Project SSR, Part Two: The Science of Singlespeeds
Posted: June 3rd, 2008 by Lath CarlsonSinglespeed bikes are the most efficient bikes in the world (except highwheelers). From the hour record , to Jennie Reed’s 2008 Keirin World Championship, when looking to getting the most return for your effort nothing beats one gear. Most hillclimb records have even been set on singlespeeds. Why? Drivetrain efficiency and weight.
What is efficiency? Scientists have a number of ways to define efficiency, for our purposes we will use “the amount of energy that the rider applies to the bike, which moves it forward.†The greater the forward movement you get out of your bike, with the same amount of effort going in, the greater it’s efficiency. A singlespeed drive bike is more efficient then a geared bike because less of your energy is “lostâ€, meaning used up by the bike without helping you move father forward. This is due to rear derailleur tension and the weight of the drivetrain components; the chain, cogs, chainrings, and derailleurs. Surprisingly, research has shown that there is very little friction added to the chain due to routing it through the pulleys of the rear derailleur. The decreased efficiency seems to have more to do with the tension applied to the chain, and the added weight. A bicycle chain is extremely efficient, about 98% of the energy applied to the chain is used to move you forward, and this is barely decreased on a geared bike. So we must look at the whole drivetrain to find the rest of the difference.
As any weight weenie will tell you, the weight called “rotational weight†is the most important on your bike. Rotational weight is the weight of the parts which spin, like your wheels and cranks. This type of weight is so critical because any effort we put into spinning it around does not help move us forward. The most important thing about rotational weight is that the further it is from what it is spinning around, the more energy it takes to spin it. You can easily feel this by taking a bike part and putting it on a board with a can, or beer bottle, under the middle of the board. First put the part near the bottle and push down on the other side a few inches away, pretty easy right? Now move the part to the end of the board and push down in the same place as before, see how much heavier it feels? When we look at wheels in a future post this will be a critical concept. In terms of a geared drivetrain this idea is important too. The chainrings, cogs, derailleur pulleys, and longer chain on a geared bike are all rotational weight and decrease its efficiency. That’s why just riding a geared bike without shifting feels nothing like riding a singlespeed bike. The added weight of the derailleurs, shifters, and cables is not rotational, but still adds to the overall weight of the bike. Overall weight is also important because a lighter bike accelerates faster and climbs with less effort. This is due to inertia, gravity and mass. Inertia is the resistance which must be overcome to get something moving. The greater the mass of the object, roughly the same as its weight, the more resistance there is. Greater mass also makes it harder to overcome gravity, the effect you encounter climbing. Yet, since all objects near earth fall at the same rate a light bike will not make you descend any slower. Without as many drivetrain components a singlespeed is more efficient, and lighter.
Being lighter and more efficient then a geared bike is a good start, but how do we get the most out of these advantages? We can start be further increasing drivetrain efficiency, reducing more weight, and adjusting our riding style to take advantage of these differences. We can also make some changes which apply equally to all types of bikes. All of these topics will be looked, in detail, in future posts.
For more about this series, see the first post here.
June 4th, 2008 at 3:36 pm
Using this logic, removing the transmission improves efficiency, so if I remove the transmission from my car and install a direct drive, then the MPGs should go up? Seems to me that cars are getting transmissions with more and more gears to improve efficiency, with the limit being the infinitely variable transmissions offered on some cars.
Admittedly, extra weight and drag do not enhance effienciency, but what about matching the characteristics of the motor to the load? Using a transmission allows a smaller motor to be used under high load, or allows the motor speed to remain in the motor’s ’sweet spot’ while the load changes.
June 4th, 2008 at 6:49 pm
Apples and oranges. When common humans begin turning thousands of RPM then a transmission will prove beneficial. Direct drives are likely more efficient at low speeds because of their lack of extraneous moving parts.
Having fewer moving parts on a bicycle drivetrain is still more efficient then running extra links through the pullies of a der.
