Bicycle Mechanics - Is aero tubing a gimmick?

So I'm sitting here and get a wild hair. Long to short my fancy super numerical airfoil calculator/ virtual wind tunnel says that a bike of round tubes has less total drag then one made of teardrop shaped tubes for the same width tube (veiwed from the front).
This program has proven acurate in several real windtunnel comparisons and just in case I confused it I ran it at a bunch of different times at various speeds, sizes, slight shape changes to both round and teardrop(including just a hint out of round), reynolds numbers, and mock numbers(sub trans-sonic) just to be sure.

This is nearing its limits as it was designed for subsonic lifting airfoils (wings) not bicycles but it should be in the ballpark. I did find a shape that is slicker than round but its more of an egg shape with the small end forward and both ends finishing fairly pointy rather than fully radiused.

humm...interesting. that's like taking the "non-fat" ice cream into the lab and finding fat. do makers of bikes with tear drop tubes claim they are more areodynamic? or is it just how they look?

So I'm sitting here and get a wild hair. Long to short my fancy super numerical airfoil calculator/ virtual wind tunnel says that a bike of round tubes has less total drag then one made of teardrop shaped tubes for the same width tube (veiwed from the front).
This program has proven acurate in several real windtunnel comparisons and just in case I confused it I ran it at a bunch of different times at various speeds, sizes, slight shape changes to both round and teardrop(including just a hint out of round), reynolds numbers, and mock numbers(sub trans-sonic) just to be sure.

Interesting findings. Fluid mechanics if a highly empirical field, and is often counter-intuitive. For example, a dimpled golf ball travels further than one with a smooth surface. You'd think a smooth surface would be best since it's the smoothest. I recall the reason had something to do with the dimples increasing energy in the air flow and delaying separation of the air along the surface - creating a smaller wake zone. The wake is a low pressure area that increases the net force against the ball in a "bad" direction. (Now this is all from memory, but I believe it is discussed in Fluid Mechanics by Fox and McDonald). Note, the wake is the drafting zone behind another cyclist (or car or truck), essentially a relatively low pressure zone.

Perhaps it is the same for the tear-drop tubing. The shape may encourage no or delayed separation; thus no or little wake zone.

There is a book out there called "Bicycling Science", by Edmund Burke (?). It probably discusses shape of tubing.

I can't explain why my 2 courses in engineering fluid mechanics from 20 years ago does not agree with your simulator. Perhaps those courses were less practical, or my memeory is going. Or perhaps your simulator is best for higher velocities and Reynolds numbers. Fluid mechanics is not linear.

Ah screw it, enough of this, I'm going for a ride!

I certainly don't have that kind of aero-engineering background, but I've often wondered about that myself. My question has to do with the effect of air coming off of the front wheel. It looks to me like before any air hits the down tube for example, it would already have been all messed up by the front wheel.

What is the difference between a single plain tube and an airfoil tube perpendicular to the airflow? All the books I remember reading said the airfoil tube would be better because there was less turbulence behind it. Is the increase in turbulence caused by the collection of tubes so close together? If what you are finding is true then what do aero rims do for you outside of creatin a stiffer rim?

First I didn't totaly finish my first post, I was being kicked of the computer and now I don't remember what I was going to add then.
Here are some results I got for better understanding
This is frame tubes only, no parts.
All tests at 35mph and 1.25 inch wide tubing, at sea level, 59f degrees.

NACA 4 digit 0050(teardrop 1.25" wide 2.5" long);
0.125 pounds clean, 0.141 pounds of drag dirty.
Round;
0.064 pounds clean, 0.081 pounds dirty.
Elipse, 1.25" wide 2.5" deep, symetrical front to back, rounded on both ends;
0.057 pounds clean 0.082 pounds dirty.


What is the difference between a single plain tube and an airfoil tube perpendicular to the airflow? All the books I remember reading said the airfoil tube would be better because there was less turbulence behind it. Is the increase in turbulence caused by the collection of tubes so close together? If what you are finding is true then what do aero rims do for you outside of creatin a stiffer rim?
They reduce the amount of spokage(A real word?... no.) in the highest velocity area and a standard square rim is not a circle. Just a stab at it but I'm talking frames in clear air.


