tire pressure does not significantly affect rolling resistance.

"wle" <[email hidden]> wrote in news:1163112881.671290.44800
@m7g2000cwm.googlegroups.com:

Quoted message said:

---Tire pressure has only a small effect on the rolling resistance of
most tires. Narrow 23-mm tires seem to roll fastest at pressures of 105
psi (7.2 bar) or more. However, running these tires at 85 psi (5.8 bar)
for improved comfort increased the test times only 2%. Wider 28-mm
tires are as fast at 85 psi as they are at higher pressures.

Your posting gave very little real data except for the above 2% number.
2% sounds like a huge amount to me. In a 100 mile time trial race, you will
be 2 miles behind due to air pressure alone. That is a difference of several
minutes, which could easily be the difference between first place and last
place in the race. Unless your math is off by a full decimal place
somewhere, I'm not going to reduce my pressure to 85psi any time soon.

"sally" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

"wle" <[email hidden]> wrote in news:1163112881.671290.44800
@m7g2000cwm.googlegroups.com:

Quoted message said:

---Tire pressure has only a small effect on the rolling resistance of
most tires. Narrow 23-mm tires seem to roll fastest at pressures of
105
psi (7.2 bar) or more. However, running these tires at 85 psi (5.8
bar)
for improved comfort increased the test times only 2%. Wider 28-mm
tires are as fast at 85 psi as they are at higher pressures.

Your posting gave very little real data except for the above 2%
number.
2% sounds like a huge amount to me. In a 100 mile time trial race,
you will
be 2 miles behind due to air pressure alone. That is a difference of
several
minutes, which could easily be the difference between first place and
last
place in the race. Unless your math is off by a full decimal place
somewhere, I'm not going to reduce my pressure to 85psi any time soon.

I'm sure the test times were the result of rolling resistance only so
the 2% increase is realistic. In reality, rolling resistance will only
account for a smaller percentage of the total drag force on a rider so
it's not going to be as big as the difference which you infer.

Phil H

In article <[email hidden]>,

wle said:

---With roughly the same power output, the rider's speed can vary by as
much as 20% depending on tire choice. For example, the rider on the
fastest tire [in this roll-down test] moved down the road at
approximately 16.4 mph (26.2 kph) while the same rider on the slowest
tire went approximately 13.6 mph (21.7 kph).

A 20% difference in speed among similar sized tires? I don't believe it.

no, they are saying it;s all rolling resistance that makes tires 'roll
faster'.

===
---Perhaps the most important result of the test is that tire pressure
does not significantly affect rolling resistance. Wide tires in
particular do not need high pressures to roll fast. But because many
current wide tires are designed to handle high pressure, they have
strong casings that lack suppleness. This results in higher rolling
resistance than necessary.
===

i guess what i don;t understand is this:

if the rolling resistance is the only thing that is different, and it;s
such a small percentage of overall force [they mention 16mph as a
ballpark speed] then how can one tire be 'faster' than another?

do these 'supple' tires magically lower air resistance somehow?

what gives?

the last paragraph...
====
---The test's findings point to a new direction for performance
bicycles. For most cyclists, wide, supple tires at low pressures offer
more speed, better comfort, increased versatility and improved safety
than today's narrow high-pressure tires. However, this type of wide,
fast tire currently is not available. Hopefully, these test results
will help persuade manufacturers to produce them.
====
....makes it sound like a thinly veiled commercial for a new magic tire
shortly to be foisted on the public

wle

Larry Dickman said:

In article <[email hidden]>,

wle said:

---With roughly the same power output, the rider's speed can vary by as
much as 20% depending on tire choice. For example, the rider on the
fastest tire [in this roll-down test] moved down the road at
approximately 16.4 mph (26.2 kph) while the same rider on the slowest
tire went approximately 13.6 mph (21.7 kph).

A 20% difference in speed among similar sized tires? I don't believe it.