June 4th, 2008 at 7:41 pm
jhl99, you have a valid point, for gas powered internal combustion engines. Gas engines have a fairly limited powerband, especially so for 4 cylinder engines. Diesels, V-twins, single cylinders, and steam engines all have different and often much longer powerbands. That’s why steam locomotives didn’t need transmissions to reach 75mph, or my Ducati for that matter! You see this with the new electric cars too, no transmissions are needed. As you point out, you must match the motor to the load. That will be a topic of a future post, when we look at the human engine. Thanks for your interest.
June 4th, 2008 at 8:44 pm
I don’t know, seems like a lot of things are being glossed over to make the claim that single speeds are most effecient…. maybe they are one theoretical operating point and set of conditions, but outside of that operating point, they become sub-optimal… Right?
June 6th, 2008 at 6:12 pm
As a SS rider (both road and mtb), there seems to be a mental advantage to riding single speeds. When I ride a geared bike I’m usually thinking about which gear is going to be most efficient for this climb, etc, etc. I usually run faster times on my fixed geared versus my geared (on the same loop) mostly because, and this is just my opinion, there’s nothing else to concentrate on except the body and not the bike. By riding a SS, it’s the rider who becomes more efficient and not necessarily the bike. In your mind you know your never in the “right” gear, so you’ve already mentally prepared yourself for the all out battle that your going to face. The geared rider knows that he/she can simply change gears at any given time, which can be an advantage and/or disadvantage since they can either up or down their game at the flick of a lever.
June 7th, 2008 at 7:51 am
In comparing a bicycle (chain and cog) transmission to an auto (intermeshing gears), I think that the drag caused by a deraileur is more analogous to carrier bearings and u-joints.
Just a thought I had while going to bed last night.
-Rob.
June 7th, 2008 at 10:01 am
As someone who was committed to geared bikes for years, I was skeptical when I first rode a singlespeed, but have definetely found them to be more efficient. For me the proof was riding porcupine rim and smoking everyone to the top, and then beating them to the car by a half an hour. Most of the people I ride with are better riders than me, and I usually hang out near the back, even on easier rides. Just goes to show that we have handicapped ourselves as riders by leaning on all those gears and all that suspension.
June 10th, 2008 at 6:13 am
jhl99, your unease with the discussion of efficiency is not misplaced. Lath mentions it parenthetically in his writeup: the most efficient bicycle setup is the “highwheeler” with no chain at all, i.e., direct drive to the wheel (well, at least from the perspective of drag introduced by the chain). But he is not advocating a race bike with a 51.5″ wheel (based on the 32/18 he used in his most recent race report). Why? Because the advantages the drivetrain offers him are worth the loss of efficiency. That’s your argument for the tranny in your car. His goal is to promote single speed bikes, so he did not attempt to address the possible advantages of having those gears.
June 10th, 2008 at 10:12 am
It’s simple folks, direct drive is always the most efficient. The only variable that you must get right is the gear ratio.
If you are doing a 200m sprint, then 52+:14 ratio. Climbing hills here in Santa Cruz on a 29er, 32:20.
The only scenario where a SS fails is when you have varied terrain. If I’m climbing hills on my mtb 32:20 and kickin ass and taking names great. Once I hit the flats though, it’s spin spin and watch the folks I just dropped pass me by.
A great example of this is racing the Sea Otter Classic. I rode SS in a geared race but at the end lost a placing in the top 10 of my class because I could out climb and maneuver my closest competitor but, once we hit the final stretch on a downhill and flat paved course, I could not compete with his highest gear and thus slipped back a placing.
