AM qualification · Evidence checklist · As of 2026-09-12
Short answer: “Qualified enough” means the defined feedstock → print → post → inspect → qualify → economics route can repeatedly produce parts that meet the acceptance and safety case for this criticality—not that a geometry printed once. Aerospace additive manufacturing is not qualified by enthusiasm. Evidence depth scales with consequence of failure, inspectability limits, and the authority that owns release.
Criticality and what “enough” means
| Criticality (illustrative) | What “enough” usually means | What is not enough |
|---|---|---|
| Non-critical / tooling / fixture | Controlled process, basic material ID, dimensional acceptance, change log | No genealogy; uncontrolled parameter drift |
| Secondary structure / moderate consequence | Material + machine + process lock, post-process records, NDT plan, FPY visibility | Coupon success without part-family link |
| Primary / fatigue- or fracture-sensitive | Allowables path (program-owned), POD-backed inspection, locked HT/HIP if used, serial conformity | Datasheet tensile numbers as design allowables |
| Pressure / rotating / crewed-spaceflight class | Full nested stack + authority approval; delta rules predefined | Demo, patent, or single hot-fire as serial proof |
Evidence boundary: Criticality labels are program- and authority-specific. This checklist does not certify parts, set design allowables, or replace customer/regulatory requirements.
—
How to use this checklist
Scope
Use this page when a team has a successful print (or a promising development campaign) and must assemble the evidence path to a released part. It is written for qualification, quality, manufacturing engineering, and supplier-management roles in industrial metal AM—especially laser powder bed fusion (LPBF)—but the gate logic applies across process categories when the route is redefined.
Authority
Addithive is research mapping, not a certification body. Controlling documents remain: the customer specification, the applicable quality system (e.g. AS9100/EN 9100 context), and the regulatory or program authority. A published standard is a framework; it is not automatic part acceptance.
Route definition (mandatory before checking boxes)
Before Gate 0, write the route in one block:
- Alloy / feedstock condition and supplier controls
- Machine model, serial/configuration, site
- Parameter set, software/firmware, build layout and orientation
- Stress relief / heat treatment / HIP (if used) / machining / finish
- Inspection methods and acceptance criteria
- Part or part-family identity and criticality class
A demo, patent or machine installation is not the same as qualified recurring output. Printer capacity is useful only when the downstream system can convert builds into accepted, released parts.
Evidence labels used below
| Label | Meaning |
|---|---|
| Fact | Supported by current standard, primary authority, or verified source cited on Addithive |
| Technical interpretation | Engineering conclusion from multiple facts; conditions stated |
| Addithive inference | Synthesis for bottleneck / readiness analysis; not a published scientific fact |
| Commercial signal | Company-reported capacity, qualification announcements, filings—not technical proof |
Standing refusals (this page)
- No investment advice
- No engineering certification or design allowables from this checklist
- No treating demos/patents as production proof
- No universal powder-reuse cycle limits
- No “HIP always”
- No equating CT voxel size with minimum detectable defect
- No AM-cheaper-by-default economics
One-page printable summary — Seven gates
Numbering crosswalk: This checklist uses Gates 0–7 (Gate 0 = application & conventional baseline; Gates 1–7 = feedstock through production control). Addithive’s production-ready page uses Gates 1–7, where Gate 6 emphasizes yield/capacity and Gate 7 documentation—here yield/FPY is folded into Gate 7 alongside travelers, SPC, and change/delta qualification.
