Aircraft design process

In short

The aircraft design process is the structured sequence by which a flight vehicle moves from a mission requirement to a certified, manufactured product: requirements definition, conceptual design, preliminary design, detail design, prototype build, flight testing and certification. It is iterative because aerodynamics, structures, propulsion, weights, stability and cost all interact — changing one drives all the others, the classic "design spiral". Indian universities examine it as a full-semester Aircraft Design course with a design project, and GATE Aerospace tests the sizing relations (weight fractions, wing loading, thrust-to-weight, V-n diagram, range equation). It is the daily work of design engineers at HAL, DRDO's ADA and ADE, CSIR-NAL, ISRO, Tata Advanced Systems, Mahindra Aerospace, Airbus and Boeing engineering centres in Bengaluru, and the many Indian drone and eVTOL startups certified under DGCA rules.

Explained

Requirements, the design spiral and types of design

• Design requirement - the mission the aircraft must fly, expressed quantitatively: payload, range, cruise speed and altitude, take-off and landing field length, rate of climb, service ceiling, endurance, manoeuvre load limits, cost and life.
• Civil requirements come from the market (seats, sector length, airport category) plus mandatory airworthiness codes; military requirements come from an Air Staff Requirement (ASR) or Qualitative Requirement issued by the services.
• Design spiral (iterative loop): requirements → first weight estimate → sizing (wing loading, thrust loading) → configuration layout → aerodynamic analysis → structural and weights analysis → performance and stability check → cost → back to weights with better data. Each lap narrows the uncertainty.
• Three classes of design effort (Raymer's classification, standard in Indian syllabi):
  • Conceptual design - what the aircraft looks like; many configurations sketched, crude but fast analysis, requirements still negotiable.
  • Preliminary design - configuration frozen ("configuration freeze"/lofting); specialist groups do wind-tunnel tests, CFD, FEM, detailed weights; ends with the decision to commit to production.
  • Detail design - every part, rib, fitting, fastener and bracket drawn and released; manufacturing tooling and assembly jigs designed; parts released to shop floor.
• Design drivers are classified as new design (clean sheet), derivative (stretch, re-engine, e.g. A320 → A320neo), and modification/retrofit (role change, avionics upgrade — HAL does many of these).
• Trade study - systematically varying one parameter (aspect ratio, engine bypass ratio, sweep) while holding the mission fixed, to find the best value against a figure of merit (DOC, empty weight, fuel burn, life-cycle cost).

Conceptual design: weight estimation and sizing

• Weight breakdown: W0 (gross take-off) = We (empty) + Wf (fuel) + Wp (payload) + Wcrew.
• Empty weight fraction We/W0 comes from statistical regressions on existing aircraft of the same class — about 0.60–0.65 for jet transports, 0.55–0.65 for general-aviation aircraft, 0.45–0.55 for jet fighters, lower still for sailplanes and some UAVs.
• Fuel fraction by the mission segment method: the mission is broken into segments (warm-up/taxi, take-off, climb, cruise, loiter, descent, land, reserve), each with a weight ratio Wi/Wi−1. Short segments use historical fractions (0.97–0.995); cruise and loiter use the Breguet equations.
• W0 is then solved iteratively (guess W0 → read We/W0 regression → compute W0 again) because We/W0 itself depends on W0.
• Constraint analysis / matching chart: each performance requirement (take-off distance, landing distance, climb gradient, ceiling, sustained turn, cruise speed, stall speed) is plotted as a curve of T/W versus W/S. The feasible region is the intersection; the design point is chosen at the lowest T/W and highest acceptable W/S inside it — low T/W means a smaller engine, high W/S means a smaller wing.
• Typical values to quote: W/S ≈ 5,000–7,000 N/m² for jet transports, 1,000–2,000 N/m² for light aircraft; T/W ≈ 0.25–0.35 for transports, 0.8–1.2 for fighters; power loading W/P for propeller aircraft instead of T/W.
• Initial geometry: wing area S = W0/(W/S); engine thrust T = (T/W)·W0; then span, chord, taper, sweep, tail areas from tail volume coefficients, fuselage length from statistical length-vs-W0 fits and from cabin layout.

