Mavics don't seem to do well in independent wind tunnel tests and reviews I notice, even though it has as you say the preferable NACA profile. Why do you think that might be?
That's a very good question. The most important consideration are the tyres, mounting a tyre that protrudes wider than the edge of the mavic rim will cause significant turbulence at the back of the wheel. It will also cause turbulence where the clincher meets the tyre bead. Speed tests are often conducted at 40km/h (50km/h in the latest bikeradar test), they do this because if they did this at a lower speed, there would be insufficient difference between the best wheel and the worst wheel to justify the horrendous price difference. The average "good" rider will probably be going at 35km/h (air density will affect this), most club riders are at 28km/h. At these "mortal" speeds a NACA aerofoil is quite effective because it's side to side vortex shedding is limited. A toroidal aerofoil is really only good in completely steady state conditions such as a wind tunnel. Out on the road which is where it matters, there are various bits of road furniture that upset the airflow coming onto the rider - technically this is called transient analysis. Items such as lamp posts, a passing car, a hedge all affect the airflow. Another factor which nobody seems to have commented on is when a rider is going full gas, their bike tends to rock slightly from left to right, only by a few degrees but this upsets the oncoming airflow. An aerofoil which sheds fewer vortices during this process will be more aerodynamic. Measuring the drag of a car in a wind tunnel is not a perfect science either but the car's velocity relative to the effects caused by street furniture is much greater. A car can be doing 120km/h and the street furniture is maybe pulsing the airflow by 2-3km/h hence the delta is massive. On a bike which is doing 35km/h and the furniture is providing 2-3km/h of pulsing the difference is much smaller. So to summarize, the rim depth has the most effect, mounting a wide(r) tyre onto a narrow rim is a big no no and a wind tunnel is a poor measure of real world (pulsing) conditions because wind is not constant in speed or direction and the speed magazine's use for their analysis is a speed that most riders do not or cannot ride at. I want to emphasise that at around 35km/h there would be little to separate any of the wheels provided the section depth was the same and the tyres were appropriately sized. The choice of aerofoil section is really only a marginal gain. My choice for the NACA aerofoil is the best choice of those marginal gains.
All good points there. Wheel makers are designing wheels that beat each other in the wind tunnel without consideration for real world conditions. As you say, a side gust of 20km/h is inconsequential to a 1500kg car travelling at 100km/h, but very significant to a 70kg rider at 30km/h. The 'rocking' bike is interesting too and will create extra turbulence. Do wheel makers test for this? No! My V-profile Lightweights are BETTER in sidewinds than the more toroidal Corima 47 S+ (same depth). Neither is a problem in winds up to 70km/h or so. Showing the difference between a 50mm and a 30mm wheel at 35km/h is hard enough, let alone the difference between 2 x 50mm wheels.
I have tried to model the rocking motion but it is extremely difficult. You basically generate a vortex to the right and then to the left but it's not consistent at all. This is coupled with micro steering corrections/movements. Your point is correct though, the wind tunnel versus real life is probably 70% accurate at best. Accounting for road furniture is difficult to do in a wind tunnel Further, the guy at bikeradar described shimano and campag wheels as "canon fodder". I'm sure he has done his credibility wonders.
I live in the US and I’m a member of a Cycling Club with 600 plus members, I’m glad I found this channel and can share all this knowledge with my Cycling buddies. Congratulations and keep up your outstanding work.
I have bike raced since the early 80's. One of my friends who bike raced had a dad who in my opinion is the founder of modern fluid dynamics, and has authored a number of college texts on fluid dynamics. I remember one day years ago running into him and we talked about aero bicycle wheels, namely Zipps. They had just come out with the dimpling. When I asked him what he thought, he replied, "Figuring out drag coefficients is a black art, even with the best computer programs. I don't know where Zipp gets their numbers from but they are really creative." Thank you for this.
This was an incredibly well done presentation Mr. Hambini. I think my students will very much enjoy watching this. So, I am going to show this to them tomorrow. (I am a physics instructor.) Thank you and kind regards.