| Gate | Evidence to assemble (headline) | What usually breaks first |
|---|---|---|
| 0 Application & conventional baseline | Criticality, AM rationale vs conventional, performance/cost case | Weak “why AM”; no baseline for delta cost/risk |
| 1 Feedstock | Spec, genealogy, acceptance, reuse rules (route-specific) | Chemistry drift; undocumented reuse; lot mix |
| 2 Machine | Qualification, calibration, maintenance, equivalency triggers | Nameplate assumed equivalent; software/optics drift |
| 3 Process lock | Parameters, software, layout, FMEA, revision control | Unlocked parameter window; build-to-build variation |
| 4 Post-processing | SR/HT/HIP/machining/finish qualified as applicable | Distortion; support scars; unqualified thermal route |
| 5 Inspection / NDT | Methods, access, acceptance criteria, POD limits | Printable but not inspectable; voxel ≠ defect size |
| 6 Part evidence | Test, analysis, conformity to intended use | Coupon ≠ part; missing load-case / fatigue link |
| 7 Production control | Travelers, SPC/FPY, change & delta qualification | Change without impact assessment; FPY opaque |
Gate 0 — Application & conventional baseline
Purpose: Prove why this part should be AM, at what criticality, and against what conventional alternative—before burning qualification budget.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Part function, interfaces, environments | Requirements owned and revision-controlled | “Print it because we can” | Fact: requirements drive acceptance |
| Criticality / consequence-of-failure class | Classification approved by design authority | Informal “non-critical” label | Authority-specific; NASA Class A/B/C logic is a mental model, not universal |
| Conventional baseline (forging, casting, machined billet, etc.) | Buy-to-fly, lead time, scrap, inspectability compared | No baseline → no economics case | Addithive inference: AM economics fail without route-specific comparison |
| AM advantage statement | Consolidation, channels, weight, sustainment—tied to measurable KPIs | Vague weight or “innovation” claims | Commercial signal ≠ technical need |
| Inspectability / powder-removal feasibility screen | Internal features have access plan | Channels printable, not cleanable or NDT-accessible | Technical interpretation: inspection often binds before printability |
| Program approval path identified | Customer / regulator / DER path known | “We’ll figure certification later” | Qualification ≠ certification |
Addithive inference: Gate 0 failures look cheap early and expensive late—especially when fatigue-critical geometry is frozen before NDT and allowables paths exist.
Gate 1 — Feedstock specification, genealogy, reuse rules
Purpose: Control material identity and history so property and defect evidence remain meaningful.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Feedstock specification (chemistry, PSD, morphology, oxygen/nitrogen where relevant) | Spec revision locked to process | “Ti-6Al-4V powder” with no revision | Alloy designation alone is rarely enough |
| Supplier qualification / CoC / lot acceptance | Incoming tests match spec; rejects dispositioned | Spot-buy powder into flight builds | Commercial signal: supplier capability ≠ lot control on your floor |
| Genealogy / digital thread | Virgin + used powder history traceable to builds | Blended lots without records | Fact: reuse is alloy-, process-, machine-, history-specific—no universal cycle limit |
| Reuse method & stop criteria | Written rules; chemistry/PSD/flow checks defined | Fixed “X cycles max” copied from a blog | Standing refusal: no universal reuse limit on this site |
| Storage, handling, humidity / contamination controls | Environment logged; containers ID’d | Open bins; unknown humidity | Boundary: polymer/metal rules differ—define process category |
| Witness / archive samples policy | Retained per plan | No retain samples for contested lots | Technical interpretation: genealogy enables MRB, not just audits |
Gate 2 — Machine qualification & equivalency triggers
Purpose: Show the equipment can repeatedly deliver the intended process—and when a second machine is not automatically qualified.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Machine acceptance / IQ–OQ evidence | Tests vs approved plan; as-found/as-left | Install photo as “qualified” | Align with aerospace LPBF machine test frameworks (see standards map) |
| Configuration baseline | Optics, lasers, recoater, gas, sensors, firmware recorded | “Same model” without config delta | Fact: nameplate ≠ equivalency |
| Calibration & maintenance | Schedule executed; out-of-tolerance dispositioned | Skipped laser power / scan-field checks | Equivalence is maintained status |
| Multi-laser / field uniformity evidence (if applicable) | Cross-field / stitch zones bounded | Edge of plate never sampled | Build-volume location matters |
| Equivalency plan for Machine B | Purpose-specific comparison build + limits predefined | Move job to sister machine without delta | Technical interpretation: same OEM/model reduces burden; does not eliminate it |
| Facility / utilities / environment | Gas, temp, power, vibration controls documented | Machine moved; no re-verify | Site is part of the route |
Decision rule (Addithive inference): Treat “same make and model” as a starting hypothesis. Call machines equivalent only when comparison is purpose-specific, measurement-backed, and under continuing change control.
