Rim brake vs Disc Brake Aero test.
#51
Senior Member
The presence of a rider is not very relevant when you're trying to isolate the effect of disc/no disc. The location of the disc brake means the rider is going to have next to no influence, and adding in a rider or mannequin would mostly serve to add an additional source of error without justification.
This is pretty basic experimental design and cost-benefit analysis. You've got an initial experiment to look for differences. Tunnel time is expensive. And adding additional sources of noise make the data much more difficult to analyze. This is a new area for bicycle aerodynamics testing. Starting simple and keeping the experiment, as much as possible, to the smallest differences with the fewest possible additional factors, is the only sensible way to start something like this. Testing with a rider could influence the airflow around the disc brakes in a way that changes the experimental outcome, but common sense should tell you that's not very likely in this case.
This is pretty basic experimental design and cost-benefit analysis. You've got an initial experiment to look for differences. Tunnel time is expensive. And adding additional sources of noise make the data much more difficult to analyze. This is a new area for bicycle aerodynamics testing. Starting simple and keeping the experiment, as much as possible, to the smallest differences with the fewest possible additional factors, is the only sensible way to start something like this. Testing with a rider could influence the airflow around the disc brakes in a way that changes the experimental outcome, but common sense should tell you that's not very likely in this case.
#52
Administrator
Join Date: Feb 2005
Location: Delaware shore
Posts: 13,558
Bikes: Cervelo C5, Guru Photon, Waterford, Specialized CX
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 1106 Post(s)
Liked 2,173 Times
in
1,464 Posts
Quick summary. Rim version of Specialized S-Works Tramac was faster than disc with all angles. The difference increased as wind moved from right to left due to rotor being on left side of wheel and is blocked more from wind of right by frame, fork and spinning spokes. When wind is 20 degrees from left, the difference in bikes is 8 watts. With other angles typical of most riding, the difference is 1-3 watts
#53
Senior Member
Join Date: Jan 2011
Location: Pacific NW
Posts: 563
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
The presence of a rider is not very relevant when you're trying to isolate the effect of disc/no disc. The location of the disc brake means the rider is going to have next to no influence, and adding in a rider or mannequin would mostly serve to add an additional source of error without justification.
This is pretty basic experimental design and cost-benefit analysis. You've got an initial experiment to look for differences. Tunnel time is expensive. And adding additional sources of noise make the data much more difficult to analyze. This is a new area for bicycle aerodynamics testing. Starting simple and keeping the experiment, as much as possible, to the smallest differences with the fewest possible additional factors, is the only sensible way to start something like this. Testing with a rider could influence the airflow around the disc brakes in a way that changes the experimental outcome, but common sense should tell you that's not very likely in this case.
This is pretty basic experimental design and cost-benefit analysis. You've got an initial experiment to look for differences. Tunnel time is expensive. And adding additional sources of noise make the data much more difficult to analyze. This is a new area for bicycle aerodynamics testing. Starting simple and keeping the experiment, as much as possible, to the smallest differences with the fewest possible additional factors, is the only sensible way to start something like this. Testing with a rider could influence the airflow around the disc brakes in a way that changes the experimental outcome, but common sense should tell you that's not very likely in this case.
all necessary components ought to be I included, and if wind is the test, do a 180* rotation. So the test is about wind at angle not bike without rider. (I know the wind to the front and significant yaw angles matter)
#54
Mr. Dopolina
Thread Starter
Join Date: May 2007
Location: Taiwan
Posts: 10,217
Bikes: KUUPAS, Simpson VR
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 149 Post(s)
Liked 117 Times
in
41 Posts
Quick summary. Rim version of Specialized S-Works Tramac was faster than disc with all angles. The difference increased as wind moved from right to left due to rotor being on left side of wheel and is blocked more from wind of right by frame, fork and spinning spokes. When wind is 20 degrees from left, the difference in bikes is 8 watts. With other angles typical of most riding, the difference is 1-3 watts
So now I wonder how we quantify that. Do we split it in half as a general statement and say 4 watts at 20 deg and 1 or so watts at shallower yaw angles? Or do we need to keep it as it is and list the data PER SIDE of the bike?