Configuration layout and aerodynamic design

• Configuration choices: wing position (high/mid/low), tail arrangement (conventional, T-tail, cruciform, V-tail, canard, all-moving tailplane), engine number and placement (wing pylon, fuselage-mounted, buried, pusher propeller), landing-gear arrangement (tricycle, tail-wheel, bicycle).
• Wing design parameters: aspect ratio AR = b²/S (high AR → low induced drag, heavier and more flexible wing), taper ratio λ = ct/cr (≈0.3–0.5 gives near-elliptic lift distribution), sweep Λ (delays compressibility drag rise, raises critical Mach number, worsens low-speed stall and tip-stall tendency), dihedral (lateral stability, typically 3–7° for low wings, 0 to −3° anhedral for high wings), twist/washout (root stalls before tip), t/c thickness ratio.
• Airfoil selection: required Clmax, design Cl at cruise, drag bucket location, pitching moment, thickness for spar depth and fuel volume; supercritical airfoils for transonic transports; laminar-flow sections for sailplanes and light aircraft.
• High-lift devices: plain, split, slotted and Fowler flaps; leading-edge slats and slots; chosen to meet the stall-speed/field-length requirement without excess weight.
• Drag build-up: CD = CD0 + K·CL², with CD0 from component build-up (skin friction × form factor × interference, plus miscellaneous and leakage) and K = 1/(π·AR·e), e ≈ 0.75–0.85.
• Area ruling (smooth cross-sectional area distribution) for supersonic designs; wave drag and wing–body blending are decided here.
• Tools: empirical methods (DATCOM, Roskam, Torenbeek, Raymer), panel methods (VLM, XFLR5), CFD (RANS solvers, ANSYS Fluent, SU2, in-house NAL/ADA codes), and wind-tunnel testing at NAL's low-speed and trisonic tunnels, IISc and IIT tunnels.

Structural design, loads, weights and balance

• Load factor n = L/W. The V-n diagram (flight envelope) bounds manoeuvre and gust load factors against airspeed, with corners at VS (stall), VA (manoeuvre speed), VC (design cruise), VD (design dive). Limits are set by the airworthiness code: transport category (CAR-25/FAR-25) typically +2.5 to −1.0; normal category light aircraft +3.8/−1.52; aerobatic +6/−3; fighters +9/−3.
• Limit load = highest load expected in service; ultimate load = limit × factor of safety (1.5 in aircraft practice); structure must not yield at limit load and must not fail for 3 seconds at ultimate load.
• Structural layout: semi-monocoque fuselage (frames, longerons, stringers, skin); wing box with front and rear spars, ribs and stressed skin carrying bending, shear and torsion; fail-safe and damage-tolerant design philosophy replaced pure safe-life for most structures.
• Materials: Al-2024 (tension, fatigue-critical skins), Al-7075 (compression, spars), Al-Li alloys, titanium (high temperature, landing-gear fittings), steel (landing gear, fasteners), CFRP composites (Tejas wing and fin, HTT-40 and Saras composite parts), honeycomb sandwich panels.
• Weight estimation methods: Class I (statistical fractions, conceptual), Class II (component equations — Torenbeek, Roskam, GD/Boeing methods), Class III/IV (detailed part-by-part from CAD models in detail design).
• Centre of gravity and balance: a weight-and-balance table of each item's weight and arm gives the CG; the CG travel with fuel burn and loading must stay within the forward and aft limits set by stability (aft limit) and control/nose-wheel lift-off and trim drag (forward limit). The wing is positioned on the fuselage to place the CG at the required fraction of the mean aerodynamic chord (MAC), typically 15–30% MAC for transports.
• Aeroelasticity must be cleared in preliminary design: flutter, divergence, control reversal, and the structural dynamic model validated by ground vibration testing.

