Hardware Products Inspection & Acceptance Execution Standards
Official Quality Assurance & Defect Classification Standard for Hardware Procurement · Companion report: Hardware Products Inspection & Acceptance Report
This document is a specialized vertical Standard Operating Procedure (SOP), independent of the General Quality Control & Technical Inspection Standards, governing accurate completion of the Hardware Products Inspection & Acceptance Report. Hardware products (stamping, die-casting, CNC machining, precision metalworking, etc.) are engineering-dependent items whose quality directly affects structural safety, mechanical assembly integrity, and corrosion resistance. When completing the report, eliminate subjective judgments such as “close enough” or “probably fits.” Enforce the mandate of Objective Record, Data-Driven, On-Site Alignment to supply clients with auditable data for release or rejection decisions.
1. Introduction & Core Hardware QA Philosophy
When completing the inspection and acceptance report, apply three non-negotiable principles: Objective Record, Data-Driven, On-Site Alignment. Expressions such as “I think,” “approximately,” or “acceptable” must not appear on the report. Every defect record must be quantifiable and traceable.
Unlike general industrial goods, hardware products require industry-specific evaluation of geometric tolerances, surface treatment durability, and thread and assembly fit. Before inspection, obtain the client-issued Specification, BOM, and approved golden sample (if applicable). For AQL sampling and disposition methodology, refer to Section 3 of the General Quality Control & Technical Inspection Standards.
2. Visual Surface Grading (Zone Classification Principles)
During cosmetic and surface-finish inspection, hardware products must not be evaluated as a single undifferentiated surface. Inspectors shall apply the industry-standard Zone A / B / C partitioning methodology, applying tiered acceptance criteria based on visual sensitivity and structural function:
Definition: Surfaces most directly exposed to the end user after assembly or installation — e.g., handle faces, decorative hardware panels, consumer-electronics outer housings.
Weighting: Any visible plating pinhole, localized burn, paint delamination, obvious scratch, or mold drag mark under standard D65 illumination at first glance shall be classified as a Major Defect (MA).
Definition: Rear faces, side faces, or mating surfaces that contact other components during installation.
Weighting: Function takes priority. Minor brush streaks, small mold impressions, or single scratches ≤ 2 mm that do not violate fit tolerances may be classified as MI; assembly-gap exceedance elevates to MA.
Definition: Internal cavities, underside faces, or areas visible only after disassembly — e.g., chassis interiors, die-cast rib backsides, master-carton exteriors.
Weighting: Where mechanical strength, assembly function, and corrosion resistance are unaffected, most cosmetic imperfections (light oil residue, sanding marks, non-sharp burrs) may be MI or accepted within process allowance.
Note: Conduct cosmetic inspection under agreed standard illuminant (typically D65, ~6500K). Avoid misjudging plating color variance and scratches under dim warehouse or color-shifted lighting.
3. Common Hardware Defects & AQL Classification Standard
Every on-site finding must be classified precisely into one of the three tiers below. Subjective or ambiguous classification is prohibited.
| Defect Class | Code | Hardware-Specific Definition | Typical Examples | AQL / Disposition |
|---|---|---|---|---|
| Critical Defect (CR) | CR | Structural Failure / Loss of Functionality / Safety Hazard | Unfinished stamped or fractured edges with cut-risk burrs; penetrating micro-cracks or failed welds on load-bearing members; Rust, Cracking, Severe Deformation, Delamination of Plating Layer; material non-conformance vs. BOM (e.g., carbon steel substituted for stainless). | AQL 0 · One finding = entire lot FAIL |
| Major Defect (MA) | MA | Functional failure or severe Zone A cosmetic damage. | Stripped threads preventing screw engagement; mounting-hole pitch error or critical GD&T exceedance preventing assembly; Zone A plating peel or large-area delamination; cross-hatch adhesion below 4B; severe porosity, cold shuts, or substrate pits affecting strength. | AQL 2.5 · Count ≥ Re = lot rejection |
| Minor Defect (MI) | MI | Cosmetic detail not affecting function. | Zone B/C scratches ≤ 2 mm; light water marks, oil, fingerprints; minor polishing lines; color variance within approved limit sample. | AQL 4.0 · Count ≥ Re = lot rejection |
CR · Critical
- Safety: sharp burrs, penetrating cracks
- Structure: failed welds, joint separation
- Compliance: material mismatch, severe corrosion
MA · Major
- Function: threads, hole pitch, tolerances
- Zone A finish: plating failure, adhesion
MI · Minor
- Zone B/C minor marks
- Limit-sample color variance, oil stains
4. Data Entry Protocols & Best Practices
4.1 General Profile & Sampling Parameters (Report Section 1)
4.2 Key Dimensional Verification (Report Section 2)
This module is not mandatory on every lot, but 5–10 pcs minimum measurement is strongly recommended when equipment permits, recording minimum, maximum, and average values.