I think those who first publish data will be the ones who set the reporting method on this.
#56
Mr. Dopolina
Thread Starter
Join Date: May 2007
Location: Taiwan
Posts: 10,217
Bikes: KUUPAS, Simpson VR
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 149 Post(s)
Liked 117 Times
in
41 Posts
And if you get some of those nifty rims with negative drag you might need to carry an anchor so you don't fly away...
#57
Senior Member
Quick summary. Rim version of Specialized S-Works Tramac was faster than disc with all angles. The difference increased as wind moved from right to left due to rotor being on left side of wheel and is blocked more from wind of right by frame, fork and spinning spokes. When wind is 20 degrees from left, the difference in bikes is 8 watts. With other angles typical of most riding, the difference is 1-3 watts
When Specialized did the leg hair test, they said shaved legs was 15W which represented 79 seconds. Assuming the tests were performed at the same wind speed the 8 sec difference would be 1.5W and only in cross-wind conditions, that's effectively zero.
It seems like the take away message is that the difference is smaller than the difference between and aero and non-aero frameset.
#58
Senior Member
Join Date: Jan 2011
Location: Pacific NW
Posts: 563
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
Or that shaving can make up the detriment of disc by double.
I guess rim brakes are going the way of the 10 speed.
.... Edit....
I do appreciate the efforts of November Dave arranging the test and contacting velo for a story. IF the story isn't available online, I'll buy a copy of the magazine, for that very story. I have always liked riding on deep carbon wheels (50mm) but I have been shopping for a disc bike for a while. So so ive been into any research on safety and performance I can find.
At 6'4", it is easy to fluctuate between 205-230lbs, which is fun for going mad speed down hills but brake performance is critical. Just as important are suitable wheels.
In in following up november's website (and bob d's) I appreciate posted suggested weight restrictions for wheels. Just like the posted weight restriction on the diving board of a backyard swimming pool, safety warnings are best heeded.
It it would seem the people that value better stopping the most are the people carrying the most emerita, if the wheels cap out at 21bs for 35mm wheels and 225lbs for 50mm wheels, it seems the products hit the majority of people, but exclude a good target market.
I guess rim brakes are going the way of the 10 speed.
.... Edit....
I do appreciate the efforts of November Dave arranging the test and contacting velo for a story. IF the story isn't available online, I'll buy a copy of the magazine, for that very story. I have always liked riding on deep carbon wheels (50mm) but I have been shopping for a disc bike for a while. So so ive been into any research on safety and performance I can find.
At 6'4", it is easy to fluctuate between 205-230lbs, which is fun for going mad speed down hills but brake performance is critical. Just as important are suitable wheels.
In in following up november's website (and bob d's) I appreciate posted suggested weight restrictions for wheels. Just like the posted weight restriction on the diving board of a backyard swimming pool, safety warnings are best heeded.
It it would seem the people that value better stopping the most are the people carrying the most emerita, if the wheels cap out at 21bs for 35mm wheels and 225lbs for 50mm wheels, it seems the products hit the majority of people, but exclude a good target market.
Last edited by BigJeff; 12-11-14 at 02:01 AM.
#59
Professional Fuss-Budget
However, we should keep in mind that because the bike is moving forward, most of the time the yaw is below 10º, and rarely goes above 20º. In addition, the faster you ride, the narrower the yaw. There is no simple formula that gives an accurate picture.
At the same time, if you're merely curious, being overwhelmed with information you don't know how to interpret is not helpful either....
AFAIK there are no industry-standard protocols for any wind tunnel tests. I'd expect to see a lot of conflicting information, where the numeric specificity of the results hides the vagueness of what you can expect in real-world use.
