University frame-flex testing: surprising results

Thomas Hood said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Thomas Hood

Hmmm...seems to me that they should have hung the weight offset from
the BB to create a moment about the BB rather than in line with the
axle centerline and better simulate actual loading...

Tom_A said:

Thomas Hood said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Thomas Hood

Hmmm...seems to me that they should have hung the weight offset from
the BB to create a moment about the BB rather than in line with the
axle centerline and better simulate actual loading...

bolting the dropouts to the table creates a restraint which isn't realistic

--
/Marten

info(apestaartje)m-gineering(punt)nl

Thomas Hood said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Thomas Hood

Dear Thomas,

As usual, the measured flex was undetectable by any rider.

The flex in the no-fork column was from 1.47 mm to 2.46 mm, a range of
0.99 mm. At most, the flex was less than a tenth of an inch.

The flex in the with-fork column was from 1.30 mm to 1.99 mm, a range
of 0.69 mm. At most, the flex was just over a sixteenth of an inch.

Damon Rinard found similar results when he tested dozens of wheels:
http://www.sheldonbrown.com/rinard/wheel/index.htm

For steel spokes, practically all wheels that Rinard tested deflected
about 1~3 mm, depending on spoke count and whether they were front or
rear.

Wheels with struts or carbon fiber spokes usually deflected another
millimeter, but the enormous effort of testing all those wheels
basically showed that there was no significant difference.

Cheers,

Carl Fogel

M-gineering said:

Tom_A said:

Thomas  Hood said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Quoted message said:

Quoted message said:

Thomas Hood

Quoted message said:

Hmmm...seems to me that they should have hung the weight offset from
the BB to create a moment about the BB rather than in line with the
axle centerline and better simulate actual loading...

bolting the dropouts to the table creates a restraint which isn't realistic

--
/Marten

info(apestaartje)m-gineering(punt)nl

The German Fatigue test agency have been doing this for sometime.
Cannondales frequently top the list of stiffest frames. True,
deflections are so small to characterize but the industry is making a
big issue out of it. Its a clear selling point.

B.D

http://cozybeehive.blogspot.com

Quoted message said:

This test is eye catching but it is not what occurs when riding. I am
not sure what parameter is best to assess frame rigidity but for my
use, pedaling while standing is the mode in which I discovered where
most of the deflection occurred. It was the Cinelli quill stem. With
a threadless steertube and tubular bar stem, nearly all the flex
vanished. That is, my handlebars no loner seemed to be attached
elastically to the bicycle.

As you've stated previously, the handlebars on your bike are narrow
steel Cinelli items from the 60's. In this respect your bike is
different to 99% of those in this group. How thick are the walls on
this bar, and how wide is it, as I'm interested in whether this bar is
dramatically stiffer than an alloy bar. The bar on my bicycle is a
relatively hefty 42cm 300 g plain-gauge (1.7mm) 3T sleeved (0.7mm)
item, and I can certainly feel the bar twist between the brake hoods
when standing.

If you had to purchase a handlebar now, you would not be able to
purchase a steel handlebar. Given this constraint, is the 31.8 mm
oversize standard ( http://tinyurl.com/28a8up ) better from a
durability point of view?

Quoted message said:

Another parameter I find interesting is the distance between front
axle and BB when braking hard. Much of that change occurs in the the
fork steertube and hardly any in bending the downtube. Rigidity in
this mode improves rider confidence in descending because front wheel
speed is consistent with the bicycle.

Again, here your bike is atypical being so large. From pictures of
your bike it appears that the distance between the head bearings is
~3/4 of the length of you forks (on my bike it is ~ 1/3)

If your frame and forks are built from materials similar to these is
the flex really mainly in the steertube rather than in the blades?
http://www.reynoldscycles.co.uk/downloads/PARTLIST.pdf (see page 4)

Assuming a 2.1mm steertube and oval 0.9mm blades of 27.5 x 20 can
anyone approximate the bending stiffness of the blades vs. the
steertube?
(I'm not being idle, but I couldn't find a the formula for 'bending
stiffness of a tube' despite Googling...)

Would a 1-1/8" steertube not have been a better choice on your bike,
or was this all that was available in 19xx ? Would you build a bike
your size of the now standard steel tubeset dimensions: (34.9mm HT,
31.8mm DT, 28.6mm TT & ST)

Thanks,

Thomas Hood

Quoted message said:

On Wed, 16 Jan 2008 10:32:04 -0800 (PST), Thomas Hood

Quoted message said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Quoted message said:

Thomas Hood

Dear Thomas,

As usual, the measured flex was undetectable by any rider.

