Bike Design
July 11, 2004 10:22 PM Subscribe
[BikeFilter] Two questions about basic bike design:
(1) I understand gear ratios, but does the actual size of the gears make a mechanical difference?
(2) Why aren't the seatposts on bikes square?
(1) Let's say I had a 2:1 ratio between the front and back gears on a bike. Now, I know that if the two gears were 6" and 3", that they would be heavier than gears that were 4" and 2" (which is a really important issue for serious cyclists), but that they would be stronger, and that past a certain point, a gear that's too small just can't have teeth that are big enough or numerous enough to work--that's not what I'm curious about. What I would like to understand better is whether there's any mechanical advantage (like work or torque, for example) that's accrued by making gears at a given ratio larger or smaller.
(2) This is much more less academic to me--why are the posts on bike seats round? I understand that there's probably a manufacturing advantage in making most of that section a round post, and I understand that a cylinder is basically a stronger shape than a long square box for the the structural span of that post. But the design of modern bikes is certainly complex enough that you could the upper opening square for the top 2 or 3 inches of the seat post, and then put a square rod coming down out of the seat, which would hold the seat square to the bike frame.
Don't you want the seat to always be as square as possible to the bike? Why have a design that allows the seat to twist back and forth, and then use a pretty ineffective collar clamp to fight that? Sure, if the seat geometry gets screwed up in a crash or something, you can whang a round post over a couple degrees to compensate for a crooked seat, but isn't that a pretty remote issue for most recreational bikers?
(This is an issue of no small interest to me--I'm a big, pretty strong guy, and the seats on any new bikes have always started to twist back and forth under me after those little pressure collars start to lose their clamping ability. I can keep tightening the screws on the clamp, but it's like the washer in an old faucet that's started to drip--once it's started to go, it's a losing battle to keep it tight. You can replace the collar, but you're still postponing the inevitable. Why not just make a square seat post, that drops directly into a square post on the bike?)
(1) I understand gear ratios, but does the actual size of the gears make a mechanical difference?
(2) Why aren't the seatposts on bikes square?
(1) Let's say I had a 2:1 ratio between the front and back gears on a bike. Now, I know that if the two gears were 6" and 3", that they would be heavier than gears that were 4" and 2" (which is a really important issue for serious cyclists), but that they would be stronger, and that past a certain point, a gear that's too small just can't have teeth that are big enough or numerous enough to work--that's not what I'm curious about. What I would like to understand better is whether there's any mechanical advantage (like work or torque, for example) that's accrued by making gears at a given ratio larger or smaller.
(2) This is much more less academic to me--why are the posts on bike seats round? I understand that there's probably a manufacturing advantage in making most of that section a round post, and I understand that a cylinder is basically a stronger shape than a long square box for the the structural span of that post. But the design of modern bikes is certainly complex enough that you could the upper opening square for the top 2 or 3 inches of the seat post, and then put a square rod coming down out of the seat, which would hold the seat square to the bike frame.
Don't you want the seat to always be as square as possible to the bike? Why have a design that allows the seat to twist back and forth, and then use a pretty ineffective collar clamp to fight that? Sure, if the seat geometry gets screwed up in a crash or something, you can whang a round post over a couple degrees to compensate for a crooked seat, but isn't that a pretty remote issue for most recreational bikers?
(This is an issue of no small interest to me--I'm a big, pretty strong guy, and the seats on any new bikes have always started to twist back and forth under me after those little pressure collars start to lose their clamping ability. I can keep tightening the screws on the clamp, but it's like the washer in an old faucet that's started to drip--once it's started to go, it's a losing battle to keep it tight. You can replace the collar, but you're still postponing the inevitable. Why not just make a square seat post, that drops directly into a square post on the bike?)
Best answer: I'm a big guy too and what I've done on my bikes is to fit a specialized seatpost clamp that has no quick release and must be tightened with an Allen wrench. These things arent usually more than $20 or so.