That said, it seems to me that if the aero penalty maxes out at a specific yaw at 10-15s over a 40k, or is that low using WAD or another averaging method, the aero penalty probably isn't going to matter.
#60
Mr. Dopolina
Thread Starter
Join Date: May 2007
Location: Taiwan
Posts: 10,217
Bikes: KUUPAS, Simpson VR
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 149 Post(s)
Liked 117 Times
in
41 Posts
Actually there are very specfic protocols at the A2 tunnel where much of this work is done. Maybe those who have done work there like Boyd or November Dave will chime in and confirm.
#61
Senior Member
Join Date: Oct 2010
Posts: 182
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 42 Post(s)
Likes: 0
Liked 26 Times
in
10 Posts
Protocols at A2:
1. You have to switch the tire from wheel to wheel, you can't bring in Conti GP4000s pre-installed on the wheels you mean to test and say "they're all the same."
2. A2 considers the data that is collected with the support fixture to be the data. They disagree with running the fixture alone and subtracting its drag.
How you weight the distribution to come up with your "x seconds saved in a 40k tt" is totally outside of the tunnel's concern. Trek did a great explanation of their process and the weighting it determined here. We used a site that I'd used a lot for sailing analysis, and ran a ton of likely scenarios through it, and came up with a fairly similar distribution to Trek. Our distribution was more similar to the one Tour Magazine uses, so we've adopted theirs for our results. FLO says you spend 80% of your time between 10* and 20*. Obviously that weighting is going to have a huge impact, so whenever you look at these things it behooves you to find out what the weighting protocol is, and then decide if you believe it.
#63
Boyd Cycling owner
Join Date: Mar 2011
Posts: 412
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 10 Post(s)
Likes: 0
Liked 3 Times
in
1 Post
On standard wheels there is a protocol that is suggested to follow. It's not set in stone and since the company who is visiting the wind tunnel is paying, they can do whatever they want with their time.
The baseline test for wheel comparisons in road wheels is to test a front wheel at a 0 to 20 degree sweep at 2.5 degree increments. When we first got to the wind tunnel to do our first tests this is what was recommended to us as the method we should follow.
A standard road front wheel is very symmetrical, which is why the run is usually only done to one side. If you were testing anything not perfectly symmetrical (bikes, disc brakes, rear wheel, rider position, most aero position to eat a slice of pizza), then you would want to run the sweep in both directions. As you can imagine the drag graphs would look pretty different on a disc brake wheel with a rotor in place facing different directions. If you wanted to sell a bunch of disc brake bikes and prove they were just as aero as rim brake bikes, you would simply rotate the bike away from the rotor side during your yaw sweep. This would be the same with disc brake wheels, so if you were testing both rim brake and disc brake wheels to get a comparison of the aero data, the yaw sweep should go in both directions to get the true picture.
In the real world, this means that if you have a crosswind coming from your right side, your disc brake bike is likely fairly close to your rim brake bike. However, if the crosswind is coming from the left, this is where the larger differences might be seen.
#64
Senior Member
Join Date: Oct 2010
Posts: 182
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 42 Post(s)
Likes: 0
Liked 26 Times
in
10 Posts
The baseline test for wheel comparisons in road wheels is to test a front wheel at a 0 to 20 degree sweep at 2.5 degree increments. When we first got to the wind tunnel to do our first tests this is what was recommended to us as the method we should follow.
A standard road front wheel is very symmetrical, which is why the run is usually only done to one side. If you were testing anything not perfectly symmetrical (bikes, disc brakes, rear wheel, rider position, most aero position to eat a slice of pizza), then you would want to run the sweep in both directions. As you can imagine the drag graphs would look pretty different on a disc brake wheel with a rotor in place facing different directions. If you wanted to sell a bunch of disc brake bikes and prove they were just as aero as rim brake bikes, you would simply rotate the bike away from the rotor side during your yaw sweep. This would be the same with disc brake wheels, so if you were testing both rim brake and disc brake wheels to get a comparison of the aero data, the yaw sweep should go in both directions to get the true picture.
