From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.racing
Subject: Re: match sprint & disk wheel
Date: 22 Nov 1996 17:28:21 GMT

Todd Marks Bauer writes:

>> That may be the motivation, but riders aren't getting any for their
>> effort.  The same goes for tying and soldering spoke crossings of
>> such wheels.  There's a lot of myth in bike racing, and some of it
>> lives on forever without a trace of truth.  The "wind up" of high
>> and low flange wheels, that can be easily computed, is so small as
>> to be practically immeasurable.  Even a radially spoked rear wheel
>> yields no perceptible compliance although it may not last long.

> Put your pencil, calculator, and micrometer away and try this:

> Tie your spokes as you would for a tied and soldered wheelset.  No
> need to solder them.  Then go out and ride those wheels.

> My real world experience is that the difference is very noticeable.

What is it that you believe you feel?  I think it is apparent that
tying can have no radial effect on the wheel and that even it could
produced a 20% change in elastic response (about 0.001 inch) no one
could detect such a difference.  That leaves lateral and torsional,
the modes that were measured in a precise setup on a machine tool bed.

As I said, low and high flange wheels were built by Wheelsmith, were
measured, were tied and soldered by Wheelsmith, and were measured
again.  Only tying and soldering was performed on the wheels between
the two measurements.  There was no measurable change detected in
torsional or lateral stiffness by micrometer dial gauges located
around the wheel.  The instrumentation for these measurements is shown
in detail in "the Bicycle Wheel".

That tying and soldering has no effect was no idle guess.  Having
inspected many wheels, it had become apparent that spokes at their
crossings, repeatedly make and break welds by fretting micro motions,
eroding the metal so that the crossing spokes nest into each other in
a practically motionless junction.  This can be detected by moving
spokes at their crossings after substantial use.  That being the case,
it was apparent that tying and soldering could not affect wheel
stiffness because there is no motion to be restrained at this point.

Later, I came across an article from the days of high wheelers, that
explained an important feature of Starley's invention of the cross
laced wheel.  The crossed spokes could be tied with twine to prevent
headers (end-overs) when one of the huge spokes broke.  Because all
racers subsequently had tied spokes, they kept tying them even after
high wheelers were replaced by today's chain driven bicycles.

Apparently no one seriously questioned the practice while the
purveyors of the art claimed it made a stronger wheel (at extra cost).
I don't doubt that they believed the claim but I'm less certain that
they believed it made a difference in racing performance.  It also
made spoke replacement something that usually required a shop visit.

It's like road and track glue for tubulars.  No one could explain the
reason for the two types of glue except in terms similar to the
defense of tying and soldering.  The difference is that there is a
good reason for hard glue but that got lost somewhere.  Tubulars have
relatively large viscous losses (up to 1/3 of the rolling resistance)
in the pressure sensitive road glue that enables a tire change on the
road.  Tire changes that have become wheel changes today.  For timed
events such as the kilometer or 4000 meters, the rolling friction loss
is important and can be completely avoided by using hard glue, glue
that makes an on-the-road tire change impossible.  This came to my
attention when reviewing rolling resistance graphic data from
clinchers and tubulars a few years ago.

Jobst Brandt      <jbrandt@hpl.hp.com>