Mudpie, you had it pretty much on with the golf ball. The reason for dimples is that they trip the laminar boundry layer early on because the resultant turbulant boundery layer is more stable and even though it is higher in drag it allows the flow to stay attached longer reducing the low pressure wake. I can force the laminar-turbulent transition prematurely if I want to with this program, to simulate dirt and bugs or dimples. It does seem to get slightly batty with real radical wing shapes at very high angles of attack but it seems to be working consistantly in this case and these arn't radical shapes anyway, very standard in fact.

Quiz: Any one know what the word GOLF originaly meant?

And my own pointy reverse egg shape(not anything like the tubes currently used);
0.031 pounds of drag clean, 0.061 pounds dirty

As MudPie hinted towards, did you try running your models with disrupted/turbulent flows ahead of the tubes? Simulate 3-D flow patterns coming off the wheels too. I've never done it myself but I imagine there's a lot of micro-vortices coming out of the front wheel. Also, instead of looking at just one tube in the flow, you may need to simulate the entire system. What happens to your flow coming off the downtube and then hitting the seat-tube?

.

...Quiz: Any one know what the word GOLF originaly meant?

From a non-golfer:

Is it: GENTLEMEN ONLY, LADIES FORBIDDEN?

or is it: all the other @#!*!$**! four-letter words had already been taken.

Interesting. I'm still trying to deduce the wind resistance of a charging rhino as that has more of a bearing on me.

It may be the benefit of smoothing flow. Besides if there were no gains to a foil shape, all the small aircraft makers wouldn't go to the extra expense of using them for struts(Round tubes are easier to make, I think)

A basic look at my book says for a circular CS, Cd(lam)=1.2, Cd(turb)=0.3 whereas a ellipse with L/D ratio of 2 (twice longer than it is high) has Cd(lam)=0.6 and Cd(turb)=0.3. A streamlined body like an airplane airfoil has a Cd of 0.04.
Once you have the size, the force of drag is Cd*A*rho*V^2/2. So you'd think the circular tube would have the most drag force.

Why you would test bike tubes at near or even beyond the MDR velocity is beyond all reason, I doubt lance armstrong will be approaching that anytime soon LOL. :)

Whats MDR?
I tested at 35mph at sea level which is a reynolds of about 35,000 for the 1.25" circle and 70,000 for those 1.25"x2.5" shapes. I do think the extremely low reynolds are affecting the results somewhat but not horribly so.
The Cd of 0.04 you refer to for a wing is based on planform area(wing veiwed from top) not frontal area, where as the circle and elipse you are refering to are being based on frontal area.
apples to apples oranges to oranges
All of mine are based on planform as thats what the its progamed for, still ends up with the same total drag at the end just a different road.
I am baffled because, like all of you, I've always known airfoils to be more aerodynamically efficient than circles.
To every reading this I cannot do full 3D shapes or closely spaced pieces, that would require the program that I've been drooling over for a while, unfortunatly my drool isn't worth just over three grand. So I use a single long tube for testing, no wheels, no brakes, no rider, no chain, no nothin'. The Aeronautical Engineering department at Embry-Riddle Aeronautical University in Daytona Beach, Florida has purchased a department-wide site license for program I am using.

Even there, the ellipse has significantly lower Cd than the circle in terms of cross section with flow being perpendicular to the axis of the cross section. Just figuring out the number for drag force, it says the ellipse is lower, which you got to some extent but the airfoil thing's kind of a mystery. Are you sure the airfoil was symmetrical and not producing lift?
The MDR number is the mach drag rise number, for some reason I got the impression that you were also testing this all the way through transonic.

Your results were with the same width tube. If you kept the cross section area or perimeter constant the width would be smaller. How would the drag numbers compare under those conditions. Its the legs that really matter in cycling performance.

The frame would lose stiffness if the width was reduced much, even with the same amount of metal.

"At 30 mph, aero funny bikes have almost 1lb less drag than standard road bicycles, giving them an enormous advantage in time trial."

p.33 High-Tech cycling second edition Edmund R. Burke, Phd. Editor

Wind tunnel test after wind tunnel test has proven this to be true.