Stability, control and systems integration

• Static longitudinal stability requires dCm/dCL < 0, i.e. the CG ahead of the neutral point; the margin (xnp − xcg)/c̄ is the static margin, typically 5–15% MAC for conventional aircraft, slightly negative (relaxed static stability) for fly-by-wire fighters like the Tejas, which the flight control computer stabilises.
• Horizontal tail sizing from the tail volume coefficient and the required static margin and trim capability; vertical tail from directional stability (Cnβ > 0), engine-out control (VMC), crosswind landing and spin recovery.
• Control surfaces sized for roll rate, trim at all CG and speed combinations, and one-engine-inoperative control; tab, horn and aerodynamic balance to limit stick forces.
• Landing gear design: track and wheelbase for tip-over and ground stability (typically a 15° tail-down/tipback angle check and an overturn angle held below about 63°), tyre and strut sizing from static and dynamic loads, retraction kinematics and stowage volume, pavement loading (ACN/PCN) for airport compatibility.
• Systems integration in preliminary and detail design: propulsion installation and inlet design, fuel system and tank volume, flight controls, electrical and avionics architecture, environmental control, ice protection, emergency equipment and cabin layout against evacuation rules.
• Design for manufacture, assembly and maintenance (DFM/DFA/DFMS): part count, access panels, MSG-3 maintenance task analysis, interchangeability, tooling and jigs decided before drawing release; concurrent engineering teams (IPTs) replace sequential hand-overs.
• Modern practice: MDO (multidisciplinary design optimisation), digital mock-up and digital twin, model-based systems engineering (SysML), CATIA/NX product data management, and reliability/safety analyses — FMEA, fault tree, and system safety assessment per ARP4754A/ARP4761.

Prototype, testing, certification and Indian practice

• Sequence after detail design: release to shop → tooling → prototype manufacture → ground tests (static test specimen to ultimate load, fatigue test specimen to several design lives, iron-bird systems rig, ground vibration test, EMI/EMC) → taxi trials → first flight → flight-test programme expanding the envelope → type certification → production and in-service support.
• Indian regulator: the Directorate General of Civil Aviation (DGCA) under the Bharatiya Vayuyan Adhiniyam, 2024 (which replaced the Aircraft Act, 1934) and the Aircraft Rules, with Civil Aviation Requirements (CARs). Design and production organisation approval and type certification are handled under CAR-21; airworthiness standards are adopted largely from FAR/CS — CAR-23 equivalents for normal-category aeroplanes, CAR-25 for transport category, CAR-27/29 for rotorcraft.
• Military aircraft are certified by CEMILAC (Centre for Military Airworthiness and Certification) with DGAQA for quality assurance, against JSS/IMTAR and relevant MIL/DEF-STAN specifications; the Tejas received Initial and Final Operational Clearance by this route.
• Drones are certified under the Drone Rules, 2021 (as amended) with DGCA-notified Certification Scheme for Unmanned Aircraft Systems (CSUAS), type certificates issued to manufacturers, and mandatory unique identification; details of the scheme and amendments change frequently, so the current CAR/notification should be checked.
• Indian design programmes worth quoting: HAL HF-24 Marut, HJT-16 Kiran, Dhruv ALH and LCH Prachand helicopters, HTT-40 basic trainer; ADA/HAL Tejas Mk1/Mk1A and the Mk2 and AMCA under development; CSIR-NAL Hansa-3 and Hansa-NG trainers and the Saras programme; the Regional Transport Aircraft study; ISRO's RLV-TD; HAL–Safran and GE engine partnerships; private design work at Tata Advanced Systems and in the eVTOL/UAV startup sector (ideaForge, NewSpace, ePlane).
• Cost and programme view: non-recurring (design, tooling, test) versus recurring (material, labour, learning curve) costs; life-cycle cost and direct operating cost are the figures of merit; roughly 70–80% of life-cycle cost is locked in during conceptual design, which is why the early phase is given disproportionate attention.

