I measure IP on all my blanks.
If you build a number of the same rods, how much deviation would you expect between blanks?
I haven't established the variance due to my technique so these are just some data points.
(I do have limited data on this but not enough, and plan to in the future).

Rod 1. IP ave. 443 stdev of 31. 5 rods. Range 415 to 479
Rod 2. IP ave 537 Stdev of 29. 7 rods. Range 515 to 595
Rod 3. IP ave 684 stdev of 26. 6 rods. Range 646 to 719

This data is from several years.
I'm a little surprised at the range.
Assuming that these are all the correct blanks can we expect such variance between batches?
Lol, I hate assuming anything!

TJB

Please clarify what you are reporting. Does "Rod 1" mean a number of examples of a certain rod model? Rod 2 a number of examples of another rod model? Etc?

Assuming that to be the case:

I think your range is too high to attribute it all to blank variation. Based my experience my most likely source of error is in measuring the exact point of 1/3 rod length deflection.

I'm not a statistition, so cannot comment with creds on the data, but my conclusion comes from the range, which seems high. On the other hand , the standard deviation seems reasonable. However, I still maintain that varitaion in data in CCS will come from both blank variation and test error.

Yes on the blanks being all the same model and manufacturers.
I use the same setup to measure so being off by more than several grams is unlikely but possible
as we all make errors from time to time.
The sample size is small as well, one mislabeled blank would through a wrench into the works.

Really just wanted to see how much variance in IP we can expect.

TJB