Dictionary
Below is a list of a few terms and abbreviations that are used in the quality work. Please e-mail us if you have comments or questions concerning the words, or if a certain word is missing. You find the address at the bottom of the page.
annotation Arbitrary information that is bound to a certain entity. It can for example be an operator name that is bound to a certain measure value, or a drawing number that is bound to a certain product. Other words that are used with similar meaning are tracefield and meta data.
CMM Coordinate measure machine. A machine that automatically measures several measure points on a detail, guided by the coordinates from the detail drawing.
Cp Process capability for stable processes. Indicates the number of times the variation fits in within the tolerance. A high Cp means a small variance and thus a good process with respect to variance. A guideline is usually Cp > 1.33, which corresponds to 4 sigma.
The formula for Cp is: Cp = (UTL-LTL)/(6*s) where s = Rbar*d2 (d2 from table).
Note: Here Cp is described as defined in QS9000, but there are also definitions of Cp and Cpk that were specified by Volvo and Ford and which are denoted as Pp and Ppk in QS9000.
Cpk Process capability index for stable processes. Process capability corrected to consider how well the distribution is centered about the target value. A high Cpk means a small variance and a well-centered distribution and thus a good process with respect to variance and centeredness. A guideline is usually Cpk > 1.33, which corresponds to 4 sigma.
The formula for Cpk is: Cpk = min(UTL-Xbar, Xbar-LTL)/ (3*s) where s = Rbar*d2 (d2 from table).
Note: Here Cpk is described as defined in QS9000, but there are also definitions of Cp and Cpk that were specified by Volvo and Ford and which are denoted as Pp and Ppk in QS9000.
failure mode and effect analysis See FMEA.
FMEA Failure Mode and Effect Analysis. A tool used during the product respective production development phases with the purpose to identify potential fault possibilities.
Gage R&R Gage Repeatability and Reproducibility. The combined estimate of measurement system repeatability and reproducibility. Gives a measure of the capability of the system.
Repeatability means the variation when an operator uses the same instrument to measure the same characteristic several times.
Reproducibility means the variation when several operators use the same instrument to measure the same characteristic. Gage R&R is a part of MSA.
LSL Lower specification limit. See LTL.
LTL Lower tolerance limit (also known as lower specification limit, LSL). Can be absolute or relative to a nominal value (and is then sometimes denoted ΔLTL or dLTL).
meta data See annotation.
MSA Measurement System Analysis. A method for deciding how much the variation within the measurement system contributes to the total variation of the process. MSA includes Gage R&R.
nom See nominal value.
nominal value Usually the value from which the tolerance limits (UTL and LTL) are specified. If the tolerance interval for a dimension is 3.0 ± 1.0, the nominal value is 3.0. Nominal value should not be mixed up with target value, even though they may coincide. There are other definitions of nominal value.
Pp Process performance for unstable processes. Indicates the number of times the variation fits in within the tolerance. A high Pp means a small variance and thus a good process with respect to variance. A guideline is usually Pp > 1.67, which corresponds to 5 sigma.
The formula for Pp is: Pp = (UTL-LTL)/(6*sigma). See also Ppk, Cp and Cpk.
Ppk Process performance index for unstable processes. Process performance corrected to consider how well the distribution is centered about the target value. A high Ppk means a small variance and a well-centered distribution and thus a good process with respect to variance and centeredness. A guideline is usually Ppk > 1.67, which corresponds to 5 sigma.
The formula for Ppk is: Ppk = min(UTL-Xbar, Xbar-LTL)/ (3*sigma).
See also Pp, Cp and Cpk.
R The range of a subgroup, i.e. the difference between the highest and lowest value in the subgroup.
Rbar R, the mean (average) of all subgroup ranges. See R.
Rmax Upper control limit in R charts.
sigma σ, standard deviation. An estimate of the spread of the measure values about the mean value. The standard deviation is mainly used if the measure values are normally distributed, since it can be misleading for other distributions.
six sigma
1. A measure of quality means that the standard deviation constitutes a maximum of one sixth of the tolerance. This implies that of one million produced details no more than 3.4 are defect.
2. A methodology that provides the techniques and tools to improve the capability and reduce variances in a process.
SPC Statistical process control. Makes it possible to produce details with zero faults by discovering abnormal variations and observing wear trends. Uses statistical methods to describe, control and illustrate processes.
SQC Statistical Quality Control. Statistical methods used in measuring, monitoring, controlling and improving quality.
Statistical Quality Control See SQC.
target value The value that is aimed for in a process. In a good process the mean value (Xbar or Xdbar) is on the target value and the spread is small.
tracefield See annotation.
USL Upper specification limit. See UTL.
UTL Upper tolerance limit (also known as upper specification limit, USL). Can be absolute or relative to a nominal value (and is then sometimes denoted ΔUTL or dUTL).
X Denotes a single value.
Xbar X, the mean (average) of all values or of the values in a subgroup. See X.
Xdbar X, the mean (average) of all subgroup Xbar values. See Xbar.