June 10th, 2008 at 10:56 am
jkoutrouba, just to clarify, I’m not looking to promote singlespeeds, I’m more interested in looking at them objectively. I will not be addressing the advantages of geared bikes directly since these advantages are generally accepted. Fire Horse has a very good point, one that I’ll be addressing in future posts. The types of courses and terrain you ride will have a big effect on whether a singlespeed will be competitive, regardless of efficiency. A good counterexample to yours is when my friend Mike Yozell won the 35-39 Expert National Championship last year at Mt. Snow on a singlespeed. That course is mostly up and down and it was muddy last year, perfect for the singlespeed. Check out my race reports for some more discussion of gearing and courses.
June 10th, 2008 at 11:26 am
Presenting anecdotal evidence as ’science’, is where things start to go sour.
Say, someone actually did some ’science’ and put PowerTap watt meters on equalivalent bikes, one geared and one single speed. And say we had a 2 magical riders (or simulated riders) that where identical and could put out 300 watts for a while, then 250 then 200 for a given a time period. And what if we through in real riding conditions: flats, hills, headwinds…etc. Who would travel farther considering the both riders would put in the same power?
If your answer is single speed, then please explain why the gear bike was invented in the first place.
If you never change speed, never get tired, never have a head wind, never draft another cyclist, never go up or down a hill then probably a single speed is more effecient than a geared bike.
June 10th, 2008 at 2:09 pm
jhl99 is correct, and the evidence (such as it is) drops way too much out of the picture. It’s when I see things like this that also “prove” what 29″ wheels are so great I become nauseated.
First, transmissions are needed because you want to go faster while maintaining certain RPM of the powerplant. For gas and diesel engines, this is needed. However, for electric motors it isn’t. Yet, in both cases you *still* have gear multipliers in most cases (just as you would with the singlespeed) and they aren’t 1:1. People also have a limited range to produce torque efficiently, and that varies greatly with the person. So in that case, a singlespeed is greatly limiting. I mean, for anyone that’s won a race on a SS while in the same class as geared bikes, do you think you couldn’t have gone *faster* if you had more gears? Really? Not even on downhill sections? My friend picked up a shop sponsorship, and while the bike he had to ride (Giant Anthem X1) was from a company that didn’t sell singlespeed bikes, he simply kept the bike in the combo that’s close enough to his SS and would only shift the big ring for descents. That alone made him faster than he would have been. Plus, he gets suspension thrown in there to boot. Hey, if you’re gonna need a fake derailleur to support SS with suspension, you might as well use one with gears (as covered later).
Of all the people I’ve seen that like singlespeeds (myself included), they all state one of four things as positive:
1. “I’m closer to the bike” (whatever, damn hippie
)
2. “It forces me to go faster up hill” (learn self discipline)
3. “It’s quieter” (I listen to music, so it’s not that much of an issue; however I use a DuraAce chain and Dumonde light so no problem there)
4. “It’s more reliable” (I’ve got over a year on my current drivetrain with simple wear items, yet my singlespeed friends blow out freewheels and crack frames regularly due to severe driveline torque and lateral frame stresses involved)
Now, I get two more things I hear from them in the negative:
1. “I hope replacement knees get better”
2. “Wait for for me/don’t pedal so fast” (when going downhill, and even on flat sections for long periods of time)
I tried singlespeed riding, and while I can understand the interest in it I would rather just have the gear selection available and push a taller gear when I choose to. That way, I can punish my knees as much as he does on the climb, yet relax with a much slower spin (or burst with a big ring and then coast) on the way down.
Oh, and going “true” singlespeed will also drop the use of long travel (well, more than 2″ really) frames. Otherwise you wind up with pseudo-derailleurs that add weight, “rotational weight” (really, you couldn’t use “rotational mass,” “initial torque,” “prevailing torque,” or any other correct terminology?), and mechanical complexity. Also, many singlespeeds use sliders that, well, slide under severe load and have to be adjusted. Or, you get the eccentric BB which can creak, get loose, and adjustments require saddle height and fore/aft adjustments. Plus, in my case, you can get track dropouts that require ungodly amounts of torque on the hex nuts and many tend to round-off internally requiring replacement.