We know that they rented a wind tunnel to check it all out so it has got to be something you're leaving out. It could be tha afore mentioned turblance from the front wheel or that they are leting the stiffness go and keep the area constant, they are using carbon I believe maybe that would make up the difference in stiffness.

Great thread though.

Joe

The golf ball thing reminds me of boat hulls is similar fluid tests. The argument was whether a smoth hull was more aerodynamic (hydrodynamic?) than one of lapstrake construction (separate boards with ridges). Seems that the ridges created areas that the surrounding water did not flow into at speed thereby reducing drag......like dimples on a golf ball. I wonder if rough or dimpled surfcaes on a bike (texured paint) would help reduce drag for TT applictions.
Tony
(aero and funny bike fan)

The golf ball thing reminds me of boat hulls is similar fluid tests. The argument was whether a smoth hull was more aerodynamic (hydrodynamic?) than one of lapstrake construction (separate boards with ridges). Seems that the ridges created areas that the surrounding water did not flow into at speed thereby reducing drag......like dimples on a golf ball. I wonder if rough or dimpled surfcaes on a bike (texured paint) would help reduce drag for TT applictions.
Tony
(aero and funny bike fan)

I've seen pictures of aero helmets that are dimpled for this purpose. I believe CWX also sells clothing designed to shave seconds off time trials.

"At 30 mph, aero funny bikes have almost 1lb less drag than standard road bicycles, giving them an enormous advantage in time trial."

p.33 High-Tech cycling second edition Edmund R. Burke, Phd. Editor

Wind tunnel test after wind tunnel test has proven this to be true.

How about a comparison number? How much drag does a regular road bike and rider have at 30 mph? What would the difference be at say 15 or 20 mph?

How about a comparison number? How much drag does a regular road bike and rider have at 30 mph? What would the difference be at say 15 or 20 mph?

Good question. Drag force is not linear to velocity. It is often a function of velocity squared (V^2).

At 20 mph, drag is proportional to 400 (=20*20)

At 30 mph, drag is proportional to 900 (=30*30)

Thus the drag force is over double at 30 than 20 mph. Realize this is only a simplistic explanation.

i've read about people doing about 2 mph faster on cervelo soloists.

Here are some numbers from the book. All tests were done at the same speed, 30MPH. These numbers are for the bike only in LBS of drag. The tests were done in 1991.

Trek, round tubes, 700c, 36 round spokes F&R 2.71
Cannondale...same 2.72
1986 Huffy Composite 700c disc R 24' disk F 1.89
Schwinn track, round tubes 700c 32 bladed 2.41
Kestral Prototype,Aero, 700c 36 round F&R 2.52
Hooker elite, 700 R disc, 24" 18 aero F 1.38

These numbers are LB of drag for the bikes only. From the chart I'm looking at standard road bikes with round tubes are around 2.5 while aero tubes bike vary greatly with most around 1.5 lbs. All tests were done at 30mph.

Retrogrouch, there were no speed comparisons. There is a graph showing drag at various speeds for the same bike with different rider positions. From the graph I can tell at speeds less than 20 mph it doesnt make a difference what you ride.

Roughly, from the graph, power output to maintain speed:

15mph 100W
20mph 220W
25mph 375W
30mph 610W

Notice the differences in Watts. From 15 to 20 mph theres a 110W difference while from 25 to 30 theres a 235W difference.
As a disclaimer I intersected the lines on the graph with a pencil and an index card as a straight edge but I think you get the general idea.

Drag coefficients of various shapes:

Round tube: 1.2
Square block: 1.3
Airfoil: 0.1 or less
2:1 ellipse: 0.6

Something else to keep in mind...I'm copying all of this from a book. I'd be more tham willing to look up anything else for you.

Jim

Here are some numbers from the book. All tests were done at the same speed, 30MPH. These numbers are for the bike only in LBS of drag. The tests were done in 1991.

