Originally Posted by
Doug Fattic
There are a couple of twists to knowing tubing thickness. 1st is manufacturing inconsistencies. I used to always weigh Columbus SL chain stays because they varied between 140 to 175 grams. I weighted them to find equal weight pairs and also to put lighter ones with lighter riders and so forth. 2nd manufacturing spaces and actual specs are often not the same. For example old SL had .9/.6/.9 wall thicknesses in their top and down tube. While generally correct, they weren't exactly those dimensions. So when looking at a tubing spec chart, it might appear that different brands had the same wall thickness for a particular model but I'm reality were a bit different.
Columbus tubes weren't even very concentric when I measured them, in the '80s. As I mic'd the tube different places around the circumference, I found different thicknesses, with a max on one side and a min on the other, like the inner and outer sufaces were truly circular, just not on the same centerline. This was repeatable, not random measurement error. So, on a couple "special" frames, I oriented the miters to put the thickest part of the tube where the forces are highest, as evidenced by where fatigue cracks are most likely to form. The differences weren't enough to make this procedure worth it generally. Sorry I don't remember now how much variance I found, but it was noticeably more than on Reynolds or Tange tubes I measured the same way.
One other framebuilder, whose name you know for sure, told me in a private email that he found the same thing, also back then. I don't know if this has any relevance to tubes made in the current millennium, probably not.