables (microlight VH typically 0.35–0.55) [STD, M]. Section S requires demonstrated longitudinal/lateral/directional stability and "suitable control feel" across the placard CG range (S 171–S 181) and spin recovery: one-turn or 3-second spin (whichever longer), recovery within one further turn, plus the moderately-mishandled and incipient cases, at critical weight/CG combinations [REG, S 221, H — this is a flight-test demonstration, but you design for it now: adequate rudder authority at full aft stick, aft-CG limit set with margin].
3.2 Loads basis — the V-n diagram (Section S Sub-Section C, all [REG, H])
- Limit manoeuvre load factors (S 337): n1 = +4.0 (at VA), n2 = +4.0 (at VD), n3 = −1.5 (at VD), n4 = −2.0 (at VA). Flaps-down: +2.0 up to VF (S 345).
- Design speeds (S 335, EAS): VA = VS1√n1 = 2.0·VS1; VF ≥ max(1.4·VS, 2.0·VSF); VD ≥ max(1.2·VH, 1.5·VA); VC may be chosen, need not exceed 0.9·VH.
- Gusts (S 333 c) / S 341): ±15 m/s at VC and ±7.5 m/s at VD — but gust cases apply only if VD > 140 kt EAS (S 301 d)), so most microlights are manoeuvre-critical and never run a gust case. This is a major simplification vs CS-VLA; don't import VLA gust cases unnecessarily. The S 341 formula is the standard alleviated sharp-edged gust with k = 0.88μ/(5.3+μ).
- Other prescribed aero-loads inputs: negative CLmax may default to −0.8; chordwise forward load = 25% of wing lift at points A and G (drag-brace design case!); CP ranges 20–60% chord (positive cases), LE–25% (negative) [REG, S 331 + AMC, H].
- Alternative: S 301 e) explicitly permits CS-VLA Appendix A (pre-computed conservative loads) instead of S 321–S 455 for conventional configurations — a legitimate effort-saver for a conventional layout, at some weight cost [REG, H; use is JC].
- Safety factor: ultimate = limit × 1.5 (S 303) — special factors stack on top (Phase 4).
3.3 Flutter and aeroelastics at this scale
Low speeds do not make you safe: fabric-covered structures and cable circuits are torsionally and circuit-soft, and several microlight-class accidents trace to tail flutter [STD/JC, M]. Compliance routes (S 629, [REG, H]): (a) flight flutter tests to VDF with deliberate excitation, showing damping margin and no rapid damping loss approaching VDF; or (b) the FAA "Simplified Flutter Prevention Criteria" (Airframe & Equipment Engineering Report 45) — wing torsional stiffness, aileron/elevator/rudder mass-balance and circuit-stiffness/free-play criteria — if your configuration qualifies (see 2.1). Design intent for a first aircraft: meet Report 45 and do the flight expansion carefully [JC — belt and braces]. Static balance of elevator and aileron to ~100% is conventional conservative practice where Report 45 demands balance [STD, M]. Mass-balance weight attachments have their own prescribed limit loads: 24g normal to surface / 12g fore-aft / 12g spanwise (S 659) [REG, H]. Divergence and control reversal are covered by the same paragraph — check aileron reversal speed on a soft fabric wing [REG requirement / analysis is STD].
GATE 3 — exit criteria (and the most important phone call of the project)
Loads report v1 covering every S 321–S 499 case applicable; V-n diagram drawn for your numbers; CG envelope with fwd limit (elevator power to flare) and aft limit (static margin + spin recovery rationale); flutter compliance route chosen and feasible; geometry frozen. Before detail design: present the concept to the BMAA Technical Office (or LAA Engineering) and agree the certification basis, test expectations, and document format. Both associations accept new designs; BMAA reviews designs "for a considerably lower fee than the CAA" [procedural, H]; the LAA chief-engineer line "run a mile from an unapproved design" cuts both ways — engage before you've welded anything [STD, H]. For SSDR none of this is required [REG, H] — doing it anyway (or buying an independent stress review) is the single best money an SSDR designer can spend [JC].
Deliverables: Aero data set; loads basis document + V-n; stability/control report with CG envelope; flutter compliance plan; frozen external geometry; written engagement with BMAA/LAA Tech Office (two-seat route).
PHASE 4 — STRUCTURAL DESIGN
Objective: Every primary load path substantiated on paper with positive margins of safety, and a test plan for everything analysis can't honestly cover.