In the real world, this means that if you have a crosswind coming from your right side, your disc brake bike is likely fairly close to your rim brake bike. However, if the crosswind is coming from the left, this is where the larger differences might be seen.
A standard road front wheel is very symmetrical, which is why the run is usually only done to one side. If you were testing anything not perfectly symmetrical (bikes, disc brakes, rear wheel, rider position, most aero position to eat a slice of pizza), then you would want to run the sweep in both directions. As you can imagine the drag graphs would look pretty different on a disc brake wheel with a rotor in place facing different directions. If you wanted to sell a bunch of disc brake bikes and prove they were just as aero as rim brake bikes, you would simply rotate the bike away from the rotor side during your yaw sweep. This would be the same with disc brake wheels, so if you were testing both rim brake and disc brake wheels to get a comparison of the aero data, the yaw sweep should go in both directions to get the true picture.
In the real world, this means that if you have a crosswind coming from your right side, your disc brake bike is likely fairly close to your rim brake bike. However, if the crosswind is coming from the left, this is where the larger differences might be seen.
Our disc brake test at A2 showed that a cross wind coming from the right has a decidedly higher penalty for the disc bike than the road bike. When we played the "what do you think will happen" game before running the test, the A2 engineer nailed it spot on. A company called Culprit ran a similar test on a road bike with similar results, and though I can't find it anywhere, a lot of people reference a test that Damon Rinard did which shows same.
#65
Senior Member
yes, and every bike must have a rider. Therefore, a test without a rider is like a test without a frame.
all necessary components ought to be I included, and if wind is the test, do a 180* rotation. So the test is about wind at angle not bike without rider. (I know the wind to the front and significant yaw angles matter)
all necessary components ought to be I included, and if wind is the test, do a 180* rotation. So the test is about wind at angle not bike without rider. (I know the wind to the front and significant yaw angles matter)
There's a persistent misunderstanding around here about what this kind of experiment can tell you. Earlier in this thread, someone has already stated that testing without a rider renders any result meaningless. That's utter nonsense. Of course the result of the experiment must be understood to have limits on its interpretation. But that's true of EVERY wind tunnel test - and every experiment ever - with a rider or without. The wind tunnel is not the real world! It is a highly controlled environment with very little resemblance to the environment in which bicycles are actually ridden. If it were not, meaningful results would be all but impossible to obtain. Are you going to tell me that, because wind tunnel tests do not include every necessary (meaning what, by the way?) component, they are not valid experiments and not useful for determining aerodynamic differences between different bicycles and components? Because I think most people would consider that a pretty loony tunes conclusion.
Everyone around here thinks they're some kind of expert. But it's really easy to just come over all smug and type "well they should always test with a rider" on an internet message board. Designing useful scientific or engineering experiments, and interpreting the data intelligently, is a lot harder and requires a fair bit more education and experience. Suffice to say there are good reasons to do this test both with a rider on the bike and without, and these tests will tell you different things. Doing the test without a rider is perfectly reasonable, and in fact is the only sensible place to start. It's simpler, cleaner, less likely to go wrong, more likely to be informative, and can inform more complex tests in the future.
#66
Mr. Dopolina
Thread Starter
Join Date: May 2007
Location: Taiwan
Posts: 10,217
Bikes: KUUPAS, Simpson VR
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 149 Post(s)
Liked 117 Times
in
41 Posts
Do you have links to articles that address the distribution of yaw angles typically seen? My guess would be that the bulk of riding would fall between 5 and 10 degrees. Is this assumption wrong?
Last edited by Bob Dopolina; 12-11-14 at 04:46 PM.
