A frame type power loss experiment
#76
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
Given that bicycle frames are so light (under 1kg for a modern carbon road frame) it makes sense to maximise lateral and torsional stiffness as much as possible at that minimal weight. Vintage lightweight race bikes were noodles in comparison to modern carbon bikes and I don’t think anyone would argue they were better for handling or power transfer.
#77
Senior Member
Join Date: May 2021
Posts: 8,808
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 4598 Post(s)
Liked 5,142 Times
in
3,177 Posts
Likes For PeteHski:
#78
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
This is the reason that late '80s Cannondales with massive down tubes were lighter than Vitus frames with all skinny tubes.
And it isn't unchallenging to make something stiff and light - that's been the game since the '80s. But it is impossible to make something as light as the lightest frames with only moderate stiffness - they would flex past their plastic deformation limit and fail. The original Columbus EL tubing does not make for long lived frames, but S3 does.
#79
Senior Member
Join Date: Jan 2005
Location: Baltimore, MD
Posts: 5,606
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2619 Post(s)
Liked 3,153 Times
in
1,797 Posts
You'll find that dodgy logic alive and well in all the lightest aluminum, steel, titanium and carbon road bikes. A 2.5 pound S3 steel road frame only gets that light because the tubing is oversized and very thin walled. If you made the walls that thin and allowed the tubing to flex, the walls would collapse from the load. Make the tubes oversize and the structure becomes stiff enough to not have to deal with with localized bending.
This is the reason that late '80s Cannondales with massive down tubes were lighter than Vitus frames with all skinny tubes.
And it isn't unchallenging to make something stiff and light - that's been the game since the '80s. But it is impossible to make something as light as the lightest frames with only moderate stiffness - they would flex past their plastic deformation limit and fail. The original Columbus EL tubing does not make for long lived frames, but S3 does.
This is the reason that late '80s Cannondales with massive down tubes were lighter than Vitus frames with all skinny tubes.
And it isn't unchallenging to make something stiff and light - that's been the game since the '80s. But it is impossible to make something as light as the lightest frames with only moderate stiffness - they would flex past their plastic deformation limit and fail. The original Columbus EL tubing does not make for long lived frames, but S3 does.
#80
Senior Member
Join Date: May 2021
Posts: 8,808
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 4598 Post(s)
Liked 5,142 Times
in
3,177 Posts
You'll find that dodgy logic alive and well in all the lightest aluminum, steel, titanium and carbon road bikes. A 2.5 pound S3 steel road frame only gets that light because the tubing is oversized and very thin walled. If you made the walls that thin and allowed the tubing to flex, the walls would collapse from the load. Make the tubes oversize and the structure becomes stiff enough to not have to deal with with localized bending.
This is the reason that late '80s Cannondales with massive down tubes were lighter than Vitus frames with all skinny tubes.
And it isn't unchallenging to make something stiff and light - that's been the game since the '80s. But it is impossible to make something as light as the lightest frames with only moderate stiffness - they would flex past their plastic deformation limit and fail. The original Columbus EL tubing does not make for long lived frames, but S3 does.
This is the reason that late '80s Cannondales with massive down tubes were lighter than Vitus frames with all skinny tubes.
And it isn't unchallenging to make something stiff and light - that's been the game since the '80s. But it is impossible to make something as light as the lightest frames with only moderate stiffness - they would flex past their plastic deformation limit and fail. The original Columbus EL tubing does not make for long lived frames, but S3 does.
I would imagine all current road race bikes are aiming for maximal torsional and lateral stiffness given the relatively small amount of material they have to play with ie 1kg or less of carbon. Motorbike frame designers are less limited in this respect and AFAIK they aim for a target torsional stiffness that is below what they could achieve if they wanted to i.e. they are not necessarily designing for maximal torsional stiffness.
#81
climber has-been
Join Date: Dec 2004
Location: Palo Alto, CA
Posts: 7,276
Bikes: Scott Addict R1, Felt Z1
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 3572 Post(s)
Liked 3,751 Times
in
1,874 Posts
You'll find that dodgy logic alive and well in all the lightest aluminum, steel, titanium and carbon road bikes. A 2.5 pound S3 steel road frame only gets that light because the tubing is oversized and very thin walled. If you made the walls that thin and allowed the tubing to flex, the walls would collapse from the load. Make the tubes oversize and the structure becomes stiff enough to not have to deal with with localized bending.
