|Gage capability and quality
Accuracy and precision
Gage capability and process capability
Statistical hypothesis testing
Manufacturing worker's role in quality
Concept of variation in quality and SPC
|Gage Capability: Reproducibility and Repeatability
excerpt from SPC Essentials and Productivity Improvement: A Manufacturing Approach
All material (C) 1996, Intersil Corporation (formerly Harris Semiconductor) or ASQC Quality Press
May not be reproduced in any form without prior permission, except as defined under "fair use" Free ActiveX program to perform gage capability (reproducibility and repeatability) calculations
Gages must be accurate (calibrated) and precise (capable) if they are to provide useful information.
We previously discussed calibration, which assures a gage's accuracy. Accuracy, however, is only one requirement for effective measurements. The other is precision. We introduced these terms in Chapter 3, which discusses a manufacturing process' ability to meet specifications. The mean of an accurate process is on the target, which is usually halfway between the specifications. Precision is the opposite of variation, and it has the same meaning in discussions of gages. Unlike a manufacturing target, a gage's target is the actual dimension of the specimen it is measuring. If a dimension is 12.0 microns, we want the gage to report 12.0 microns. Table 5-1 shows how gage is like a manufacturing process. The gage's product (the measurement) has unavoidable variation. Table 5-1. Similarities between Gages and Processes
Process Gage Product The workpiece A measurement Target Usually the midpoint of the specification The specimen's actual dimension Variation Process variation Gage variation (reproducibility and repeatability)
An accurate gage will, on average, report the specimen's actual dimension. The words "on average" are unsettling, and they should be. A non-capable (imprecise) gage will return widely differing measurements from the same specimen. If the piece's dimension is slightly out of specification, the gage may pass it. If the piece is slightly within specification, the gage may reject it. Figure 5-1 shows an almost-perfect gage. If a part is bad, it will almost certainly fail. If it is good, the gage will almost certainly pass it. The figure also shows a non-capable gage that can easily pass bad pieces and reject good ones.
[Figure 5-1 appears here. Figure 5-16, which is for advanced readers,
appears later in the chapter.]
Reproducibility and repeatability are the elements of gage variation. (They actually refer to "lack of reproducibility" and "lack of repeatability.") If a gage has good reproducibility, the measurement will not depend on the person who uses it. If a gage has good repeatability, it returns the same number each time we measure a given specimen. A gage study, or reproducibility and repeatability (R&R) study, measures the gage variation.
[The chapter shows how to perform a gage capability study by using the General Motors "long form" method.]Gage Capability and Process Capability
[This section explains how a mediocre gage can significantly depress an excellent process' process capability index. It shows how to relate the true process capability to the observed process capability, given the gage's percent tolerance ratio. For a gage with a 30% P/T ratio, a process whose true Cp is 1.605 will report a Cp of only 1.4. The QS-9000 standard, incidentally, says that P/T ratios should be 10% or less.]
Figure 5-15. Inherent or Actual Process Capability [PTCC = percent tolerance consumed by (lack of) capability, or the gage's P/T ratio]
Send Mail to(mail to webmaster@ is discarded due to abuse by spammers.)