Drive Train Effecieny

Werehatrack said:

BTW, my engineer contact confirms that the sparse data he's got indicates a smaller-cog penalty
that shouldn't be anywhere near this severe if the two cogs are otherwise the same in design
characteristics, so the small-cog-effort theory remains unproven as the *sole* cause in this
instance.

I ran some rough numbers through the excellent calculator at www.analyticcycling.com and the
difference is even more dramatic than I suspected.

For a "normal" ride in a normal position, and riding on a flat road, the difference between power
required to achieve the two velocities you obtained with the different gear ratios represent an
almost 2/1 ratio.

The model showed it should take 203.5 watts to go 22mph, and only
103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and 17mph
with the same effort, over 100 watts of energy is being "absorbed" by the chain. In that case, I
would actually expect it to get hot enough to notice - maybe not smoking hot, but certainly well
above ambient temperature (not that there's much of a difference between the two here in Arizona
in the summer!).

Thoughts to ponder...

Mark Hickey Habanero Cycles habcycles.comhabcycles.comOpen ↗ Home of the $695 ti frame

Quoted message said:

The model showed it should take 203.5 watts to go 22mph, and only
103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and 17mph
with the same effort, over 100 watts of energy is being "absorbed" by the chain. In that case,
I would

The model doesn't take air resistance into effect?

--
Phil, Squid-in-Training

"Peter Cole" <[email hidden]> wrote in message news:<PZ8Xa.54695$uu5.5496@sccrnsc04>...

Quoted message said:

Larger sprockets are more efficient, since the chain goes through a smaller bend angle as it
dis/engages <snip>

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed (360
degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension is
higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that means chain
wear will be more or less the same with an 11 as with a 16 for the same gear.

Andrew Bradley

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

Quoted message said:

"Peter Cole" <[email hidden]> wrote in message

news:<PZ8Xa.54695$uu5.5496@sccrnsc04>...

Quoted message said:

Quoted message said:

Larger sprockets are more efficient, since the chain goes through a

smaller

Quoted message said:

Quoted message said:

bend angle as it dis/engages <snip>

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed (360
degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension is
higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that means chain
wear will be more or less the same with an 11 as with a 16 for the same gear.

I can't follow most of this, but what I can is wrong.

Andrew Bradley said:

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed (360
degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension is
higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that means chain
wear will be more or less the same with an 11 as with a 16 for the same gear.

Andrew Bradley

Okay, but higher tension means more friction. Also, with the smaller sprocket, there is more rapid
bending as the links go through the 360d.
--
Cheers! OliverS When replying personally, remove "_removespam_"

"When I see an adult on a bicycle, I do not despair for the future of the human race." HG Wells

Mark Hickey <[email hidden]> wrote in message
news:<[email hidden]>...

Quoted message said:

(Steven) said:

And hearing all the noise in the Giro etc. about bike manafactures being able to make a bike
signifigantly less than the 16.5 pounds that the UCI set's as the lower limit of weight. My
thought is essentially this, why not make that bike but throw extra weight back in by making the
drive train more efficient.

Lets say that for arguments sake the average pro cylist puts out 400 watts during a climb up alp
d'huez and his drive train is 90% effecient.

Thing is, the data I've seen suggest a clean, well-adjusted drivetrain is closer to 98% efficient.
That doesn't leave a lot of room for improvement (at least not without spending a fortune on a
chain with itty bitty little bearings in every link).

That number hasn't changed in recent history, either - probably won't in the near future.

Mark Hickey Habanero Cycles habcycles.comhabcycles.comOpen ↗ Home of the $695 ti frame

I take that back. That previous post was for a hub drive. The peak that a shimano 27 speed ultegra
(mtn bike) drivetrain produces is
97.2%. This is however at high power output (370 watts) and in the perfect gear. Other
gearing/wattages are somewhat less.

However, again this data could be wrong, and I think I have proved I am less than perfect at reading
charts. However, it certainly does infer that if drive train effeciency is now at 98% accross the
bored (no matter what gear or company) The number has at least changed in recent history.

Of interesting note. Effeciency was not notable determined by lateral stress (i.e little front,
little rear).

SPA

Steven said:

Mark Hickey <[email hidden]> wrote in message
news:<[email hidden]>...

Quoted message said:

(Steven) said:

And hearing all the noise in the Giro etc. about bike manafactures being able to make a bike
signifigantly less than the 16.5 pounds that the UCI set's as the lower limit of weight. My
thought is essentially this, why not make that bike but throw extra weight back in by making the
drive train more efficient.

Lets say that for arguments sake the average pro cylist puts out 400 watts during a climb up alp
d'huez and his drive train is 90% effecient.

Thing is, the data I've seen suggest a clean, well-adjusted drivetrain is closer to 98% efficient.
That doesn't leave a lot of room for improvement (at least not without spending a fortune on a
chain with itty bitty little bearings in every link).