Hey amazing video and I really enjoyed it! Just a disclaimer, I have no degree in engineering but as I understand some of the research out there is that the reason the wheel industry moved away from the NACA profile and into toroidal shapes is due to stability considerations. Apparently, the average yaw angles that cyclists encounter are between 5-10 degrees and this is where the toroidal shape performs better than NACA profiles (and some claim to have a "sailing" effect" akin to sailing into the wind). Also, I read somewhere that the discontinuity between the tire and the rim actually creates a beneficial vortex that causes airflow to reattach to the rim (if the tire is narrower than the wheel). And one last thing about toroidal shapes is that it creates a similar profile between the leading edge and the trailing edge of the wheel-tire system. For sure some of these is marketing as some of the research where than by the wheel manufacturers but I would like to hear your thoughts on this. Thanks again for the great video!
Hello Franz, thankyou for the comments. The key thing to realise from this analysis is the type of aerofoil profile has very little effect in comparison to the length of the aerofoil chord (ie the depth). The difference between a toroidal and a naca profile will not be much but the difference between a box that is 80mm and a Zipp 404 will be massive - the box will win most of the time. When you get to a yaw angle where large degrees of separation take place, it is more advantageous to have a shallower section wheel (especially in the front). In the range of 5-10 degrees, a NACA 0024 profile should be better than a toroidal shape because it would be no where near separating. I estimate at the Reynolds number for the top of the wheel, the stall angle is about 12 degrees. Link here airfoiltools.com/airfoil/details?airfoil=naca0024-il
Ok, I love your enthusiasm toward this aspect of airflow. Obviously you're extremely bright and into bicycling. While I realize the complexity of including the hub and spokes would be astronomical...they represent a MASSIVE amount of area which cause a significant amount of turbulence ~ I didn't consider that road "furniture" would play into this equation either...that was shocking;-) Thanks for the upload!!!
Years back, I had 30mm aluminum clincher and 44mm carbon tubular wheels in the classic "V" shape, which was popular back then (10-15 years ago). I don't really recall having to fight the wind very much and having my bike blown all over the place with these wheels. With my HED Belgium C2's (24mm) and Enve 3.4's (35mm front and 45mm rear), the wind is brutal and my front wheel gets incredibly sketchy with the "U" shape.....even my 24mm HED's get blown around in gusty crosswinds - and the tire profile fits much better/cleaner with my current wheels than my older (and narrower) wheels, which had the bulb shape going on back then.
the amount of twitchyness is related to how much separation you get. it might be a particular combination that sets it off. largely the separation is related to the Reynolds number of the wheel and fluid
Hambini: that's true to a degree but twitchy behaviour also is affected by centre of pressure both for the front wheel in isolation (with the imbalance causing a steering torque) plus the bike as a whole. E.g. If the entire bike centre of pressure was at the back axel (approx) the bike would not be turned.
Jof Arnold Also, the degree of head rake has a dramatic effect on the tire staying straight...less angle twitchier steering...too much angle and it "flops" into turns. In the quest for overall weight reduction those are negleted b/c of target riders, target weights, cost and deadlines~ It's a black art that many manufacturers overlook.
Fascinating insight Mr Hambini. The big take away for me was how quickly, given the number of major assumptions/simplifications made, you got into deep water! Despite this, there are still calls for you to delve deeper into the murky world beyond haha! Good luck to your PC with that! For the mortal rider constrained inside the rules of competition, the potential gains seem too small to bother about, outside, stream liner bodywork would seem less bother than individual attention to each and every component, if of course you're prepared to compromise the look of your bicycle in that way.
