One of the questions in bicycle designs is what parameters are desirable in a bike. Unfortunately, there is little agreement on what makes a well-riding frame.
For decades, riders and builders alike have proclaimed that frame flex is detrimental - a 1930s catalog for the Bates Cantiflex claimed that "whip, the power thief, is eliminated on a Bates" - yet riders consistently have chosen the lightweight, and thus in the days of standard tube diameters, flexible frames. Even on flat roads, few racers preferred super-stiff drainpipe frames over the top-of-the-line flexy "noodles." When my Bianchi SL was involved in a crash in 1990, a bike dealer lent me a mid-line Japanese-made (?) Bianchi made from thick-wall, straight-gauge tubing. It shifted fine, the brakes worked OK, even the tires stuck to the road fine, but I found that this bike did not respond to my input like the SL frame, even on flat roads. I used it for training, but for racing, I borrowed my girl-friend's Peugeot made from Reynolds 531, which, even though it was 3 cm too small for me, worked much better (with a very long seatpost to get the saddle to the correct height).
In fact, many modern production frames made from "medium weight" OS steel have very stiff frames, yet many riders with a lot of experience on various bikes don't find their performance very impressive.
As I see it, we know which frames we like, but we don't know why. Empirical studies can help determine which design parameters give us the bikes we like.
For example, I have one bike with an OS top tube that otherwise seems to be pretty classic and lightweight, yet I and others have found that it doesn't respond well to our efforts to make it go fast. But with numerous variables, such as profile, weight and length of the chainstays, front-end geometry and more, I am hard-pressed to claim the OS top tube is the culprit. In all these comparisons, there are too many variables to isolate a factor.
The theoretical analyses can be useful if you know what you are trying to achieve. If you want the frame with the least feasible sideways BB deflection, the analyses can help you design this. But the analyses cannot tell you whether that frame will ride better than others. To simplify things, most current designers simply have assumed that you want the lightest bike with the most lateral stiffness and the most vertical compliance. Obviously, this already means everybody will weigh the inherent compromises differently, but the bigger question is whether this paradigm even is useful. What if neither the flexy noodle nor the super-stiff drainpipe bikes climbs best, but the bike that has just the right amount of lateral flex? Or what if some other parameter matters more, like the frequency of the oscillations of the top tube? I am not aware of any tests that have tried to figure this out.
When I was a scientist, we called models that were based on dubious assumptions "garbage in - garbage out models." It seems to me that many analyses of bicycle frames fall into that category.
Maybe that is why people rave about the ride of bikes that are decades old, despite the engineering resources that have been expended in the quest to make better bikes.
Jan Heine Editor/Publisher Vintage Bicycle Quarterly c/o Il Vecchio Bicycles 140 Lakeside Ave, Ste. C Seattle WA 98122 http://www.vintagebicyclepress.com
At 9:33 AM -0700 6/16/06, dave bohm wrote:
>The science of finite analysis, structural stiffness and stress is
>certainly too involved to adequately discuss on this list.
>
>I respectfully add my opinion that a test of bicycle frames, identical
>in all ways except the top tube is not the best use of financial
>resources or physical effort.
>
>Certainly, various previous studies have shown the generalized average
>of stresses as they concern the top tube and although only empirical
>data would substantiate this completely, in my own studies, actual
>deflections have closely mirrored that of the finite analysis
>conclusions.
>
>Would the top tube diameter and thickness add or subtract to the overall
>stiffness of a bicycle frame...absolutely.
>
>For unit weight would it contribute to the overall stiffness and load as
>efficiently as the downtube. No.
>
>Theses could be written on this and they have. I personally have been
>part of three graduate studies of similar items and this has shown the
>generalized stresses fairly conclusively. Both through finite analysis
>and empirical deflection testing.
>
>Maybe, an effort like this would be better put to use studying other
>aspects of frame design as the top tube is just not a major design
>element in modern bicycles.
>
>I would normally back this argument with empirical data, but it would
>require many pages of explanation.
>
>Sincerely,
>
>Dave Bohm
>Bohemian Bicycles