I think the square post thing doesn't work because a circular post can have equal pressure all along the circumference of the post whereas with something square, the corners would have more pressure - i.e. I think it is harder to build and keep tight.
posted by gen at 10:34 PM on July 11, 2004
I think the square post thing doesn't work because a circular post can have equal pressure all along the circumference of the post whereas with something square, the corners would have more pressure - i.e. I think it is harder to build and keep tight.
posted by gen at 10:34 PM on July 11, 2004
What Gen said. You would probably want to use a different system for securing the post in the seat tube. I'm not an engineer, but a tiny increase in seat tube circumference allows repositioning of a round post, but with a sqaure one you may practically need a notch and clamp on two non-parallel sides, as the loosening of one will only give you increased tolerance in one dimension.
posted by planetkyoto at 11:36 PM on July 11, 2004
posted by planetkyoto at 11:36 PM on July 11, 2004
Best answer: (1) Ah, the sprocket size enigma. Obviously, all else being equal (ie, assuming a perfect imaginary chain), equal sprocket ratios will yield equal efficiencies regardless of sprocket size. In the real world, however, it turns out that there are measurable differences in the efficiencies of large vs. small drive train sprocket pairs.
Obviously chain/sprocket interplay is the culprit, introducing variables like articulation angle, chain speed, chain tension, friction and something-else-I'm-sure-I'm-forgetting. I've read or heard about a couple of studies, and their conclusions were maddeningly contradictory. So I just chose the conclusion which made me happiest; ie, that there is a very tiny advantage to large sprocket pairs.
I remember that the study which chafed me indicated that higher chain tension (which accompanies smaller sprocket sizes) resulted in greater efficiencies; while both studies confirmed this odd bogglebit, the one I happily swallowed found that larger sprocket sizes were still and in spite of slightly more efficient. But it really is a tiny little advantage, a lot less than 1% as I recall, and not really worth fretting about...
...unless you are a superstitious nonrational cyclist, like me, who would swear on his dear grandma's golden braids that larger sprocket pairs make for a significantly more efficient transfer of energy.
(2) The seatpost is round so that you can neatly wind a bit of duct tape around it for emergency repairs. Whassamaddawidya?
posted by Opus Dark at 3:38 AM on July 12, 2004
Obviously chain/sprocket interplay is the culprit, introducing variables like articulation angle, chain speed, chain tension, friction and something-else-I'm-sure-I'm-forgetting. I've read or heard about a couple of studies, and their conclusions were maddeningly contradictory. So I just chose the conclusion which made me happiest; ie, that there is a very tiny advantage to large sprocket pairs.
I remember that the study which chafed me indicated that higher chain tension (which accompanies smaller sprocket sizes) resulted in greater efficiencies; while both studies confirmed this odd bogglebit, the one I happily swallowed found that larger sprocket sizes were still and in spite of slightly more efficient. But it really is a tiny little advantage, a lot less than 1% as I recall, and not really worth fretting about...
...unless you are a superstitious nonrational cyclist, like me, who would swear on his dear grandma's golden braids that larger sprocket pairs make for a significantly more efficient transfer of energy.
(2) The seatpost is round so that you can neatly wind a bit of duct tape around it for emergency repairs. Whassamaddawidya?
posted by Opus Dark at 3:38 AM on July 12, 2004
Response by poster: See, that's what I'm talking about--I knew there had to be some kind of arcane debate about gear size going on in the cycling/nerd community. (I mean that as an affectionate compliment.)
[And that "duct tape" argument is totally convincing. Totally.]
posted by LairBob at 5:45 AM on July 12, 2004
[And that "duct tape" argument is totally convincing. Totally.]
posted by LairBob at 5:45 AM on July 12, 2004
Response by poster: Oh, and gen, thanks for the pointer on the better clamp--I'll definitely look for it.
I had been wondering whether square seat posts were more likely to rattle, and I can see they might, but there are so many high-tech advances on the average bike nowadays (like shock absorbers in the seat posts) that I have a hard time believing that _something_ couldn't be done to lock the seat square to the frame.
So let me throw that larger question back out to anyone in the cycling commumity who's interested in responding--square post or not, in something as technically sophisticated as the modern bicycle, why hasn't some means evolved of keeping the seat dead square to the frame? Is there any advantage to having a seat that can rotate against the frame, or is it just something that hasn't been dealt with?
posted by LairBob at 5:53 AM on July 12, 2004
I had been wondering whether square seat posts were more likely to rattle, and I can see they might, but there are so many high-tech advances on the average bike nowadays (like shock absorbers in the seat posts) that I have a hard time believing that _something_ couldn't be done to lock the seat square to the frame.