Singlespeed riding is fun for those that love it, but please, don’t believe that you will be that much slower with gear selection than you would be without. If you’re going to race with a carbon frame, then use components befitting a frame of that cost and do actual per-item and system drag/torque analysis. I don’t see how it would be that much worse (granted, it will be) and completely worthless for racing and rider/bike efficiency as this series would suggest.
June 10th, 2008 at 3:00 pm
It’s great to see so much discussion on this.
In response to jhl99, in the scenario you are proposing the answer would be the singlespeed rider, but it’s a bit more complicated then that. In your example the only difference between the two bikes would be the efficiency of the drivetrains, and as we’ve seen a singlespeed is more efficient in pure terms. But, that assumes that the riders could pedal at a constant wattage. That would be tough on a geared bike and nearly impossible on a singlespeed. If their bikes are identical weight then theoretically their average wattage over a course, riding at the same pace, will be equal. The big difference will be in a graph of their wattages over the course. A geared bike allows for a more constant wattage, but that is not necessarily better. Doing some real world tests like this is the kind of thing I will be trying out and I’m interested in seeing the results.
In response to “student driver”, you bring up some good points and some of which are exactly what we will be looking at in this series. I’m not looking to prove anything, as a long time geared and singlespeed rider, am just curious. I’m not sure if I would be faster or not on a geared bike, my race results so far don’t lean either way, I’ll need to finish out a season on the singlespeed before I try and draw any concussions. As mentioned before, just as geared bikes make sense in some conditions, singlespeeds do in others. In my case, riding in the PA and mid Atlantic region, a singlespeed seems to work well. I have noticed an increase in climbing speed, decrease in maintenance (I went through 3 derailluers in one month last year due to sticks) and even with past knee problems have not had any issues with the right gear on. The biggest disadvantage so far has been on flat sections, but I’m working on some high RPM drills to help with that. You might not like this, but with 29″ wheels running tubeless I see no reason for rear suspension, and I live in one of the rockiest parts of the country. I rode a 26″ Titus Racer X before . As far as terminology goes, I’m trying to make this all as non-technical as I can, my day job is building interactive exhibits for science museums and I’ll be using a similar style here. And finally, I’m sure there are more advanced engineers and physicists out there then me, and I welcome your comments and corrections.
June 10th, 2008 at 4:57 pm
Interesting article. I don’t find it difficult to believe that there is some additional efficiency in a single speed setup compared with a derailleur setup. I don’t think that it is enough to make a noticeable increase in rider speed. Overall, I believe I do ride faster since switching to single speed riding and racing exclusively last summer. I’ve attributed it to the fact that I have to ride differently. Climbs require an attack rather than gearing down and spinning up. (Try racing single-speed and getting stuck behind a geared rider on some hilly single-track and you’ll know exactly what I’m talking about.) The other factor I think plays into being faster on a single-speed is that I tend to use my whole body on the single-speed. I grew noticeably stronger in my upper body after the first couple of months after switching over. As others have mentioned, I also noticed that I’m more focused (or less distracted) since there’s no shifting to consider – I just have to think about putting down power and picking a line. I’d say that it has made me a stronger rider that I’ve ever been and I started racing as an Expert back in ‘90.
As for knees, I’ve found that I spin more riding the single-speed than I did when riding with gears. I live and ride in Vermont so that’s not from a lack of hills. I guess I must have a tendency to shift up and mash the pedals. I have plenty of knee damage from years of ski racing when I was younger so I’m a bit cautious about the issue now.
June 11th, 2008 at 10:30 am
Lath_C, god bless ya. You are much, much more of a patient man then I ever will be. The bummer about posting messages is that not every one has the same amount of knowledge, experience, or objectivity about a matter, especially cycling. This is why I don’t even…bother with Craigs List Discussions or mtbr.com any more.