#67
Senior Member
Join Date: Jul 2013
Location: Somewhere in TX
Posts: 2,266
Bikes: BH, Cervelo, Cube, Canyon
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 212 Post(s)
Likes: 0
Liked 6 Times
in
6 Posts
As usual, the disc brake nuthuggers have to inject conjecture into their deranged ranting. The weight penalty will never evaporate. It's just physics.
#68
Boyd Cycling owner
Join Date: Mar 2011
Posts: 412
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 10 Post(s)
Likes: 0
Liked 3 Times
in
1 Post
[QUOTE=Bob Dopolina;17380637]
DJ Connel did a blog about this back in 2011. It gets pretty technical with a lot of math that involves letters instead of numbers :-)
On Bicycles, and.... what else is there?: yaw angle: variable wind speed and variable rider heading
Basically it tells you what everybody should have learning in statistics 101. Yaw angle distribution will be a bell curve with the highest probability at 0 degrees and going out from there.
FLO says you spend 80% of your time between 10* and 20*[,/QUOTE]
That doesn't seem right to me.
Do you have links to articles that address the distribution of yaw angles typically seen? My guess would be that the bulk of riding would fall between 5 and 10 degrees. Is this assumption wrong?
That doesn't seem right to me.
Do you have links to articles that address the distribution of yaw angles typically seen? My guess would be that the bulk of riding would fall between 5 and 10 degrees. Is this assumption wrong?
On Bicycles, and.... what else is there?: yaw angle: variable wind speed and variable rider heading
Basically it tells you what everybody should have learning in statistics 101. Yaw angle distribution will be a bell curve with the highest probability at 0 degrees and going out from there.
#69
Senior Member
Join Date: Oct 2010
Posts: 182
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 42 Post(s)
Likes: 0
Liked 26 Times
in
10 Posts
Between the Trek thing I linked and the thing that Boyd linked, that's a pretty compelling argument.
A "windy" US city sees an average of <9 mph wind during the day. That's measured at an airport weather station, high up, in clean air. Down where your wheels are, there's less wind than that. People always toss around "oh yeah, I ride in 25mph wind all the time..." When it's actually blowing 25, you go outside and you notice that it's windy before you notice that it's pi**ing down rain - that's how windy 25 is.
A "windy" US city sees an average of <9 mph wind during the day. That's measured at an airport weather station, high up, in clean air. Down where your wheels are, there's less wind than that. People always toss around "oh yeah, I ride in 25mph wind all the time..." When it's actually blowing 25, you go outside and you notice that it's windy before you notice that it's pi**ing down rain - that's how windy 25 is.
#71
Senior Member
Join Date: Jan 2011
Location: Pacific NW
Posts: 563
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
.... meaningful results would be all but impossible to obtain. Are you going to tell me that, because wind tunnel tests do not include every necessary (meaning what, by the way?) component, they are not valid experiments and not useful for determining aerodynamic differences between different bicycles and components? Because I think most people would consider that a pretty loony tunes conclusion.
.
.
1. Wheel alone. (disc vs non disc)
2. Wheel on bare frame
3. " with frame and complete group (no cables)
4. Same with brake cables only
5. Repeat with manaquinn rider or a human rider
Every run will introduce more data/noise, but at the same time it becomes more accurate since it gets closer to "reality", which is the true objective of "what is the measurable difference between riding disc vs rim brake."
Yes, The accuracy of the measuring system while testing a complex system may be a challenge, but it also contributes to one possible conclusion that "the difference is not measurable in the wind tunnel used".
november dave made the perfect point, they looked at using him buy he wasn't standardized in pedaling enough to be used as a constant in the test. (But the dynamics of pedaling and wheel rotation may be appropriate factors to include)
Some say the cyclist is "necessary" for the bike to function.
#73
Senior Member
[QUOTE=Bob Dopolina;17380637]
The yaw angle distribution is going to depend heavily on the rider speed. Tony Martin is probably going to be under 10deg the vast majority of his rides, while a Sunday tri-athlete will have a much higher angle spread.