So let me throw that larger question back out to anyone in the cycling commumity who's interested in responding--square post or not, in something as technically sophisticated as the modern bicycle, why hasn't some means evolved of keeping the seat dead square to the frame? Is there any advantage to having a seat that can rotate against the frame, or is it just something that hasn't been dealt with?
posted by LairBob at 5:53 AM on July 12, 2004
the seatpost is round because its lighter than a square seatpost of equal strength.
posted by Fupped Duck at 5:59 AM on July 12, 2004
posted by Fupped Duck at 5:59 AM on July 12, 2004
A square seatpost also has a higher drag coefficient than a round one.
posted by mbd1mbd1 at 6:52 AM on July 12, 2004
posted by mbd1mbd1 at 6:52 AM on July 12, 2004
Best answer: Seatposts are round because that's the strongest and lightest form there is. Square seatposts of equal tube thickness would break. Stresses concentrate in corners when the part is formed and during use. Sharp corners are bad news. Square seat posts would constantly be failing, tearing along one of the corners. The alternative would be to make the post unacceptibly heavy (tens of pounds more, probably).
You can make the seatpost oval, and you can buy bikes that way. There's no standard oval post sizes though, so you pay through the nose for it.
Still, you shouldn't be having twisting problems with your post. Olympic track sprinters seem to make do somehow. If your post turns and/or slips, you've probably got a sizing problem on your bike. The first thing to check is that your post is the right size for the bike. Post are graded in 0.1 mm increments. It's easy to get a post a size down. You may even find that you need a size up over the nominal dimensions of your frame. You may not need to buy a new post, though posts are cheap, a thin shim will probably do the job, plus maybe a new clamp.
If that's not it, check to make sure that the post is being clamped in the region specced by the manufacturer. Some higher-end posts have regions that are thinner-walled than the ends (to save weight). If you have an expensive ATB post with the saddle really low, then you may have this problem. You need to buy a shorter road post to solve it.
Big gear have less friction and chain wear than small gears---the links don't have to bend as far when engaging the teeth. The real choice for small gears varsus big gears is operational: do big gears get in the way? For mountain bikes, the answer is yes. A 54-tooth chain-ring will get bent on your first ride. A much smaller (and stronger) 36-tooth ring will not. So road bikes tend to have large front cogs because efficientcy is important to roadies, and ATBs have "micro-drives" and bash-rings because they want their gear systems to survive their rides.
posted by bonehead at 7:56 AM on July 12, 2004
You can make the seatpost oval, and you can buy bikes that way. There's no standard oval post sizes though, so you pay through the nose for it.
Still, you shouldn't be having twisting problems with your post. Olympic track sprinters seem to make do somehow. If your post turns and/or slips, you've probably got a sizing problem on your bike. The first thing to check is that your post is the right size for the bike. Post are graded in 0.1 mm increments. It's easy to get a post a size down. You may even find that you need a size up over the nominal dimensions of your frame. You may not need to buy a new post, though posts are cheap, a thin shim will probably do the job, plus maybe a new clamp.
If that's not it, check to make sure that the post is being clamped in the region specced by the manufacturer. Some higher-end posts have regions that are thinner-walled than the ends (to save weight). If you have an expensive ATB post with the saddle really low, then you may have this problem. You need to buy a shorter road post to solve it.
Big gear have less friction and chain wear than small gears---the links don't have to bend as far when engaging the teeth. The real choice for small gears varsus big gears is operational: do big gears get in the way? For mountain bikes, the answer is yes. A 54-tooth chain-ring will get bent on your first ride. A much smaller (and stronger) 36-tooth ring will not. So road bikes tend to have large front cogs because efficientcy is important to roadies, and ATBs have "micro-drives" and bash-rings because they want their gear systems to survive their rides.
posted by bonehead at 7:56 AM on July 12, 2004
Best answer: Opus Dark covered the sprocket-size question pretty well. A friend who was a mechanical engineer and cyclist once told me that, with a roller-chain drive, 12 teeth (gears are customarily measured in tooth count, not diameter) is considered the lowest you can go before you start getting undesirable mechanical losses from the chain links bending/unbending as they come on/off the gears. The smallest cogs I've ever seen for a bike were 10 teeth (probably limited by the size of the freewheel body).
