Where's it mandated to use a 300 piece run for a capability study? I have been going through the AIAG books and cannot find the reference?

If I remember correctly, the PPAP manual requires a significant production run of 300 pieces to produce PPAP samples. The capability study requirement is 100 pieces from that 300 piece run.

The default requirement (found in the 3rd Edition PPAP manual) is a minimum 300-piece production run. For the "initial process study" the default requirement (found in the same place) is x-bar/R charting with a minimum of 25 subgroups and 100 data points with sampling done during the PPAP production run. Either requirement may be changed to suit the circumstances, with customer approval. The expectations should always be discussed with the customer when the default requirements don't make sense.

Thanks, but where is it stated in the PPAP manual? My eyes have glazed over trying to find the "shall" or direct reference.

Thanks, but where is it stated in the PPAP manual? My eyes have glazed over trying to find the "shall" or direct reference.

Sorry, but at the moment I don't have a manual handy. Look for "Significant Production Run" and "Initial Process Study" in the first 7 or 8 pages of the manual.

Sorry, but at the moment I don't have a manual handy. Look for "Significant Production Run" and "Initial Process Study" in the first 7 or 8 pages of the manual.

The 100 piece requirement was on page 6. The significant production run of 300 was before that.

There are in fact 2 separate requirements
First a significant production run which is defined on page 2 as "shall be from one to eight hours of production ....a minimum of 300 consecutive(emphasis is original) parts.
Second page 6 at least 100 readings

The only connection between the 2 numbers is in Note 4 to I.2.2.9 where the 100 pieces for the capability study are from the "significant production run"

There are in fact 2 separate requirements
First a significant production run which is defined on page 2 as "shall be from one to eight hours of production ....a minimum of 300 consecutive(emphasis is original) parts.
Second page 6 at least 100 readings

The only connection between the 2 numbers is in Note 4 to I.2.2.9 where the 100 pieces for the capability study are from the "significant production run"

Howard,

Thanks. :thanx: It was just TOO obvious to find on page 1. Often I find the AIAG manuals are a bit opaque and feel more like I am reading the federal register. I am dealing with low volume production parts for automotive aftermarket accessories sold to the OEM and 300 piece runs pretty much equal annual production volumes so we're going to revert to 30 piece initial studies using XmR charts to plot Ppk.

Howard,

Thanks. :thanx: It was just TOO obvious to find on page 1. Often I find the AIAG manuals are a bit opaque and feel more like I am reading the federal register. I am dealing with low volume production parts for automotive aftermarket accessories sold to the OEM and 300 piece runs pretty much equal annual production volumes so we're going to revert to 30 piece initial studies using XmR charts to plot Ppk.

May not be relevant, but the Truck OEM rules within the PPAP manual Section II.4.3 & .4 might be an aid to negotiations to the customer as "reasonable" in this circumstance?

Martin
Older but not getting any Wiser!

Howard,

Thanks. :thanx: It was just TOO obvious to find on page 1. Often I find the AIAG manuals are a bit opaque and feel more like I am reading the federal register. I am dealing with low volume production parts for automotive aftermarket accessories sold to the OEM and 300 piece runs pretty much equal annual production volumes so we're going to revert to 30 piece initial studies using XmR charts to plot Ppk.
:bigwave:
In general and not specific at all, you are just the opportunity to say it
The answers are all there it is just a matter of looking. It is amazing what you find when you actually read the standard/handbook etc.

:bigwave:
In general and not specific at all, you are just the opportunity to say it
The answers are all there it is just a matter of looking. It is amazing what you find when you actually read the standard/handbook etc.

I knew I had read it before and tried to find it, but it was "hidden" on page 1.

.....is x-bar/R charting with a minimum of 25 subgroups and 100 data points with sampling done during the PPAP production run....

I have never understood the requirement for x-bar/r with subgroups of 4.

It seems to ignore rational subgrouping. In my process at least we collect all the PPAP run parts, so I use X and mR charts. I lost a pitched battle with a Chrysler SQE and finally had to do it by the book.

I did extensive comparisons with the data using both methods and the conclusions drawn were the same using either method.

I could not get an explanation why I should arbitrariy grouping 4 parts together that were not collected as a sample of 4. The best he could do was "the book says so".

Anyway the customer is always right I guess.

I have never understood the requirement for x-bar/r with subgroups of 4.

It seems to ignore rational subgrouping. In my process at least we collect all the PPAP run parts, so I use X and mR charts. I lost a pitched battle with a Chrysler SQE and finally had to do it by the book.

I did extensive comparisons with the data using both methods and the conclusions drawn were the same using either method.

I could not get an explanation why I should arbitrariy grouping 4 parts together that were not collected as a sample of 4. The best he could do was "the book says so".

Anyway the customer is always right I guess.
The book says nothing specific about subgroup size other than it can't be less than 4. It also doesn't say that the number of data points must be 100. If you want subgroups of 5, it just means that you increase the number of individuals to 125 and do 25 5-piece subgroups.

I have never understood the requirement for x-bar/r with subgroups of 4.

It seems to ignore rational subgrouping. In my process at least we collect all the PPAP run parts, so I use X and mR charts. I lost a pitched battle with a Chrysler SQE and finally had to do it by the book.

I did extensive comparisons with the data using both methods and the conclusions drawn were the same using either method.

I could not get an explanation why I should arbitrariy grouping 4 parts together that were not collected as a sample of 4. The best he could do was "the book says so".

Anyway the customer is always right I guess.

I am afraid its as scientific as 100 measurements divided by 25 subgroups equals sample size of four. They may not understand rational subgrouping, but you cannot fault their math.