The importance of crank length to the cyclist.

[coapman]

http://www.plan2peak.com/files/32_article_JMartinCrankLengthPedalingTechnique.pdf

J Martin is not aware of the fact that additional maximal torque can be applied at 12 and 1 o'c, so his research results on pedal power delivery and pedaling efficiency are incomplete, which makes the situation not as clear cut as you believe it to be. Frank takes note of the research of others but is still prepared to do his own research.

 

[CoachFergie]

You have been perpetuating that nonsense for years and have never provided any evidence beyond what you think you observe from Jacques Anquetil. Think I will stick with published evidence over unproven theories thank you very much.

 

[blutto]

It was optimal because we either got them so aero that power was compromised or optimal power that drag increased or speed on track or road based testing decreased.

...is it just me or does anyone else smell the stench of rancid hypocrisy in the air.....

Cheers

blutto

 

[FrankDay]

It was optimal because we either got them so aero that power was compromised or optimal power that drag increased or speed on track or road based testing decreased.

Thank you very much Coach Fergie for making this point because it exactly makes my point. You, of course, are not alone in your thinking. You are convinced that what you are doing for your athletes is optimal and you make the above statement to confirm this belief. But, did you ever ask the question, what was it that compromised power as you got them "so aero"? If you (or the hundreds of other cycling coaches out there) ever bothered to ask that question you might conclude it was because they get "too scrunched up". Then, you might ask, is there anything that we could do about that? But, it appears you (and everyone else) never bothered to ask the question, let alone look for a solution. So, even though you know about Martin, and you have been part of this thread for over 1000 posts, it seems you still can't come to believe that there might actually be something that can be done about that power loss when your athletes get "so aero". At least you shouldn't feel to bad about this because you definitely are not alone in this way of thinking.

 

[CoachFergie]

Thank you very much Coach Fergie for making this point because it exactly makes my point.

You are assuming that the only consequence of being too low is a reduced hip angle. And that changing crank length is the only way to address the problem.

 

[coapman]

You have been perpetuating that nonsense for years and have never provided any evidence beyond what you think you observe from Jacques Anquetil. Think I will stick with published evidence over unproven theories thank you very much.

The nonsense is on the side of you and other coaches who never asked the question, what prevents me from applying max torque at 12, and then setting about solving it. All that's involved is a simple adjustment in the use of muscles. I missed this earlier, how did raising the front of your saddle by 2 degrees enable you to lower your stem by 4cm.

 

[Tapeworm]

The nonsense is on the side of you and other coaches who never asked the question, what prevents me from applying max torque at 12, and then setting about solving it. All that's involved is a simple adjustment in the use of muscles. I missed this earlier, how did raising the front of your saddle by 2 degrees enable you to lower your stem by 4cm.

How do you know you are applying max torque at these angles?

 

[FrankDay]

You are assuming that the only consequence of being too low is a reduced hip angle. And that changing crank length is the only way to address the problem.

I am not assuming anything. I, as do you, KNOW that being too low can compromise power. I, at least, have asked the question as to what might cause this and whether something might be done about it. I have, as a result, concluded that hip angle at TDC is a potential explanation. I have seen nothing to suggest this analysis is wrong.

 

[coapman]

How do you know you are applying max torque at these angles?

Because of the direction in which the force is directed and the amount of resistance that has to be created to counteract this force.

 

[coapman]

Where we may differ is my experience that there are more effective ways to achieve an optimal position in terms of aerodynamics and power delivery than changing crank length, putting the seat up, shifting it back, raising the bars and using a shorter stem. As mentioned previously I was able to lower my stem 4cm just by tilting my saddle up 2 degrees.

All riders have their own recognized standard saddle height, how can you raise the saddle height, (if that's what you mean by putting the seat up) and expect improvement without shortening your crank length. A person could interpret that statement of yours in two completely different ways

 

[Tapeworm]

Because of the direction in which the force is directed and the amount of resistance that has to be created to counteract this force.

Measured in any way or just assumed?

 

[Night Rider]

No, I don't care about power for this test. The rider will be pedaling but we will not be testing power, which is typical for wind tunnel testing. I am only looking at rider aerodynamics. Once we establish what the aerodynamics of various positions are any rider can test their sustainable power in those positions outside of the wind tunnel which should allow them to assess what is best for them. It is not perfect but it is more information than they have now. I only have 2 hours. We are using a bike that will allow us to assess the full range of positions and rapid changing of components. The bike will not have any cables or brakes and I expect to be able to change crank length in less than 2 minutes. I am expecting to be able to test about a 9-10 inch difference in seat to handlebar drop. I also hope to get both road bike positions and aerobar positions. Again, we have a limited amount of time but I hope to get one run with just the bike alone so those of you concerned about this can assess the effects of the bike as it modifies the results. Again, I am primarily interested in assessing the effects of rider position on drag, generally agreed upon as the major component of drag in bicycle riders. This means the bike being used is not important.I will do my best. Not sure what the arrangement is.
Looking forward to it.