Trek's whitepaper showed an average of 10 deg on a Kona course (notoriously windy), ~3 deg ave for Pheonix, and 6 deg for Wisconsin. Those tests were likely done with a domestic pro caliber rider. The DJ Connel calculation agrees with Trek's numbers. It's possible Flo's calculations are based on a combination of very windy conditions (Kona) and weaker riders. Barring that, I have absolutely no idea where Flo's 10-20 deg remark came from.
Edit: Found an actual comment from Flo about the 10-20 remark. It's from something Zipp used back in 2011. https://forum.slowtwitch.com/cgi-bin/...327589#5327589
My guess is that the yaw angles are based on average cyclist speeds of ~20 mph, and/or potentially not corrected for the difference between 10m wind-speed (weather reported) and 1m wind-speed (cyclist)
Edit: And another link to comments from Zipp stating 9-17deg yaw angles.
https://nyvelocity.com/content/equipm...-josh-poertner
One other comment is that the wheels profile almost doesn't matter near zero yaw, since the tire shields the rim. The aero testing basically confirms this, all wheels look similar head on. So it makes sense to optimize wheels at a slightly higher yaw angle, since that's where you see any differences.
FLO says you spend 80% of your time between 10* and 20*[,/QUOTE]
That doesn't seem right to me.
Do you have links to articles that address the distribution of yaw angles typically seen? My guess would be that the bulk of riding would fall between 5 and 10 degrees. Is this assumption wrong?
That doesn't seem right to me.
Do you have links to articles that address the distribution of yaw angles typically seen? My guess would be that the bulk of riding would fall between 5 and 10 degrees. Is this assumption wrong?
Trek's whitepaper showed an average of 10 deg on a Kona course (notoriously windy), ~3 deg ave for Pheonix, and 6 deg for Wisconsin. Those tests were likely done with a domestic pro caliber rider. The DJ Connel calculation agrees with Trek's numbers. It's possible Flo's calculations are based on a combination of very windy conditions (Kona) and weaker riders. Barring that, I have absolutely no idea where Flo's 10-20 deg remark came from.
Edit: Found an actual comment from Flo about the 10-20 remark. It's from something Zipp used back in 2011. https://forum.slowtwitch.com/cgi-bin/...327589#5327589
My guess is that the yaw angles are based on average cyclist speeds of ~20 mph, and/or potentially not corrected for the difference between 10m wind-speed (weather reported) and 1m wind-speed (cyclist)
Edit: And another link to comments from Zipp stating 9-17deg yaw angles.
https://nyvelocity.com/content/equipm...-josh-poertner
One other comment is that the wheels profile almost doesn't matter near zero yaw, since the tire shields the rim. The aero testing basically confirms this, all wheels look similar head on. So it makes sense to optimize wheels at a slightly higher yaw angle, since that's where you see any differences.
Last edited by gsa103; 12-11-14 at 12:54 PM.
#74
Senior Member
Join Date: Apr 2013
Location: Rep. of Dallas
Posts: 1,062
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Perfect, expensive stuff to buy that almost no one needs.
#75
Senior Member
Join Date: May 2010
Location: Antioch, IL
Posts: 2,330
Bikes: 2013 Synapse 4
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 70 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
Quick summary. Rim version of Specialized S-Works Tramac was faster than disc with all angles. The difference increased as wind moved from right to left due to rotor being on left side of wheel and is blocked more from wind of right by frame, fork and spinning spokes. When wind is 20 degrees from left, the difference in bikes is 8 watts. With other angles typical of most riding, the difference is 1-3 watts
so, is 1-3watts of power loss worth it for better braking?? is it better braking? who really cares who has disc and who doesn't? FTR - I think disc is nice tech, but my next bike will be rim brakes, I simply don't have a need for the added cost. I live in IL (no hills to speak of), I don't ride in the rain or even in the threat of rain. so the tangible benefits simply aren't there for me so I cna't justify the cost.