Key points

• Design spiral - the iterative loop through weights, aerodynamics, structures, propulsion and performance that converges on a consistent aircraft.
• Three phases - conceptual (what it looks like), preliminary (configuration frozen, detailed analysis, commit decision), detail (every part drawn and released).
• Requirements - payload, range, speed, field length, climb, ceiling, load factor, cost and life, stated quantitatively before any sketch.
• Empty weight fraction - We/W0 from statistics; about 0.6 for jet transports, around 0.5 for fighters.
• Mission segment method - fuel fraction built from segment weight ratios, with Breguet for cruise and loiter.
• Iterative W0 solution - because the empty weight fraction itself depends on the gross weight being solved for.
• Wing loading W/S - sets wing size, stall speed, field length, ride quality and cruise efficiency.
• Thrust-to-weight T/W - sets engine size, climb, acceleration and turn performance.
• Constraint (matching) diagram - all performance requirements as T/W versus W/S curves; design point in the feasible region at minimum engine and wing size.
• Aspect ratio trade - high AR lowers induced drag but raises structural weight and flexibility.
• Sweep - raises critical Mach number and delays drag rise; penalises low-speed lift and promotes tip stall.
• V-n diagram - the manoeuvre and gust envelope fixing limit load factors from the airworthiness code.
• Factor of safety 1.5 - ultimate load = 1.5 × limit load; no yielding at limit load.
• Static margin - distance of CG ahead of the neutral point as a fraction of MAC; negative only with a stability-augmenting flight control system.
• CG travel - must remain between forward and aft limits through all loading and fuel states; the wing is located to achieve it.
• Class I / II / III weights - statistical, component-equation, and detailed part-by-part estimates matched to the design phase.
• Certification - DGCA CAR-21 and CAR-23/25/27 for civil aircraft in India; CEMILAC and DGAQA for military; Drone Rules 2021 and CSUAS for UAS.
• Early cost lock-in - most of the life-cycle cost is committed in conceptual design, when very little has yet been drawn.

Formulas and examples

Example 1 — Gross weight sizing of a 19-seat turboprop commuter. Requirement: 19 passengers + 2 crew, payload 19 × 100 kg = 1,900 kg (passenger + baggage), crew 2 × 90 = 180 kg. Assume We/W0 = 0.60 and estimated Wf/W0 = 0.18.

• W0 = (1,900 + 180)/(1 − 0.18 − 0.60) = 2,080/0.22 = 9,455 kg.
• We = 0.60 × 9,455 = 5,673 kg; Wf = 0.18 × 9,455 = 1,702 kg.
• Check: 5,673 + 1,702 + 1,900 + 180 = 9,455 kg ✔
• Sensitivity: if the structure comes out heavier, We/W0 = 0.63, then W0 = 2,080/0.19 = 10,947 kg — a 3-point rise in empty weight fraction inflates gross weight by 16%. This amplification is why weight control is policed from day one.

Example 2 — Mission fuel fraction and cruise range check. Segments: warm-up/take-off 0.980, climb 0.985, cruise (to be found), descent 0.990, landing 0.995. Cruise: V = 140 m/s, L/D = 14, TSFC c = 0.000113 1/s (≈0.4 lb/lbf·h), required cruise range 1,200 km.

• Breguet: ln(Wi−1/Wi) = R·c/(V·(L/D)) = (1.2 × 10⁶ × 1.13 × 10⁻⁴)/(140 × 14) = 135.6/1,960 = 0.06918.
• Cruise fraction Wi/Wi−1 = e^−0.06918 = 0.93315.
• Wn/W0 = 0.980 × 0.985 × 0.93315 × 0.990 × 0.995 = 0.88730.
• Wf/W0 = 1.06 × (1 − 0.88730) = 1.06 × 0.11270 = 0.1195, i.e. about 12% — still short of the 18% assumed in Example 1, so the real design mission must carry a longer cruise or a loiter/diversion reserve. The lesson for the design spiral: feed 0.119 back only if the segment list truly represents the design mission.