1. Power Meter, fair enough given your limitations in time.

2. Why not leave the cables and brakes on? You want to measure the effect of position changes of the rider on a bike that is to be ridden on the road. If you don't want brakes and cables test a track bike. I would be putting helmet and cycling gear on the rider and testing 1. In the drops 2. on the hoods 3. Sprinting out of saddle

3. I predict you are going to find minimal differences (either way) if you are using a rider in an optimal position to begin with. If you are using a 'Fred' type rider sitting upright to begin with then results could be all over the place and open to interpretation.

 

[FrankDay]

1. Power Meter, fair enough given your limitations in time.

2. Why not leave the cables and brakes on? You want to measure the effect of position changes of the rider on a bike that is to be ridden on the road. If you don't want brakes and cables test a track bike. I would be putting helmet and cycling gear on the rider and testing 1. In the drops 2. on the hoods 3. Sprinting out of saddle

3. I predict you are going to find minimal differences (either way) if you are using a rider in an optimal position to begin with. If you are using a 'Fred' type rider sitting upright to begin with then results could be all over the place and open to interpretation.

Well, the test is completed and the results are VERY interesting and, I am sure, not what anyone would have predicted, including me. The results leave one particular unanswered question in my mind. And, we did use a track bike with a disk wheel in back and a zipp in front and an adjustable stem. We also tried to look at whether there was any benefit to using an aero helmet when very low and if so, how large. I am traveling right now and very tired so I will wait until I am home and have some time to review the data before posting the results. Stand by to be amazed.

 

[FrankDay]

Regarding the testing: We made one crucial error in our "design" that makes the data really look strange but in my 12 hour ride home I think I have figured it out such that it now makes sense to me and I think I learned a lot. Anyhow, it is complicated and I need a little time to sit down and put it all together. So I beg everyone to give me a little time.

 

[sciguy]

Anyone else hearing "I'm a spin doctor" playing in the background?

Hugh

 

[FrankDay]

While I continue to try to compose my thoughts into something logical, understandable, and short I will give everyone this teaser. My conclusions regarding the effects of crank length on aerodynamics can be distilled down to two words: It depends. Stay tuned.

 

[sciguy]

Frank,

Thanks for investing in the wind tunnel time. It will be interesting to see the results especially if photo side views are included. How about first showing the raw data without any analysis that might bias our thinking?

Cheers from South Africa,

Hugh

 

[FrankDay]

Frank,

Thanks for investing in the wind tunnel time. It will be interesting to see the results especially if photo side views are included. How about first showing the raw data without any analysis that might bias our thinking?

Cheers from South Africa,

Hugh

I would do that but the raw data is essentially uninterpretable. It is because we made a critical error and did what we thought would be our "worst case" postion in an essentially unridable position, so it tested much better than it should have. We went to the Velodrome the morning of the test to prove o ourselves that the positions we were testing were ridable. We had some issues with the "long crank" position and had to raise the bars and he eventually said, "I think I could ride this", so we left it there. As soon as we were in the tunnel it was clear this was not a good position. We did the testing on PowerCranks and it was clear he was having troubple getting is leg over the top, even though this was the first run of the day when he was fresh, and his is a guy who has been riding PowerCranks 40 miles per day for several years. At the end of the day, we repeated this test, and he could barely pedal the bike.

His handlebars should have been at least 2-4 inches higher than they were. As a result, over all of our runs that encompassed a change in seat to handlebar height of about 25 cm we saw no almost no change. And, the long crank, "high body" data gave us the lowest drag of the day. It was bizarre and it was like there was nothing we could do to change anything. The rider before us was able to see a drop in drag of about 150 gm with about a 1-2 cm drop in bar height and changing the way his helmet sat on his head. (And all the time his coach kept asking him how his power was) We never saw more than 100 gm change despite a 25 cm drop (and we did our testing at 30 mph, while this other guy was at 25 mph, as we were looking for higher speeds to try to uncover subtle differences). We were totally baffled until I finally put it together why we had this outlier and why all the changes when low resulted in almost no change. If we throw out the first data point, the lowest drag we got was at 142.5 crank length and the lowest stem position. But, it wasn't by much.