Thank you so much! I have been an aircraft enthusiast all my life,and what your saying makes so much sense. There's so much marketing hype out there! Really grateful to you for blowing these myths apart!!! Subbed,
Awesome great vid as usual! You have my appreciation and thanks for busting these myths and marketing hype! I've always been searching for unbiased independent test results on aero drag and I find comfort in knowing that the cord length is what makes it count mostly. Would be great if moving forward there could be an analysis on the differences in drag between a disc and non-disc aero wheel. Looking forward to more of your word Hambini. Thank God for people like you =)
I do agree with you about the dimples as well!-) Also the sheer PRICE difference between the top tier equipment and the intermediate range matched to "aero advantage". The KWs needed to achieve higher velocity are reduced over hours...BUT at what cost and how many non-pro riders ACTUALLY have the leg power/stamina to seriously take advantage of it? I understand the appeal (like super cars etc) but....😱💥😁
My UA-cam watching is now within 3 years of realtime! As a casual observer, are you certain you modeled the 2nd wheel with rotation? The leading edge, especially on the top slice looked too nice to me. As a TV consumer, MythBusters did a show with golf ball-like dimples on the surface of a car, and found a noticeable difference in gas mileage. The dimples in the first wheel may be more than snake oil if you have interest in following up!
Hambini Sir, I summise that, to use a 23mm tyre then a rim of x would be sufficient for minimum drag and a 25 tyre will have a deeper rim depth (y) because the teardrop/NACA shape of the rim&tyre system needs to be bigger to incorporate the bulge or balloon of the tyre. A 28mm tyre would require, an even deeper rim. Would that be right enough?
Thanks for the modelling. This draws a question (from my non-engineering profession background)... why would more people not incorporate a carbon fender with good aerodynamic shape characteristics to cover the top half or more of the front wheel (assuming you weren't worried about crosswinds as much and its a calm day)? Wouldn't that reduce a lot of the front trailing edge turbulence effects by creating a continuous surface (more or less) to flow into the frame better? Obviously the speeds aren't comparable, but you see now in some Moto2 bikes where the front fenders cover a large portion of the fork/brake surface now, which probably aides in not having wind hit all those surfaces in weird ways. Has anyone ever experimented with front carbon fenders/shapes to test the affects on this with good results?
This is the first of your videos I’ve watched and it was awesome, thank you so much! I studied engineering (a long time ago) and its great to have such a clear explanation of the fundamentals of drag forces on wheels. Have you calculated the CdA of the deep section 808 versus say a 404? I’ve just invested in an 808 for my TT’s and keen to understand the CdA improvement likely from the deeper section. Also, I saw in the comments you didn’t see much benefit in a rear disc, at least in side wind conditions. I’ve been riding a disc this season compared to a rear 404 last season. It is definitely faster. I’ve calculated my CdA using the formulae for cycling resistances (inputting other known parameters such as power, weight, temperature, humidity, air pressure, and reasonable estimates for mechanical efficiency and rolling resistance). Based on my calcs across many efforts this season and last, my CdA has been lowered by 0.014. The disc is the only difference between this season and last, so I had attributed that to the disc. Interested in your thoughts as to whether you agree or not. Again, awesome video, I’ll now be looking at your other content.
Hi Briony, It's a bit difficult to say for certain that it's the disc that is causing the difference. I have never found these generic formulas to be particularly accurate - they get you in the ball park but to get to 0.014 of precision would be pushing it. However, parking that to one side. The disc will give you an Aero improvement, the impact would be a function of your body and bike in relation to it.
Thanks Hambini. When you say you don't find the generic formulas accurate, what generic formulas are you referring to? I mean the physic equations that allow you to calculate speed based on gravitational resistance, air resistance, rolling resistance, and mechanical resistance. While there is some variability day to day (probably because of differing wind conditions), you can minimise these errors by choosing a loop course. And by measuring CdA numerous times for each configuration, I think you can generate reasonable average CdA figures for each configuration over time. The chart below is an example of CdA with different clothing, keeping all other parameters constant. If I was proficient in CFD I'd try that also. But without CFD, the field test is the best I've got. Hey, does CFD allow you to quantify the CdA benefit of a deep dish Zipp 808 versus say a Zipp 404 or 202? I'd be very interested. Thanks.