It doesn't say they are similar sized. I also note that the fastest
tire in the test is a 700x23. That's pretty skinny.

Pete

Quoted message said:

A 20% difference in speed among similar sized tires? I don't believe it.

ya
doesn;t make sense

plus they talk about a 'roll down test' but then also mention
'different riders with the same power'
doesn;t roll down mean just coasting?

since they say different riders, and coasting, surely they can;t be
comparing riders of different weights?

i tried to write to the magazine with some questions

wle.

wle who? said:

from this week's "road bike rider" email distribution....

Quoted message said:

---------

Quoted message said:

1. Weekly Dispatch o^o o^o o^o o^o o^o o^o

Quoted message said:

We heard the buzz about a surprising new tire test in Bicycle
Quarterly, a nifty magazine published in Seattle and edited by Jan
Heine. The name was Vintage Bicycle Quarterly until recently, but
"vintage" has been axed because it implied the mag was about old,
collectible bikes and equipment.

Quoted message said:

Well, plenty of pages are devoted to arcane and interesting gear,
history and randonneur-style riding, but Bicycle Quarterly also
publishes cutting-edge material. The tire test is testimony, appearing
in the Autumn 2006 issue.

Quoted message said:

Heine gave RBR permission to summarize several major findings.
Interestingly, they confirm lots of what Uncle Al has been ranting
about for years regarding tire width and inflation pressure.

Quoted message said:

Some test conclusions will be particularly enlightening if you're
riding on narrow, high-pressure clinchers seeking more speed via lower
rolling resistance. Your skinny tires may not be as fast as you think.

Quoted message said:

For the full eight-page report on tire performance, order the Vol. 5
No. 1 issue from http://www.bicyclequarterly.com. The test included
nine 700C tires, seven 650B tires and two tubulars. The protocol and
results were reviewed by industry experts. These are eight findings:

Quoted message said:

---With roughly the same power output, the rider's speed can vary by as
much as 20% depending on tire choice. For example, the rider on the
fastest tire [in this roll-down test] moved down the road at
approximately 16.4 mph (26.2 kph) while the same rider on the slowest
tire went approximately 13.6 mph (21.7 kph).

Quoted message said:

---Many longtime riders believe tires with a cotton casing are faster
than modern casings made from nylon. Testing seems to confirm this. The
best-performing tire in the test, the Deda Tre Giro d'Italia 700x23C
(actual width 24.5 mm), has a cotton casing.

Quoted message said:

---Tire pressure has only a small effect on the rolling resistance of
most tires. Narrow 23-mm tires seem to roll fastest at pressures of 105
psi (7.2 bar) or more. However, running these tires at 85 psi (5.8 bar)
for improved comfort increased the test times only 2%. Wider 28-mm
tires are as fast at 85 psi as they are at higher pressures.

Quoted message said:

---Tubular tires perform worse at very high pressure. At 130 psi (9
bar), the narrow Clement Criterium rolled slower than it did at a more
comfortable 105 psi. The wider Clement Campione del Mundo rolled
slightly faster at 85 psi than at 105 psi.

Quoted message said:

---Wide tires do not roll slower at lower pressures. In fact, testing
indicated that a wide tire at lower pressures rolls faster than a
narrow tire at high pressures, if all other factors remain the same.
Even narrow tires can be ridden at comfortable pressures with only very
small concessions to performance.

Quoted message said:

---Tires rolled slightly slower with Michelin's relatively thick latex
tubes than with butyl tubes. Thinner latex tubes, like used in tubular
tires, may offer better performance, but when used in clinchers they
are more prone to punctures caused by friction between tire and tube.
Latex tubes do improve comfort.

Quoted message said:

---Perhaps the most important result of the test is that tire pressure
does not significantly affect rolling resistance. Wide tires in
particular do not need high pressures to roll fast. But because many
current wide tires are designed to handle high pressure, they have
strong casings that lack suppleness. This results in higher rolling
resistance than necessary.

