A regional airliner OEM we mentored in 2024 had a titanium bracket with a 47-day lead time from PO to First Article delivery: 14 days for forging, 6 days queue at a Nadcap-certified machine shop, 11 days for 5-axis milling, 4 days for Type II anodizing, 5 days for FPI and CMM inspection, 2 days for FAI documentation compilation, and 5 days buffer. Later, a design change removed two weight-saving pockets, and the engineering change notice took 28 days to pass through AS9100 configuration management alone—longer than half of the manufacturing processes. The part itself wasn't the issue; it was the audit trail surrounding it. This is a microcosm of aerospace manufacturing. The metal cutting, composite layup, and printing done in the shop are not fundamentally different from other industries; what truly differentiates it is the surrounding layer—AS9100 Rev D, FAA 14 CFR Part 21, EASA Part 21, Nadcap qualifications for each process, complete material traceability from melt lot to installed part, and a change control system where adjusting a 0.2 mm fillet can re-trigger First Article Inspection. The truly important design actions: every decision you make in the first week either simplifies or inflates the paperwork you'll deliver in the 18th month.
Which Stack of Regulations Are You Actually Designing To?
Before selecting a process or material, you need a certification basis. Commercial transport aircraft fly under 14 CFR Part 25 (FAA) or CS-25 (EASA); the supply chain operates under AS9100 quality management, AS9102 First Article Inspection, and AS9145 APQP. Each special process—heat treatment, welding, NDT, chemical processing, additive manufacturing—is separately certified by Nadcap. A supplier can have AS9100 but still be unable to bid on your chemical milled bracket simply because they lack the Nadcap AC7108 accreditation. Confirm bidding eligibility first, then confirm the part.
| Standard | Scope | Who performs | What breaks without it |
|---|---|---|---|
| AS9100 Rev D | Aerospace QMS baseline | IAQG / Registrar | Cannot bid for most primes |
| AS9102 Rev C | First Article Inspection | Prime Contractor | FAI rejection, rework |
| AS9145 | Aerospace APQP & PPAP | Prime Contractor | Phase gates stuck |
| AS9146 | FOD Control Plan | Prime Contractor | Shipment freeze |
| Nadcap AC7004 | Special Process QMS Foundation | PRI | Loss of process qualification |
| Nadcap AC7108 | Chemical Processing | PRI | Anodize/Chem-mill disqualification |
| Nadcap AC7110 | Heat Treating | PRI | Certification coupons invalid |
| 14 CFR Part 21 / EASA Part 21 | Certification Procedures | FAA / EASA | Cannot obtain airworthiness |

The Prototyping Staircase from Concept to Flight-Representative Part
Aerospace prototyping is a staircase, not a sprint. Each step answers different questions and builds different levels of trust with certification authorities. Skipping one step to save four weeks usually costs an extra sixteen weeks later when the flight-representative part fails a test—a failure that could have been exposed in the first step with a USD 380 concept print. Plan this staircase, budget for it, and refuse to let marketing use an SLA concept model for flight part photography.
| Step | Goal | Common Processes | Typical Unit Cost | Certification Weight |
|---|---|---|---|---|
| Concept | Appearance, fit, interference | SLA / FDM | USD 80–400 | None |
| Functional | Mechanical behavior under load | SLS / MJF Nylon, CF-FDM | USD 200–900 | Supports DFMEA |
| Flight-Representative | Final material, final process | 5-axis Ti / DMLS / Layup | USD 2k–40k | Provides certification test data |
| Production-Intent | Final tooling, final factory, FAI | Production line, small batch | Same as production cost | AS9102 required |
Production Process Selection in a Low-Volume Reality
An aerospace program delivering 480 airframes annually is not automotive. A 5-axis machined monolithic titanium part will win on a geometrically complex project with a total demand of 2,400 units over five years. DMLS is competitive when parts are small, geometrically complex, and already fall within approved material-process combinations (e.g., Ti-6Al-4V to AMS 7010). Layup with autoclave still owns large composite primary structures. Investment casting with HIP still owns turbine hot sections.
| Process | Sweet Spot Volume | Typical Material Spec | Lead Time | Notes |
|---|---|---|---|---|
| 5-Axis Machining | 50–5,000 / year | AMS 4928 Ti, AMS 4027 Al | 6–12 weeks | Monolithic structures, brackets |
| DMLS (LPBF) | 10–1,500 / year | AMS 7010 Ti, AMS 5662 Inconel | 4–10 weeks | Fuel nozzles, heat exchangers |
| Investment Casting + HIP | 200–10,000 / year | AMS 5383 CoCr, AMS 5397 | 14–22 weeks | Turbine hot sections |
| Manual / Automated Layup + Autoclave | 20–2,000 / year | Hexcel 8552 IM7, Cycom 5320 | 8–16 weeks | Primary structures, skins |
| Sheet Metal + Chemical Milling | 500–20,000 / year | AMS-QQ-A-250/12 | 5–9 weeks | Fuselage skins, ribs |
| Superplastic Forming + Diffusion Bonding | 100–2,000 / year | AMS 4911 Ti | 9–14 weeks | Engine ducts, complex casings |

