Unequal Bilateral Tolerance spreadsheet wanted

rnsvasan

Involved In Discussions
Can someone provide the formula / sheet for unequal bilateral tolerance. Is it different from the one we use for equal bilateral tolerance ? pl. clarify.

Thanks in advance
 

Miner

Forum Moderator
Leader
Admin
Please provide some additional context. It is unclear what type of worksheet you desire.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
Also, you specify the type of tolerance, which is one of the least important aspects of an SPC chart (which I assume you are looking for, since you posted in the SPC forum.) We need to know the process you have an the type of characteristic - length, diameter, form (roundness, parallelism). The distribution of the process - not the tolerance - is the most key question to answer. Tolerance only comes into play in SPC if you have a continuous uniform distribution. It also relates to capability - not control.
 

Ronen E

Problem Solver
Moderator
Can someone provide the formula / sheet for unequal bilateral tolerance. Is it different from the one we use for equal bilateral tolerance ? pl. clarify.

Thanks in advance

When the two sides of the tolerance are unequal, use the smaller one for Cpk calculations.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
When the two sides of the tolerance are unequal, use the smaller one for Cpk calculations.

Be careful that is not always true. In fact in metric tolerances it is rarely true, as the nominal can be simply the closest whole number rather than a "target". I recommend reading this thread.
 

rnsvasan

Involved In Discussions
Also, you specify the type of tolerance, which is one of the least important aspects of an SPC chart (which I assume you are looking for, since you posted in the SPC forum.) We need to know the process you have an the type of characteristic - length, diameter, form (roundness, parallelism). The distribution of the process - not the tolerance - is the most key question to answer. Tolerance only comes into play in SPC if you have a continuous uniform distribution. It also relates to capability - not control.

Sorry, I couldn't reply immediately as i was out of station.
Yes for your question, we need to plot for Outer diameter of the piston having tolerance of 50 + 0.10 / -0.15. The target value is 50 - acceptable error on plus side is 0.10 and minus side is 0.15. Then which method is to be used for calculating Cpk value.
 

Miner

Forum Moderator
Leader
Admin
The Cp/Cpk indices do not take off-center targets into consideration. They are based solely on the specifications, not the target. You simply calculate both the upper (Cpku) and lower (Cpkl) values and use the smaller of the two as Cpk.

In your situation, the Cpm index may be more relevant as it takes the actual target value into consideration.

There are also a lot of discussions on the relative merits and failings of capability indices in general if you search the cove.
 

bobdoering

Stop X-bar/R Madness!!
Trusted Information Resource
The Cp/Cpk indices do not take off-center targets into consideration. They are based solely on the specifications, not the target. You simply calculate both the upper (Cpku) and lower (Cpkl) values and use the smaller of the two as Cpk.

In your situation, the Cpm index may be more relevant as it takes the actual target value into consideration.

There are also a lot of discussions on the relative merits and failings of capability indices in general if you search the cove.

I agree, using Cpk is likely what a customer will want to see, Cpm may be academically more beneficial. My guess is if you are doing precision machining of piston ODs (grinding, CNC turning, etc.), neither is applicable - as if you are correctly controlling the process, it should not be normal - and normality is an important basis for Cpk and Cpm (see 2.2.11.5 Processes with One-Sided Specifications or Non-Normal Distributions). I recommend reading Statistical process control for precision machining. When properly controlled, you should find the the capability to be (USL-LSL)/(UCL-LCL).
 
Top Bottom