P-FMEA CNC Machining Example: Complete Process FMEA with Real Data
This is a complete Process FMEA example for a CNC machining operation, following the AIAG-VDA FMEA methodology. Every process step, failure mode, and risk rating uses realistic values from actual manufacturing environments.
📋 This example covers 10 sequential steps from raw material receiving through final shipping — typical for a precision-machined aluminum or steel component in automotive/aerospace.
Why CNC Machining Needs a P-FMEA
CNC machining involves tight tolerances, expensive tooling, and multi-step processes where a single failure mode can cascade. A structured FMEA process identifies where failures are most likely and most severe—before they reach the customer.
Complete P-FMEA Table
S = Severity, O = Occurrence, D = Detection. AP = Action Priority (High / Medium / Low) per AIAG-VDA methodology.
| Process Step | Function | Failure Mode | Effect | Cause | Prevention | Detection | S | O | D | AP |
|---|---|---|---|---|---|---|---|---|---|---|
| Raw Material Receiving | Verify material certificate & dimensions | Wrong alloy grade received | Part fails under load / warranty claim | Supplier labeling error | Incoming material cert review | XRF alloy verification | 9 | 3 | 3 | High |
| CNC Program Setup | Load correct G-code program | Wrong program revision loaded | All parts out of spec / scrap batch | Operator selects outdated file | Program version control system | First-article dimensional check | 8 | 4 | 3 | High |
| Workpiece Clamping | Secure workpiece in fixture | Insufficient clamping force | Part shifts during cut → scrap + tool breakage | Worn fixture / wrong torque | Fixture maintenance schedule | Visual & torque verification | 8 | 3 | 4 | High |
| Rough Machining | Remove bulk material to near-net shape | Excessive tool wear undetected | Dimensional drift across batch | No tool-life monitoring | Tool-life counter in CNC controller | In-process dimension probe | 6 | 5 | 4 | Medium |
| Finish Machining | Achieve final dimensions and surface finish | Surface roughness out of spec (Ra > 1.6 µm) | Assembly interference / customer rejection | Worn insert / incorrect feed rate | Cutting parameter validation sheet | Surface roughness tester (profilometer) | 7 | 4 | 3 | High |
| Deburring | Remove sharp edges and burrs | Residual burrs on internal features | Assembly damage / safety hazard | Manual process inconsistency | Standardized deburring procedure | Visual inspection with magnification | 7 | 5 | 5 | High |
| CMM Inspection | Verify critical dimensions per drawing | Measurement error due to probe calibration drift | Non-conforming parts shipped as conforming | Skipped calibration cycle | Automated calibration reminder | Gauge R&R study / MSA | 9 | 2 | 3 | High |
| Surface Treatment (Anodizing) | Apply protective coating per spec | Coating thickness below minimum | Corrosion in service → field failure | Bath chemistry out of range | Daily bath titration log | Eddy-current thickness gauge | 8 | 3 | 3 | High |
| Final Packaging | Protect parts for shipping | Inadequate packaging → transit damage | Customer receives damaged parts | Wrong packaging material selected | Packaging work instruction | Outgoing visual inspection | 5 | 3 | 3 | Medium |
| Shipping & Documentation | Include CoC and ship to correct address | Missing Certificate of Conformance | Customer rejects shipment / production delay | Document not attached to shipment | Shipping checklist procedure | Final documentation audit | 6 | 3 | 2 | Low |
How to Read This FMEA Table
Each row represents one potential failure scenario for a specific process step. The combination of S, O, and D determines the Action Priority, which replaces the legacy RPN approach in the AIAG-VDA standard.
⚠️ Key finding: Raw Material Receiving and CMM Inspection carry the highest severity (9) because failures here can result in undetected non-conformance reaching the customer.
Key Findings from This Example
- Deburring is a high-risk step due to manual process variability.
- Shipping & Documentation is low-risk because failures are caught by the final documentation audit.
- Most high-AP items involve either material mix-ups or dimensional control failures.
Recommended Actions
For every High-AP item, the FMEA team should define specific actions, assign responsibility, and set a target date. The Qhubio FMEA tool generates these recommendations automatically based on your process data.
Using This Example as a Template
You can use this P-FMEA as a starting point for your own CNC machining risk analysis. For a faster approach, try the FMEA generator. Compare with Excel-based templates or explore the template comparison.
Common Mistakes in CNC Machining FMEAs
- Listing generic failure modes ("part out of spec") instead of specific ones
- Using the same S/O/D values for every row
- Ignoring setup and material-handling steps
- Not distinguishing between prevention controls and detection controls
- Treating the FMEA as a one-time document instead of a living risk record
Process Context, Scope, and Assumptions
This PFMEA represents a typical CNC machining cell producing a precision aluminum or steel component for automotive or aerospace. Defining scope tightly is the single most important decision — broad scopes always produce generic failure modes.