Quoted message said:

---The test's findings point to a new direction for performance
bicycles. For most cyclists, wide, supple tires at low pressures offer
more speed, better comfort, increased versatility and improved safety
than today's narrow high-pressure tires. However, this type of wide,
fast tire currently is not available. Hopefully, these test results
will help persuade manufacturers to produce them.
______________________________________

I think the testing method and instrumentation is probably the culprit
in these erroneous findings that go contrary to most RR tests done
without involving bicycles or riders but only rolling tires.

Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll notice
that this is a family of curves of identical shape.

Jobst Brandt

"Latex tubes do improve comfort"? Once I got to that one, I knew the
results were questionable.
--
Typoes are a feature, not a bug.
Some gardening required to reply via email.
Words processed in a facility that contains nuts.

Werehatrack said:

"Latex tubes do improve comfort"? Once I got to that one, I knew the
results were questionable.

After having used Michelin latex tubes for about a year, I believe I
can feel a noticeable improvement in ride quality versus using butyl
tubes. Sounds strange I know. Can't quantify it but in my mind it
exists.

Quoted message said:

I think the testing method and instrumentation is probably the culprit
in these erroneous findings that go contrary to most RR tests done
without involving bicycles or riders but only rolling tires.

Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll notice
that this is a family of curves of identical shape.

With a smooth roller, yes. Add some real world surface features, which
are often sized at ~20% of the tyre section height even for fairly
smooth pavement, and your "infinite pressure" pneumatic tyres will
roll like iron tyred cart wheels. Bicycle tyre tests which don't
involve bicycles or riders are only slightly interesting, and almost
completely useless.

Kinky Cowboy*

*Batteries not included
May contain traces of nuts
Your milage may vary

In article <[email hidden]>,

Quoted message said:

Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

Here's the same data plotted on a more appropriate scale for
extrapolating.

http://www.ocf.berkeley.edu/~tee/rbt/rolres.gif

Quoted message said:

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll notice
that this is a family of curves of identical shape.

The limit at infinite pressure for the best tire is about half
that of the glued tubulars: significantly better, but certainly not zero.

The 28c Specialized Touring (white line) has an offset that's
nearly as large as the tubulars. I'm not familiar with this tire, but
would guess that it has a raised tread rib, and this loss is from squirming
of that rib.

-Luns

someone said:

Quoted message said:

I think the testing method and instrumentation is probably the
culprit in these erroneous findings that go contrary to most RR
tests done without involving bicycles or riders but only rolling
tires.

Quoted message said:

Quoted message said:

Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

Quoted message said:

Quoted message said:

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll
notice that this is a family of curves of identical shape.

Quoted message said:

With a smooth roller, yes. Add some real world surface features,
which are often sized at ~20% of the tyre section height even for
fairly smooth pavement, and your "infinite pressure" pneumatic tyres
will roll like iron tyred cart wheels. Bicycle tyre tests which
don't involve bicycles or riders are only slightly interesting, and
almost completely useless.

There is no point in testing riders or bicycles when the merit of tire
RR is being assessed. Maybe you can explain what advantage you see in
doing so.

RR comes from tread and casing deformation which is dependent on
inflation pressure. If these tests are done on rougher surfaces all
values will be higher but the characteristic will remain unchanged.
You are suggesting that the lab at IRC that constructed this tire
test, that is more credible and apparent than any others performed,
does not know how to correctly test tires.

I don't think you understand what is being tested and what the
mechanisms that cause rolling losses are. Beyond that, tread patterns
only make more RR as you can see in those curves. Those are the ones
that don't flatten as much as one would expect as inflation pressure
was increased.

Jobst Brandt

Luns Tee said:

Quoted message said:

Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

Quoted message said:

Here's the same data plotted on a more appropriate scale for
extrapolating.

http://www.ocf.berkeley.edu/~tee/rbt/rolres.gif

Quoted message said:

Quoted message said:

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll
notice that this is a family of curves of identical shape.