To strengthen a cylindrical tube against local buckling, you either increase the wall thickness, decrease the tube diameter, or both.
To stiffen a cylindrical tube, you increase wall thickness, increase the tube diameter, or both.
But if all you do is make the tube larger while keeping the same wall thickness, you've made it more prone to buckling.
Last edited by terrymorse; 02-05-24 at 12:27 PM.
Likes For terrymorse:
#82
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
That's a bit confusing. Other things equal, the larger the tube diameter, the more susceptible it is to local buckling.
To strengthen a cylindrical tube against local buckling, you either increase the wall thickness, decrease the tube diameter, or both.
To stiffen a cylindrical tube, you increase wall thickness, increase the tube diameter, or both.
But if all you do is make the tube larger while keeping the same wall thickness, you've made it more prone to buckling.
To strengthen a cylindrical tube against local buckling, you either increase the wall thickness, decrease the tube diameter, or both.
To stiffen a cylindrical tube, you increase wall thickness, increase the tube diameter, or both.
But if all you do is make the tube larger while keeping the same wall thickness, you've made it more prone to buckling.
What you're talking about is the beer can problem, where the tube gets so wide and thin that an external force will damage it. This is the reason you can't make a steel bike as light as aluminum - the higher density makes the tube wall too thin. And it's the reason you can't sit on the top tube of certain very stiff carbon bikes.
#83
climber has-been
Join Date: Dec 2004
Location: Palo Alto, CA
Posts: 7,276
Bikes: Scott Addict R1, Felt Z1
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 3572 Post(s)
Liked 3,751 Times
in
1,874 Posts
What you're talking about is the beer can problem, where the tube gets so wide and thin that an external force will damage it. This is the reason you can't make a steel bike as light as aluminum - the higher density makes the tube wall too thin. And it's the reason you can't sit on the top tube of certain very stiff carbon bikes.
You also can't make a steel bike as stiff as aluminum bike of the same weight because an aluminum tube of the same weight will always be stiffer. Resistance to bending vs weight is more important than strength when stiffness is a design requirement. This is the case with airplanes -- and bike frames -- and it's why you don't see airplane structures made out of steel.
#84
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
You can't make a steel bike as light as an aluminum bike with the same strength, but it's not because of steel's higher density. It's because steel's specific strength (strength-to-weight ratio) is lower than aluminum.
You also can't make a steel bike as stiff as aluminum bike of the same weight because an aluminum tube of the same weight will always be stiffer. Resistance to bending vs weight is more important than strength when stiffness is a design requirement. This is the case with airplanes -- and bike frames -- and it's why you don't see airplane structures made out of steel.
You also can't make a steel bike as stiff as aluminum bike of the same weight because an aluminum tube of the same weight will always be stiffer. Resistance to bending vs weight is more important than strength when stiffness is a design requirement. This is the case with airplanes -- and bike frames -- and it's why you don't see airplane structures made out of steel.
You know what else is stronger than aluminum by weight? Pine. But pine has such low density that attaching components to it would be impractical. Which is the other side of the density coin from steel. For airplanes, pine does work because of the size of the structure. Steel doesn't again because of both wall thickness and cost. Aluminum is a much more cost efficient high strength material.
Last edited by Kontact; 02-05-24 at 06:45 PM.
#85
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
I've wondered about the longevity of the "S3" frames. (Thanks for mentioning that descriptor; I hadn't come across it before.) Knowing how the bike market works, my guess is that those 2.5-pound steel frames are built with an exceedingly small safety margin with respect to potential stresses and damage. The people building them are probably figuring that they'll get away with building them that way, though, banking on the purchasers being almost exclusively older riders who will use them little and treat them gently.
But compared to ultralight steel frames that were standard tubing diameters, the oversized heat treated steel tubes are harder, stronger, thinner, stiffer and lighter. We know those old light tubesets wore out fairly quickly because they flexed themselves to death, but if the frame is stiff it can't flex enough to begin wearing out.
Likes For Kontact:
#86
Senior Member
Join Date: Jul 2009
Posts: 6,067
Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 4018 Post(s)
Liked 7,530 Times
in
3,031 Posts
#87
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
It would be more accurate to say that frames of similarly high stiffness are equally unlikely to wear out. Obviously a frame made of solid tungsten would not wear out either.
