Capability Calculation on 1 Population Multiple Tolerances

[frantz]

Hi,
Does anyone knows how to calculate process capability for a population with multiple tolerances?
Can we consider using deviation from Aim and then calculate capabilities with a tolerance of +/- 1.0 and How?
Example temperature.
Below is a data sample:
Temp Aim LSL USL
889.0 904 854 954
883.4 882 832 932
887.6 882 832 932
890.5 904 854 954
887.2 904 854 954
888.4 882 832 932
883.9 882 832 932
892.3 882 832 932
891.0 882 832 932
891.2 882 832 932
888.8 904 854 954
899.1 904 854 954
891.6 904 854 954
902.7 910 860 960
908.5 910 860 960
906.9 910 860 960

Thanks

 

[howste]

Are you mixing data from multiple process streams, or are you changing the settings with different batches?

 

[qualitytrec]

I do not really understand your question. Are you asking how to track capability on an oven at various settings or how to track capability on a process with multiple settings or...???

It sounds like you are asking capability on ovens set at different temps. Perhaps heat treat ovens.

Mark

 

[frantz]

Are you mixing data from multiple process streams, or are you changing the settings with different batches?

Howste,
Here is the problem, we have a continuous process and the temperature tolerances is dependent of the product, therefore when the product is going to the furnace the temperature and the tolerances change. In my opinion there is no way to calculate capability on the whole population. I used to do a deviation from aim (Delta) and calculate the capability for the process base on the deviation. Someone told me that calculating capability from deviation from the aim is wrong.
Hope that answer your question.

Thanks
Frantz

 

[frantz]

I do not really understand your question. Are you asking how to track capability on an oven at various settings or how to track capability on a process with multiple settings or...???

It sounds like you are asking capability on ovens set at different temps. Perhaps heat treat ovens.

Mark

Mark,
I am sending you the answer I send to Howste.
Here is the problem, we have a continuous process and the temperature tolerances is dependent of the product, therefore when the product is going to the furnace the temperature and the tolerances change. In my opinion there is no way to calculate capability on the whole population. I used to do a deviation from aim (Delta) and calculate the capability for the process base on the deviation. Someone told me that calculating capability from deviation from the aim is wrong.
Hope that answer your question.

Thanks
Frantz

 

[Bev D]

'someone' is mis-informed. there is no reason that a capability analysis can't be performed based on delta to target. There are special considerations when calculating a Capability index, but these are useless anyway.

Basically plot your data first and see what it tells you. dont' automatically combine the different temperature levels but look to see if you have difficulty with any particular setting.