Quoted message said:

The limit at infinite pressure for the best tire is about half that
of the glued tubulars: significantly better, but certainly not zero.

Quoted message said:

The 28c Specialized Touring (white line) has an offset that's nearly
as large as the tubulars. I'm not familiar with this tire, but would
guess that it has a raised tread rib, and this loss is from
squirming of that rib.

Elegant!

Although the RR values at infinity seem high there must be an
explanation for their separation and high value. Are we seeing the
correct vertical scale? At infinite pressure, only thin tread rubber
is still flexing as tread squirm. That the logarithmic curves are
parallel rather than converging seems odd at first inspection.

Jobst Brandt

In article <[email hidden]>,

Quoted message said:

http://www.ocf.berkeley.edu/~tee/rbt/rolres.gif

Quoted message said:

Quoted message said:

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll
notice that this is a family of curves of identical shape.

Quoted message said:

The limit at infinite pressure for the best tire is about half that
of the glued tubulars: significantly better, but certainly not zero.

Quoted message said:

The 28c Specialized Touring (white line) has an offset that's nearly
as large as the tubulars. I'm not familiar with this tire, but would
guess that it has a raised tread rib, and this loss is from
squirming of that rib.

Elegant!

Although the RR values at infinity seem high there must be an
explanation for their separation and high value. Are we seeing the
correct vertical scale? At infinite pressure, only thin tread rubber
is still flexing as tread squirm. That the logarithmic curves are
parallel rather than converging seems odd at first inspection.

These aren't log-scale. The vertical scale is linear, and
exactly the same scale as in your graph - you can even check the data
points visually and see that they agree. The horizontal scale is
actually -1/pressure with the origin at the right, and the tickmarks
then re-labelled according to pressure.

1/pressure is just area per force, so dimensionally, the
k/pressure term is proportional to the area of the contact patch for a
given wheel load.

That the curves straighten out into lines in this scale shows that
the data fits a relationship of

resistance = offset + k/pressure

and that the curve for each tire can actually be captured by two
constants - the offset, and the scaling factor. I have a scatter plot
of what these constants are for the various tires, but it doesn't
really tell anything that can't be seen from the image I gave above.

-Luns

Luns Tee said:

Quoted message said:

http://www.ocf.berkeley.edu/~tee/rbt/rolres.gif

Quoted message said:

These aren't log-scale. The vertical scale is linear, and
exactly the same scale as in your graph - you can even check the data
points visually and see that they agree. The horizontal scale is
actually -1/pressure with the origin at the right, and the tickmarks
then re-labelled according to pressure.

1/pressure is just area per force, so dimensionally, the
k/pressure term is proportional to the area of the contact patch for a
given wheel load.

That the curves straighten out into lines in this scale shows that
the data fits a relationship of

resistance = offset + k/pressure

and that the curve for each tire can actually be captured by two
constants - the offset, and the scaling factor.

Interesting graph. I'd assume that for the clinchers the offset value
would be primarily determined by the losses due to the tread rubber at
the contact patch whereas the 'k' value would be primarily due to flex
losses in the casing and sidewall. In that case a combination of the
casing used in the Avocets and the tread rubber of the Michelin Hilite
Comp would seem to be a worthwhile tire design.

Quoted message said:

someone said:

Quoted message said:

I think the testing method and instrumentation is probably the
culprit in these erroneous findings that go contrary to most RR
tests done without involving bicycles or riders but only rolling
tires.

Quoted message said:

Quoted message said:

Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

Quoted message said:

Quoted message said:

All these tires nearly approach zero RR at infinite pressure with a
constant offset for tubulars that have rim glue losses. You'll
notice that this is a family of curves of identical shape.

Quoted message said:

With a smooth roller, yes. Add some real world surface features,
which are often sized at ~20% of the tyre section height even for
fairly smooth pavement, and your "infinite pressure" pneumatic tyres
will roll like iron tyred cart wheels. Bicycle tyre tests which
don't involve bicycles or riders are only slightly interesting, and
almost completely useless.