Trek, round tubes, 700c, 36 round spokes F&R 2.71
Cannondale...same 2.72
1986 Huffy Composite 700c disc R 24' disk F 1.89
Schwinn track, round tubes 700c 32 bladed 2.41
Kestral Prototype,Aero, 700c 36 round F&R 2.52
Hooker elite, 700 R disc, 24" 18 aero F 1.38

These numbers are LB of drag for the bikes only. From the chart I'm looking at standard road bikes with round tubes are around 2.5 while aero tubes bike vary greatly with most around 1.5 lbs. All tests were done at 30mph.

Retrogrouch, there were no speed comparisons. There is a graph showing drag at various speeds for the same bike with different rider positions. From the graph I can tell at speeds less than 20 mph it doesnt make a difference what you ride.

Roughly, from the graph, power output to maintain speed:

15mph 100W
20mph 220W
25mph 375W
30mph 610W

Notice the differences in Watts. From 15 to 20 mph theres a 110W difference while from 25 to 30 theres a 235W difference.
As a disclaimer I intersected the lines on the graph with a pencil and an index card as a straight edge but I think you get the general idea.

I look at the same data and get an entirely different impression.

It looks to me like there's a modest benefit from using bladed spokes and a very significant benefit from using disc wheels. The frame tubes don't seem to make much difference which is kind of what I intuitively expected. The Aero Kestrel, for example, was only marginally better than the round tube Cannondale. Once you put a rider into the airstream, the percent difference is going to get really small. On the other hand, the faster you go the more benefit you get from even a small aero improvement.

I broke my right collar bone so it's overlapped and now my right shoulder is about 2" narrower than my left. If I was really serious about getting aero, I could have my left collarbone broken and repaired the same way and reduce my frontal area.

I think I figured out the why behind my results. I have an older version of the program and while it appears to be giving accurate results for the aero shapes the circle is making it wiggy for some reason, I decided to try a few things with just the round shape to test things out and it gave some obviously wrong answers.

I also figured out why the elipse has a much lower drag than the teardrop. It has much more laminer flow and at these low reynolds numbers friction/boundary-layer drag is much more significant than it is at the higher numbers found on bigger objects and higher speeds. Under these conditions boundary-layer drag is likely much higher than pressure drag. So, at this low Reynolds number, limiting boundary layer growth is more important than reducing the front-rear pressure difference. At high reynolds, pressure difference becomes more important, hence the teardrop shapes. This is from books.(I useta know this, even I have to refresh once in a while.) This being the case the round tubes should be fairly competitive with teardrop shapes as the round tubes have thiner B.L.s and just as much laminar B.L. as the teardrop shapes and less turbulent B.L. which would cancel out the higher pressure drag.
What do ya all think.

i've read about people doing about 2 mph faster on cervelo soloists.
It would seem that way for me...My old bike Vs my Cervelo, although I have a better wheelset on the new bike. (same tyres). I can go out and average 29-30kmh on certain types of rides with less training than with the old bike where I would be about 26-27kmh.

I beleive there is a bit of info at www.cervelo.com

Brendon

Just thought I'd throw this up here. 1980 Huffy Aerowind. Squished, not drawn oval tubing. I haven't noticed it being faster than my other bikes, lol.,,,,BD

Bikedud, I thought I'd never see one of those again!!!! Its not fast because it takes all of your energy just to get it moving!!

It's not "that" heavy. It weighs less than my 71 Continental, that's for sure. I had an Aerowind back when I was 16. The first bike I ever paid for with my own money. Less than a week later it was stolen from school, yes it was locked. I think the gym teacher liked it, he was the teacher in charge of the lock cutters. Gone in less than four hours, from a locked parking lot, the first day I rode it there. Sounds fishy to me? Finally found another one last year, was a store demo that was put up after they were discontinued. It's really in almost new condition.,,,,BD

Okay, so about the whole golf ball issue - I have always wondered why no bike or car manufacturers have never attempted something like this. Does it just not work on the scale of bikes or cars, or do they think it would be too ugly, or is it too expensive, or have they just never thought of doing it?

Any ideas, anyone? I always thought it would be really cool to have a bike frame with golf ball dimples all over it! or maybe I've just been smoking too much, I dunno...