Gate 3 — Process lock (parameters, software, layout, FMEA)
Purpose: Freeze the manufacturing recipe that material and machine evidence actually support.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Parameter set / process window | Revision-controlled; approved range stated | Operator “tweaks for density” | Key process variables must be named |
| Build processor / CAM / firmware versions | Exact versions on traveler | Silent auto-update | Software change can be major even if “small” |
| Orientation, supports, nesting / layout rules | Standard layouts or approved variants | Ad-hoc packing for throughput | Thermal mass and location affect properties |
| Process FMEA / risk file | Linked to controls and inspection | FMEA as shelfware | Technical interpretation: FMEA should drive witness locations |
| In-process monitoring (if used) | Alarm limits and disposition rules | Sensors collecting unused data | Monitoring ≠ qualification by itself |
| Witness coupon scheme | Locations represent critical features | Only center-plate bars | Coupon-to-part link must be argued |
| Build record completeness | Start/stop, interruptions, atmosphere, recoater events | Missing interruption logs | Genealogy of the build, not only the powder |
Gate 4 — Post-processing qualification (SR, HIP, machining, finish)
Purpose: Qualify the thermal and subtractive steps that set final properties and geometry—only where they are part of the route.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Stress relief / heat-treatment procedure | Furnace qualification, load thermocouples, recipes locked | “Standard HT” without load survey | Boundary: HT is route-specific |
| HIP (if used) | Cycle qualified; containers/loads controlled | Assumed mandatory for all metal AM | Standing refusal: no “HIP always” |
| Cut-off, support removal, peening | Damage criteria; trained process | Support scars into fatigue sites | Often the slow failure mode |
| Machining / datums | Datum strategy accounts for distortion | Print-net-shape assumed final | Dimensional stack must include post |
| Cleaning / powder removal / FOD | Cleanliness method validated | Visual “looks empty” | Critical for channels and heat exchangers |
| Surface finish specification | Ra/Rz or process-based finish tied to fatigue case | Cosmetic polish only | Fatigue-critical surfaces need explicit criteria |
| Distortion / residual-stress control | Compensation or machining stock verified | First-article scramble every lot | Technical interpretation: AM creates geometry and material state together |
Gate 5 — Inspection / NDT / acceptance criteria & POD limits
Purpose: Prove relevant defects and dimensions can be found, measured, and dispositioned—before serial volume.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Inspection plan by feature criticality | Methods matched to defect types | 100% CT by default with no rationale | Cost and throughput are constraints |
| Dimensional / metrology MSA | Capability vs tolerance | CMM without fixture control | Include as-built and post-machined states |
| NDT method qualification | Technique sheet; reference standards | Method chosen for availability only | Access geometry can veto the method |
| Defect acceptance criteria | Written size/type/location rules | “No defects” without definition | Must exist before serial production |
| POD / detectability basis | Stated for critical defect populations | Voxel size treated as min defect | Standing refusal: CT voxel ≠ minimum detectable defect |
| CT / XCT process controls | Reconstruction, thresholding, analyst competence | Uncontrolled windowing | Technical interpretation: POD is method- and feature-specific |
| Leak / pressure / functional tests (as applicable) | Tied to operating envelope | One successful demo test | Demo ≠ recurring acceptance logic |
| MRB / disposition authority | Named, trained, recorded | Shop-floor “use as is” | Nonconformance path is production control |
Hidden bottleneck (aligned to Addithive guides): Complex geometry can be printable before it is inspectable. CT resolution, NDT access, and defect acceptance criteria often become binding constraints—especially for internal channels and safety-critical features.
Gate 6 — Part evidence (test, analysis, conformity)
Purpose: Connect the manufacturing route to the part’s intended performance and safety case.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Design analysis for AM condition | Orientation, anisotropy, surface, residual stress addressed | Analysis assuming wrought isotropic allowables without justification | Program owns allowables—this page does not issue them |
| Material / coupon data linked to route | Chemistry, microstructure, static/fatigue as required | Generic OEM datasheet as allowables | Standing refusal: datasheet ≠ design allowables unless route + statistics support use |
| Feature- or part-level tests | Load, fatigue, fracture, leak, thermal as required by criticality | Coupon tensile only for fatigue part | Scale evidence to consequence |
| Pre-production / first-article package | Conformity to drawing + PPP / traveler | Pretty photos of first build | Configuration must match serial intent |
| Traceability to serial numbers / lots | Digital thread from powder to release | Orphan parts | Required for MRB and field issues |
| Flight heritage or prior art (if claimed) | Heritage bounded to same route/family | Heritage from different alloy/machine/HT | Technical interpretation: heritage does not auto-transfer |
| Acceptance rationale document | Why this evidence set releases this part | Scattered emails | Authority must be identifiable |
Fact / boundary: Qualification establishes that the manufacturing route can meet defined requirements. Certification or customer/authority approval uses that evidence inside a program-specific framework. Nested layers (material, machine, process, part, supplier/site) are not interchangeable.
