Statistical Prerequisites For Valid Cp / Cpk Calculations

Hello Dave, First you must make sure your measurement system is calibrated for accuracy, repeatability and reproducibility. We perform a standard GR&R Study on our CmController system on a regular basis to ensure that the system itself is capable to make accurate measurements and statistical evaluations. When measuring a particular machine, whether a printer, dispenser or placer, you must define what the performance specifications are and what the expected statistical results are going to be based on this performance spec. Only then can we evaluate and improve a machine's performance based on the statistical results it provides. If a machine meets the expected statistical results, i.e. consistent low standard deviation values (indicates stable performance), then and only then will we begin to optimize the machine with adjustments (mean displacement values to machine software code) before a full length Cp/Cpk evaluation is executed. Each tested machine will yield different expected results because of their specific quality performance specs which are always set by the individual manufacturers. Nonetheless, you can only be certain the calculations are valid if you have a proven and suitable measurement system for validating that a machine meets performance specification. Let me know if you require additional assistance. Thanks!

This message was posted the OnBoard Forums

reply »