Does zero defects prove the lot is perfect?

No. Zero observed defects gives a probability bound under the model. The upper limit still allows defects below that bound and depends on random, independent sampling.

What happens if I find one defect?

The zero-defect calculation no longer applies. Use an approved analysis or acceptance plan that accounts for the observed defective count.

Is this an AQL sampling plan?

No. AQL acceptance plans include defined quality points, acceptance rules and producer and consumer risks. This calculator supplies only the zero-defect binomial confidence bound.

Can I inspect consecutive stamped parts?

Consecutive parts may share tool drift or a common material defect. Use the approved random sampling method; correlated observations do not meet the independent-trial assumption.

Sheet Metal Stamping & Press Lines calculator

Zero-Defect Inspection Sampling Calculator

Find the random sample size needed when your check requires zero observed defects. You need a maximum defect percentage, one-sided confidence requirement and planned sample count.

What this calculator does

Size a random sample requiring zero observed defects for your defect limit and confidence requirement.

Formula used

Probability of zero defects = (1 − defect fraction)^sample count
Required sample = ceil(log(1 − confidence fraction) ÷ log(1 − defect fraction))
Detection probability (%) = 100 × (1 − (1 − defect fraction)^planned sample)
Sample gap = planned sample − required sample
Zero-defect upper limit (%) = 100 × (1 − (1 − confidence fraction)^(1 ÷ planned sample))

Inputs explained

Defective Part Limit: Maximum defect percentage from your quality requirement.
Required Confidence: One-sided confidence requirement approved for this quality check.
Planned Random Sample: Number of independent random parts scheduled for inspection.

How to use the result

Best suited to zero-defect quality demonstration, random sample planning.
This binomial estimate suits a continuing process or large population, not exact finite-lot sampling. It is not an AQL acceptance plan or certification. Clustered defects, drift and imperfect inspection violate the probability model.

Current U.S. benchmarks

The producer price index for steel mill products stands at 381.162 (BLS, Aug 2026), up 23.4% from a year earlier. Quotes priced off last quarter's material cost miss this move.
U.S. iron and steel import customs value ran $2.2B in Aug 2026 (Census International Trade). The U.S. ran a trade deficit of $0.6B in the category that month. This dollar total mixes price, quantity, product mix, origin, and timing; it does not measure physical import volume or prove a tariff or reshoring effect.
The U.S. has 53,790 fabricated metal products establishments employing about 1,441,471 workers (Census County Business Patterns, 2023).

Common questions

Does zero defects prove the lot is perfect? No. Zero observed defects gives a probability bound under the model. The upper limit still allows defects below that bound and depends on random, independent sampling.
What happens if I find one defect? The zero-defect calculation no longer applies. Use an approved analysis or acceptance plan that accounts for the observed defective count.
Is this an AQL sampling plan? No. AQL acceptance plans include defined quality points, acceptance rules and producer and consumer risks. This calculator supplies only the zero-defect binomial confidence bound.
Can I inspect consecutive stamped parts? Consecutive parts may share tool drift or a common material defect. Use the approved random sampling method; correlated observations do not meet the independent-trial assumption.

Last reviewed 2026-10-06.