4.1 The factor stack — what Section S prescribes vs leaves to you
Ultimate strength requirement = limit load × 1.5 (S 303) × applicable special factors (S 619) [REG, H]:
| Factor | Value | Applies to | Source |
|---|---|---|---|
| Basic safety factor | 1.5 | everything | S 303 [REG, H] |
| Fitting factor | 1.15 | every fitting (part joining one member to another), its attachment, and bearing on joined members — unless that joint is proven by limit+ultimate test; none needed for continuous metal joints, welds, wood scarfs backed by comprehensive data | S 625 [REG, H] |
| Casting factor | 2.0, reducible to 1.25 with 3-sample static tests + 100% radiographic/equivalent NDI of production | castings | S 621 [REG, H] |
| Bearing factor | 2.0 | bolted/pinned joints with relative motion or free-fit joints subject to pounding/vibration | S 623 [REG, H] |
| Control-surface hinge factor | 6.67 on ultimate bearing strength of softest bearing material (plain hinges); ball/roller: stay within approved rating | hinges | S 657 [REG, H] |
| Push-pull control joint factor | 3.33 bearing (2.0 for cable-system joints) | control circuit joints in angular motion | S 693 [REG, H] |
| Cable factor | 2.0 on nominal breaking strength | structural + primary control cables (min Ø 2 mm in primary controls, S 689) | S 626 [REG, H] |
| Two-hinge surfaces | extra 1.5 on hinges and local structure | control surfaces/flaps with only two hinges | S 625 e) [REG, H] |
| Composite default | 2.25 overall factor of safety unless otherwise agreed (i.e. accept a 50% knockdown or negotiate via test evidence) | composite structure of uncertain/variable strength | S 619 b) [REG, H] |
| Textile primary members (non-lifting) | 5.0 | e.g. fabric straps as structure | S 626 b) [REG, H] |
| Harness attachments | 1.33 × S 561 emergency loads | belt/harness load path | S 625 d) [REG, H] |
Where Section S is silent (wood variability, weld efficiency, bonded metal), the designer chooses allowables such that understrength is "extremely remote" (S 613, statistical basis) [REG, H] — in practice: ANC-18 / Forest Products Lab minimum values for selected aircraft-grade spruce/ply with standard moisture and duration-of-load factors [STD, M], MMPDS A/B-basis values for metals [STD, H], 80–85% weld-zone efficiency for normalised 4130 unless tested [STD, M].
4.2 Governing load cases to run (typical conventional microlight) [REG list, H; criticality ranking JC]
Wing up-bending/torsion at +4 limit (A and D points — D adds torsion via lower CL/higher q and the Cm0 floor); forward-chord case (25% lift) for drag bracing; n4 = −2 down case (strut-braced wings: strut goes to compression — buckling, and lift-strut attachment fittings see reversal); asymmetric rolling cases (S 349 area: 100%/70% semispan); tail balancing + manoeuvre loads (S 421–S 447); control circuits at 125% surface hinge moments capped by the pilot-force table: pitch 75 daN stick / roll 30 daN stick / rudder 90 daN per pedal (design minimum 60% of these; dual controls ×0.75 each, together and opposing) (S 395–S 411) [REG, H]; ground cases — limit descent velocity 0.51(W/S)^¼ m/s clamped to 2.13–3.05 m/s (W/S in N/m² as in the CS-VLA parent rule [REG formula H; units M — confirm against AMC]), inertia n ≥ 2.67 / ground reaction ≥ 2.0, wing lift relief ≤⅔W, level/tail-down/one-wheel/side (1.33 vert + 0.83 side)/braked-roll (μ = 0.8) + nosewheel 2.25×-static supplementary cases (S 473–S 499) [REG, H]; emergency landing ultimate 9.0g fwd / 4.5g up / 4.5g down / 3.0g side on occupants, items of mass, and fuel tanks, 15g engine retention if engine is above/behind occupants or tanks (S 561) [REG, H]; engine mount torque (S 361) and side load (S 363) [REG, H].