#88
Senior Member
Join Date: Oct 2015
Posts: 15,623
Bikes: 2015 Workswell 066, 2017 Workswell 093, 2014 Dawes Sheila, 1983 Cannondale 500, 1984 Raleigh Olympian, 2007 Cannondale Rize 4, 2017 Fuji Sportif 1 LE
Mentioned: 144 Post(s)
Tagged: 0 Thread(s)
Quoted: 7702 Post(s)
Liked 3,615 Times
in
1,906 Posts
When it comes to wear you cannot compare aluminum and steel, because the causes of wear are different.
Also, just because one bike is heavy and weak does not mean all bikes wear according to weight. That is bad science.
The idea that the lightest frames wear least is actually pretty absurd. With less material to absorb vibration, a lighter frame should suffer more fatigue wear than a frame with thicker tubes.
In any case .... the thread is about frame stiffness and power transmission. This whole light weight digression is at best questionably related, and also full of very questionable ideas. Lighter bikes are not necessarily stiffer because they would fail otherwise ... they might be exactly as stiff, overall, as a heavier frame, or nearly as stiff .... but based on weight alone, you are merely guessing.
St8ffer frames transmit power more efficiently ... that is the whole point, and pretty basic. When one tries to extrapolate "faster" then one has to define "faster' in this equation. A stiffer frame Should be a more stable platform for the fork and wheels and thus handle more precisely, but this again involves too many variables to make generalizations.
For a long time the goal was lateral/torsional stiffness (to resist BB flexing---maximum power transfer) and vertical compliance, to suck up bumps (tiny seat stays.) Then with the double-edged sword of (relatively) fat tires and aero frames, the old model of "racy" frames was slain. Aero frames were heavier, and usually very stiff---but measurably faster in terms of time over race distance. Wider tires and wheels lessened the road shock so the frames were comfortable enough for pros to tolerate for the duration of a race,
As far as I can figure, the only time maximum stiffness matters is on sprints (which not all riders even do) and attacks on climbs, and I doubt any but the very strongest riders are generating the kind of wattage which is noticeably flexing a very stiff pro-level CF frame.
Nothing wrong with riders wondering where their bikes fit in on a performance scale measured against other bikes, but as I and others have said, I doubt the results would be much more than random. if one guy on one frame puts out 500 watts anat the pedals and 485 at the hub, and another 510 at the pedals and 400 at the hub .... maybe the figures would be reversed at twice the power. Maybe the one bike deflected almost as much as it was going to and the other would keep twisting or bending after a certain power level ... no one would know. plus, if the difference between pedal and hub was only a few watts, then you would have to look at everything in the drive train---too many variables, no control, shaky data.
Do it if you like though.
Also, just because one bike is heavy and weak does not mean all bikes wear according to weight. That is bad science.
The idea that the lightest frames wear least is actually pretty absurd. With less material to absorb vibration, a lighter frame should suffer more fatigue wear than a frame with thicker tubes.
In any case .... the thread is about frame stiffness and power transmission. This whole light weight digression is at best questionably related, and also full of very questionable ideas. Lighter bikes are not necessarily stiffer because they would fail otherwise ... they might be exactly as stiff, overall, as a heavier frame, or nearly as stiff .... but based on weight alone, you are merely guessing.
St8ffer frames transmit power more efficiently ... that is the whole point, and pretty basic. When one tries to extrapolate "faster" then one has to define "faster' in this equation. A stiffer frame Should be a more stable platform for the fork and wheels and thus handle more precisely, but this again involves too many variables to make generalizations.
For a long time the goal was lateral/torsional stiffness (to resist BB flexing---maximum power transfer) and vertical compliance, to suck up bumps (tiny seat stays.) Then with the double-edged sword of (relatively) fat tires and aero frames, the old model of "racy" frames was slain. Aero frames were heavier, and usually very stiff---but measurably faster in terms of time over race distance. Wider tires and wheels lessened the road shock so the frames were comfortable enough for pros to tolerate for the duration of a race,
As far as I can figure, the only time maximum stiffness matters is on sprints (which not all riders even do) and attacks on climbs, and I doubt any but the very strongest riders are generating the kind of wattage which is noticeably flexing a very stiff pro-level CF frame.
Nothing wrong with riders wondering where their bikes fit in on a performance scale measured against other bikes, but as I and others have said, I doubt the results would be much more than random. if one guy on one frame puts out 500 watts anat the pedals and 485 at the hub, and another 510 at the pedals and 400 at the hub .... maybe the figures would be reversed at twice the power. Maybe the one bike deflected almost as much as it was going to and the other would keep twisting or bending after a certain power level ... no one would know. plus, if the difference between pedal and hub was only a few watts, then you would have to look at everything in the drive train---too many variables, no control, shaky data.