The flex in the no-fork column was from 1.47 mm to 2.46 mm, a range of
0.99 mm. At most, the flex was less than a tenth of an inch.

The flex in the with-fork column was from 1.30 mm to 1.99 mm, a range
of 0.69 mm. At most, the flex was just over a sixteenth of an inch.

Damon Rinard found similar results when he tested dozens of wheels:
http://www.sheldonbrown.com/rinard/wheel/index.htm

For steel spokes, practically all wheels that Rinard tested deflected
about 1~3 mm, depending on spoke count and whether they were front or
rear.

Wheels with struts or carbon fiber spokes usually deflected another
millimeter, but the enormous effort of testing all those wheels
basically showed that there was no significant difference.

Cheers,

Carl Fogel

Carl,

You're preaching to the choir... It was the relative placings I found
surprising, particularly the Alan/ GT.

Regards,

Thomas Hood

Thomas Hood said:

Quoted message said:

This test is eye catching but it is not what occurs when riding. I
am not sure what parameter is best to assess frame rigidity but for
my use, pedaling while standing is the mode in which I discovered
where most of the deflection occurred. It was the Cinelli quill
stem. With a threadless steertube and tubular bar stem, nearly all
the flex vanished. That is, my handlebars no loner seemed to be
attached elastically to the bicycle.

Quoted message said:

As you've stated previously, the handlebars on your bike are narrow
steel Cinelli items from the 60's. In this respect your bike is
different to 99% of those in this group. How thick are the walls on
this bar, and how wide is it, as I'm interested in whether this bar
is dramatically stiffer than an alloy bar. The bar on my bicycle is
a relatively hefty 42cm 300 g plain-gauge (1.7mm) 3T sleeved (0.7mm)
item, and I can certainly feel the bar twist between the brake hoods
when standing.

Quoted message said:

If you had to purchase a handlebar now, you would not be able to
purchase a steel handlebar. Given this constraint, is the 31.8 mm
oversize standard ( http://tinyurl.com/28a8up ) better from a
durability point of view?

That's not germane, the pint is that I achieved a great improvement in
rigidity, one that was entirely the quill stem replacement. If I had
ridden with wider bars, the effect would probably been greater. I
found this improvement especially good because I undertook it to get
rid of frozen stems in the steertube, something that occurred even
though there was an O-ring seal on the head bearing lock nut.

Quoted message said:

Quoted message said:

Another parameter I find interesting is the distance between front
axle and BB when braking hard. Much of that change occurs in the
the fork steertube and hardly any in bending the downtube.
Rigidity in this mode improves rider confidence in descending
because front wheel speed is consistent with the bicycle.

Quoted message said:

Again, here your bike is atypical being so large. From pictures of
your bike it appears that the distance between the head bearings is
~3/4 of the length of you forks (on my bike it is ~ 1/3)

Quoted message said:

If your frame and forks are built from materials similar to these is
the flex really mainly in the steertube rather than in the blades?

http://www.reynoldscycles.co.uk/downloads/PARTLIST.pdf (see page 4)

Probably, considering the fork blade fore and aft width compared to
the steer tube. The point is that you can rock a bicycle fore and aft
with the front brake clamped and watch that motion. Viscount bicycles
were miserable at that, and I never saw a tall frame of these.

Quoted message said:

Assuming a 2.1mm steertube and oval 0.9mm blades of 27.5 x 20 can
anyone approximate the bending stiffness of the blades vs. the
steertube?

Quoted message said:

(I'm not being idle, but I couldn't find a the formula for 'bending
stiffness of a tube' despite Googling...)

Timoshenko has all those formulas in profusion, however for rectangular
cross sections the bending stiffness goes as the third power of the
bending height and I'm sure that oval cross sections are somewhere in
between the second and third power depending on shape.

Quoted message said:

Would a 1-1/8" steertube not have been a better choice on your bike,
or was this all that was available in 19xx ? Would you build a bike
your size of the now standard steel tubeset dimensions: (34.9mm HT,
31.8mm DT, 28.6mm TT & ST)

I don't need not steenkin fat tube. Mine works well enough as I brake
hard enough to raise the rear wheel on my long wheelbase bicycle.