4.3 Material system trade study (decision locks the rest of the project)
| System | Structural character & failure modes | QC/inspectability | Amateur suitability | Section S interaction |
|---|---|---|---|---|
| Wood (spruce/fir + birch ply, epoxy or resorcinol) | Benign fatigue, good damping; fails by moisture cycling, glue-line starvation, compression creep at fittings | Glue lines invisible — process control only (coupon per glue batch [STD]); rot needs access holes | High if patient; cheap tooling | Variability handled via conservative ANC-18 allowables + S 613; no extra prescribed factor [REG/STD, H/M] |
| Aluminium (6061-T6 or 2024-T3 sheet/tube, pulled or driven rivets) | Predictable, best data; fails by fatigue at fittings, buckling; corrosion manageable | Fully visual-inspectable; rivet quality visible | High; the class default for kit-style airframes; pulled structural rivets (e.g. Avex) are accepted class practice [STD, M] | Cleanest route through S 613/S 619 — published allowables, no negotiated factors [STD, H] |
| 4130 steel tube truss + fabric fuselage | Crashworthy cage (helps S 561); fails by weld defects, internal corrosion | Welds visually inspectable + cheap dye-pen; internal corrosion hidden (oil-fog tubes [STD]) | Requires real welding skill — inspector will scrutinise; practice coupons expected [STD] | Weld process falls under S 605 process-spec requirement [REG, H] |
| Composites (glass/epoxy, mouldless or moulded) | Highest shape freedom, stiffness tailoring; fails by voids, disbonds, hot-soak Tg loss, UV; no visual warning | Tap-test only at amateur level; cure/mix-ratio logs are the QC | Moderate-high effort; foam/glass methods well documented (Rutan lineage) | FoS 2.25 unless otherwise agreed (S 619 b)) — "otherwise agreed" in practice means element/full-scale testing to ultimate with temperature allowance [REG H; practice STD, M] |
| Aluminium tube + Dacron sailcloth (X-Air/Skyranger style) | Lightest cost/effort path; bolted joints, fabric life-limited by UV/tension | Everything visible; fabric strength-tested at revalidation [STD, M] | Highest — fastest amateur route | Watch S 627 (no flexible coatings, inspectability) and fabric strength retention [REG, H] |
[Whole-table classification: characteristics STD/H–M; recommendations JC.]
4.4 Margin bookkeeping
For every element: MS = allowable / (limit × 1.5 × special factors) − 1 ≥ 0, tabulated with load case, method, and allowable source. This stress report is the document the BMAA/LAA reviewer works through; format it per their templates [STD, H]. Section S's blunt rule hangs over all of it: "The strength of any part having an important bearing on safety and which is not amenable to simple analysis must be established by test" (S 601) and analysis alone is acceptable only for structure types where experience shows analysis reliable (S 307 a)); substantiating tests are "normally taken to ultimate design load" with corrections for material/dimensional variation (AMC S 307 a)) [REG, H]. Plan accordingly: for a conventional metal wing, analysis + a proof-to-limit test may be negotiable; for composites or novel geometry, budget a full ultimate static test of wing and tail — likely on a dedicated test article, since a structure taken to ultimate (held 3 s, S 305 b)) is generally not flown afterwards [REG H; programme choice STD/JC, M].
GATE 4 — exit criteria
Stress report with positive margins on every primary path; test plan agreed in principle with the engineering office; mass statement v2 (now bottom-up from member sizing) still closes against MTOM with growth allowance ≥5% remaining; materials sourced with traceability (release certs / receipts for primary-structure material — expected at MAAN stage [STD, M]).
Deliverables: Stress report; structural test plan; updated mass/CG; material traceability file.
PHASE 5 — SYSTEMS DESIGN
Objective: Powerplant installation, fuel, electrics, controls and instruments designed to the specific Section S subpart requirements — which are short, concrete, and all verified below [REG, H throughout unless tagged].
- Powerplant: install per engine manufacturer's installation manual; where impracticable, per proven aircraft practice (AMC S 901 b)). Electrical bonding across powerplant components (S 901 c)). Compatibility evidence: 3-hr ground test schedule (AMC S 901 b)1)) + 25 hr flight (AMC S 903). Propeller clearances: ≥180 mm ground (nosewheel types; 230 mm tailwheel), ≥25 mm radial to structure (100 mm to wires), 13 mm longitudinal growing to 50 mm at tip under test (AMC S 925). Two-stroke note: paper fuel filters <10 µm are explicitly disallowed with premix oil (AMC S 977 d)). Carb-heat/induction icing falls under S 1091 — at minimum follow the engine maker's installation requirement [M on detail]. Firewall/shroud must be fireproof, sealed with fire-resistant grommets (S 1191); cowling fire-resistant, fireproof near exhaust (S 1193).
