Purpose

A control chart is a graphical tool used by quality technicians to control, analyze and document the processes involved in production and other quality-relevant areas.

Control charts are used to detect systematic deviations of a quality characteristic from a target value (signals) against the background of inevitable, random fluctuations in individual measured values (interference). Today, this form of control chart is the central element of statistical process control (SPC).

Implementation Considerations

Control charts are primarily used in inspections during production to monitor and manage controlled production processes . A process is controlled or under statistical control if the observed process parameters only vary randomly from sample to sample. The control chart can be used to detect special influences on the process. You correct the disturbances by intervening in the process. You can use the control chart to determine whether the correction was successful.

The control chart is also a suitable medium for determining whether a process is or was in control or not, even if it is too late for corrective interventions. The control chart, therefore, can also be used for procurement and dispatch (for example, for incoming inspections based on a scheduling agreement with periodic release orders).

Integration

You define how to run a control chart for an inspection characteristic during your inspection planning activities. You can run control charts within an inspection lot/production order or for several inspection lots/production orders.

You can display and update control charts during results recording or using a separate transaction. You can have the system calculate the action limits and warning limits (if necessary) in the control chart window. The algorithms required to do this are stored in the Customizing application and can be replaced if necessary.

You can use a control chart to valuate the inspection results .

If the relevant valuation rule is used, a sample is rejected if at least one of the action limits defined in the associated control chart was violated.

If action limits are violated, the Workflow can be triggered automatically, provided that you have defined a suitable defect code in the inspection plan characteristic and the system has been set accordingly.

A run-chart is a graphic that is similar to a control chart. The run chart displays a chronological curve of the measured values or sample mean values for a characteristic. In contrast to control charts, a run chart does not contain action limits; instead, it contains limits for the tolerance range. As with a histogram , you can call up a run chart during results recording and when making evaluations, without making any prior settings in the basic data.

Features

In statistical process control, one or more control variables of the observed characteristic are selected and determined by taking samples from the process at set time intervals if possible. These statistics are entered in the chart in chronological order. The most important control variables are:

Two control variables of a characteristic are often run in parallel as two tracks on a control chart (for example, the mean value and the standard deviation). In this example, the location and dispersion of the process can be observed at the same time.

Apart from the control variable, each track on the control chart also contains control limits for the process. You must intervene in the process if these limits are violated. In addition to these action limits, you can also define warning limits (only when using SAP Statistical Graphics) or a mean line for individual chart types.

The limits are generally determined from the current process data or the results of a preliminary run, using statistical methods. The limits are calculated using various algorithms that are based on different mathematical approaches.

The standard control charts are:

Acceptance charts are based on the specified tolerance and control the share of scrap in the process. With these charts, the limits are extended if a long-term reduction in the process dispersion can be achieved by means of technical or organizational changes.

With Shewhart charts , the limits contract in this case. These charts only take internal process parameters into account and not external tolerance specifications. In a constantly recurring statistical test, the hypothesis that the defined "in control" status of the process has not (yet) changed, is tested.

Other types of control chart are currently used in industry in addition to these standard types. These include charts with extended limits, pre-control charts, cusum charts and various moving-average charts.

The following control chart types are available in the Quality Management component for inspection characteristics: