Engineering study

One seat fitting. Every way to make it.

The pivot fitting that carries an aircraft seat's backrest, designed from one set of requirements for six manufacturing processes and four alloy families at once, then checked for strength, stiffness, ductility, mass and cost.

mass: 725 g today, 279 g recommended
−62 %
configurations considered, 19 designed and verified
24
meet every requirement
5
per part at 2,000 a year, recommended design
$59

A study we ran on a representative part, not a client project: the interfaces and loads follow a typical certified economy seat; the cost figures come from simple, stated models.

01 · Requirements

What the part has to do.

Two M8 bolts to the seat's side plate, a Ø12 backrest pivot, a Ø8 lug for the recline lock, and a keep-out zone for the tray-table mechanism. Inside a 120 × 180 mm envelope, everything else is free.

Load caseForce on the pivotChecked against
Backrest recline12 kN, cyclicfatigue, safety factor 1.25
16 g forward crash28 kNyield, safety factor 1.5
Downward push9 kNyield, safety factor 1.5
  • Stiffnesspivot deflection ≤ 1 mm
  • Ductilityelongation at break ≥ 5 % (crash)
  • Mass≤ 400 g (today: 725 g)
  • Cost≤ $250 per part at 2,000 a year
02 · Method

Twenty-four designs, not one.

  1. 01
    Rule outCombinations with no production route are dropped before any computing: magnesium is not printed, investment cast or bent as sheet in production; titanium and steel are not die cast.
  2. 02
    DesignFor every remaining process and alloy, topology optimisation finds where material carries the three load cases, with the volume and the weight of each case tuned to that alloy's strength.
  3. 03
    Make it makeableEach result is reshaped by its process: tool radius for milling, draft and fillets for casting and forging, constant thickness and bends for sheet, the thinnest printable members for powder-bed printing.
  4. 04
    VerifyThe final shapes are checked again with finite elements: fatigue and yield with safety factors, pivot deflection, ductility, then mass and cost.
  5. 05
    RankQualified designs are ranked by price plus the value of mass over the aircraft's life ($2,700 per kg).
03 · Results

Process × material.

Material ↓ / Process →CNC millingMetal 3D printing (LPBF)Investment castingLaser-cut and bent sheetClosed-die forgingHigh-pressure die casting
Aluminum
Al 7075-T7351279 g · $59Recommended
Scalmalloy242 g · $258Fails: cost
A357-T6370 g · $51Qualified
Al 2024-T3 sheet261 g · $21Fails: strength
Al 7075-T73 forging326 g · $52Qualified
Al A380 die cast412 g · $27Fails: strength, ductility, mass
Titanium
Ti-6Al-4V251 g · $307Fails: cost
Ti-6Al-4V (HIP)259 g · $340Fails: cost
Ti-6Al-4V (cast, HIP)266 g · $174Qualified
Ti grade 4 sheet420 g · $44Fails: strength, mass
Ti-6Al-4V forging379 g · $145Qualified
No production route
Steel
AISI 4130 Q&T525 g · $116Fails: mass
17-4PH H900395 g · $272Fails: cost
17-4PH (cast, H1025)458 g · $65Fails: mass
301 stainless ½ hard492 g · $25Fails: mass
AISI 4340 forging656 g · $57Fails: mass
No production route
Magnesium
Mg Elektron 43476 g · $66Fails: mass
No production routeNo production routeNo production route
Mg Elektron 43 forging454 g · $70Fails: mass
Mg AM60B die cast275 g · $28Fails: strength

Mass of the finished part and price per part at 2,000 a year. Most of the fourteen failures are not strength but mass: steel and magnesium designs that hold the loads end up over the 400 g limit. Milled titanium and every printed design pass the engineering checks and fail on price; bent sheet and die casting fail on strength.

04 · Recommendation

Milled aluminum.

Al 7075-T7351, machined from plate: 279 g instead of 725 g, $59 a part, the lowest price plus lifetime value of the five that pass.

#ConfigurationMassPricePrice + mass value
1CNC milling, Al 7075-T7351279 g$59$812
2Investment casting, Ti-6Al-4V (cast, HIP)266 g$174$891
3Closed-die forging, Al 7075-T73 forging326 g$52$932
4Investment casting, A357-T6370 g$51$1,051
5Closed-die forging, Ti-6Al-4V forging379 g$145$1,168
05 · What it is and isn't

A first answer in hours.

The whole design space, 24 configurations, ran in about two and a half hours of compute. It is a planar model with stated cost models: the right tool to decide which two or three directions deserve full 3D analysis, supplier quotes and a test article, not a replacement for them.

We are now running it in pilot projects with engineering teams, on their own parts, with their loads, standards and suppliers. If one of yours costs you weeks, we would like to run it.

Bring us a part