Three Aerospace Parts and the Decisions that Shaped Them
Fuel Nozzle Converged from 20 Welded Pieces to 1 Printed Part
A turbofan combustor fuel nozzle, originally made of 20 machined, brazed, and TIG-welded parts, was redesigned as a single DMLS Inconel 718 (AMS 5662) piece. Mass decreased from 0.94 kg to 0.67 kg, and field durability at the inner nozzle tip increased by 5x—because the original weld points, which were failure sources, were eliminated. The number of parts per engine dropped from 20 to 1. The project absorbed USD 2.3 million in qualification work—test coupons, witness parts every 20 units, and CT scans for every outgoing part—before flight approval. The truly important design action: on day one, the team locked the external flange geometry and fuel inlet seat to the existing engine interface, so the certification basis only needed incremental changes for the interior. If they had altered the flange, the entire fuel system ICD would have had to be re-issued, and qualification costs would have tripled.
Lightning Strike Tested Composite Control Surface
An aileron, redesigned from aluminum skin and ribs to IM7/8552 carbon fiber with an added copper mesh for lightning protection, saved 4.7 kg per surface. It passed tests in SAE ARP5412 Zone 2A with a 100 kA waveform. The project spent USD 620,000 on layup tooling and 22 weeks on qualification. Calculated over 240 shipsets annually and a 12-year production cycle, the weight-corresponding endurance CO2 offset alone covered the tooling cost by the third year.
Titanium Bracket Remaining CNC due to Calculated Support
A landing gear attach bracket, with 1,400 units per year, was examined for conversion to DMLS. The topology-optimized DMLS version saved 0.6 kg and cost USD 1,950 per part; the 5-axis version cost USD 1,180 per part and met the same strength requirements. The project chose to maintain CNC because saving 0.6 kg per landing gear, at 1,400 units annually, was not worth an additional USD 1.08 million in annual part costs plus USD 900,000 in qualification fees. Not every aerospace bracket should become a printed part—calculate honestly.
Material Traceability Is the Real Deliverable
Customers don't buy parts; they buy the certified paper trail behind the parts. Every melt lot, every powder batch, every layup batch must be traceable upstream to the mill test report and downstream by serial number to the installed airframe. An airworthiness directive issued three years from now will ask: which airframes contain material from melt lot 7749L, and the answer must be found within one hour. Before the first chip flies, plan your data architecture.
Aerospace Project Mistakes that Burn Years
| Mistake | Why it failed | How to avoid |
|---|---|---|
| Hiring a non-Nadcap supplier to save 12% | Prime contractor audit directly rejects PO | Verify Nadcap accreditation for each process before RFQ |
| Freezing geometry before certification basis is defined | FAA demands a delta explanation you can't provide | Lock certification basis and ICD at PDR, not CDR |
| Skipping the flight-representative prototype step | DMLS test coupons cannot predict full-scale fatigue | Budget for 1–3 full-scale test articles early |
| Treating DMLS configuration management like CNC | Every parameter tweak is a qualification event | Freeze print parameters according to AMS material spec |
| No FOD plan in the workshop | Assembly rejected during source inspection | Implement AS9146 from day one and audit weekly |
| Using commercial-grade fasteners for flight parts | Loss of batch traceability leads to direct scrap | Procure to NAS/MS/AN specs with full certs |
Habits to Keep Aerospace Projects on Track
| Do | Avoid |
|---|---|
| Lock certification basis before detailed design | Let design drift to redraw cert delta |
| Qualify one material-process pair per part | Change heat treatment spec mid-production |
| Audit Nadcap scope for each special process supplier | Assume AS9100 covers Nadcap processes |
| Allocate 15–25% of NRE for FAI & qualification coupons | Treat FAI as an afterthought |
| Save parameter snapshots for each DMLS build | Update machine firmware without re-qualification |
| Reserve 4–8 weeks for paperwork after final delivery | Assume project closure after final shipment |

Pre-Certification Readiness Checklist
- Certification basis (14 CFR / CS sections and revisions) specified and cited on every drawing
- AS9100 Rev D supplier survey completed, including evidence of flow-down to sub-suppliers
- Nadcap accreditation letters for each special process archived
- AS9102 FAI plan written before any cutting of ship parts
- Material traceability architecture recording melt lot, heat treat lot, and serial number for each part
- AS9146 FOD prevention plan deployed in every unit handling parts
- Build parameters for each DMLS material-process pair baselined and signed off
- Full-scale test article schedule aligned with flight test gates, not optimism
Design Takeaways
Aerospace manufacturing is less about finding clever processes and more about respecting the surrounding certification wrapper. Choose material-process pairs already on AMS specifications, lock your certification basis at PDR, fully plan the prototyping staircase instead of directly rushing to flight-representative parts, and treat material traceability as a deliverable. Projects that land on time are those that account for paperwork and metal together, not as an afterthought.
0 comments