- Fuel system: flow demonstration 150% of takeoff consumption (gravity) or 125% (pump) (S 955); arranged so vapour lock cannot occur (S 951 c) — and the AMC flags flat-bottomed/low-dihedral integral tanks as a known starvation trap); unusable fuel established, ≤5% of capacity (S 959); tank withstands 1½ psi (S 965); sump 0.10% of capacity or 120 cm³ (or 25 cm³ accessible sediment bowl/gascolator arrangement); tank compartment vented/drained, no engine-fire impingement, no leak path onto occupants, tanks retain contents at 9g/4.5g/3g and survive heavy-landing deformation (S 967 / S 561 f)). Ethanol: Rotax 912-series approved to E10; hose and composite-tank degradation is a live fleet issue — spec ethanol-rated hose throughout [STD, H per TIL 064 §7].
- Electrical: Section S is light here; standard practice governs [STD]: single battery + master contactor, fusing at the source, engine ignition independent of the aircraft battery (magneto/self-powered CDI — Rotax 912 ignition is self-generating), load analysis if EFIS-dependent — and if one electronic display replaces required instruments it must run ≥20 min after alternator/generator failure, with any stall warner independent of it (S 1301 c)) [REG, H].
- Flight controls: sized in Phase 4 to S 395–S 411 forces; details prescribed: positive stops sized to circuit design loads; cables ≥2 mm with inspectable terminals, guarded pulleys matched to cable size, fairleads ≤3° deflection, turnbuckles non-binding (S 689); anti-jam/foreign-object protection, anti-mis-assembly features or permanent marking (S 685 area); trim with direction + position indication, irreversible unless balanced (S 677 area); functional test of full circuit under load for jamming/friction/deformation (S 683). Free play limits tie back to your flutter case [REG + STD, H].
- Instruments — regulatory minimum is startlingly short: ASI and altimeter (S 1303); engine instruments as the engine manufacturer requires plus fuel quantity per tank visible strapped-in, oil quantity (dipstick acceptable), MAP if independently-controllable VP prop (S 1305); lap strap and upper-torso restraint for all occupants (S 1307). No compass required by Section S [REG, H] — fit one anyway, plus slip ball, CHT/EGT appropriate to the engine, and (judgement) a stall margin cue for the test programme [JC]. ASI system must be calibrated in flight to ±5 mph/±5% from 1.3VS1 to VNE (S 1323) — this is also how you prove the 35/45 kt stall legally, in CAS [REG, H].
- SSDR divergence: none of the above is required (no minimum instrument fit at all — TIL 045 §3.4 [REG, H]); a 3-point harness is required by the ANO (Art 77/Sch 5) [REG, H]. Treat the Section S systems rules as your free checklist [JC].
GATE 5 — exit criteria
Systems schematics complete (fuel, electrical, controls); engine installation design reviewed against the engine maker's manual; fuel flow/vapour analysis done on paper (test in Phase 8); instruments list fixed (drives panel, electrical load, weight); failure-modes pass on fuel + controls (informal FMEA) [JC but expected by reviewers in practice, M].
Deliverables: Schematics pack; powerplant installation drawings; instrument & equipment list; updated mass/CG (systems are where the growth allowance dies — re-check).
PHASE 6 — DETAIL DESIGN
Objective: Drawings an inspector can check the aircraft against, down to fastener callouts.
- Hardware standards: AN/MS/NAS hardware (or LN/DIN metric equivalents) with traceable sourcing; the inspector expects recognised aircraft hardware in primary structure and controls [STD, H]. Section S hard rules: locking devices on all primary-structure and control connections; rotating joints need a non-friction lock (castellated nut + pin) in addition to any self-locking nut (S 607) [REG, H]. Standard practice on top [STD, H]: no threads in bearing (grip-length selection), ≥1 thread protruding, nyloc temperature limits keep them out of the hot firewall-forward zone (all-metal locknuts there), torque per AC 43.13-1B tables, safety wire on turnbuckles.
- Joint sizing method (worked once, then templated): joint limit load → ultimate = ×1.5 ×1.15 fitting factor (×2.0 bearing factor if it rotates) → check bolt shear, bearing on each ply, net-tension, and shear-out at edge distance 2D (rivets: pitch ≥3D, edge ≥2D) [REG factors H; geometry rules STD, H]. Example: an AN3 (Ø 3/16") steel bolt in single shear carries ≈2,100 lbf ultimate; in 1.6 mm 2024-T3 the joint is bearing-critical well before bolt shear — so the sheet, not the bolt, sizes the joint [STD method H; table values M — take finals from MMPDS/AC 43.13, not memory].
