Control chart
The quality control chart ( QCC ) or short control chart (English " [ quality] control chart", where " chart" not really, " map ", but rather " graph " or "Certificate" means ) is used in quality management for the evaluation of test data. In their statistical sampling characteristics are shown graphically (eg, sample mean and sample standard deviation of a workpiece dimension ). On quality control charts are warning and action limits and tolerance limits shown.
- 2.1 warning limits and control limits
- 2.2 Tolerance Limits
Types of control charts
Basically, control charts are distinguished by the type of features to be examined in control charts for variable characteristics and control charts for attribute characteristics.
Control charts for variable characteristics
The control charts for variable characteristics include, among others
- The original value chart
- Shewhart control chart (IMR Card by Walter A. Shewhart )
- The pre -control control chart ( process control charts ) and
- Acceptance control chart
Process control charts
The process control chart is a control chart, which is not based on predetermined limits. The upper and lower warning limit and the upper and lower control limits are calculated from the existing process data; they do not reflect the tolerance range, but only the observed frequency distribution of the monitored with the respective graph sampling parameter. The warning and action limits are periodically recalculated based on the latest process data. Collected on process control charts process data form the basis for the process capability study in which the frequency distribution of the observed feature is compared to the tolerance range.
The main process control charts are:
Control charts are also used for analysis of location and dispersion.
Acceptance control charts
The acceptance control chart is a control chart, in which the action and warning limits are calculated on predetermined tolerance limits. The tolerance limits indicate which deviations can be up there in a product to still be useful. The use of acceptance control charts is contrary to the principle of continuous improvement.
Control charts for attribute characteristics
The essential attributive control charts are:
Limits
Warning limits and control limits
Limits in quality control charts are shown by horizontal, by color or line thickness highlighted lines. A distinction between warning and action limits, which are respectively above and below the defined optimal mean value of the process to be controlled.
The distance between the two warning limits ( ±) and the two control limits ( ±) from the mean value is equal to, the following relationships apply when the measured value distribution of the Gaussian normal distribution obeys:
The eleventh measurement point (fifth from right) in the control chart shown above the upper warning limit. When a control limit would be exceeded, it is possible that the process is out of control at this point. In just 3 of about 1000 cases, however, exceeded the action limit for statistical reasons ( in the above- defined 3 -sigma range), without this necessarily means that the process or its parameters have changed (). When exceeding the warning limits possible, unintended changes in the process are to be sought and, if necessary, to take appropriate corrective actions to bring the process back to its proper condition. Thus the process can be corrected in the ideal case, even before this gets out of control and possibly defective parts are produced.
Tolerance limits
Tolerance limits ( Upper limit ( USL ) and lower limit ( LSL) ) are never marked on process control charts, as they apply to individual feature values and not for the parameters displayed on the control charts ( sample means, sample spans, etc.).
Indicator of the process
The quality control chart is also an indicator of the process in and of itself. In the evaluation of a control chart, a distinction between random and systematic influences. Random influences lead to a spread of test data on the control chart, they are caused by factors such as small temperature fluctuations or characteristics of the material and are to be regarded as normal, always existing part of the process. Systematic influences can lead to a slow shift of test data on the quality control chart or even sudden, drastic process changes; they are caused by special factors such as tool wear or incorrectly adjusted machines.
Indicator of the product
The course of the measurement points of the sampled parts shows the quality of the parts from the sample. It can be concluded on the quality of the total quantity of parts.
Evaluation of control charts
Systematic deviations are subject to laws. From the course of the measurement points on the control chart can infer these laws.
Thus we speak of a " trend " when at least seven data points show an almost linear slope in the direction of a boundary. There may be a strong increase in tool wear, which soon caused an overrun of the intervention or warning limit.
A " pattern" ( law ) is a non-random curve, such as the periodic " swing " around the mean value line. It may mean temperature variations that cause the production times greater, sometimes smaller parts. If there are 7 drawn points above or below the average line, the process mean is likely to be delayed. It is spoken by one pass ( or "Run "). This can indicate, for example, that a cutting tool has suffered damage and the parts made from now on larger or smaller.
The control limits are therefore not the only signs of potential problems; the arrangement of the measuring points should also be noted. If more than 90 % of the plotted points in the middle third of the range between the control limits or less than 40 % of the points in this period, is also assumed that a systematic ( non-random ) effect may have occurred.