Camilo

When I'm looking at gearing available - for instance most recently when I was looking at Sram AXS gearing options - all I did was do a simple calculation of the front/back gear ratio of the highest gear and the lowest gear and compare that to the high and low gears I've been satisfied withor that my wife has been satisfied with. Whether the low is as low or lower is the most important part. The high gear just needs to be close.

I really don't know (repeat: I really don't know, not implying that nobody would know) why one would have to be more precise than that. Oh, and crank length - for me - 170-175 - is A-OK and I notice no difference, not only in how it feels, but would be meaningless for how the gearing works. Repeat: FOR ME. I have no doubt that people who compete and have the motor to be truly competitive would want to optimize these things, but for me and probably the other 99.9% of enthusiastic riders and weekend competitors... ? Probably not.

I used to play around with the online calculators, but they didn't give me any more functional information than a simple division exercise.

Road Fan

Not worrying about Sheldon Brown for the moment, sorry.

If you have discovered that 1:2.7 is an ideal ratio for you, more power to you! I assume you are riding a single speed primarily. If you have read Whitt and Sharp, you know that the frictional loss in chain/sprocket drive systems decreases as the sprockets get larger, though it is not a large effect. Yes, leverage increases when the pedal arm gets longer, but I don't think that makes any difference in rear wheel power or speed if the sprocket ratio is kept 1.2.7. The energy lost decreases as the sprocket pair gets bigger, but the change in output power is small.

PeteHski

Power is a function of how hard you pedal (pedal force) x cadence. Power is not a direct function of gearing other than choosing an appropriate gear that gives you a sensible ratio of pedal force vs cadence for a given power level. A longer pedal arm has the same effect as a lower gear i.e. it reduces your pedal force for a given power output, but it also effectively reduces your cadence since your pedalling circumference is larger. So you have to move your feet faster to maintain the same cadence and hence power. Power = Pedal Force x Crank Arm Length x Cadence.

ThermionicScott 

Pedantic note: your preferred gear ratio is 2.7:1, not 1:2.7.

I've read that industry prefers a minimum sprocket size of 16T for efficiency and wear. So either of your two ratios are just fine. The 52/19 would be slightly smoother and longer-lasting, but at the penalty of extra weight.

My fixed-gear has been 45/16 for years and years. In the beginning, I was bummed that I couldn't fit a chainring bigger than 45T on that bike, but it has proven to be plenty durable and smooth, so I'm happy with it.

Road Fan

I entirely agree. I don't know what I said that would make you think I don't agree with the basic physics here.

PeteHski

Sorry, it was just an elaboration of your point "I don't think that [crank arm length) makes any difference in rear wheel power or speed if the sprocket ratio is kept 1.2.7"
I'm agreeing with you here.

RChung

The range in drivetrain efficiency for derailleur bikes can be much larger than that; even more if the drivetrain is dirty. For single-speed track bikes, we've measured efficiency in the range of 98%; for triple-ring MTB drivetrains we've seen efficiencies in the range of low 90's up to 96%, depending on the ring-cog combination. If you're putting out 250 watts at the crank on a steep hill, that's like 10 watts difference, which is, um, a drag.

When we do aero testing on the road, the differences in drivetrain losses are large enough to be annoying and affect the precision of our estimates.

PeteHski

I wasn't really thinking of the extreme ends of the drivetrain spectrum or even differing efficiencies across the cassette and chainrings in the same drivetrain. I was thinking more about the effect of changing the size of the chainring and cassette gear for a given gear ratio in the same drivetrain - related back to the OP's question. i.e. would a 52/19 be significantly more efficient than a 43/16 on the same chain-line?

RChung

As in all things, it depends on what you mean by "significantly."

I just looked it up, I gave some data and analysis in this thread several years ago. You may have to scroll down a bit. In that thread there are also links to some articles on drive train efficiency that appeared in the IHPVA journal.

PeteHski

I had a look at your link, but I can't see any data. It looks like you compared 54/15 vs 42/12 on the same chain line, but I can't see the actual results.

My idea of "significant" would be around 2.5% i.e 5W difference at 200W output - but that's obviously completely arbitrary. I expect the OP would not be looking for the odd Watt here or there in this analysis.

RChung

Hmmm. Can you see the image in post #12 in that thread?

But you're right, I didn't post a link to the actual data used to make that chart. I just looked, and if you want I can give you a link to the raw data files. As an oddity, I see that yesterday was the 14th anniversary of when the data were collected. Must be a winter thing.

The image file is here.
The original SRM file is here.
The original PT file is here.

beng1

The pedal cranks make one revolution for every 2.7 the rear wheel makes, the cranks are the first mechanism in the driveline of the bike, so you put them first. Just as in an automobile with a 4:11 to 1 rear axle, the driveshaft, which is before the rear axle, turns 4.11 times for each revolution of the rear wheels. So you are 100% wrong.

ThermionicScott 

Think of it however you want, but the convention for bicycles is otherwise. If you say 1:2.7 on a bicycle forum, most people will assume you're talking about a very low gear for climbing hills.