- Castings/bearings/hinges: apply the S 621 / S 623 / S 657 / S 693 factors from the Phase 4 table at the detail level — hinge bearing at 6.67 on the softest material is what actually sizes piano-hinge and bushed hinges [REG, H].
- Drawing standards: numbered drawings with revision blocks; parts list with material + spec per part; process notes invoking your written process specs (S 605 requires defined process specifications for glue/weld/heat-treat/composite work [REG, H]); assembly drawings showing locking method per fastener group. The BMAA inspector's job is explicitly conformity against approved data (TIL 064 §5.6, with the MAAN as the as-built reference) — if it isn't on a drawing, it can't be conformed [procedural, H].
- Marking/anti-misassembly on control elements (S 685 d)) and rigging idiot-proofing (S 612) are detail-design items, not afterthoughts [REG, H].
GATE 6 — exit criteria
Released drawing set under change control (issue log); every primary joint has a margin entry; build sequence sketched; long-lead purchases (engine, spar material, canopy) ordered against frozen drawings.
Deliverables: Drawing set + parts list; process specifications; fastener/locking schedule; updated stress report annexes.
PHASE 7 — MANUFACTURING
Objective: Build the aircraft so that what exists equals what was approved, with evidence.
- Project registration & supervision (two-seat UK): register the project with BMAA (Form AW022 — signed by owner and inspector) before building; construction must be ≥51% amateur effort and run start-to-finish under BMAA supervision per TIL 039, with stage inspections by a BMAA-authorised inspector at key stages — in practice: before closing any box structure or covering (wings, fuselage, tail), control systems complete, fuel system, engine installation, and final [procedural REG-equivalent, H on the framework; exact stage list per TIL 039, M]. Book the inspector at project start and agree hold points; an uninspected closed structure may have to be opened [STD, H].
- Jigging & tooling: flat, stable wing jig (spar datum + rib stations), fuselage table/rotisserie; calibrate the torque wrench and the scales you'll use for weighing [STD/JC].
- Process control per material system: glue coupon per mixing session (wood); weld test coupons + dye-pen on cluster welds (steel); cure log + traveller coupon per layup batch, cure-temperature control (composite); rivet practice strips per gun/setting (aluminium). Keep them — they are your S 605 evidence [STD implementing REG, H].
- Records: build log with dated photos of every cavity before closing; deviations/concession register (anything not-per-drawing gets a numbered concession reviewed against stress); material certs filed against drawing part numbers [STD, H].
- Weighing: performed or witnessed by a BMAA inspector (Form AW028) [procedural, H]; empty weight & CG determined per S 29 in a defined, repeatable condition [REG, H].
- SSDR: none of this is required [REG, H]. Replicate the stage-inspection points with any experienced second pair of eyes (a BMAA/LAA inspector acting privately as a "qualified person", an engineer friend) and keep the same records — partly for safety, partly because an undocumented aircraft is unsellable and uninsurable in practice [JC, strongly held].
- US E-AB: keep the builder's log photo-rich — it is your 51%-rule evidence and the DAR will sample it [REG/STD, H].
GATE 7 — exit criteria
All stage inspections signed; as-built conforms to drawings or has closed concessions; aircraft weighed — actual empty weight & CG inside the Phase 4 predictions (if not: stop and re-run the loading/stress cases before any flight); registration applied for (G-reg; ANO Art 24 applies to SSDR too [REG, H]).
Deliverables: Signed inspection records; build log; weighing report (AW028); concession register; registration.
PHASE 8 — TESTING & CERTIFICATION
Objective: Demonstrate compliance, expand the envelope, and convert a machine into a permitted aircraft.
8.1 Structural & ground testing (sequence matters)
- Structural substantiation tests per the Phase 4 plan: proof to limit (no detrimental permanent deformation, controls operable — S 305 a)) and, where required, ultimate held 3 s (S 305 b)), normally to ultimate per AMC S 307 a) [REG, H]. Whiffletree or distributed sandbags on an inverted wing is the classic rig [STD, H]. Control circuits: functional test under S 397 loads — no jamming/excess friction/deformation (S 683) [REG, H].
- Fuel flow test (150%/125% as applicable) at worst attitude; tank 1½ psi pressure check; unusable-fuel determination (can partly run in flight test) [REG, H].