That number hasn't changed in recent history, either - probably won't in the near future.

Mark Hickey Habanero Cycles habcycles.comhabcycles.comOpen ↗ Home of the $695 ti frame

I take that back. That previous post was for a hub drive. The peak that a shimano 27 speed ultegra
(mtn bike) drivetrain produces is
97.2%. This is however at high power output (370 watts) and in the perfect gear. Other
gearing/wattages are somewhat less.

The 97.2% was the figure uncorrected for efficiency losses caused by the ergometer drive wheel. Fig.
14 shows a curve with an approximate correction for these losses and shows efficiency going up to
98%. The chainwheels used were for a mountain bike (44, 32, 22 teeth) and generally showed improved
efficiency for the larger chainwheels. A typical road bike arrangement would have more of the gears
in the range with relatively high efficiencies.

Quoted message said:

However, again this data could be wrong, and I think I have proved I am less than perfect at
reading charts. However, it certainly does infer that if drive train effeciency is now at 98%
accross the bored

I'm sure some readers are at that stage by now. 🙂

Quoted message said:

(no matter what gear or company) The number has at least changed in recent history.

I don't think it has ever been claimed that efficiency is always at 98% and above - just that such a
level is achievable under good conditions. A dirty drivetrain will certainly reduce efficiency as
will use of gears with small cogs (at least below about 13 teeth).

Quoted message said:

Of interesting note. Effeciency was not notable determined by lateral stress (i.e little front,
little rear).

But that ratio was significantly less efficient than an equivalent gear ratio using the large
chainring (91.9 vs. 93.9% uncorrected efficiency). Using the large/large combination was slightly
more efficient (94.2%) than an equivalent gear with the small chainring (93.9%) despite the lateral
misalignment. From an efficiency standpoint it's generally best to use larger chainrings (and
correspondingly large cogs) as much as possible. [But the effect is pretty small and does not
explain the 17 - 22 mph difference claimed in another post in this thread.]

Phil said:

Quoted message said:

The model showed it should take 203.5 watts to go 22mph, and only
103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and 17mph
with the same effort, over 100 watts of energy is being "absorbed" by the chain. In that
case, I would

The model doesn't take air resistance into effect?

Yes it does - hence the ~50% reduction in effort against a 23% reduction in velocity.

Mark Hickey Habanero Cycles habcycles.comhabcycles.comOpen ↗ Home of the $695 ti frame

OliverS may have said:

Andrew Bradley said:

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed (360
degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension is
higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that means
chain wear will be more or less the same with an 11 as with a 16 for the same gear.

Andrew Bradley

Okay, but higher tension means more friction. Also, with the smaller sprocket, there is more rapid
bending as the links go through the 360d.

Smaller bend radius = greater incident angle per link, therefore greater amount of flex per link. It
still shouldn't make this much of a difference. Not even counting the fact that the der has wound
itself up tighter, and is flexing the chain more at each of its bends as well.

Hmm. Aren't the wheels on the der about the same size as an 11? Of course, there's no load on them
to speak of.

--
My email address is antispammed; pull WEEDS if replying via e-mail. Yes, I have a killfile. If I
don't respond to something, it's also possible that I'm busy.

Mark Hickey may have said:

For a "normal" ride in a normal position, and riding on a flat road, the difference between power
required to achieve the two velocities you obtained with the different gear ratios represent an
almost 2/1 ratio.

I was mentally guesstimating a 35 to 40% loss since the velocities are both relatively low, but that
figure is not surprising.

Quoted message said:

The model showed it should take 203.5 watts to go 22mph, and only
103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and 17mph
with the same effort, over 100 watts of energy is being "absorbed" by the chain. In that case,
I would actually expect it to get hot enough to notice - maybe not smoking hot, but certainly
well above ambient temperature (not that there's much of a difference between the two here in
Arizona in the summer!).

Thoughts to ponder...

Pondering continues. For now, pondering is all I'm going to do about it; the cursed 11 has been
relegated to the pile of unfinished project materials until after the taxes have been filed. It's
going to be hard enough to slip the daily ride in by itself; no time for tinkering at all.

--
My email address is antispammed; pull WEEDS if replying via e-mail. Yes, I have a killfile. If I
don't respond to something, it's also possible that I'm busy.

Squid-in-Training' may have said:

Quoted message said:

The model showed it should take 203.5 watts to go 22mph, and only
103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and 17mph
with the same effort, over 100 watts of energy is being "absorbed" by the chain. In that
case, I would

The model doesn't take air resistance into effect?

It would have to be considering air resistance, or the difference wouldn't be as sharp.

--
My email address is antispammed; pull WEEDS if replying via e-mail. Yes, I have a killfile. If I
don't respond to something, it's also possible that I'm busy.