So, without context, the raw data is pretty useless and I think it will be better delivered after I give my analysis. If I can't explain the data then I don't have a clue. But, there was a lot of discussion going on as we were gathering this pretty bizarre data set and after a lot of reflection I think I can explain it. I think I have a much better understanding of what is going on now and how crank length affects the various parts of the aerodynamic drag/power puzzle. It was time and money well spent as far as I am concerned. Oh, and, of course, as a result of this new knowledge I now have some new questions I want to look at to see if it is possible to "solve" one finding I got wrong in my predictions.

Edit: I might add, I think my understanding can now explain the Cervelo data also, why one rider got worse, one rider got better, and two stayed about the same.

I am about ready to put it out. Probably tomorrow.

 

[sciguy]

As they say "Very interesting"

Thanks,

Hugh

I would do that but the raw data is essentially uninterpretable. It is because we made a critical error and did what we thought would be our "worst case" postion in an essentially unridable position, so it tested much better than it should have. We went to the Velodrome the morning of the test to prove o ourselves that the positions we were testing were ridable. We had some issues with the "long crank" position and had to raise the bars and he eventually said, "I think I could ride this", so we left it there. As soon as we were in the tunnel it was clear this was not a good position. We did the testing on PowerCranks and it was clear he was having troubple getting is leg over the top, even though this was the first run of the day when he was fresh, and his is a guy who has been riding PowerCranks 40 miles per day for several years. At the end of the day, we repeated this test, and he could barely pedal the bike.

His handlebars should have been at least 2-4 inches higher than they were. As a result, over all of our runs that encompassed a change in seat to handlebar height of about 25 cm we saw no almost no change. And, the long crank, "high body" data gave us the lowest drag of the day. It was bizarre and it was like there was nothing we could do to change anything. The rider before us was able to see a drop in drag of about 150 gm with about a 1-2 cm drop in bar height and changing the way his helmet sat on his head. (And all the time his coach kept asking him how his power was) We never saw more than 100 gm change despite a 25 cm drop (and we did our testing at 30 mph, while this other guy was at 25 mph, as we were looking for higher speeds to try to uncover subtle differences). We were totally baffled until I finally put it together why we had this outlier and why all the changes when low resulted in almost no change. If we throw out the first data point, the lowest drag we got was at 142.5 crank length and the lowest stem position. But, it wasn't by much.

So, without context, the raw data is pretty useless and I think it will be better delivered after I give my analysis. If I can't explain the data then I don't have a clue. But, there was a lot of discussion going on as we were gathering this pretty bizarre data set and after a lot of reflection I think I can explain it. I think I have a much better understanding of what is going on now and how crank length affects the various parts of the aerodynamic drag/power puzzle. It was time and money well spent as far as I am concerned. Oh, and, of course, as a result of this new knowledge I now have some new questions I want to look at to see if it is possible to "solve" one finding I got wrong in my predictions.

Edit: I might add, I think my understanding can now explain the Cervelo data also, why one rider got worse, one rider got better, and two stayed about the same.

I am about ready to put it out. Probably tomorrow.

 

[42x16ss]

His handlebars should have been at least 2-4 inches higher than they were. As a result, over all of our runs that encompassed a change in seat to handlebar height of about 25 cm we saw no almost no change. And, the long crank, "high body" data gave us the lowest drag of the day. It was bizarre and it was like there was nothing we could do to change anything.

Interesting. Maybe it was a case of a cleaner rider profile performing better than a smaller/lower one?

Have you done any testing yet to see which of the positions were the most sustainable in terms of rider output? Some photos (if possible) would be interesting too.

 

[FrankDay]

Here is a link to a video taken by Mikes wife in the wind tunnel with Mike on 85 mm cranks and the flat stem. It is on facebook so I hope he has set access so everyone can see it. Looking at that I would have expected if we could get the head a little lower we would save quite a bit. Perhaps so if we had lowered a little but lowering the head a lot didn't do much. Dropping the stem 4 inches from this position gave us a savings of 11 gm from a total of about 3000.

 

[FrankDay]

So, here is what we did and what I have concluded.

First, let me say I was not trying to find the best aerodynamics for each position. I was simply trying to see how much benefit one could expect from increasing the drop from seat to handlebars as it related to crank length and I also wanted to look at the effect of an aero helmet, as to whether it was less as the head was low compared to the back. We made great big changes and were strapped for time.

We tested the following crank lengths and stem positions (distance from baseline in inches).

