Free, in-browser, no sign-up
Ten quality engineering calculators: AQL sampling plans to ANSI/ASQ Z1.4, process capability (Cp and Cpk), Gage R&R, X̄ and R control limits, sigma level, OC curves, FMEA risk priority and PPM conversion. Everything runs in your browser. Nothing is uploaded, and nothing is behind a sign-up.
Ten calculators, standards-backed
Each card implements a published method - ANSI/ASQ Z1.4, AIAG MSA, the NIST handbook constants - and shows its work in the result register.
Lot size in, sampling plan out: sample size, code letter, and the accept/reject numbers for your major and minor AQL. The plan is only defensible if it is tied to the customer specification, so set the AQLs to the contract, not to habit.
How well the process fits the specification. Cpk under 1.33 means any drift in the mean starts shipping defects; Cpk without demonstrated stability is a number about a process that is not behaving the same way twice.
Defects per million opportunities and the sigma level it implies. Count opportunities honestly: inflating them flatters the sigma without changing a single outcome the customer sees.
Centre line and control limits for a means chart, from the process itself. If you find yourself choosing limits to match the specification, you are building a chart that agrees with you, not one that warns you.
Control limits for subgroup ranges using the D3 and D4 constants, n from 2 to 10. Run the range chart before trusting the means chart: if spread is out of control, the X̄ limits are built on sand.
How much of your variation is the measurement system itself. Under 10% of tolerance is acceptable; over 30% means the gage, not the process, is what you have been charting. This is the study that decides whether the rest of your data means anything.
How many units to inspect for a given confidence and margin of error, assuming the worst-case proportion unless you know better. Halving the margin quadruples the sample: precision is bought, never wished for.
The operating characteristic curve for a plan of n samples and acceptance number c: the probability a lot passes, across true defect rates. This is the honest picture of what your sampling plan actually protects against.
Severity × occurrence × detection, each scored 1-10. The number ranks where to act first; its value is entirely in being computed early, which is exactly when nobody wants to spend the time.
Parts-per-million and percent defective, converted both directions, with the sigma level the rate implies. Enter either side; the other follows.
These calculators implement published methods; the standards are the authority.