Do it if you like though.
#89
climber has-been
Join Date: Dec 2004
Location: Palo Alto, CA
Posts: 7,276
Bikes: Scott Addict R1, Felt Z1
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 3572 Post(s)
Liked 3,751 Times
in
1,874 Posts
Okay, but all forms of steel still lose out to aluminum on stiffness. And stiffness in a bike frame is a bigger design constraint than strength.
#90
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
#91
climber has-been
Join Date: Dec 2004
Location: Palo Alto, CA
Posts: 7,276
Bikes: Scott Addict R1, Felt Z1
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 3572 Post(s)
Liked 3,751 Times
in
1,874 Posts
Aluminum tube is much stiffer by weight because its tube thickness is much greater than a steel tube of the same weight and outer diameter.
#92
Senior Member
Join Date: Apr 2011
Posts: 7,349
Mentioned: 43 Post(s)
Tagged: 0 Thread(s)
Quoted: 4585 Post(s)
Liked 1,739 Times
in
1,138 Posts
Aluminum has a similar relationship with carbon fiber - by the time you've made the structure as stiff and light as CF, the walls would be too thin to be useful.
Which is why there is some interest in magnesium, since it is also in the same ballpark of stiffness to weight as the other bike metals, but even lower density, so you can balloon out the tube diameters even more until you are in the CF range for ultralight weight. Read the comments by Rod Morton in this article:
https://road.cc/content/tech-news/24...-aluminium-and
The flip side to this is that it isn't a bad thing that the lightest steel and titanium bikes (Ti is currently the lightest metal frame) are more flexible than fat aluminum. That's kind of what this thread is about. But for a metal with no fatigue limit like aluminum, it really helps that it does its best work at very high stiffness.
Anyone remember when Klein offered frames with boron reinforcements?
#93
Senior Member
The high stiffness of aluminum tubes is achieved by giving them larger diameters than typically used for steel: this is practicable because the lower density means you can use thicker walls for the same weight, which reduces susceptibility to denting.
#94
Grupetto Bob
Join Date: Sep 2020
Location: Seattle-ish
Posts: 6,468
Bikes: Bikey McBike Face
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 2699 Post(s)
Liked 5,974 Times
in
3,063 Posts
You're stretching with that "entire industry" business. Yes, there are some companies in the industry that make marketing claims that stiffness = fast, but that in itself isn't evidence that all companies believe this nor is it evidence that the basic premise is even true.
__________________
Road 🚴🏾♂️ & Mountain 🚵🏾♂️
Road 🚴🏾♂️ & Mountain 🚵🏾♂️
#95
Senior Member
Join Date: Jun 2004
Location: Torrance, CA
Posts: 3,061
Bikes: Homebuilt steel
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2213 Post(s)
Liked 428 Times
in
340 Posts
Would you rather have threads on the same topic endless repeated? Personally, I'm okay with someone bumping an old thread. Doing so is evidence that they are SEARCHING for information, which is something many don't do.
Likes For Nessism:
#96
Senior Member
Join Date: Jul 2009
Posts: 6,067
Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 4018 Post(s)
Liked 7,530 Times
in
3,031 Posts
#97
climber has-been
Join Date: Dec 2004
Location: Palo Alto, CA
Posts: 7,276
Bikes: Scott Addict R1, Felt Z1
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 3572 Post(s)
Liked 3,751 Times
in
1,874 Posts
Techno gibberish: Bending stiffness is a function of a tube cross section’s moment of inertia, which increases with the outer radius to the fourth power, and decreases with the inner radius to the fourth power.
Bending stiffness rises rapidly as you increase the wall thickness (decrease inner radius).
Some back of envelope math suggests that for a 1” OD tube, an aluminum tube’s moment of inertia would be about 3.5 times greater than a steel tube of equal mass. Quite a bit stiffer.
#98
Senior Member
Similar stiffness in tension and compression, but I’m pretty sure the bending stiffness is markedly higher for aluminum tube of the same outer diameter and mass.
Techno gibberish: Bending stiffness is a function of a tube cross section’s moment of inertia, which increases with the outer radius to the fourth power, and decreases with the inner radius to the fourth power.