180 0
142.5 0
110 0
85 0
85 -4
85 -4 (Landis "bars up" position)
0 -4 (Landis "bars up" position)

Of course, we did not pre-ride the zero crank length position. And we made a mistake in pre-riding the 180 position. We were trying to get good ridable positions at each crank length. We moved the bars up twice before Mike said, "I think I could ride this" and we set that as our base-line position. After he rode all the other positions he commented that they were all way more comfortable than the 180 position and he could even see himself racing on the 85's even though he had never experienced them before. Unfortunately, we should have done more testing because the 180 position was unridable for more than a few minutes, which we discovered in the wind tunnel. But, I thought it was still high enough that we would see a big difference.

So, here are the raw results (drag in grams, not a force number but what they gave us).

180 road helmet 3095 grams
180 aero helmet 2947 and 2991 grams (two runs)
142 aero helmet 2949 grams
85 road helmet 2984 grams
85 aero helmet 2973 grams
85 aero helmet, bars down additional 4 inches 2962 grams
85 aero helmet, bars down 4 inches and angled 45º 3028 grams
0 aero helmet, bars down 4 inches and angled 45º 3062 grams

What have I concluded from this.

1. Everyone seems to think that the benefits of an aero helmet are the same regardless of your position (what the manufacturers would like you to believe). My data does not support this. If it is a hot day and one has a very good aerodynamic position then one might consider an ordinary road helmet for the additional cooling.

2. Presuming that my 180 numbers are very low for what should have been the best position for this rider I have concluded that there are two competing interests in the aerodynamic equation when it comes to crank length, upper body/head and the legs. Shorten the cranks and the upper body aerodynamics will generally improve and the leg aerodynamics will generally worsen. If the body is already low that aspect won't improve much and the overall drag will actually increase. I see no other explanation to explain this data. Therefore, if you have a poor, pretty upright position (as most do) I would expect your aerodyamics to improve with shorter cranks. If you already have a very good position, then I would expect no aerodynamic benefit from shorter cranks and I would only look at shorter cranks as to how they might affect other aspects of racing: power, comfort, efficiency, etc. If going shorter does not improve any of those then stay with the longest cranks you can ride that doesn't deteriorate any of those things.

So, as I alluded to, my new mantra regarding crank length and aerodynamics is: IT DEPENDS

I will post pictures of the positions we tested when I am able to upload them to a server with annotations as to what each one is.

 

[FrankDay]

Here are the images in the order they were tested. The forum restricts me to 4 images at a time so there will be another post to complete the series

First, 180 mm cranks with road helmet.

Next, 180 mm cranks with aero helmet.

Edit: I did do an 85 mm cranks with road helmet here but didn't get a picture. Look at the next one and substitute a road helmet. Drag 2984 gms
next 85 mm cranks with aero helmet

next 85 mm cranks with aero helmet and 4" drop in bar height from base

 

[FrankDay]

to continue
next, 85 mm cranks with aero helmet and 4" drop in bar height from base plus angled at 45º, which drops elbows even further.

next 0 mm cranks with aero helmet and 4" drop in bar height from base plus angled at 45º, which drops elbows even further. Can't imagine a bigger drop than this.

next 110 mm cranks with aero helmet and 4" drop in bar height from base

next 142.5 mm cranks with aero helmet and 4" drop in bar height from base

Let the comments and criticisms begin!

 

[sciguy]

2. Presuming that my 180 numbers are very low for what should have been the best position for this rider I have concluded that there are two competing interests in the aerodynamic equation when it comes to crank length, upper body/head and the legs. Shorten the cranks and the upper body aerodynamics will generally improve and the leg aerodynamics will generally worsen. If the body is already low that aspect won't improve much and the overall drag will actually increase. I see no other explanation to explain this data. Therefore, if you have a poor, pretty upright position (as most do) I would expect your aerodyamics to improve with shorter cranks. If you already have a very good position, then I would expect no aerodynamic benefit from shorter cranks and I would only look at shorter cranks as to how they might affect other aspects of racing: power, comfort, efficiency, etc. If going shorter does not improve any of those then stay with the longest cranks you can ride that doesn't deteriorate any of those things.
.

Thanks for the excellent data set!!!!!! Wasn't there some guy that predicted the outcome on this thread earlier? Hmmmmmm Didn't he say mean leg length exposed will increase with shorter cranks and increase overall drag?

In addition, a quick view of the 180mm position shows the rider in what many would feel to be a very ridable position with regular cranks where one doesn't need to fight the cranks "to get them over the top" as one does with PCs. This is really a pretty plain vanilla position for 170mm cranks in my experience.

YMMV,

Hugh