Example 3 — Wing sizing and stall speed from wing loading. A light trainer has W0 = 1,200 kg (11,772 N), chosen W/S = 850 N/m², AR = 7.2, CLmax (flaps down) = 1.9.

• S = 11,772/850 = 13.85 m².
• b = √(7.2 × 13.85) = √99.7 = 9.99 m; c̄ = 13.85/9.99 = 1.39 m.
• Vs at sea level (ρ = 1.225): Vs = √(2 × 11,772/(1.225 × 13.85 × 1.9)) = √(23,544/32.24) = √730.2 = 27.0 m/s = 97.3 km/h ≈ 52.6 kt — comfortably inside the 61 kt (113 km/h) stall-speed ceiling that normal-category single-engine rules impose.

Example 4 — Drag polar and best L/D. CD0 = 0.022, AR = 7.2, e = 0.80.

• π·AR·e = 3.1416 × 7.2 × 0.80 = 18.09; K = 1/18.09 = 0.05527.
• CL at (L/D)max = √(CD0/K) = √(0.022/0.05527) = √0.3980 = 0.631.
• CD = 0.022 + 0.05527 × 0.631² = 0.022 + 0.0220 = 0.0440 → (L/D)max = 14.3.
• Check by formula: ½·√(18.09/0.022) = ½ × √822.3 = ½ × 28.68 = 14.34 ✔

Example 5 — V-n corner point and ultimate load. Normal-category aeroplane, W = 11,772 N, S = 13.85 m², CLmax (clean) = 1.45, n_limit = +3.8.

• Vs(clean) = √(2 × 11,772/(1.225 × 13.85 × 1.45)) = √(23,544/24.61) = √956.7 = 30.9 m/s.
• Manoeuvre speed VA = Vs·√n = 30.9 × √3.8 = 30.9 × 1.949 = 60.2 m/s (≈117 kt).
• Limit load on the structure = 3.8 × 11,772 = 44,734 N; ultimate = 1.5 × 44,734 = 67,100 N — the static test article must hold this without failure.

Where it is used

• Military aircraft design: ADA (Tejas Mk1A, Mk2, AMCA), HAL Aircraft Research and Design Centre and Rotary Wing R&D Centre (HTT-40, LCH, LUH), DRDO's ADE (Rustom/TAPAS, Archer UAVs) — roles in configuration design, loads, weights, stress and flight sciences.
• Civil and regional aircraft: CSIR-NAL (Hansa-NG, Saras Mk2, RTA studies), with DGCA CAR-21 design organisation approval processes and type certification campaigns.
• Global design centres in India: Airbus, Boeing, Collins, Safran, GE Aerospace, Honeywell and Rolls-Royce engineering offices in Bengaluru and Hyderabad doing stress, loads, aero, systems and certification work on live programmes.
• Tier-1 and private industry: Tata Advanced Systems, L&T, Mahindra Aerospace, Dynamatic, Bharat Forge — design-to-build of structures, assemblies and tooling to customer specifications.
• Drones and eVTOL: ideaForge, Garuda, NewSpace Research, ePlane, Asteria designing to the Drone Rules, 2021 and the DGCA UAS certification scheme — the fastest-growing Indian entry point for fresh design engineers.
• Academic and project work: AICTE-curriculum Aircraft Design courses and capstone design projects, SAE Aero Design and student UAV competitions, and GATE Aerospace sizing and performance questions.