Some folks have their biases and no matter of convincing or “scientific” evidence will change their opinion weather it be the advantages of 29ers or singlespeeds. Look at all the anti-29er and anti-SS articles in that rag MTB Action. But still the market keeps growing and is here to stay in both sectors. They called it a fad but more people and companies are becoming and staying believers.
As for “scientific” evidence, studies can be just as wacked out as some of the uninformed or emotional opinions on here. For the uninitiated, so called “scientific” studies are with their flaws as well. Why, you involve people. Common flaws:
-vetting the peer review process: the opinion or unpaid, anonymous, and unaccountable individuals whose opinions are kept confidential making it impossible to know the parameters of the reviews.
-research articles are often…either contradicted, exaggerated, or seriously misquoted.
-reviewers are usually not allowed to contact the researchers directly to ask questions, and they do not try to replicate the research..wtf?
-granted, the system has often had successes but, peer review also lacks consistent standards.
-procedures vary among the world’s 10,000 or so journals. A peer reviewer often spends about four hours reviewing research that may have taken months or years to complete.
I could go on ad nauseum but I’ll spare us all. This brings us back to plain ol’ tired and true…common sense folks. The most elegant, reliable, and efficient designs are the most simple.
The more factors/processes involved in a mechanical process the more opportunity for friction, resistance, damage, reduced reliability, increased maintenance, increased cost, etc.
Addressing “Why was the derailleur created?” it was to make the cycling experience more accessible to masses which translates into more sales. The old timers, even on the Tour de France, originally rode fixed gears. This obviously wouldn’t appeal to your weekend warrior type. So enter the human factor.
You can either make the machine easier to ride, to do a particular task (ie ride over a mountain range) and thus appeal to more people. Or you can….evolve the rider. This transformation however involves sweat, pain, will power, and don’t say it…change.
Anyway Henry Sturmy (developer of the Sturmy-Archer hub) said it best “The only mechanically correct place for a gear in a pedal cycle is the hub …”.
June 11th, 2008 at 8:32 pm
AntEater, well said! Your experience just about mirrors my own, which is what led me to starting writing about this. All of the points you mention we will be taking up.
Fire Horse, I agree with you regarding MTBR type flame wars, if only some of those people would ride as much as they post. So far the discussion here has been more insightful, just what I would expect from the readers of the Rag. My plan here is to start with the “scientific” and theoretical and move on to real world testing. I agree with most of what you bring up regarding this type of research, but it’s a good place to start the conversation at least. Try not to get too upset by my next post! Thanks for your comments regarding the reason for derailleurs, I heard one from an old tour rider that they were only good for riders over 45.
June 11th, 2008 at 10:34 pm
Recently i was coming back from a quick blast with my road bike and just glanced at this “totally cool” wannabe hipster girl riding down the sidewalk.She glared at me (I imagine-i was going up a hill,she wasn’t pedaling),briefly.
I imagined my bike wasn’t “cool enough” and i probably didn’t “get it”.I think “fixies” are getting like that.I will probably re-build my ‘98 Stumpjumper single ’cause I want to ride it without making it something that would lack against my other stuff.I love the frame too.It’s that or mount it on a wall.I shift often and go thru gears fast.No way you could hold my wheel.Shouldn’t be “too cool” and mis-represent data to support what we all love.
June 18th, 2008 at 12:21 am
When discussing mechanical technical issues it is easy to just approach things from an engineering standpoint , IE “fewer parts equals greater efficiency “. However if the vehicle in question is human powered it is valuable to consider the psycological aspects and advantages to the system in question .
Riding a singlespeed trains you to work harder . No bailout gear means you have to pedal harder and or stand up more of the time , period . Everybody knows that standing up uses more energy , so you end up riding seated longer than you would have before . Then you stand on the pedals . While lacking ultimate efficiency , standing does crank up your cardio workout and still provides a benefit you will miss when you just shift another gear .
Riding a singlespeed gives you a psycological advantage over your competitors . They look at your bike and they KNOW your bike weighs 1 or 2 or 3 pounds less than theirs . They think that if you are wiling to ride the same course as them with only one gear that you must be tough . They know that if they snap a shifter cable or smash their derailler on a rock , they are toast and you will just motor on by .