(Steven) may have said:

www.ihpva.org/pubs/HP52.pdf

Interesting.

For those who find the rest too esoteric, there's a cutaway of the Rohloff on page 11.

--
My email address is antispammed; pull WEEDS if replying via e-mail. Yes, I have a killfile. If I
don't respond to something, it's also possible that I'm busy.

OliverS said:

Andrew Bradley said:

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed
(360 degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension is
higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that means
chain wear will be more or less the same with an 11 as with a 16 for the same gear.

Andrew Bradley

Okay, but higher tension means more friction.

Of course, that's why larger sprockets are more efficient.

Quoted message said:

Also, with the smaller sprocket, there is more rapid bending as the links go through the 360d.

How do you mean more rapid?

Andrew Bradley

"Peter Cole" >

Quoted message said:

Quoted message said:

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed
(360 degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension is
higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that means
chain wear will be more or less the same with an 11 as with a 16 for the same gear.

I can't follow most of this, but what I can is wrong.

But you can't/won't explain why.

Andrew Bradley

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

Quoted message said:

Phil said:

Quoted message said:

The model showed it should take 203.5 watts to go 22mph, and only
103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and
17mph with the same effort, over 100 watts of energy is being "absorbed" by the chain. In
that case, I would

The model doesn't take air resistance into effect?

Yes it does - hence the ~50% reduction in effort against a 23% reduction in velocity.

But you said that the chain absorbs 100 watts of energy out of the 203.5 output by the rider... I'm
confused. Doesn't the air absorbe some of the energy?

--
Phil, Squid-in-Training

Phil said:

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

Quoted message said:

Phil said:

>The model showed it should take 203.5 watts to go 22mph, and only
>103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and
> 17mph with the same effort, over 100 watts of energy is being "absorbed" by the chain. In
> that case, I would

The model doesn't take air resistance into effect?

Yes it does - hence the ~50% reduction in effort against a 23% reduction in velocity.

But you said that the chain absorbs 100 watts of energy out of the 203.5 output by the rider...
I'm confused. Doesn't the air absorbe some of the energy?

W claimed that he puts out the same amount of effort to ride at 17 mph with the 11-tooth cog as he
does to ride at 22 mph with a larger cog (but same overall gear ratio). Riding at 22 mph and
assuming normal amounts of rolling resistance, air resistance, and about 2% loss to the drivetrain
requires 203 W of energy output by the rider. Riding at 17 mph and assuming normal amounts of
rolling resistance, air resistance, and 2% loss in the drivetrain should only require 103 W of
energy - mainly because of the much lower air resistance when riding slower.

If W's claim is correct that he puts out just as much energy in this second case (i.e. he still puts
out 203 W) then the extra 100 W of energy must be wasted somehow since only 103 W should be needed
to ride 17 mph. Since the same bike is used in both cases the only difference is in the drivetrain
gears so the wasted 100 W must be due to very high chain friction around the smaller gears (esp. the
11-tooth cog). This is the part that seems unbelievable unless there is something clearly wrong with
the cog, chainwheel, and/or chain.

(Steven) said:

Mark Hickey <[email hidden]> wrote ...

Quoted message said:

[email hidden] (Steven) wrote:

Quoted message said:

Quoted message said:

Quoted message said:

Lets say that for arguments sake the average pro cylist puts out 400 watts during a climb up alp
d'huez and his drive train is 90% effecient.

Thing is, the data I've seen suggest a clean, well-adjusted drivetrain is closer to 98%
efficient. That doesn't leave a lot of room for improvement (at least not without spending a
fortune on a chain with itty bitty little bearings in every link).

That number hasn't changed in recent history, either - probably won't in the near future.

www.ihpva.org/pubs/HP52.pdf

This study puts the effecieny of Shimano 7 speed derailer hubs at
94.5%-90.3% effecient. I am not sure what has changed since 2001....I would imagine effeciency has
improved but probably not to 98%, although I have been wrong a few times in my life so....Where
did you get this number? I would love to read read the research you got this number from.

This is not backed up by the research I could find. The best effecieny rating I could find was
98.6% but it was for a large rear cogged single speed drive train, taking the

The Tour riders do NOT ride internally geared hubs. It's common knowledge that the efficiency of
internally geared hubs is a lot lower than a well-adjusted, well-maintained "normal" drivetrain. You
seem to have answered your own question, citing the 98.6% efficiency of a single speed drivetrain.
Throwing in a rear derailleur won't hurt it all that much.

Mark Hickey Habanero Cycles habcycles.comhabcycles.comOpen ↗ Home of the $695 ti frame

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

Quoted message said:

"Peter Cole" >

Quoted message said:

Quoted message said:

I don't think that's quite right. All sprockets bend chain at the same rate at a given speed
(360 degrees per wheel turn), it's just that the small ones do it at higher tension.