Gate 7 — Production control & change / delta qualification
Purpose: Keep making the same part—and define what happens when something changes.
| Evidence to assemble | Pass signal | Red flag / incomplete | Notes / boundaries |
|---|---|---|---|
| Quality system & AM procedures | Procedures live; audits passed | AS9100 certificate without AM work instructions | Certificate ≠ route control |
| Traveler / digital thread | Every gate recorded per serial | Paper gaps at post or NDT | End-to-end, not print-only |
| FPY, scrap, rework metrics | Stable FPY; causes coded | Unknown true yield | Printer capacity without FPY is not throughput |
| SPC / recurring qualification builds | Trends reviewed; stops defined | Capability study from one campaign | Maintenance of state |
| Training & personnel qualification | Roles for build, HT, NDT, release | Key-person dependency | Skills are a bottleneck (Brief Atlas) |
| Change-control procedure | Four-question impact assessment used | “Minor” defined by effort, not effect | See response levels below |
| Delta / full requalification triggers | Pre-approved matrix by change type | Case-by-case invention under schedule pressure | Addithive inference: predefined beats improvised |
| Supplier control (if outsourced steps) | Flow-down of route; incoming verification | Blind trust of bureau CoC | Who owns evidence vs who sells capacity |
Change response levels (aligned to Addithive requalification guide)
| Response | Appropriate when | Evidence may include |
|---|---|---|
| Documented review | Administrative or already bounded | Config comparison, rationale, approvals |
| Verification | Narrow, plausible product risk | Calibration, witness coupons, dimensional/material checks |
| Delta qualification | Only part of the chain affected | Targeted builds, NDT, mechanical tests, capability comparison |
| Full requalification | Prior evidence cannot represent new route | New plan, builds, material data, part evidence, approvals |
Decision rule: Never infer qualification continuity from process similarity alone. Trace the change to affected product characteristics, existing evidence, and the approval authority.
Red-flag sheet (demo ≠ serial; datasheet ≠ allowables)
| Claim you hear | Why it fails | What to demand instead |
|---|---|---|
| “We printed it successfully.” | Prototype ≠ production-ready | Seven-gate evidence; stable FPY |
| “We have a patent / white paper.” | IP ≠ process control | Locked route + serial records |
| “Machine is installed / capacity available.” | Capacity ≠ accepted parts | Downstream post + NDT + release balance |
| “OEM datasheet properties.” | Not design allowables by default | Program allowables path; route-matched data |
| “Same machine model elsewhere.” | Nameplate ≠ equivalency | Configuration + output comparison |
| “HIP’d, so porosity is solved.” | HIP is not universal or magic | Defined cycle + inspection still required if critical |
| “CT voxel is 50 µm, so we catch 50 µm defects.” | Voxel ≠ POD / min detectable | Method qualification + POD argument |
| “Powder reused 10× max—industry standard.” | No universal reuse limit | Alloy/process/history-specific rules |
| “AM is cheaper.” | Not by default | Route-specific cost vs conventional + yield |
| “Qualified for Ti-6Al-4V.” | Incomplete sentence | Qualified for which machine, parameters, site, post, inspection, part family? |
Aerospace additive manufacturing is not qualified by enthusiasm.
Mapping to ISO/ASTM 52920, 52941:2026, NASA-STD-6030
Editorial note: Standards are edition-controlled. Confirm the current controlled edition before engineering or contractual use. Addithive cites designations verified as of 2026-09-12; where publication status is in transition, that is stated.
| Document (as of 2026-09-12) | Role in this checklist | Gate touchpoints |
|---|---|---|
| ISO/ASTM 52920:2023 — Additive manufacturing — Qualification principles — Requirements for industrial additive manufacturing processes and production sites | Broad industrial framework for AM process and production-site quality assurance; complements a QMS; technology-agnostic principles | Gates 0–7 (site, feedstock management, system operations, process finalization, change assessment) |
| ISO/ASTM 52941:2026 — Additive manufacturing for aerospace — System performance and reliability tests for laser metal powder-bed fusion machines for metallic materials (2nd edition; replaces 52941:2020). Secondary catalogs list published 2026-08-11 (ISO stage 60.60); iso.org direct fetch was Cloudflare-blocked at draft time—confirm the controlled edition with ISO/ASTM before contractual citation. Scope: qualification and re-qualification tests for laser metal PBF machines for aerospace (contractual non-aerospace use allowed); also usable after maintenance. | Machine performance / reliability test methods for LPBF aerospace context | Gate 2 (and Gate 7 after maintenance/repair) |
| NASA-STD-6030 (dated 2021-04-21) — Additive Manufacturing Requirements for Spaceflight Systems | Rigorous public mental model for nested AM control (material, machine, process, part, facility); QMP / part-process concepts; crewed hardware emphasis with tailoring guidance for other missions | All gates; especially criticality, foundational process control, part production control. Related: NASA-STD-6033 for equipment/facilities control—confirm current editions. |
Fact: A published standard is not the same as a program approval; customers and regulators can add stricter requirements.