- Engine ground running: the AMC S 901 3-hour schedule; cooling margins vs limits (S 1041); prop clearance under start/run per AMC S 925 [REG, H].
- Taxi tests: low→high speed, brakes, directional control (no ground-loop tendency — S 233) [REG, H].
- Pitot-static: leak check; plan the in-flight calibration method (GPS box pattern or pacer) for S 1323 [REG requirement, H; method STD].
8.2 Flight test (UK two-seat route)
- Authority to fly: Certificate of Clearance for Flight for Test Purposes (Form AW029) — signed by owner + BMAA inspector + the project test pilot [procedural, H]. Test-pilot acceptability for a new type is agreed with the BMAA; for a genuinely new design expect them to require demonstrated test experience, not just the builder's enthusiasm [procedural, M].
- Programme content = Sub-Section B demonstrations at worst-case weight/CG (S 21 / S 25): stalls in all configurations and powers with the 1 kt/s technique, CAS-corrected, proving VS0 ≤ 45 kt at MTWA (S 49 / S 201–S 207); spin programme per S 221 (with recovery parachute and escape route planned — this is the highest-risk block [JC]); climb ≤4 min to 1000 ft (S 65); takeoff/landing distances determined (S 51 / S 75); trim, stability, control-feel across envelope (S 143–S 181); vibration/buffet free to VDF (S 251) and flutter excitation expansion to VDF (S 629 b)); crosswind investigation (S 234 area); ASI calibration (S 1323); 25 hr engine/prop compatibility (AMC S 903); fuel system behaviour incl. low-fuel/slip cases. [All REG, H; ordering JC.]
- Limitations fall out of test: VNE = 0.9 × VDF (and VDF ≤ VD) (S 1505); VFE ≤ 0.9·VF (S 1511); manoeuvring speed placard ≤ VA (S 1507); ASI markings (white/green/yellow arcs, red line at VNE — S 1545 area) [REG, H].
- Documentation to finish: Pilot's Handbook with mandatory content (S 1581–S 1585: limitations, weights/CG, performance, procedures, weighing-amendment procedure), maintenance schedule, placards. The BMAA Tech Office issues the MAAN (the aircraft's approved-design baseline) and, for a type others may build, a HADS; the Permit to Fly (non-expiring) is issued by BMAA/CAA, kept alive by an annual Certificate of Validity (inspector airworthiness review + check flight per the check flight schedule) [procedural, H]. Permit conditions: non-aerobatic (≤60° bank), day VMC, no icing [REG, H]. BMAA then formally owns continued airworthiness of amateur-built types (defect reporting goes to them) [procedural, H].
- Who signs what, summarised: builder (work + logbooks); BMAA inspector (stage inspections, weighing witness, AW029, annual CoV inspection); BMAA Tech Office engineer (design approval/MAAN, permit recommendation); test pilot (AW029 + flight test schedule); CAA (oversight of BMAA's A8-26 approval; issues/backs the Permit) [procedural, H].
8.3 US comparison
- E-AB: register; airworthiness application + notarised eligibility statement (Form 8130-12); DAR/FAA inspection → special airworthiness certificate with operating limitations; Phase 1 = 40 hr (25 hr with type-certificated engine/prop combination) or the task-based programme (AC 90-89C ch. 2: 17 tasks, test cards, Aircraft Operating Handbook produced; no separate FAA approval needed to elect it) [REG, H]; then Phase 2 normal ops; builder can obtain the repairman certificate for that aircraft (condition inspections) [REG, H]. The FAA never reviews your stress report — freedom and exposure in one sentence [REG/H, framing JC].
- Part 103: no certification events at all; AC 103-7 is the substantiation method if challenged on weight/speed/stall [REG, H].
8.4 Timeline & cost (all [JC, L] — wide bands, stated for budgeting, not quotation)
- SSDR own design (simple, single-seat, 2-stroke or small 4-stroke): design 300–800 hr; build 600–1,500 hr; cash £10–30k (engine £3–12k dominating); calendar 2–5 yr.
- Two-seat new type via BMAA, amateur prototype: design + substantiation 1,000–3,000 engineering hours; build 1,500–4,000 hr; structural + flight test campaign 6–24 months elapsed; cash £40–120k including engine (new 912ULS of order £20–25k [L]), materials, instruments, test consumables, association fees (engineering review + permit, order £2–10k [L — get the current BMAA fee schedule; it is published]); insurance for test flying is a real and variable line item. Calendar realistically 4–10 yr.