Bending stiffness rises rapidly as you increase the wall thickness (decrease inner radius).
Some back of envelope math suggests that for a 1” OD tube, an aluminum tube’s moment of inertia would be about 3.5 times greater than a steel tube of equal mass. Quite a bit stiffer.
Techno gibberish: Bending stiffness is a function of a tube cross section’s moment of inertia, which increases with the outer radius to the fourth power, and decreases with the inner radius to the fourth power.
Bending stiffness rises rapidly as you increase the wall thickness (decrease inner radius).
Some back of envelope math suggests that for a 1” OD tube, an aluminum tube’s moment of inertia would be about 3.5 times greater than a steel tube of equal mass. Quite a bit stiffer.
Let's say that we're using 1" OD tubing in steel with a .4mm (.01575") wall thickness. Then (1^4 - (1-.01575)^4)) = .06152.
Now, suppose we're using an aluminum alloy that's 1/3rd as dense and 1/3rd as stiff. So, we thicken those walls by a factor of 3 (bringing them to .047244").
Now: (1^4-(1-.047244)^4) = .202795.
Edit per choddo's post below: Then we divide by 3 for the stiffness factor = .058667.
So it's .06152 versus .058667, with the steel tube being stiffer.
//==================================
The above example is simple, but the way it scales for weight isn't quite right. If I fudge the inner diameter of the aluminum tube downward so that the mass-per-length of the tube is equal (wall thickness ends up ~.048024"), then the stiffness of the aluminum tube ends up calculating to .0596.
Still less stiff than the steel tube of equivalent outer diameter and mass density.
Last edited by HTupolev; 02-07-24 at 02:32 AM.
#99
Senior Member
Join Date: Apr 2019
Location: UK
Posts: 1,405
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 682 Post(s)
Liked 456 Times
in
340 Posts
As I said before, google. It’s much more likely that a new poster having just found something here will zombiefy an old post that they’ve been linked directly into and don’t even notice the date among all the cruft that forums litter around, than a regular visitor would.
#100
Senior Member
Join Date: Apr 2019
Location: UK
Posts: 1,405
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 682 Post(s)
Liked 456 Times
in
340 Posts
Could you show the math? I'm electrical and not mech, so it's very possible that I have this totally wrong, but: some of this seems to linearize quite a bit when the wall thickness is small relative to the diameter, which seems to be true for bicycle tubing.
Let's say that we're using 1" OD tubing in steel with a .4mm (.01575") wall thickness. Then (1^4 - (1-.01575)^4)) = .06152.
Now, suppose we're using an aluminum alloy that's 1/3rd as dense and 1/3rd as stiff. So, we thicken those walls by a factor of 3 (bringing them to .047244").
Now: (1^4-(1-.047244)^4) = .202795.
Then we divide by 1/3 for the stiffness factor = .058667.
So it's .06152 versus .058667, with the steel tube being stiffer.
//==================================
The above example is simple, but the way it scales for weight isn't quite right. If I fudge the inner diameter of the aluminum tube downward so that the mass-per-length of the tube is equal (wall thickness ends up ~.048024"), then the stiffness of the aluminum tube ends up calculating to .0596.
Still less stiff than the steel tube of equivalent outer diameter and mass density.
Let's say that we're using 1" OD tubing in steel with a .4mm (.01575") wall thickness. Then (1^4 - (1-.01575)^4)) = .06152.
Now, suppose we're using an aluminum alloy that's 1/3rd as dense and 1/3rd as stiff. So, we thicken those walls by a factor of 3 (bringing them to .047244").
Now: (1^4-(1-.047244)^4) = .202795.
Then we divide by 1/3 for the stiffness factor = .058667.
So it's .06152 versus .058667, with the steel tube being stiffer.
//==================================
The above example is simple, but the way it scales for weight isn't quite right. If I fudge the inner diameter of the aluminum tube downward so that the mass-per-length of the tube is equal (wall thickness ends up ~.048024"), then the stiffness of the aluminum tube ends up calculating to .0596.
Still less stiff than the steel tube of equivalent outer diameter and mass density.
I thought this was an interesting effect of using imperial because of course 1^4=1
If you do the same calculation in metric I get
almost twice the value, for your densities and stiffness, for steel (1.96e-7) vs alu (1.21e-7)
I might have made a mistake though.
This site suggests aluminium is indeed about 1/3 the density of steel. Not sure about stiffness
https://www.random-science-tools.com...beams/tube.htm