Jobst Brandt

Thomas Hood said:

Quoted message said:

This test is eye catching but it is not what occurs when riding.  I am
not sure what parameter is best to assess frame rigidity but for my
use, pedaling while standing is the mode in which I discovered where
most of the deflection occurred.  It was the Cinelli quill stem.  With
a threadless steertube and tubular bar stem, nearly all the flex
vanished.  That is, my handlebars no loner seemed to be attached
elastically to the bicycle.

As you've stated previously, the handlebars on your bike are narrow
steel Cinelli items from the 60's. In this respect your bike is
different to 99% of those in this group. How thick are the walls on
this bar, and how wide is it, as I'm interested in whether this bar is
dramatically stiffer than an alloy bar. The bar on my bicycle is a
relatively hefty 42cm 300 g plain-gauge (1.7mm) 3T sleeved (0.7mm)
item, and I can certainly feel the bar twist between the brake hoods
when standing.

If you had to purchase a handlebar now, you would not be able to
purchase a steel handlebar.

http://www.yellowjersey.org/nitto.html

Given this constraint, is the 31.8 mm

Quoted message said:

oversize standard (http://tinyurl.com/28a8up) better from a
durability point of view?

Quoted message said:

Another parameter I find interesting is the distance between front
axle and BB when braking hard.  Much of that change occurs in the the
fork steertube and hardly any in bending the downtube.  Rigidity in
this mode improves rider confidence in descending because front wheel
speed is consistent with the bicycle.

Again, here your bike is atypical being so large. From pictures of
your bike it appears that the distance between the head bearings is
~3/4 of the length of you forks (on my bike it is ~ 1/3)

If your frame and forks are built from materials similar to these is
the flex really mainly in the steertube rather than in the blades?http://www.reynoldscycles.co.uk/downloads/PARTLIST.pdf(see page 4)

Assuming a 2.1mm steertube and oval 0.9mm blades of 27.5 x 20 can
anyone approximate the bending stiffness of the blades vs. the
steertube?
(I'm not being idle, but I couldn't find a the formula for 'bending
stiffness of a tube' despite Googling...)

Would a 1-1/8" steertube not have been a better choice on your bike,
or was this all that was available in 19xx ?  Would you build a bike
your size of the now standard steel tubeset dimensions: (34.9mm HT,
31.8mm DT, 28.6mm TT & ST)

Thanks,

Thomas Hood

Quoted message said:

Thomas Hood said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Thomas Hood

Dear Thomas,

As usual, the measured flex was undetectable by any rider.

I'm having trouble finding where it says this.

-snip-

Thomas Hood said:

If you had to purchase a handlebar now, you would not be able to
purchase a steel handlebar.

-snip-

Why is that?
--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971

Thomas Hood said:

Quoted message said:

This test is eye catching but it is not what occurs when riding. I am
not sure what parameter is best to assess frame rigidity but for my
use, pedaling while standing is the mode in which I discovered where
most of the deflection occurred. It was the Cinelli quill stem. With
a threadless steertube and tubular bar stem, nearly all the flex
vanished. That is, my handlebars no loner seemed to be attached
elastically to the bicycle.

As you've stated previously, the handlebars on your bike are narrow
steel Cinelli items from the 60's. In this respect your bike is
different to 99% of those in this group. How thick are the walls on
this bar, and how wide is it, as I'm interested in whether this bar is
dramatically stiffer than an alloy bar. The bar on my bicycle is a
relatively hefty 42cm 300 g plain-gauge (1.7mm) 3T sleeved (0.7mm)
item, and I can certainly feel the bar twist between the brake hoods
when standing.

If you had to purchase a handlebar now, you would not be able to
purchase a steel handlebar. Given this constraint, is the 31.8 mm
oversize standard ( http://tinyurl.com/28a8up ) better from a
durability point of view?

for the same wall thickness, of course.

Quoted message said:

Quoted message said:

Another parameter I find interesting is the distance between front
axle and BB when braking hard. Much of that change occurs in the the
fork steertube and hardly any in bending the downtube. Rigidity in
this mode improves rider confidence in descending because front wheel
speed is consistent with the bicycle.

Again, here your bike is atypical being so large. From pictures of
your bike it appears that the distance between the head bearings is
~3/4 of the length of you forks (on my bike it is ~ 1/3)

If your frame and forks are built from materials similar to these is
the flex really mainly in the steertube rather than in the blades?
http://www.reynoldscycles.co.uk/downloads/PARTLIST.pdf (see page 4)

Assuming a 2.1mm steertube and oval 0.9mm blades of 27.5 x 20 can
anyone approximate the bending stiffness of the blades vs. the
steertube?
(I'm not being idle, but I couldn't find a the formula for 'bending
stiffness of a tube' despite Googling...)

