How to assess capability for parameters which has got either Min values or Max values

R

raju9619

I work in a cable industry. Here all the parameters such as dimensions, conductor resistance, Tensile strength values are mentioned either in minimum values or Maximum values

Eg: .The standards Says

1."Tensile strength of the PVC shall me Min 12.5 N/mm²" which means any thing above 12.5 is OK.

2. Conductor Resistance of Cable Shall be Maximum 7.98 Ω/KM Which means anything less than 7.98 is OK


I need to do CP CPk study for these . Please help me out ..
 

Jim G

Involved In Discussions
Re: How to do SPC for parameters which has got either Minimum values or Maximum value

Hi Raju
Cpk is an index which measures a process performance to its specification limits, relative to the natural variability of the process. Cpk tells you how the process is capable of performing in the future. Cpk is calculated using an estimate of the standard deviation using R-bar/d2. Remember, the calculation requires the process to be in a state of statistical control. You can have a good Cpk and still have data outside the specification.
Hope this helps
Jim G


We learn more by looking for the answer to a question and not finding it than we do from learning the answer itself".-Lloyd Alexander
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Re: How to assess capability for parameters which has got either Min values or Max va

I work in a cable industry. Here all the parameters such as dimensions, conductor resistance, Tensile strength values are mentioned either in minimum values or Maximum values

Eg: .The standards Says

1."Tensile strength of the PVC shall me Min 12.5 N/mm²" which means any thing above 12.5 is OK.

2. Conductor Resistance of Cable Shall be Maximum 7.98 Ω/KM Which means anything less than 7.98 is OK


I need to do CP CPk study for these . Please help me out ..

Oh, boy...here we go again.

First, are you satisfied with the gage R&R of your test method? If not, your capability will be artificially bad. Tensile testing has some pretty wide variation by its nature.

Second, the whole point of Cpk versus Cp was to show if your variation was centered. So, how does one show "centeredness" to a unilateral tolerance. You don't. So, why calculate it?

Finally, you will find that people will suggest you use one or the other "half" of the Cpk calculation (Cpu or Cpl). It's nice, but what does it prove? It shows "distance to the specification" - which as a totally different question than "is your process centered?" So, now that you calculated oranges, what do you do with it in a world of apples? Beats me...except give customers useless data.
 
Top Bottom