I once had an aerodynamic bike with oval seat tube and post. It would never stay put, and eventually a framebuilder had to tap a set-screw into the seat tube. Round tubes are not only stronger, they're a lot easier to machine accurately. I'm pretty sure that it would be impossible to get uniform clamping tension around the perimeter of any non-round post, so while you might avoid the side-to-side rock, you'd be more likely to get that sinking feeling.
posted by adamrice at 8:05 AM on July 12, 2004
I once had an aerodynamic bike with oval seat tube and post. It would never stay put, and eventually a framebuilder had to tap a set-screw into the seat tube. Round tubes are not only stronger, they're a lot easier to machine accurately. I'm pretty sure that it would be impossible to get uniform clamping tension around the perimeter of any non-round post, so while you might avoid the side-to-side rock, you'd be more likely to get that sinking feeling.
posted by adamrice at 8:05 AM on July 12, 2004
Response by poster: Thanks, thanks, thanks, bh and ar...those answers totally make sense. I assumed that "seat wiggle" shouldn't be a problem on a well-set-up bike--there's a lot of big, strong serious bikers who put their machines through a lot more than I ever would. I'll look into the post size as well, before I even try and change the clamp, and that overview on gear sizes explains a lot.
posted by LairBob at 8:49 AM on July 12, 2004
posted by LairBob at 8:49 AM on July 12, 2004
The other advantage to round is that it's more tolerant of poor tolerance. Which is to say it's easier to clamp down on a too-small round post than a square or oval one.
Which, when you're building a half-million bikes in a Chinese factory, is pretty important.
posted by five fresh fish at 9:05 AM on July 12, 2004
Which, when you're building a half-million bikes in a Chinese factory, is pretty important.
posted by five fresh fish at 9:05 AM on July 12, 2004
Best answer: Like everything else in mechanical design, chainring and rear sprocket size are the result of compromise. Smaller parts are considered desirable from a weight reduction perspective, but there's two limits to how small cogs can be made. There's an obvious size limit introduced by the position and attachment of the sprocket to the other parts of the transmission, and there's an equally important limit introduced by the consideration of strength and wear.
Rate of wear is proportional to the stress applied between the parts in contact and the and the distance that the contact surfaces travel across each other. In the case of a chain drive this is determined by the torque applied to the transmission and the angle through which the individual chain links must articulate. For a fixed gear ratio and torque, the link articulation angle increases for each reduction in the number of teeth. If you do the maths, it turns out that the increase in wear with reduction in teeth is non-proportional - the chain wears out much quicker, the smaller you go. Similarly, the force applied by the chain is distributed amongst a fewer number of teeth and the sprocket teeth wear quicker as well, although in a linear relationship. End result is that everything wears much quicker with fewer teeth.
Where there's wear, there's heat, of course, so, as mentioned above, there's a correspondingly higher loss in efficiency, as well. This maybe shouldn't be over-emphasised, however. Even with fewer teeth and the lateral flex demanded by a derailleur system, the bicycle chain drive remains remarkably efficient compared to any alternative system.
Commercially, it's been tried. It's a known problem with Moulton small-wheel bikes, where smaller rear sprockets are required to give similar gear size to larger wheel bikes. Moultons at one time (still do?) used a specially made 9-tooth sprocket for their highest gear, which was notoriously fast wearing (and expensive to replace). In the late 80s/early 90s Sun Tour promoted their "Microdrive" system for mountainbikes. They offered smaller chainrings and sprockets for a weight advantage. It wasn't a low-budget line and only survived a couple of seasons before being rejected by its users because of unacceptably rapid wear. This and other commercial faliures contributed to the ultimate withdrawl of Sun Tour as a competitor to the omnipresent Shimano in the bike component market.
Weight reduction in bike transmissions is currently generally approached from a materials, rather than mechanical, perspective. Consumer-level sprockets are almost universally steel, as they should be. For racing at the professional level, aluminium has been tried, and often found to be insufficiently strong and too fast-wearing. Now titanium is favored. Some sprocket cassettes in titanium are sold as "single-use", one-race-only parts. Outrageously expensive, but no surprise that the likes of Mr Armstrong and others sometimes use them for the marginal advantage they offer.
posted by normy at 10:41 AM on July 12, 2004
Rate of wear is proportional to the stress applied between the parts in contact and the and the distance that the contact surfaces travel across each other. In the case of a chain drive this is determined by the torque applied to the transmission and the angle through which the individual chain links must articulate. For a fixed gear ratio and torque, the link articulation angle increases for each reduction in the number of teeth. If you do the maths, it turns out that the increase in wear with reduction in teeth is non-proportional - the chain wears out much quicker, the smaller you go. Similarly, the force applied by the chain is distributed amongst a fewer number of teeth and the sprocket teeth wear quicker as well, although in a linear relationship. End result is that everything wears much quicker with fewer teeth.