The thing I like most about riding a singlespeed is it teaches you new skills . I started riding a singlespeed when I was 46 years old ( Im 50now ). I was burning out on riding – commuting , trail riding , I wasnt progressing in my fitness any more . When you have a job and a family there is only so much time and motivation in a day , you look for the next advantage . Since I cant afford a titanium anything , I downsized my drivetrain . Within a year , I felt 10% stronger than before . I spent less time wrenching than before . And people notice and make positve comments about your riding just because you left your gears at home . The best part is that now when I ride my geared bike I have skills and fitness I never would have developed without my singlespeed . And I can ride uphill AND downhill 2 or 3 gears higher than I did before .
SO , do you want to scrape together more money for some carbon fiber crap or doyou want to build a bigger motor ?
June 19th, 2008 at 1:33 am
I thought of an interesting experiment around drivetrain efficiency. I hypothesize that a large proportion of efficiency loss can be attributed to chain line. A straight chainline being the most efficient of course, but how much does being 1mm, 2mm etc affect it. I came up with this theory based on chain wear due to cross chaining (poor chainline.)
The experiment i have in mind to prove/disprove this theory involves a powermeter type cassette hub, a trainer, a derailleur and a bike with horizontal drop outs.
Step 1. Set the bike up an approriate gear (i.e; 70″) the actual gear needs to be the same throughout the experiment but the experiment can be repeated with various gear combinations to see if perhaps a higher gear ratio is more affected by chainline than a lower gear ratio. We are simply looking at the effects of chainline. Set the chainline using a single gear with spacers on either side and a derailleur lined up vertically with the gear. this is the ‘Zero’
Step 2. Pedal the bike at constant speed or power. Record observed power or speed.
Step 3. Move the gear one spacer in or out (+/- 1). Realign deraileur and repeat step 2.
Step 4. Repeat Step 3 until the last gear is in the last position and then repeat step 3 in the opposite direction from the zero position.
Step 5. Remove derailleur and tension chain by shortening chain and moving the wheel. Try to keep the chain stay length as close as possible to that used in steps 1-4.
Step 6. Repeat Step 2.
Step 7. Move the gear one spacer in or out (+/- 1). Realign deraileur and repeat step 2.
Step 8. Repeat Step 3 until the last gear is in the last position and then repeat step 3 in the opposite direction from the zero position.
Now if we graph the results of power or speed while the other is held constant versus gear position (+/-1, +/-2 etc) we shoud see the effects of chainline efficiency for both single speed and geared set ups and what effect a derailleur has on efficiency and if it that effect varies with chain line.
Hope this interests somebody. I don’t have a power meter of any type so I am unable to do the experiment.
June 19th, 2008 at 8:10 am
That’s a great idea for a study, I don’t believe it’s been done yet. Although because of the very small differences it is probably more realistic to do it with a constant torque motor drive suppling the power and a load cell at the cassette (Powertap) to measure differences in delivered power. Any engineering grad students out there? High schoolers needing a science project?
The closest thing I’ve seen is a study that found that:
“Chain drive efficiency has been studied to understand energy loss mechanisms in bicycle drive trains, primarily for derailleur-type systems. An analytical study of frictional energy loss mechanisms for chain drives is given along with a series of experimental measurements of chain drive efficiency under a range of power, speed and lubrication conditions. Measurements of mechanical efficiency are compared to infrared measurements indicating that frictional losses cannot account for the observed variations in efficiency. The results of this study indicate that chain tension and sprocket size primarily affect efficiency and that non-thermal loss mechanisms dominate overall chain drive efficiency.”
From: Effects of Frictional Loss on Bicycle Chain Drive Efficiency
Journal of Mechanical Design — December 2001 — Volume 123, Issue 4, pp. 598-605