There is no point in testing riders or bicycles when the merit of tire
RR is being assessed. Maybe you can explain what advantage you see in
doing so.

RR comes from tread and casing deformation which is dependent on
inflation pressure. If these tests are done on rougher surfaces all
values will be higher but the characteristic will remain unchanged.

Jobst, was the data linked above measured by rolling on a smooth drum?

IIRC, the test in [Vintage] Bicycle Quarterly was a roll-down test on
pavement. The justification given was that tests on perfectly smooth
surfaces miss certain effects, alluded to above. On a perfectly smooth
surface, no "tire suspension" (that is, flexibility in the contact
patch and the tire just above it) is needed, so rolling resistance
always decreases with increased pressure.

But on a rough surface, a tire with infinitely high pressure, like a
steel wheel coated with rubber, would launch the bike into thousands of
tiny vertical hops. Those consume energy, and represent lost forward
motion. In fact, the vertical bounces would probably consume more
energy than the tire hystresis, the main effect measured on a smooth
test surface.

Perhaps testing rolling resistance by an outdoor roll-down test isn't
perfect. I can see error being introduced by minor changes in rider
position, clothing, temperature, wind speed & direction, etc. But the
testers seemed aware of those problems, made efforts to control them,
and the data in the original article shows good precision.

Another point is, IIRC, they didn't describe the exact nature of the
asphalt. (Is there some roughness measurment in common use, similar to
the RMS method of describing metal finishing smoothness?) ISTM that
rougher asphalt might rank tires differently than smoother asphalt.

In any case, until shown otherwise, I'd expect rolling resistance
measured _somehow_ using an appropriate mass on an appropriately rough
surface - not smooth steel - would be more valid for real riding.

Quoted message said:

You are suggesting that the lab at IRC that constructed this tire
test, that is more credible and apparent than any others performed,
does not know how to correctly test tires.

Well, it wouldn't be the first time a manufacturer made a big mistake!

- Frank Krygowski

Werehatrack said:

"Latex tubes do improve comfort"? Once I got to that one, I knew the
results were questionable.

[Vintage] Bicycle Quarterly - largely a one-man operation, it seems -
does contain a lot of subjective judgements stated as fact. Many of
them seem questionable at best.

But ISTM that one can be skeptical about those judgements, yet not toss
out other data resulting from measurements made with decent technique.

- Frank Krygowski

Frank Krygowski said:

Quoted message said:

Quoted message said:

> I think the testing method and instrumentation is probably the
> culprit in these erroneous findings that go contrary to most RR
> tests done without involving bicycles or riders but only rolling
> tires.

Quoted message said:

Quoted message said:

Quoted message said:

> Please review the chart at:

http://www.sheldonbrown.com/brandt/rolling-resistance-tubular.html

Quoted message said:

Quoted message said:

Quoted message said:

> All these tires nearly approach zero RR at infinite pressure with
> a constant offset for tubulars that have rim glue losses. You'll
> notice that this is a family of curves of identical shape.

Quoted message said:

Quoted message said:

Quoted message said:

With a smooth roller, yes. Add some real world surface features,
which are often sized at ~20% of the tyre section height even for
fairly smooth pavement, and your "infinite pressure" pneumatic
tyres will roll like iron tyred cart wheels. Bicycle tyre tests
which don't involve bicycles or riders are only slightly
interesting, and almost completely useless.

Quoted message said:

Quoted message said:

There is no point in testing riders or bicycles when the merit of
tire RR is being assessed. Maybe you can explain what advantage
you see in doing so.

Quoted message said:

Quoted message said:

RR comes from tread and casing deformation which is dependent on
inflation pressure. If these tests are done on rougher surfaces
all values will be higher but the characteristic will remain
unchanged.