- Base rate honesty: a majority of clean-sheet amateur designs are never completed, and of those completed, first-of-type approval campaigns frequently double their planned test duration [JC, M — consistent with association commentary; no solid published statistic].
GATE 8 — exit criteria
Permit to Fly (or SSDR: your own signed-off test programme completion) + current Certificate of Validity; Pilot's Handbook issued; maintenance schedule in force; weight & CG schedule current.
PHASE 9 — CROSS-CUTTING DISCIPLINES (run from Phase 1, audited at every gate)
- Weight & CG ledger: one owner, one spreadsheet, every part weighed as bought/made vs estimate; growth-allowance burn-down chart reviewed at each gate. CG checked against the S 23 loading rectangle (55 kg pilot ↔ max occupants × zero ↔ full fuel) at every revision [REG envelope H; discipline STD].
- Change control: after the MAAN exists, any modification needs approval before embodiment (TIL 064 §5.4; AW002 forms; major changes need BMAA engineering) [procedural, H]. Before approval, run the same discipline privately: drawing revision + stress check + mass update or it doesn't happen [STD].
- Iteration loops, and where rework bites: weight ↔ stall-W/S ↔ structure is the master loop (each pass changes loads, hence sizing, hence weight — converge it on paper in Phases 2–4, because converging it in hardware costs a re-spar). Engine change is a restart of Phase 5 + mount + CG + flutter mass distribution. Late aft-CG discovery is the classic killer: it surfaces at first weighing (Gate 7) and the fixes (ballast, engine relocation, tail enlargement) range from ugly to structural [STD/JC, H as a failure pattern].
- Most common failure points [JC/M, ranked]: (1) empty-weight growth eating payload/stall margin; (2) project abandonment from undocumented scope creep and life events — mitigated by sub-assembly milestones and club/Strut community; (3) stall speed busting the class limit in CAS during test (ASI position error optimism); (4) documentation debt discovered at permit application — record as you go or pay double; (5) inspector/engineering-office disengagement after long silences; (6) flutter/free-play findings at envelope expansion forcing balance retrofits; (7) fuel starvation/vapour issues from flat tanks and unproven flows (the AMC singles this out for a reason); (8) regulatory drift across a decade-long project — re-baseline at every gate.
DEPENDENCY MAP — what locks earliest and costs most to change
| Order | Decision | Locked at | Cost to change after lock | Ripples into |
|---|---|---|---|---|
| 1 | Regulatory class (SSDR / S two-seat / 103 / E-AB) | Gate 1 | Total restart — different MTOM, stall, oversight | Everything |
| 2 | Seats + engine class | Gate 2 | Near-restart (mass fractions, CG architecture, S 25 payload floor) | Sizing, structure, systems |
| 3 | Configuration (wing/tail/gear/tractor-pusher) | Gate 2 | Major redesign; tail choice can silently forfeit the simplified flutter route (S 629 c)); pusher buys the 15g case (S 561 e)) | Loads, flutter, crash cases |
| 4 | Material system | Gate 2–3 | Restart of Phases 4, 6, 7 and the test plan (composites swap analysis for testing via the 2.25 factor) | Stress, QC regime, inspector skill match |
| 5 | Wing area / W/S (stall compliance) & span/spar architecture | Gate 3 | New spar = new wing; strut vs cantilever flips the down-load case | Structure, performance, flutter stiffness |
| 6 | CG architecture (engine/crew/tank longitudinal arrangement) | Gate 3 | Ballast at best; structural relocation at worst | Stability, spin, loading envelope |
| 7 | Exact engine model & fuel type | Gate 5 (class fixed Gate 2) | Mount, cowl, CG, fuel system, 25-hr compatibility re-run | Systems, test programme |
| 8 | Control circuit type (cable/pushrod) & surface balance | Gate 5–6 | Moderate; but flutter substantiation may need repeating | Flutter, detail design |
| 9 | Instruments/avionics, fairings, paint, interior | Gate 6–7 | Cheap (within electrical load & weight ledger) | Little |
Rule of thumb: anything above line 6 changed after Gate 4 costs an order of magnitude more than changing it on paper would have [JC, H as a pattern].
READING & REFERENCE LIST BY PHASE
Phase 1 (regimes): CAP 482 Issue 8 (the primary text — read S 2, S 21–S 29 first); BMAA TIL 045 (SSDR) and TIL 600 (600 kg Light Sport Microlights); ANO 2016 Arts 24/33/77/226 + Sch 5; 14 CFR Part 103 + AC 103-7; AC 20-27G (E-AB); FAA/AOPA MOSAIC materials (if US-relevant).