Would a 1-1/8" steertube not have been a better choice on your bike,

of course. oversize tube is absolutely the way to go. it helps with
fatigue and it helps with stiffness, [and therefore shimmy] - good for
bikes with dished rear wheels.

Quoted message said:

or was this all that was available in 19xx ? Would you build a bike
your size of the now standard steel tubeset dimensions: (34.9mm HT,
31.8mm DT, 28.6mm TT & ST)

didn't have dished rear wheels back in those days, yet jobst's frame
shimmys "at will". stupid to keep riding that when a better mechanical
solution is only a few hundred dollars away.

[email hidden] wrote:
<snip irrelevance>

Quoted message said:

I don't need not steenkin fat tube. Mine works well enough as I brake
hard enough to raise the rear wheel on my long wheelbase bicycle.

wow, frame tube stiffness affects braking on your single pivot
calipers??? spectacular jobst!!! absolutely freakin' spectacular.

Andrew Muzi said:

-snip-

Thomas Hood said:

If you had to purchase a handlebar now, you would not be able to
purchase a steel handlebar.

-snip-

Why is that?

Palo Alto is a special ferrous handlebar free zone, with existing
handlebars "grandfathered" in. 😉

--
Tom Sherman - Holstein-Friesland Bovinia
"And never forget, life ultimately makes failures of all people."
- A. Derleth

John Forrest Tomlinson said:

Quoted message said:

Thomas Hood said:

http://materials.open.ac.uk/bikeframes/bikeframe.htm

Thomas Hood

Dear Thomas,

As usual, the measured flex was undetectable by any rider.

I'm having trouble finding where it says this.

Dear John,

That's because I said that the measured flex was undetectable by any
rider, not the report.

The maximum measured flex amounted to less than a tenth of an inch.

As Damon Rinard's tests show, the same load will deflect the rims of
each of a rider's wire-spoked wheels roughly as much.

In actual cornering, the rubber tires will deform even more.

No rider can detect a frame smoothly flexing less than a tenth of an
inch at the bottom bracket amidst all the other things happening
during cornering.

The road surface is likely to vary considerably more than a tenth of
an inch. The rider's shoes and socks are likely to squash more than a
tenth of an inch.

The rider cannot steer to within a tenth of an inch through corner.

This does not, however, stop bicyclists from believing that they can
sense all sorts of infinitesimal differences.

Cheers,

Carl Fogel

jim beam said:

Quoted message said:

or was this all that was available in 19xx ? Would you build a bike
your size of the now standard steel tubeset dimensions: (34.9mm HT,
31.8mm DT, 28.6mm TT & ST)

didn't have dished rear wheels back in those days, yet jobst's frame
shimmys "at will". stupid to keep riding that when a better mechanical
solution is only a few hundred dollars away.

For what it's worth...

The two bikes I've been putting the most miles on these days are both
built on old-fashioned 68cm lugged steel frames with 25.4mm top tubes
and 28.6mm down tubes. One of them (a '70s Nishiki) has a clunky
straight gauge tubeset, but the other (a pre-index '80s Fuji Sagres)
is double butted and respectably light. I haven't been able to induce
shimmy in either one of them, even zooming downhill at automotive
speeds. I think the lack of shimmy is probably more attributable to
the rigid and robust componentry (Primo Powerbite BMX cranks, hefty
steel-quilled stems, dishless 48 spoke wheels with 700x40 tires) than
to any characteristics of the frames.

As a point of comparison, the last bike I had that shimmied was built
on a massive straight gauge 68cm Cannondale touring frame with a 2"
downtube and oval seat stays-- probably the stiffest commercially
available road bike frame I have ever seen, let alone ridden. That
bike had much more flexible parts attached to it, though (aluminum
quill stem, drop bars, 36 spoke wheels with 126mm 7 speed rear, square
taper crank).

So if Jobst's bike really does shimmy chronically, my guess is that it
has at least as much to do with the MA2 rims he uses as with the
diameters of tubing in his frame.

