Gage R&R Measurement System Analysis
Before any process improvement, verify your measurement system is trustworthy. Gage R&R > 30% means the "improvement" you see may just be measurement noise.
Step 1: Set Study Parameters
Total 60 measurements needed (order: part→operator→replicate)
Before a new gauge goes live, an engineer sets up a Gage R&R study with 3 operators, 10 parts, 2 replicates each (60 measurements total) to evaluate whether repeatability (consistency of repeated measurements by the same operator) and reproducibility (consistency between operators) meet measurement system requirements. Per the AIAG MSA manual, %GRR < 10% is excellent, 10-30% is acceptable, and > 30% means the gauge is unacceptable and must be improved or replaced.
10 parts, 3 operators, 2 replicates each (60 measurements total) Result summary: EV (equipment variation / repeatability) = 0.023 mm AV (appraiser variation / reproducibility) = 0.015 mm GRR = √(EV² + AV²) = √(0.023² + 0.015²) = 0.027 mm TV (total variation) = 0.095 mm %GRR = (GRR/TV) × 100% = (0.027/0.095) × 100% = 28.4% → Marginally acceptable; ndc = 4 (insufficient discrimination, recommend ≥ 5)
FAQ
Related Tools
Gauge Bias Analysis
Gage R&R evaluates gauge variation; bias analysis confirms systematic error. Together they form a complete MSA study.
Open toolGauge Linearity Analysis
Linearity analysis verifies whether bias is consistent across the measurement range (small vs. large dimensions) — the third element of MSA.
Open toolProcess Capability Index
CPK is only meaningful after MSA confirms the gauge is valid. Large measurement error distorts CPK results.
Open tool