Phase 2 (sizing): Gudmundsson, General Aviation Aircraft Design (2nd ed.) — the workhorse; Raymer, Aircraft Design: A Conceptual Approach (methods, with the statistical caveat in §2.2); Stinton, The Design of the Aeroplane; Hiscocks, Design of Light Aircraft; Pazmany, Light Airplane Design; your comparator database.
Phase 3 (aero/loads): Abbott & von Doenhoff, Theory of Wing Sections; UIUC airfoil database + XFLR5/XFoil; Hoerner, Fluid-Dynamic Drag; CS-VLA (esp. Appendix A, usable via S 301 e)); FAA Airframe & Equipment Engineering Report No. 45, "Simplified Flutter Prevention Criteria" (invoked by S 629 c)).
Phase 4 (structures): Bruhn, Analysis & Design of Flight Vehicle Structures; MMPDS (or free MIL-HDBK-5J) for metal allowables; ANC-18 Design of Wood Aircraft Structures + FPL Wood Handbook; Gordon, Structures (conceptual grounding); LAA/BMAA stress-report format guidance from the engineering office.
Phases 5–7 (build): AC 43.13-1B/2B Acceptable Methods, Techniques and Practices (torque tables, hardware, practices — UK associations treat it as reference-grade [STD]); Bingelis, The Sportplane Builder / Sportplane Construction Techniques / Firewall Forward / On Engines; Standard Aircraft Handbook for Mechanics and Technicians; engine installation manual (Rotax manuals are free online); BMAA TIL 039 (amateur build process), TIL 044 SIGMA (inspection standards), TIL 064 (airworthiness system), TIL 012 (weighing).
Phase 8 (test): AC 90-89C Amateur-Built Aircraft and Ultralight Flight Testing Handbook (the task-based Phase 1 lives in ch. 2 — excellent test-card material even for UK use); Askue, Flight Testing Homebuilt Aircraft; BMAA TIL 075 (limiting speeds & ASI calibration); BMAA check flight schedules; FAA-H-8083-1B Weight & Balance Handbook.
UNCERTAINTY REGISTER — verify these yourself before relying on them
- Post-January 2026 regulatory changes: CAP 482 Issue 8 confirmed current on the compilation date via the CAA publications page, but BMAA/LAA TIL/TL issue states and fee schedules were not all date-checked; TIL 045 (2018) predates the 600 kg reform (its SSDR content remains consistent with current BMAA web guidance).
- Ground-load descent-velocity units in S 473 b): formula matches CS-VLA 473 (W/S in N/m²); the clamp (2.13–3.05 m/s) is verified text; confirm units via AMC/engineering office before using in anger.
- BMAA stage-inspection list and new-type flight-test schedule length: framework verified (TIL 064 / TIL 039 references); the precise hold points and test-hour expectations for a first-of-type are set case-by-case by the Tech Office — get them in writing at Gate 3.
- All costs, fees, hours, and the Rotax price — [JC, L] throughout; obtain current BMAA fee schedule and engine quotes.
- MOSAIC airworthiness provisions take effect 24 July 2026; if the US path matters, re-read the final rule then.
- AN-hardware allowable values quoted in the Phase 6 example are from memory [M] — take design values from MMPDS/AC 43.13 tables.
SOURCES
- CAA CAP 482 publication page (and the Issue 8 PDF itself, read directly)
- CAA Section S consultation and Comment Response Document CAP2548
- BMAA SSDR page and TIL 045 — Single-seat Microlights
- BMAA TIL 064 — Guide to Airworthiness
- BMAA TIL 044 — SIGMA Standard Inspection Guidelines
- BMAA 600 kg pages and TIL 600 — 600 kg Light Sport Microlights
- BMAA New Aircraft and HADS index
- BMAA Inspectorate
- LAA Technical Leaflets index
- FLYER — New 600 kg microlight class is now law and Francis Donaldson on unapproved designs
- FAA AC 90-89C and EAA task-based Phase 1
- FAA MOSAIC final rule issuance and AOPA MOSAIC FAQ
- CAA microlight reform — implementation and key decisions
Working copies of CAP 482 Issue 8 (PDF + extracted text) are saved at C:\Users\ciana\AppData\Local\Temp\aircraft-refs\.