SPC Control Charting - Target Charts

 

Last Revised:  2/05/02         By: CRS      

 

Key words: Target, charts, control chart, data entry, runs and trends based on target, specific configuration, Monitoring modes, Nominal charting, frozen control limits, fixed control limits, runs based on target

 

Summary

There are many specific configurations for monitoring processes using DataLyzer® control charts.  The 2 most frequently used modes are the default mode where the control limits are left to calculate +/- 3 sigma of the averages each time new data is entered.  The last 25 subgroups of data are considered in this calculation and the control limits usually vary slightly each time a subgroup of data is entered.  Run and trend flags are based on the process grand average.  In this method, the control limits respond to the data being entered and the run and trend flags offer additional evidence the process’ statistical location is changing.

 

A second method is “Target charting” or “Nominal charting”.  This method usually calls for the control limits to be fixed or frozen at “historical” levels (+\- from a nominal or target value).  Run and trend flags are then calculated from this target or nominal rather than from the grand average of the process (as would be the case in the previous example). In Nominal charting, the goal is to fix the process at a particular statistical location and allow the run, trend and out of control flags to alert the user when the process begins drifting away from a desired statistical location on the xbar chart.  In the chart below, the green line is the average line and the blue line with arrows at each side is the target line.

 

 

Setup

DataLyzer® Spectrum will allow target chart input whether using enter/edit mode for one characteristic or a sequence for multiple or automated data entry.  The specs for the characteristic should be set for +\- at the desired target or nominal when creating the characteristic. 

 

To setup a characteristic to calculate runs based on the target, open the characteristic file you wish to change.  Select PREFERENCES/SET PREFERENCES from the main menu of the characteristic creation screen.  Select the CONTROL CHART CALCULATIONS tab and then check the box labeled “Runs based on Target”.  When data is entered in the characteristic, the run and trend analysis will then be based on the target.  Once the target chart is enabled, the process will not be allowed to shift from its desired center.

 

NOTE: Calculations will not be made on data that was entered before the preferences were changed.

 

Control limits for xbar should be frozen at historical levels. The control limits can be accessed by selecting CHARACTERISTIC/CONTROL LIMITS from the characteristic setup screen.

 

Advanced Applications

Functions can be created to subtract the nominal value from each full measurement entered. If the nominal will be the same each time you can simply use “R1-2.30”.  For example if 2.30 is the nominal. 

 

You can also combine different products from the same process. These products can be of different sizes, but their specifications in terms of deviation from nominal must be the same.  An example of this is a cutoff process, which cuts 9”, 12” and 15” pieces.  The pieces are different sizes but their specs all similarly deviate from their respective nominals +\- .010.  You can combine these different products in to one control chart, if they are reflecting the same basic process. 

 

To add or subtract a value that may vary between subgroups, define a parameter called “nominal” as numeric.  You can use this nominal parameter in a function so the variable value can be edited every time a subgroup is added.  The operator will be prompted during data entry to enter the current nominal value.  This value is automatically subtracted from each sample in the subgroup as defined in the characteristic’s function.

 

By using target charts, the process is pegged to a particular nominal value.  The run, trend and out of control flags will appear anytime the process’ statistical location drifts.   

 

For more information on characteristic setup, see the DataLyzer® Spectrum User’s Manual.

 

 

 

DataLyzer® Spectrum Application Note

courtesy of DataLyzer International, Inc. 

 

Control Number: 45