Kurt Sperry writes:
>
> Wasn't the list told a year or two back that lugs perform no
> real significant structural function once a frame has been
> built up and they were essentially used only to facilitate
> building it up?
There's truth to that, but the alternative to lugs is lugless, where there will always be a fillet of brazing material, a curved transition from one tube to the next, that almost always is feathered for a smooth transition. Lugs on the other hand are 'supposed' to have edge definition, and pretty much always do, so the lug edges are stress risers that lugless frames don't have.
In a normal lug, the edge is far enough away from the joint that it doesn't cause a problem in practice. Thin tubing requires thin lugs to keep the stress riser in the safe zone, and this thinning is one of the hallmarks of an excellent framebuilder, but feathering the edge down to nothing isn't needed.
When the lug edge occurs almost right at the joint, I think we have a problem. The lug edge can't flex enough and so makes sort of a fulcrum that the tube flex gets concentrated at. I don't have any finite-element analysis or numbers to prove it, it's kind of a feeling, but it seems to be borne out in the real world. Or I could be full of it, certainly my sample size is too small to be real scientific. Other framebuilders have said similar things about lugs and stress risers, but that could just be "lore".
I think it is useful though, to think of lugs as more "evolved" than "designed", with almost-random variations, and only the fittest ones kept, over a period of many decades. Think of the forward-pointing points of the Prugnat I-series lugs, or lots of ornate English shapes that had points going forward on the HT part of the DT lug. I can think of Italian and French examples also. If the evolutionary theory of lugs has any validity, then the prevalence of these types of points might be telling us something.
Mark Bulgier
Seattle WA USA