Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.

Cpk Calculation - I analyse a double seam cans

#1
I need help in setting up my process CPk. I analyse a double seam cans and data taken using a software KUHNKE is been used to set-up my CPK which gives negative value after inputing my USL,LSL AND AVERAGE.Kindly advise me on how best to get my CPK with the data( eight seaming heads,generating different data on a particular machine)
Thanks in anticipation.
Uwalasky
 

Jim Wynne

Super Moderator
#2
What is being measured? Are you trying to combine data from multiple sources? Have you verified that the process is statistically stable for each data source? A description of the process would also be helpful.
 

Bill Levinson

Involved In Discussions
#3
The only way to get a negative Cpk is to have the process mean outside the specification limits. Also, if the heads generate different data while doing similar jobs (correlated) we could conceivably be looking at a multivariate situation, but I don't know enough to say.
 
#4
Thanks @Jim Wynne & @ Bill Levinson. The process as i earlier said is a can making factory, i need formula on how to calculate the CPK because what am using is not correct. Clue on the data i have: Average value, LSL & USL.
Thanks
 
#5
Thanks @Jim Wynne & @ Bill Levinson. The process as i earlier said is a can making factory, i need formula on how to calculate the CPK because what am using is not correct. Clue on the data i have: Average value, LSL & USL.
Thanks
As @Jim Wynne pointed out, have you verified that the data is statistically stable?

Also, if you use any statistical software package, for example: Minitab or JMP, it can help you calculate the Cpk, if that is what you are asking.
 

Jim Wynne

Super Moderator
#8
There are variation in data,Kindly see the manual calculation in excel. sheet as attached
It would be good if you could answer the questions I asked earlier. I suspect that there are few people here who are familiar with the machinery and details of your process.
 
#9
The machine in question is a canopy seaming machine, with 8 seaming heads ported producing cans at a speed of 800 cans per mins. Am trying to monitor the effectiveness of the machine using one of its critical parameters seam thickness with Lower limit of 0.90 and upper limit of 1.03
The cans in question are used for the production of evaporator milk, meaning there is no room for micro leakage.
Thanks for all the contributions
 

Welshwizard

Starting to get Involved
#10
Without knowing your machine and process its impossible to make a judgement about how to organise the data in context. Whether you can isolate the data from each head or group the data some other way is a rational decision that you have to make based upon your process knowledge, the steps below are the bare essentials and fundamentals of what you seem to be asking, apologies if you already know this.


Crucial - Demonstrate homogeneity of the process data stream by plotting the data in time series order on an individual moving range chart. If the data demonstrates a lack of stability then the process is unpredictable and any characterisation is wishful thinking.

If the data demonstrates predictability from the past then you can predict in to the future with reasonable confidence.

The best description of Cpk for me is; effective space available/ one-half of the space required = Distance to the nearer spec/ Average - Lower Natural Process Limit.

Worked Example:

Process demonstrates stability/predictability

Process Average = 0.14063"
Design specs on a fictitious dimension are an upper limit of 0.155" and Lower limit of 0.125"
Distance between the natural process limits is 0.1517 - 0.1296 = 0.0221"

Calculate Cpk
- Distance to the nearer spec = 0.155 - 0.1406 = 0.01437"

- One half of space required = Half the distance between the natural process limits 0.0221/2 = 0.01105"

Cpk = 0.0144 (rounded)/0.0115 = 1.30
 
Top Bottom