PeteHski

For some reason I can't see the image in that thread and the link here doesn't work for me either.

RChung

How odd. Anyway, here's the image. You can go back to that thread to read the explanation (and the criticisms).

beng1

Single-speed has nothing to do with it. When I ride a multi-speed bike, which I do often, I pay attention to the ratios, and I have noticed the last few years I am very comfortable in the 1:2.7 range on level ground. This is why when I built up a single-speed bike the first gearing I tried was a 52/20, which is in the 2.6 range, and it worked well. Now with the 50/20 I am down in the 2.5 range, which is still fine for running around town and grades up to about 5% for me. If I go for a special ride on a level route, or run a TT on the single speed on a level course, then I will try the 18T on the back and get the bike back up in the 2.7 range so it will go 20mph or so with a cadence in the 80s.

I always clean the derailleurs on my multi-speed bikes, I think it is one of the most neglected areas on the average road bike, and those little wheels like to gum up. if anything does.

My coaster-braked single-speeder probably has a lot of drag from the grease-packed rear hub and braking mechanism, the rear wheel does not like to spin as easily as a freewheel mechanism, but after a breaking-in period it seems very loose and free and I don't think it has any more penalty than a derailleur system. If my health holds out maybe I will get to test it around a known circuit with the 2.7 gearing and see if it gives the same average speed over distance as my old 70s road bikes.

A clydesdale rider like myself, who is well over six-feet tall and weighs over 200 pounds, can use longer pedal cranks if anyone can. I have tested a lot of bikes in a short period of time over an old level TT course for average speed and the fastest times always come from the bikes with the longest pedal cranks. With the long cranks a rider pedaling at their top cadence can either hold the same gear up a grade easier than with shorter cranks, or on the level can possible hold a gear higher than with shorter cranks for a longer period of time. Big, long heavy legs are just harder to spin higher cadence with, but the longer cranks will let a big and tall rider have more torque and the same power at a low cadence, as a small, lighter rider will at a higher crank rpm with short cranks. It is no different than getting power from an internal combustion engine where rpm x torque = power. If you can not get the rpm, a longer stroke can make the same power at a lower rpm.

PeteHski

The thing to remember with crank length vs cadence is that for longer cranks it is harder to maintain cadence as your foot speed will be higher for a given cadence (pedalling in a larger circumference). So the increased torque is cancelled out by the effective reduction in cadence. It's worth noting that Brad Wiggins ( who is 190 cm tall) set his hour record on 170 mm cranks after switching from 177.5 mm cranks. The 2016 men's Olympic GB track team used 165 mm cranks.

Crank length really comes down to personal preference based on your stroke length, but there is no inherent advantage to running longer cranks - especially on a multi-geared bike where you can manipulate torque vs cadence at will. However, there are several potential advantages to running shorter cranks e.g. allowing a more aerodynamic position and less demanding on your range of joint articulation. I've discussed this with two very experienced fitters and they both advocate relatively short cranks, even for tall riders.

Germany_chris

While I'm not a fan of GCN the did a video about this not too long ago and video does show the difference in crank length in relation to pedaling dynamics. <br /><br />At least I thought it was neat

GamblerGORD53

There is no doubt that for instant acceleration, just upping cadence is the way to go.
I have used all the crank lengths.
So going from 165 to 180 is 9% diff. At the same road SPEED, 180 will do a 91 cadence at say 100GI, while the 165 will have 100 cadence with 91 GI.
When you calculate the sum of foot travel on the 2 circumferences, the total will be exactly the SAME. Foot speed is exactly the SAME. What is gained in FORCE is LOST in distance. Simple physics 101. FACT.
I had to replace my 180 mm White Ind. crank because the stump was too thick for my chaincase and their rings are stupid 3/32".

livedarklions

beng1

It is also a fact that what is a long distance to a small or average rider is not to a large, tall rider, so your comments mean nothing at all.

Erzulis Boat 

For a minute, I thought you had him! But, 93-97 is awful close to 100. What I would do if I were you, is go back and edit the post to claim the RPM was like 75 or something.

livedarklions

Why? If we're discussing crank size we're discussing marginal differences,. I just was correcting a factual error, not trying to make some larger point. I think crank size really doesn't matter much, but I'm not a racer, so I don't know how the marginal differences play out in different contexts.

His range was 100-110, btw, which pushing a big gear non-stop for 1 hour is significantly faster than 93-97. The middle of those ranges is exactly 10 rpm apart.

livedarklions

That's absurd. X% increase in the distance of a single revolution is X% regardless of the height of the rider. Your comment is the one that's meaningless. Your point may go to comfort, but it has nothing to do with speed.

ofajen

On the whole, being large and tall is a disadvantage in cycling. For similar level of training, W/kg scales with an exponent of about 0.7. The work against gravity when climbing obviously scales with an exponent of 1.0. That’s why smaller, lighter riders tend to be the ones who are the fastest climbers.

Larger riders have a smaller relative advantage on level ground, since the increase in frontal surface area scales with an exponent a slight bit less than 0.7.

Otto