Technical interpretation: Use 52920 for production-site/process qualification architecture; use 52941-class machine tests where LPBF aerospace machine acceptance is in scope; use NASA-STD-6030 as the strictest widely cited system checklist for spaceflight-class control—even when your program is not NASA—without copying NASA class rules into non-NASA contracts without authority.
| Checklist gate | 52920 lens | 52941 lens | NASA-STD-6030 lens |
|---|---|---|---|
| 0 Application | Manufacturing assessment / feasibility | — | Part classification / consequence |
| 1 Feedstock | Feedstock management | — | Material process / QMP inputs |
| 2 Machine | System set-up / equipment | Performance & reliability tests | Equipment controls |
| 3 Process | Build cycle / parameters | — | Qualified material & part processes |
| 4 Post | Process finalization | — | Thermal / finishing within PPP |
| 5 Inspection | QA measures along route | — | NDE / acceptance in production control |
| 6 Part | Part-specific vs process-generic qual | — | Preproduction article / part evidence |
| 7 Production / change | Documented requalification assessment | Re-qual after maintenance | Change control / recurring evidence |
Company-exposure note (research mapping only)
This is research mapping, not a recommendation, buy/sell signal, or endorsement. Inclusion means a company appears in the public AM value chain relative to capacity or evidence ownership—not that AM is financially material to the company.
| Exposure type | Examples (illustrative, from Addithive aerospace guide mapping) | What matters for this checklist |
|---|---|---|
| Machine / process platforms | EOS, Nikon SLM Solutions, 3D Systems, Velo3D | Who sells machines vs who holds your machine qualification file |
| Production sites / bureaus | Sintavia, Materialise, Nikon AM Synergy, 3D Systems, AAMC | Quality approvals, production organization approvals, customer programs—who owns the evidence package |
| Powder suppliers | Carpenter Additive, Sandvik Osprey, IperionX | Lot traceability and aerospace documentation depth |
| Inspection / metrology | ZEISS, Nikon, Hexagon, Waygate / Baker Hughes | CT/NDT capability and defect-detection evidence—not equipment catalogs alone |
Addithive inference: Capacity sellers and evidence owners are often different legal entities. A service bureau may print; the design authority may still own allowables and release; an NDT lab may own POD studies. Map contracts to Gates 1–7 explicitly.
Commercial signal caution: Qualification announcements, installed-base claims, and capacity expansions are signals of activity. They do not by themselves prove recurring accepted-part output for your alloy–machine–post–inspection route.
Sources & review
Primary Addithive alignment (fetched 2026-09-12)
- What makes an additive manufacturing part production-ready?
- Aerospace Additive Manufacturing Qualification Guide
- How does aerospace additive manufacturing qualification actually work?
- Scientific Evidence & Editorial Standard
- When does an AM process change require requalification?
- What makes two LPBF machines equivalent?
- About Addithive
- AM Bottleneck Brief (product voice)
External standards / authorities (verify current controlled edition before use)
- ISO/ASTM 52920:2023 — Qualification principles for industrial AM processes and production sites
- ISO/ASTM 52941:2026 — Aerospace LPBF machine system performance and reliability tests (2nd ed.; secondary catalogs: published 2026-08-11 / stage 60.60; confirm controlled edition with ISO/ASTM before contracts)
- ISO/ASTM 52941:2020 — Prior edition (supersession path)
- NASA-STD-6030 (2021-04-21) — Additive Manufacturing Requirements for Spaceflight Systems
- NASA-STD-6033 — Equipment and facilities control (related; confirm edition)
- FAA materials on AM for aviation/TSO contexts (program-specific; see Addithive requalification & equivalency source lists)
- NIST publications on AM part qualification / data management (as cited on related Addithive pages)
Disclaimer
This article is engineering-led research for structuring evidence. It is not investment advice, not a certification, and not a substitute for design allowables, customer specifications, or regulatory approval. Controlling specifications and qualified production data remain authoritative.
Last reviewed: 2026-09-12 Approved with fixes; published 2026-09-12. Next-review trigger: New edition or status change of ISO/ASTM 52920, ISO/ASTM 52941, or NASA-STD-6030; major FAA/EASA AM guidance update; or revision of the Addithive Scientific Evidence & Editorial Standard.
Trust root: Scientific Evidence & Editorial Standard
Related: Production-ready · Aerospace Qualification Guide · Tools · Research Hub.