- Part family: precision machined housings, 6061-T6 / 1.2379 steel, forged or bar stock.
- Equipment: 3-axis and 5-axis CNC machining centers with automated tool change and in-process probing.
- Volume: 50k–500k pcs/year, 2–3 shifts.
- Customer baseline: IATF 16949 or AS9100, plus customer-specific requirements.
- Special characteristics: Ø8 H7 locating bore, sealing surface Ra ≤ 1.6 µm, anodized layer 15–25 µm.
- Out of scope: incoming raw material (supplier PFMEA), downstream assembly.
Rating Justification — Why These S, O, D Values?
Every rating in the table above is anchored to AIAG-VDA scale rows and to capability or field data. A rating without recorded justification is the single most common audit finding.
Raw Material Receiving — S=9, O=3, D=3 → AP High
- S=9 — wrong alloy reaches the customer and fails under load; potential safety impact, AIAG-VDA SE row 9.
- O=3 — supplier mislabel incident rate < 1 per 100 receipts over 5 years.
- D=3 — XRF alloy verification 100% on receipt, MSA validated.
- AP=High — driven by S=9 regardless of low O, per AIAG-VDA AP table.
CNC Program Setup — S=8, O=4, D=3 → AP High
- S=8 — wrong program scraps the batch and stops the customer line.
- O=4 — observed operator selection error ~0.5% pre-controls.
- D=3 — first-article dimensional check before run-on, MSA validated.
- Action: MES-enforced program lock by part number, target O=2.
Finish Machining — S=7, O=4, D=3 → AP High
- S=7 — Ra failure causes assembly interference / customer rejection.
- O=4 — Cpk 1.20 on Ra, insert wear is the dominant cause.
- D=3 — profilometer 100% with locked sampling plan, GR&R 9%.
Deburring — S=7, O=5, D=5 → AP High
- S=7 — internal burrs damage mating assembly.
- O=5 — manual deburring of internal features, scrap rate 1.2%.
- D=5 — visual + magnification; internal channels not fully inspectable.
- Action: automated thermal/abrasive deburring, target O=3, D=3.
For the full scale logic see the Severity, Occurrence, and Detection rating guides, plus the Action Priority logic.
Prevention vs Detection Controls
AIAG-VDA separates Prevention controls (lower Occurrence) from Detection controls (lower Detection rating). Mixing them in one cell is the most common audit finding on CNC PFMEAs.
| Step | Prevention (lowers O) | Detection (lowers D) |
|---|---|---|
| Raw Material Receiving | Approved-supplier list, dual labeling rule | XRF alloy verification 100% |
| CNC Setup | MES program lock by part number | First-article dimensional |
| Rough Machining | Tool-life counter, validated parameter table | In-process probe every 25 parts |
| Finish Machining | Insert change at 90% of validated life | Profilometer sampling, locked plan |
| Deburring | Automated deburring fixture (planned) | Borescope inspection |
| Anodizing | Supplier SPC on bath chemistry | Eddy-current thickness 100% |
See FMEA vs Control Plan for how each control flows into the Control Plan.
Auditor Perspective — What They Will Challenge
| Auditor question | Strong answer |
|---|---|
| "Why D=3 on Finish Machining?" | 100% profilometer, GR&R 9%, MSA validated, locked sampling plan. |
| "Show me the action for High AP rows." | Action log: owner, target date, residual S/O/D, Cpk post-action. |
| "Where is this control in the Control Plan?" | Same characteristic ID and step ID, identical reaction plan. |
| "How is Occurrence justified?" | 12-month Cpk, scrap rate, supplier history, similar-process data. |
| "When was this PFMEA last reviewed?" | Revision history, triggered by latest 8D or engineering change. |
Use the full FMEA audit checklist 48 hours before the audit.
How Teams Score CNC PFMEAs Wrong
- D=2 on sampling inspection — sampling can never deserve D≤3.
- O=2 with no Cpk evidence — Occurrence requires capability data.
- Severity downgraded "because we'd catch it" — controls do not change Severity.
- Prevention and Detection mixed — must be in separate columns per AIAG-VDA.
- No residual rating after action — every closed action requires re-rating and evidence.
- PFMEA step IDs do not match Process Flow / Control Plan — guaranteed audit finding.
Full list in Common FMEA mistakes.
Implementation Notes
- Start from the current Process Flow Diagram, not from an old PFMEA.
- Add a rating-rationale column next to each S, O, D — biggest single jump in audit readiness.
- Synchronize step IDs and characteristic IDs across PFMEA, Process Flow, and Control Plan.
- Define a written rule: visual D≥7, sampling D≥6, 100% automated with GR&R<10% can reach D=2–3.
- Re-rate after every closed action and attach Cpk / ppm / audit evidence.
- Trigger PFMEA review on every 8D, engineering change, supplier change, tooling change.
- Use FMEA software or the FMEA generator to enforce AP calculation, revisions, and Control Plan linkage.