Chalo

Chalo said:

jim beam said:

Quoted message said:

or was this all that was available in 19xx ? Would you build a bike
your size of the now standard steel tubeset dimensions: (34.9mm HT,
31.8mm DT, 28.6mm TT & ST)

didn't have dished rear wheels back in those days, yet jobst's frame
shimmys "at will". stupid to keep riding that when a better mechanical
solution is only a few hundred dollars away.

For what it's worth...

The two bikes I've been putting the most miles on these days are both
built on old-fashioned 68cm lugged steel frames with 25.4mm top tubes
and 28.6mm down tubes. One of them (a '70s Nishiki) has a clunky
straight gauge tubeset, but the other (a pre-index '80s Fuji Sagres)
is double butted and respectably light. I haven't been able to induce
shimmy in either one of them, even zooming downhill at automotive
speeds. I think the lack of shimmy is probably more attributable to
the rigid and robust componentry (Primo Powerbite BMX cranks, hefty
steel-quilled stems, dishless 48 spoke wheels with 700x40 tires) than
to any characteristics of the frames.

As a point of comparison, the last bike I had that shimmied was built
on a massive straight gauge 68cm Cannondale touring frame with a 2"
downtube and oval seat stays-- probably the stiffest commercially
available road bike frame I have ever seen, let alone ridden. That
bike had much more flexible parts attached to it, though (aluminum
quill stem, drop bars, 36 spoke wheels with 126mm 7 speed rear, square
taper crank).

So if Jobst's bike really does shimmy chronically, my guess is that it
has at least as much to do with the MA2 rims he uses as with the
diameters of tubing in his frame.

the two factors, wheels and frame, work together. wheels that will
shimmy on a flexible frame, won't on a stiff one. flexible wheels on a
shimmy frame make it shimmy bad. stiff wheels on a shimmy frame make
shimmy much harder to induce.

Chalo Colina said:

Quoted message said:

Quoted message said:

or was this all that was available in 19xx? Would you build a
bike your size of the now standard steel tubeset dimensions:
(34.9mm HT, 31.8mm DT, 28.6mR?m TT & ST)

Quoted message said:

Quoted message said:

didn't have dished rear wheels back in those days, yet jobst's
frame shimmys "at will". stupid to keep riding that when a better
mechanical solution is only a few hundred dollars away.

Quoted message said:

For what it's worth...

Quoted message said:

The two bikes I've been putting the most miles on these days are
both built on old-fashioned 68cm lugged steel frames with 25.4mm top
tubes and 28.6mm down tubes. One of them (a '70s Nishiki) has a
clunky straight gauge tubeset, but the other (a pre-index '80s Fuji
Sagres) is double butted and respectably light. I haven't been able
to induce shimmy in either one of them, even zooming downhill at
automotive speeds. I think the lack of shimmy is probably more
attributable to the rigid and robust componentry (Primo Powerbite
BMX cranks, hefty steel-quilled stems, dishless 48 spoke wheels with
700x40 tires) than to any characteristics of the frames.

Quoted message said:

As a point of comparison, the last bike I had that shimmied was
built on a massive straight gauge 68cm Cannondale touring frame with
a 2" downtube and oval seat stays-- probably the stiffest
commercially available road bike frame I have ever seen, let alone
ridden. That bike had much more flexible parts attached to it,
though (aluminum quill stem, drop bars, 36 spoke wheels with 126mm 7
speed rear, square taper crank).

Quoted message said:

So if Jobst's bike really does shimmy chronically, my guess is that
it has at least as much to do with the MA2 rims he uses as with the
diameters of tubing in his frame.

I shimmies when I let it, while riding no-hands at around 23MPH and
up. That the wheels are not the source was determined by filling the
front tier completely with water, which made no difference, and
changing the tire and tube onto a Rolf Vector Pro, and a Bontrager
paired spoked wheel. None of these made any difference. When I
reported this to Damon Rinard, he repeated the experiment at his lab
with instrumentation that showed that shimmy buildup was the same and
that frequency was identical. He used an accelerometer on the side of
the headtube, the part of the bicycle that shimmies.

http://www.sheldonbrown.com/brandt/shimmy.html

Jobst Brandt

Quoted message said:

Chalo Colina said:

Quoted message said:

> or was this all that was available in 19xx? Would you build a
> bike your size of the now standard steel tubeset dimensions:
> (34.9mm HT, 31.8mm DT, 28.6mR?m TT & ST)

Quoted message said:

Quoted message said:

didn't have dished rear wheels back in those days, yet jobst's
frame shimmys "at will". stupid to keep riding that when a better
mechanical solution is only a few hundred dollars away.