The generic formula used for drag is CDA * 1/2 rho V^2 to give the drag force and then apply a sin or cos correction for angle. In my view this is not accurate and especially bad for a wheel, Wheels are rotating so the CD value changes as the wheel angle (yaw) changes. The formula you use above assumes a Fixed CDA. I guess at the end of the day, it's how much of an accurate answer you want. Hope that helps Hambini
OK, I understand, thanks. Yes, the calculated CdA for a field test is an average of various yaw angles, and is also for the entire and unique bike + rider system. By isolating one change in equipment and doing numeous field tests I’ve been able to determine statistically significant differences in the CdA for different configurations. But there is variability day to day and I see that the CFD method has some major advantages. Thanks again for the insights Hambini.
If you're chasing that little of a gain it might be better to understand what your true intention is behind it. Max speed? Overall avg speed? What chassis is best suited for that? The number 1 most important question of all...how do you train? Do you have the leg power/endurance to get the full advantage out of your mathematical endeavors? Math is a great start...but out on the road and in competition are the TRUE tests of what works. I love your attention to detail! I'm sure your one hell of a competitor!!! 😁😀👍
Good Morning Sir, As i understand it, since the wheel is at any one time 50% aerodynamic to the direction of travel [please correct me if im wrong] Would there be any benefit to attaching a LMP1 style wing or bodywork type idea to the fork, to mask/protect the trailing edge (6 o clock to 12 o clock) of the wheel to prevent those vortices creating drag? (I understand that the first thing F1 designers would do in regulation-free rules, is encase the wheels). Cheers Hambini!
Awesome software and great explanation, thanks! Mavics are certainly great in a perfect world but in reality you rarely hit the wind head on. Also the rough surface is used in many disciplines including aviation to reduce early separation and the results are spectacular.
Hambini, your videos keep restoring my faith in humanity. I have a question, though. I understand that spinning up wheels require power. But once wheels are in motion, their kinetic energy is constant and there should be no need to use power (i.e. add kinetic energy to them) -- unless they somehow interact with the air and drag it along. Is that what you had in mind -- that the air resistance of spinning wheels is different from the resistance of wheels at rest? Thank you for your answer.
Also, vortex isn’t drag but indicative of drag. Thrust energy is charged into volume of air molecules by compression during the frontal section that is followed by energy return from air molecules in the expansion phase by the trailing section, provided that we have lamina flow during expansion. Energy in vortex is harnessed from energy in the expansion flow. Suppress vortex generating can conserve the expansion energy back on forward thrust. The end effect is what we call low drag.
The dimples: Schuberth helmets used these to reduce airsound. The theory behind this is a controlled airstream without random stalls. The efficiency of the helmet got worse but the difference between silence and tornado feelings became less. And this is why Zipp put the dimples on the rim. The efficiency got worse but the crossover performance became more repeatable. BTW: NACA was the predecessor of the NASA and developed wing profiles for airplanes. You referred to the fully symmetric ones (0009 or 0012), but they made a lot more. However for bikes we need the fully symmetric ones only. Nobody came to the idea to use a RG 15 profile for crosswind situations (would be funny if the wind direction changes).
Right, finished it. Jeez, that was good mate. You seriously need to do more stuff like this. Phenomenal. Question; the fork, hub and spokes will obviously affect air flow, so does that improve things for the top half of the wheel or add to the messy flow?
It makes it more messy, One of the things that you can do is when you disturb the airflow you can put a structural member immediately behind it to effectively have it operating in a drafted section. For example My cervelo S5 has a "dropped" downtube, whoever designed the bike realised they were going to get turbulence generated by the front wheel which they could do little to avoid and put the downtube as close to the wheel because it has to go somewhere so lets put it in an area where there is little velocity. This is similar to a bike rider drafting a car, they are sitting behind the car, the car has disturbed the air and they are riding along in an area with no appreciable wind. If they slip slightly back then the air has a chance to straighten itself and they get the full brunt again. The type of spoke has less effect as the rim depth gets bigger because the spoke is shorter and the relative velocity reduces.
