I doubt any professional frame builders actually use the intersection of chainstay and seatstay, because it is very impractical. For just one example of a problem with it, on most bikes it is impossible to put the axle there. It is a rare coincidence when those three lines (the third being the center of the dropout slot) meet at a point. Another problem is in calculating the cut length for the chainstay. Imagine a size run of a stock model: Same dropout, but you have to cut the chainstay to a different length for each framesize, due to the varying seatstay angle moving the intersection point? Not practical. The alternative is cut the stays all the same length, but then the measurement (by the intersection method) is different for each framesize - looks funny on the geometry chart - hard to explain to the customer why the small frames have longer chainstays than the large frames.
Framebuilders will always (in my experience) choose a datum or "measure-to" point for each brand/model of dropout, and always measure there regardless of the seatstay angle. The two most popular options are (1) the center of the slot and (2) the point that most people will set the axle at in use, which is forward of the halfway point and roughly above the rear derailleur mount (where "above" is taken perpendicular to the slot). Usually about a centimeter further forward than halfway, for long slot Campy 1010 dropouts.
I prefer option (2) myself but I understand it is just a convention and there's nothing magical about it. It does have the advantage of being where most people will put the axle, thus it will end up being the "real" chainstay length more often than option (1).
Precision here only matters to the framebuilder, who has to miter the tubes and set the jigs. To the user, a centimeter one way or the other is of little matter, as Jan Heine and others have stated.
Mark Bulgier Seattle WA USA
Dale wrote:
> > << I tend to like the stay intersection method for both chainstay
> > and wheelbase, but that's really just because it gives me a way to
> > find the measurement point, and to be consistent in my own work. >>
> >
> > That makes sense.
> > The only problem I could possible see with that seat & chain
> > stay-intersection idea, is that on extremely large and small frames
> > those seat stays may point to strange places (so to speak).
> > But I haven't heard a better approach!