Quoted message said:

For what it's worth...

Quoted message said:

The two bikes I've been putting the most miles on these days are
both built on old-fashioned 68cm lugged steel frames with 25.4mm top
tubes and 28.6mm down tubes. One of them (a '70s Nishiki) has a
clunky straight gauge tubeset, but the other (a pre-index '80s Fuji
Sagres) is double butted and respectably light. I haven't been able
to induce shimmy in either one of them, even zooming downhill at
automotive speeds. I think the lack of shimmy is probably more
attributable to the rigid and robust componentry (Primo Powerbite
BMX cranks, hefty steel-quilled stems, dishless 48 spoke wheels with
700x40 tires) than to any characteristics of the frames.

Quoted message said:

As a point of comparison, the last bike I had that shimmied was
built on a massive straight gauge 68cm Cannondale touring frame with
a 2" downtube and oval seat stays-- probably the stiffest
commercially available road bike frame I have ever seen, let alone
ridden. That bike had much more flexible parts attached to it,
though (aluminum quill stem, drop bars, 36 spoke wheels with 126mm 7
speed rear, square taper crank).

Quoted message said:

So if Jobst's bike really does shimmy chronically, my guess is that
it has at least as much to do with the MA2 rims he uses as with the
diameters of tubing in his frame.

I shimmies when I let it, while riding no-hands at around 23MPH and
up. That the wheels are not the source was determined by filling the
front tier completely with water, which made no difference,

filling tires with water doesn't affect wheel stiffness jobst, so of
course not!!!

Quoted message said:

and
changing the tire and tube onto a Rolf Vector Pro, and a Bontrager
paired spoked wheel. None of these made any difference.

see above.

Quoted message said:

When I
reported this to Damon Rinard, he repeated the experiment at his lab
with instrumentation that showed that shimmy buildup was the same and
that frequency was identical. He used an accelerometer on the side of
the headtube, the part of the bicycle that shimmies.

of course not. see above.

Quoted message said:

http://www.sheldonbrown.com/brandt/shimmy.html

yet more drivel that needs correction. when jobst? when?

Chalo Colina said:

Quoted message said:

Quoted message said:

or was this all that was available in 19xx? Would you build a
bike your size of the now standard steel tubeset dimensions:
(34.9mm HT, 31.8mm DT, 28.6mR?m TT & ST)

Quoted message said:

Quoted message said:

didn't have dished rear wheels back in those days, yet jobst's
frame shimmys "at will". stupid to keep riding that when a better
mechanical solution is only a few hundred dollars away.

Quoted message said:

For what it's worth...

Quoted message said:

The two bikes I've been putting the most miles on these days are
both built on old-fashioned 68cm lugged steel frames with 25.4mm top
tubes and 28.6mm down tubes. One of them (a '70s Nishiki) has a
clunky straight gauge tubeset, but the other (a pre-index '80s Fuji
Sagres) is double butted and respectably light. I haven't been able
to induce shimmy in either one of them, even zooming downhill at
automotive speeds. I think the lack of shimmy is probably more
attributable to the rigid and robust componentry (Primo Powerbite
BMX cranks, hefty steel-quilled stems, dishless 48 spoke wheels with
700x40 tires) than to any characteristics of the frames.

Quoted message said:

As a point of comparison, the last bike I had that shimmied was
built on a massive straight gauge 68cm Cannondale touring frame with
a 2" downtube and oval seat stays-- probably the stiffest
commercially available road bike frame I have ever seen, let alone
ridden. That bike had much more flexible parts attached to it,
though (aluminum quill stem, drop bars, 36 spoke wheels with 126mm 7
speed rear, square taper crank).

Quoted message said:

So if Jobst's bike really does shimmy chronically, my guess is that
it has at least as much to do with the MA2 rims he uses as with the
diameters of tubing in his frame.

It shimmies when I let it, while riding no-hands at around 23MPH and
up. That the wheels are not the source was determined by filling the
front tire completely with water, which made no difference, and
changing the tire and tube onto a Rolf Vector Pro, and a Bontrager
paired spoked wheel. None of these made any difference. When I
reported this to Damon Rinard, he repeated the experiment at his lab
with instrumentation that showed that shimmy buildup was the same and
that frequency was identical. He used an accelerometer on the side of
the headtube, the part of the bicycle that shimmies.

http://www.sheldonbrown.com/brandt/shimmy.html

Jobst Brandt