Digital Caliper Protocol: Before each measurement session, fully close the jaws and zero the display. Confirm zero before use. Plastic calipers are prohibited on-site for hardware tolerances typically held at 0.01–0.05 mm; inadequate gauge resolution causes false pass/fail.
Thread Go/No-Go Gauge Protocol: When testing the No-Go gauge, do not force engagement — apply natural finger torque only. Excessive force damages workpiece threads, wears gauges, and creates false rejects. Light rotation; stop beyond two pitches and record F.
Zero digital calipers before use. Micrometers, plug gauges, and block gauges must be within calibration validity; clean measuring faces before use.
Measure bore, OD, thickness, and hole pitch at multiple points; record min, max, and average. Pitch error can prevent assembly — mark F in item result column.
Thread dimensions govern assembly. Compare against Specification tolerance; combine with Section 3 go/no-go results — do not rely on hand-feel estimation.
Any critical dimension outside spec → F for that row; if MA-class, count in Section 4 MA tally for AQL disposition.
4.3 On-Site Functional & Destructive Tests (Report Section 3)
Note (destructive test sample size): Cross-hatch adhesion, load/destructive, and carton drop tests are destructive. To protect bulk inventory, typically 1–3 pcs per lot — not the full cosmetic sample n. Record actual destructive piece count in the findings column.
This module is conditional: where factory equipment exists (e.g., Salt Spray Tester, Hardness Tester, Torque Wrench & Gauge, Coating Thickness Gauge), the inspector must execute per SOP and record data. If unavailable, mark “No Equipment On-Site” — do not fabricate readings.
Thread Go/No-Go Test
- Go gauge must thread smoothly to full depth without binding.
- No-Go engagement must not exceed two pitches; beyond that = thread non-conformance (F).
- Never force the No-Go gauge — finger torque only.
- Record go/no-go status factually in the findings column.
Coating Cross-Hatch Adhesion Test
- Using a Cross-Hatch Cutter, score 25 grids of 1 mm × 1 mm (through coating, not substrate).
- Apply 3M 610 tape, exclude air, peel at 90° rapidly — three consecutive cycles.
- Inspect grid edges for paint lift; below 4B = F.
Assembly Verification Protocol
Pull 5–10 pcs for full mating assembly; verify smooth operation, uniform gap, no jamming. Document gap and damping objectively in findings.
Additional test entry points
- Salt spray / corrosion: Prefer factory batch or recent third-party reports; if no Salt Spray Tester on-site, mark accordingly and record duration/rating if available.
- Hardness: Rockwell (HRC) or Vickers (HV) with Hardness Tester — three readings, average vs. Specification.
- Load / destructive: Apply design static load; record deformation or fracture.
- Carton drop: Execute 1 corner / 3 edges / 6 faces per agreed height; document carton and product damage.
4.4 Defect Tally & Final Disposition (Report Sections 4 & 5)
Correct example: “Of 80 sampled units, 3 exhibit [severe plating burn and peel on Zone A handle bend (location)] [peel area approx. 3 mm × 5 mm, classified MA (data)] (see photo 02).”
Incorrect example: “Handles look bad, a few defective.” — Non-auditable; unacceptable entry.
When completing Section 4:
- Tally CR, MA, MI separately in Actual Count Value; enter MA/MI Ac/Re per sample size n.
- CR ≥ 1 → must select FAIL/HOLD regardless of MA/MI.
- Any class ≥ Re → lot FAIL; all within Ac → PASS.
On-Site Alignment & Sign-off Protocols
When CR ≥ 1, or MA/MI counts reach or exceed Re, the final conclusion must be marked FAIL/HOLD on-site. The inspector shall present non-conforming units to the factory QA manager or plant director and obtain signature on the report.
If the factory states that “this is how we always ship” or “it does not affect use,” respond professionally and firmly: “Manager Zhang, I acknowledge the operational pressures on your production floor. However, my mandate is to document objective measurements and evaluate them against the client’s approved specification. The recorded data exceed the acceptance criteria. I respectfully request your signature to acknowledge these findings on the report; final release disposition remains solely with the Canton Buying Desk client authority.” If the factory declines to sign, document the refusal and report immediately to the client Desk; retain the full report and photographic evidence.
5. Synergy with General Quality Control Standards
Lot sizing, code-letter lookup, and Ac/Re tables follow ISO 2859-1 per Section 3 of the General Quality Control & Technical Inspection Standards. Apply the standard five-step field methodology from that document.
For stamping, die-casting, CNC, forging, and welded hardware PSI, apply this zone methodology and hardware defect dictionary, and issue the Hardware Products Inspection & Acceptance Report as the on-site authorization record.