Quoted message said:

Jobst, was the data linked above measured by rolling on a smooth
drum?

All RR tests by the auto industry are done on smooth drums, pavement
granularity only being undefined and not comparable from one tester to
the next. These tests need to be verifiable at different labs, so
they use standard smooth drums of known diameter. Roller roll-down
tests by Micheline at InterBike are also done on steel rollers.

Quoted message said:

IIRC, the test in [Vintage] Bicycle Quarterly was a roll-down test
on pavement. The justification given was that tests on perfectly
smooth surfaces miss certain effects, alluded to above. On a
perfectly smooth surface, no "tire suspension" (that is, flexibility
in the contact patch and the tire just above it) is needed, so
rolling resistance always decreases with increased pressure.

I think they need to assess what these "certain" effects are and how
they do not correlate to smooth drum testing. I propose that the
order of RR quality does not change with true pavement compared to a
steel drum, the losses coming from the same effects.

Quoted message said:

But on a rough surface, a tire with infinitely high pressure, like a
steel wheel coated with rubber, would launch the bike into thousands
of tiny vertical hops. Those consume energy, and represent lost
forward motion. In fact, the vertical bounces would probably
consume more energy than the tire hysteresis, the main effect
measured on a smooth test surface.

That is only true if the roughness is greater than the rubber
thickness and uneven at that. Bouncing does not in itself absorb
energy, it is the hysteretic losses in the tread and casing. There is
no purpose in testing tires on RR ballast for instance.

Quoted message said:

Perhaps testing rolling resistance by an outdoor roll-down test
isn't perfect. I can see error being introduced by minor changes in
rider position, clothing, temperature, wind speed & direction, etc.
But the testers seemed aware of those problems, made efforts to
control them, and the data in the original article shows good
precision.

The rider has far more influence on drag than any tire so the whole
experiment falls apart.

Quoted message said:

Another point is, IIRC, they didn't describe the exact nature of the
asphalt. (Is there some roughness measurement in common use,
similar to the RMS method of describing metal finishing smoothness?)
ISTM that rougher asphalt might rank tires differently than smoother
asphalt.

You can't describe that. That is why you don't do that.

Quoted message said:

In any case, until shown otherwise, I'd expect rolling resistance
measured _somehow_ using an appropriate mass on an appropriately
rough surface - not smooth steel - would be more valid for real
riding.

I don't believe so. If you could show me where the energy goes that
makes the pavement test produce a different hierarchy among samples, I
would be interested.

Quoted message said:

Quoted message said:

You are suggesting that the lab at IRC that constructed this tire
test, that is more credible and apparent than any others performed,
does not know how to correctly test tires.

Quoted message said:

Well, it wouldn't be the first time a manufacturer made a big
mistake!

Well? Tell me where the big mistake might be. As you see, these are
the first RR tests that revealed the dirty secret of tubular rim glue
and explained why there is pressure sensitive road glue and hard non
resilient track glue.

Jobst Brandt

Quoted message said:

Frank Krygowski writes:

All RR tests by the auto industry are done on smooth drums, pavement
granularity only being undefined and not comparable from one tester to
the next. These tests need to be verifiable at different labs, so
they use standard smooth drums of known diameter. Roller roll-down
tests by Micheline at InterBike are also done on steel rollers.

Couldn't that be interpreted to mean that the auto industry,
similarly, is just dodging the issue?

[snip]

Quoted message said:

Quoted message said:

Another point is, IIRC, they didn't describe the exact nature of the
asphalt. (Is there some roughness measurement in common use,
similar to the RMS method of describing metal finishing smoothness?)
ISTM that rougher asphalt might rank tires differently than smoother
asphalt.

You can't describe that. That is why you don't do that.

The people who make it their busines to monitor road surface quality
certainly can, see:

http://www.wdm.co.uk/pdf/S_RAV.pdf

....where roughness and texture, as well as cracks and wheel ruts are
some of the quality parameters that are assessed.