Where there's wear, there's heat, of course, so, as mentioned above, there's a correspondingly higher loss in efficiency, as well. This maybe shouldn't be over-emphasised, however. Even with fewer teeth and the lateral flex demanded by a derailleur system, the bicycle chain drive remains remarkably efficient compared to any alternative system.
Commercially, it's been tried. It's a known problem with Moulton small-wheel bikes, where smaller rear sprockets are required to give similar gear size to larger wheel bikes. Moultons at one time (still do?) used a specially made 9-tooth sprocket for their highest gear, which was notoriously fast wearing (and expensive to replace). In the late 80s/early 90s Sun Tour promoted their "Microdrive" system for mountainbikes. They offered smaller chainrings and sprockets for a weight advantage. It wasn't a low-budget line and only survived a couple of seasons before being rejected by its users because of unacceptably rapid wear. This and other commercial faliures contributed to the ultimate withdrawl of Sun Tour as a competitor to the omnipresent Shimano in the bike component market.
Weight reduction in bike transmissions is currently generally approached from a materials, rather than mechanical, perspective. Consumer-level sprockets are almost universally steel, as they should be. For racing at the professional level, aluminium has been tried, and often found to be insufficiently strong and too fast-wearing. Now titanium is favored. Some sprocket cassettes in titanium are sold as "single-use", one-race-only parts. Outrageously expensive, but no surprise that the likes of Mr Armstrong and others sometimes use them for the marginal advantage they offer.
posted by normy at 10:41 AM on July 12, 2004
Response by poster: Verrry interesting...thanks, normy.
posted by LairBob at 11:37 AM on July 12, 2004
posted by LairBob at 11:37 AM on July 12, 2004
I've wiped out on a bike and twisted the seat in the process. If the seat hadn't rotated because the square tube was locking it, I might not be the man I am today.
posted by smackfu at 12:34 PM on July 12, 2004
posted by smackfu at 12:34 PM on July 12, 2004
Response by poster: Point well-taken (as it were), smackfu. [Sorry.]
That's actually probably an important safety consideration, though, since it allows the structure to fail predictably and safely once forces reach a dangerous threshold.
posted by LairBob at 1:23 PM on July 12, 2004
That's actually probably an important safety consideration, though, since it allows the structure to fail predictably and safely once forces reach a dangerous threshold.
posted by LairBob at 1:23 PM on July 12, 2004
Yeah, seems to me that weight is consideration #1 when thinking about gear sizes, and then it should be a problem of wear and efficency after that.
I'm reminded of Sheldon Brown's comment on bicycle frame design...
'Billions of diamond-frame bikes have been made from tubing for over a century, and during that time, hundreds of thousands of very smart people have spent billions of hours riding along and thinking about ways to fine-tune the performance of their bikes. The tubular diamond frame has been fine tuned by an evolutionary process to the point where it is very close to perfection, given the basic design and materials.'
(Though this doesn't mean that we smart people who spend thousands of hours on bicycles should stop thinking about such things.)
His site is, as usual, a great place to go for more info...
posted by kaibutsu at 2:00 PM on July 12, 2004
I'm reminded of Sheldon Brown's comment on bicycle frame design...
'Billions of diamond-frame bikes have been made from tubing for over a century, and during that time, hundreds of thousands of very smart people have spent billions of hours riding along and thinking about ways to fine-tune the performance of their bikes. The tubular diamond frame has been fine tuned by an evolutionary process to the point where it is very close to perfection, given the basic design and materials.'
(Though this doesn't mean that we smart people who spend thousands of hours on bicycles should stop thinking about such things.)
His site is, as usual, a great place to go for more info...
posted by kaibutsu at 2:00 PM on July 12, 2004
in fact... brown invented the NANODRIVE gear system...
posted by kaibutsu at 3:40 PM on July 12, 2004
posted by kaibutsu at 3:40 PM on July 12, 2004
That Nanodrive page is great. I've got to get me one of those Citroën Monospoke Wheelsets.
posted by normy at 4:08 PM on July 12, 2004
posted by normy at 4:08 PM on July 12, 2004
Check the seat setup on this rig. They focus on it in the last half of the piece.
posted by NortonDC at 6:17 PM on July 12, 2004
posted by NortonDC at 6:17 PM on July 12, 2004
This thread is closed to new comments.
posted by LairBob at 10:28 PM on July 11, 2004