Interview questions

• Q: Why is the aircraft design process iterative rather than sequential? A: Because weights, aerodynamics, propulsion, structures and performance are mutually dependent — the gross weight you need depends on the empty weight and fuel, which depend on the size and shape, which depend on the gross weight; so the loop is traversed until it converges.
• Q: What exactly is frozen at the end of preliminary design? A: The external configuration and major interfaces — planform, fuselage lines, engine choice and placement, structural layout and system architecture — so that detail design and tooling can proceed without rework.
• Q: How do you choose the design point on a constraint diagram? A: Plot every performance requirement as a T/W versus W/S boundary, take the feasible region satisfying all of them, and pick the point with the lowest T/W and the highest W/S inside it — the smallest engine and smallest wing that still meets every requirement.
• Q: What happens to the design if the empty weight estimate is 5% optimistic? A: Gross weight grows by far more than 5% because of the weight-growth amplification in the sizing equation — more structure needs more fuel and thrust, which needs more structure. Weight control is therefore a programme-level discipline.
• Q: Why is a high aspect ratio not always better? A: It cuts induced drag but increases wing bending moment, structural weight, flexibility and flutter risk, and can exceed airport gate span limits — so there is an optimum for the mission.
• Q: What does the V-n diagram tell a designer? A: The combinations of airspeed and load factor the aircraft must survive — stall boundary, manoeuvre load limits from the airworthiness code, gust lines, and VA, VC and VD — which set the design loads for the structure.
• Q: Why is the factor of safety in aircraft structures only 1.5? A: Because loads and material properties are known with unusual precision, structures are tested to ultimate load, and weight is critical; the margin is backed by statistical loads data, fatigue and damage-tolerance analysis and inspection programmes rather than by brute material.
• Q: How is the wing's longitudinal position on the fuselage decided? A: By weight-and-balance: the wing is placed so that the CG with all loading and fuel states stays between the forward and aft limits, giving the required static margin aft and adequate control and nose-wheel lift-off capability forward.
• Q: Can an aircraft be designed with negative static margin? A: Yes — relaxed static stability designs like the Tejas are aerodynamically unstable and rely on a quadruplex fly-by-wire flight control computer for artificial stability, which cuts trim drag and tail size but demands extremely reliable, redundant control software and hardware.
• Q: Which Indian authorities certify a new civil aircraft and a new military aircraft? A: DGCA, under CAR-21 and the applicable CAR-23/25/27/29 airworthiness standards, for civil aircraft; CEMILAC with DGAQA for military aircraft, leading to Initial and Final Operational Clearance.
• Q: Why is most of the life-cycle cost committed in conceptual design? A: Because configuration, technology level, engine choice and maintainability — which govern manufacturing, fuel burn and maintenance cost for decades — are all chosen there, while the cost of changing them later is enormous.
• Q: What is the difference between Class I and Class II weight estimation? A: Class I uses statistical fractions of gross weight for the whole aircraft, suitable for conceptual sizing; Class II uses empirical equations for each component (wing, fuselage, gear, systems) based on geometry and loads, used after a configuration exists.

Common mistakes

• Treating conceptual sizing as a one-shot calculation and not iterating W0 until the empty weight fraction and gross weight are consistent.
• Mixing mass and force: using kg in the stall-speed or drag equations instead of newtons, or forgetting the 9.81 factor.
• Using the maximum CLmax with flaps for the clean-configuration V-n corner, or the clean value for landing field length.
• Confusing limit and ultimate load, or applying the 1.5 factor twice.
• Choosing a design point outside the feasible region of the matching chart, or picking a low W/S "for safety" and ending up with an oversized, draggy, heavy wing.
• Forgetting CG travel: checking balance only at take-off weight and not at zero-fuel, maximum-payload and ferry conditions.
• Quoting design requirements qualitatively ("long range, good performance") instead of as numbers that can be designed and tested against.

Related topics

Aircraft systems: hydraulics, fuel and landing gear · Aircraft performance and flight mechanics · Aircraft structures and aeroelasticity · Aerodynamics and compressible flow · Aircraft stability and control · Composite materials in aerospace