Okay, so about the whole golf ball issue - I have always wondered why no bike or car manufacturers have never attempted something like this. Does it just not work on the scale of bikes or cars, or do they think it would be too ugly, or is it too expensive, or have they just never thought of doing it?

Any ideas, anyone? I always thought it would be really cool to have a bike frame with golf ball dimples all over it! or maybe I've just been smoking too much, I dunno...

Structural integrity, shot peening can increase the fatigue life of a product but all those dimples are a lot bigger and can be dramatically increase stress risers in a frame. I've seen time trial helmets with dimples on it like a golf ball though.

EDIT: Apparently the solid wheels are also dimpled. http://www.roadcycling.com/news/printer277.shtml

Lexus also had a commercial touting a that their luxury car's underbody has a dimpled panel to reduce buffeting from pressure and noise.

DID YOU KNOW: Golf ball dimples were discovered by accident when it was realized that old used balls with a lot of nicks on them traveled farther than new smooth balls. These "experienced" balls became the basis for the pre-nicked dimpled surface.

ALSO: People with dimpled cheeks given the same frame and body and run, can do a 10km time trial 2 seconds quicker due to the aerodynamic advantage of their face. ;)

Straight from the wind tunnel....

http://www.slowtwitch.com/mainheadings/techctr/frametube.html

Structural integrity, shot peening can increase the fatigue life of a product but all those dimples are a lot bigger and can be dramatically increase stress risers in a frame...

Damn physics, always confounding me! First I find out I can't gold plate myself, now this. What next, are the boffins going to tell me it's impossible to build a ladder to the moon? Aarrgh!

Oh well, thanks for the confirmation.

Straight from the wind tunnel....

http://www.slowtwitch.com/mainheadings/techctr/frametube.html

Ahh HA! My reverse egg pointy shape was better!
Actualy sharp leading edges are known to produce less drag, by those in the know. The reason airplanes use round nosed wings is for stall characteristics relating to angle of attack and drag in a variety of wind directions, the wider your wind angle window the rounder the leading edge must be to avoid boundery layer seporation (causes a dramatic drag increase. AKA stalling.)


Okay, so about the whole golf ball issue - I have always wondered why no bike or car manufacturers have never attempted something like this. Does it just not work on the scale of bikes or cars, or do they think it would be too ugly, or is it too expensive, or have they just never thought of doing it?
Golf balls need dimples because they have to be round those other objects are better off using different shapes and maintaining laminer flow as long as possible. Dimples turn laminer into turbulent flow which sticks better for reduced pressure drag but adds friction drag.

First I didn't totaly finish my first post, I was being kicked of the computer and now I don't remember what I was going to add then.
Here are some results I got for better understanding
This is frame tubes only, no parts.
All tests at 35mph and 1.25 inch wide tubing, at sea level, 59f degrees.

NACA 4 digit 0050(teardrop 1.25" wide 2.5" long);
0.125 pounds clean, 0.141 pounds of drag dirty.
Round;
0.064 pounds clean, 0.081 pounds dirty.
Elipse, 1.25" wide 2.5" deep, symetrical front to back, rounded on both ends;
0.057 pounds clean 0.082 pounds dirty.


They reduce the amount of spokage(A real word?... no.) in the highest velocity area and a standard square rim is not a circle. Just a stab at it but I'm talking frames in clear air.


Mudpie, you had it pretty much on with the golf ball. The reason for dimples is that they trip the laminar boundry layer early on because the resultant turbulant boundery layer is more stable and even though it is higher in drag it allows the flow to stay attached longer reducing the low pressure wake. I can force the laminar-turbulent transition prematurely if I want to with this program, to simulate dirt and bugs or dimples. It does seem to get slightly batty with real radical wing shapes at very high angles of attack but it seems to be working consistantly in this case and these arn't radical shapes anyway, very standard in fact.

Quiz: Any one know what the word GOLF originaly meant?

The origin of the word GOLF is debatable; the medieval Dutch word "kolf" or "kolve" translates as club.