An interesting question is the chain wear implications of this. With small sprockets tension
is higher but the wear on the pins is spread over a correspondingly wider arc. Maybe that
means chain wear will be more or less the same with an 11 as with a 16 for the same gear.

I can't follow most of this, but what I can is wrong.

But you can't/won't explain why.

You are equating friction/efficiency with wear. Wear comes primarily from contaminants.

Smaller sprockets will have higher power losses since the articulation angle is greater, and the
force is greater. This is partially offset by the lower chain speed. A 3x smaller sprocket will have
a 3x greater tension and 3x greater articulation angle, but a 3x smaller chain speed, with a
proportional 3x greater articulation friction loss overall.

The efficiency of a roller chain drive doesn't seem to go down with chain tension though:
*****************************************************************
jhu.edubike.htmlOpen ↗

" The researchers found two factors that seemed to affect the bicycle chain drive's efficiency.
Surprisingly, lubrication was not one of them.

"The first factor was sprocket size," Spicer says. "The larger the sprocket, the higher the
efficiency we recorded." "

" The second factor that affected efficiency was tension in the chain. The higher the chain tension,
Spicer says, the higher the efficiency score. "This is actually not in the direction you'd expect,
based simply on friction," he says. "It's not clear to us at this time why this occurs." "

*******************************************************************

It may be that smaller sprockets cause more chain wear because the greater articulation angle
increases contaminant intrusion. That's only conjecture. The only thing that seems known for sure is
that smaller sprockets themselves wear faster.

Phil said:

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

Quoted message said:

Phil said:

> The model showed it should take 203.5 watts to go 22mph, and only
> 103.3 watts to go 17mph. That means that if you're actually going 22mph with one ratio, and
> 17mph with the same effort, over 100 watts of energy is being "absorbed" by the chain. In
> that case, I would

The model doesn't take air resistance into effect?

Yes it does - hence the ~50% reduction in effort against a 23% reduction in velocity.

But you said that the chain absorbs 100 watts of energy out of the 203.5 output by the rider... I'm
confused. Doesn't the air absorbe some of the energy?

Peter explained it well in his post - if one assumes the same power output, the extra energy that
would have propelled the rider to 22mph has to be going somewhere if he's now only going 17mph. In
this case, the "missing power" is ~100 watts - a HUGE number.

If I could get Lance to ride the OP's 11, I might have a chance! ;-)

Mark Hickey Habanero Cycles habcycles.comhabcycles.comOpen ↗ Home of the $695 ti frame

Peter Cole:

Quoted message said:

You are equating friction/efficiency with wear. Wear comes primarily from contaminants.

I was assuming there must be some relation to force and movement. Grinding paste requires these to
grind, no?

Quoted message said:

Smaller sprockets will have higher power losses since the articulation angle is greater, and the
force is greater. This is partially offset by the lower chain speed. A 3x smaller sprocket will
have a 3x greater tension and 3x greater articulation angle, but a 3x smaller chain speed, with a
proportional 3x greater articulation friction loss overall.

Here you have shown (by arcane means) that losses increase with chain tension which is what I said
(equal speed and gear assumed). "Articulation angle" is not the reason since the total articulation
of the chain per unit of time is the same for all sprockets.

360 degrees per revolution of wheel (external angles of a polygon add to 360)

Quoted message said:

jhu.edubike.htmlOpen ↗ " The second factor that affected
efficiency was tension in the chain. The higher the chain tension, Spicer says, the higher the
efficiency score. "This is actually not in the direction you'd expect, based simply on friction,"
he says. "It's not clear to us at this time why this occurs." "

This is poor reporting. How do you marry this report with what you have just shown yourself? What
chain tension do they mean?

The one we want is the tension at constant bike speed (or constant power output) AND _constant
cadence_ (or gear)

Neglecting this last requirement is the only explanation I can come up with for the confusion
surrounding "chain tension" in this experimentation.

That big gears are more efficient than small gears with a given sprocket is easy to accept, and
bigger gears involve more chain tension.

What do you make of this snip of a posted snip of a post? :

<< A googled post from alt.rec.bicycles.recumbent:

However, in an article in the fall 2000 "Human Power" called "Efficiency of bicycle chain drives:
results at constant velocity and supplied power" John and Claire Walton did an analysis comparing
chain and sprocket efficiency at a constant supplied power and vehicle speed.

Using data from the previous Spicer HP article, they found that at constant power and vehicle
speed, the efficiencies were 92% for the 11 tooth,
90.5% for the 15 tooth, and 88.5% for the 21 tooth.

Another cycling myth (of many) bites the dust!

Tailwinds (and nine tooth sprockets) Rich Pinto Bacchetta Bicycles >>

Andrew Bradley