That aside, spelled backwards GOLF is FLOG which is a good description of my game. I could get better, but that would mean less time in the saddle....unacceptable.

Damn physics, always confounding me! First I find out I can't gold plate myself, now this. What next, are the boffins going to tell me it's impossible to build a ladder to the moon? Aarrgh!

Oh well, thanks for the confirmation.

Oh but you CAN gold plate yourself...

humm...interesting. that's like taking the "non-fat" ice cream into the lab and finding fat. do makers of bikes with tear drop tubes claim they are more areodynamic? or is it just how they look?


Actually, it's a bit of both. I spent about 20 years in the bike industry. One of the puzzles that companies had to solve was "If everyone's hanging Shimano or Campy on their bike, how do we make our bikes different, so customers are buying off of something other than how much Dura-Ace Bike X has.

The result was mixed. Some companies came up with visual things that intrigued the customer but offered no advantages or actually performed worse than the design they replaced (e.g.rigid straight forks on mountain and road bikes). Some designs were dictated by the material used. The best way to get a strong and light aluminum bike is a fat tube with a thin wall (Cannondale). This last one was genius because the American mentality will generally gravitate towards bigger, all things being equal. Skinny tubed steel bikes were perceived as wimpy next to the "Dales". Throw in the lighter weight and you've made steel bikes an endangered species, which is exactly what happened.

The current compact road frame design incorporating a sloping top tube actually began with mountain bikes (Kona), the idea being that you could get more clearance for off-road use so as not to bang your cajones in an emergency dismount. Also, you could fit more folks with fewer sizes (just like the road models). Better for inventory control. Because the top tube slopes, the theory was that vibration from off road riding was reduced since it the angles made the riding more compliant. Lastly, with a shorter seat tube, you should get a stiffer seat to bottom bracket connection which would improve energy transfer.
In truth, this design does work and it does give a manufacturer a frame that looks different which is the best of both worlds.

Re: Aerodynamics, they make a bike look fast sitting still. To really obtain a fast aerodymanic bike, you need a company with a big enough R&D budget to design a bike that actually is faster. Of the aerodynamic offerings out there, all look fast but not all are faster. Obviously, as well you need a "Lance-type" motor as well.

One example of a time-tested aerodynamic design: the front wheel on Team Postal's TT bike was originally designed by Specialized over 10 years ago in conjunction with Dupont. They used a Cray computer to design the wheel as well as extensive wind-tunnel testing. Early editions of this wheel were too heavy, but as these were ridden, it was discovered that less carbon fiber was needed in the design to maintain integrity, also the foam core of the wheel was replaced by an lighter advanced foam used in space shuttle applications. The result was a wheel that was light and aerodynamic. Eventually, Specialized sold the rights to produce and sell this wheel to HED which now markets it as the HED3.
Even though some of the Team Postal TT bike specs (frame & parts) were changed from last year, this wheel remains, because they feel it is the best aero front made.

Okay, so about the whole golf ball issue - I have always wondered why no bike or car manufacturers have never attempted something like this. Does it just not work on the scale of bikes or cars, or do they think it would be too ugly, or is it too expensive, or have they just never thought of doing it?

Any ideas, anyone? I always thought it would be really cool to have a bike frame with golf ball dimples all over it! or maybe I've just been smoking too much, I dunno...

Re: Dimples, No bike frames with dimples, but TDF Phonak Team using ZIPP 808 wheels with dimpled surface on rims. In the discussion on the development of these wheels, golf balls were mentioned. Evidently, it was very difficult to extrude the metal with this patter. See www.zipp.com

Re: Dimples, No bike frames with dimples, but TDF Phonak Team using ZIPP 808 wheels with dimpled surface on rims. In the discussion on the development of these wheels, golf balls were mentioned. Evidently, it was very difficult to extrude the metal with this patter. See www.zipp.com

http://www.cyclingnews.com/road/2004/tour04/tech/?id=zipp

Above link has nice picture and story about dimpled technology..

... Evidently, it was very difficult to extrude the metal with this patter...
I'll say. My arm tired out pretty quick swinging that ballpeen drag reducer that I bought at the 'As seen on TV' store. :p