That's a waste of time, real air outside of a windtunnel is never laminar, the various yaw angles are htting at the same time. Pick any angle and you will get numbers, all of which are meaningless.
Hi, great video. I’m wondering if there’s any new learnings since you made this video 3 years ago. Specifically, I’m wondering if you still think the NACA profile is optimal in the real world where there is constant fluctuations in yaw and wind velocity. Thanks!
I can definitely understand a leading edge / leading edge discussion, however, can a discussion of the trailing edge of the front edge really take place without the spoke, hub, and fork structures in place?
A fucking BMC time machine. I appreciate the Hambini reviews but would love to see the tests and data points collected on more down to earth frames and platforms.
Love your videos. This is the best one of them. Thank you for taking the time to make it. But I'm still confused I get it that 1. Wheel depth is the overwhelming determinant of how aero the wheels are. 2. But steep yaw angles cause stall, and stall makes the wheel less stable and causes you to lose the aero benefits of the deep wheel. 2. A toroidal wheel is *supposed* to help solve that problem by deferring the stall to steeper angles, right. Is that all correct? If so, well how often does a toroidal make a difference? How often does stall matter? Because in your example, you only used a head-on wind model. For 3 months I've been trying to decide on a pair of aero wheels for my road bike and even after months of googling every aero wheel in existsnce, I'm still confused. Ignoring weight and price, what really matters for real world aerodynamics? Should one just go deep and ignore the shape? How often am I going to hit wind angles where that toroidal shape matters in the real world? Perfect example I was looking at today. Deep-but-flat Fast Forward F6R (60mm/60mm depth) and shallower-but-toroidal Enve SES 4.5s (45/55 depth) . I've been through many such comparisons with many different wheels but after 3 months I have no idea what really matters between them in real-world winds.
I love your videos. I am confused though. You said that at the bottom of the wheel the drag is much less important because the wind is moving in the same direction as the wheel. But then when you started to talk about chord length, you demo'd it at the bottom of the wheel?
Great vid, clearly explained how the deeper section is more aerodynamic. I've had a (quick) look to see if anyone has asked this question already. Simply which wheel is going to be faster up a climb, a wheel that is heavier with a larger section or lighter wheel but has a shallower section. If the lightweight wheel has a section of 33mm and weight of 1250g and the more aero wheel is 55 mm at 1400g, climbing at a speed of 14kmh. Climb has gradient of 7% (I don't know if that's relevant). Which wheel would be quicker on the climb? Thanks
Really good job at tackling a very complex subject. Thanks. The one question/surprise is why you didn't comment much on the stability effect of the toroidal shape at the average yaw angle. I bought into that argument in my latest wheelset purchase and I do think that i notice the benefit, versus my other bike which still has naca profile rims/tyres on it. Do you think that there is a stability benefit and, for you, might that outweigh pure aero cut through out on the open road?
Hi Hambini and thanks for sharing all this amazing knowledge. Aren't you doing the air simulation on the wheel you design with the airflow perpendicular to the wheel instead of inline with iit as you do with the shimano one? Thanks again for all the knowledge
I'd love to see how the airflow is different w/ the inclusion of a bottom tube (esp tight clearance back there), fork and esp the crown area. Most of the disrupted flow was on the top half of the wheel which is where those items would have a great effect. Not to mention spokes, but I'm sure those would be very difficult to model.
Going down the road the top of the wheel is double the speed and the bottom of the wheel is standing still. Freewheeling in a bicycle stand yes it would be going backwards. Not sure if this would change the airflow?? Thanks for your channel, I’m going to be purchasing some ntn bearings for my bicycle
Excellent rim review however I do have a couple conversation points. I think the spokes and hubs would have a significant impact on the airflow of the wheels, specifically the rim to spoke conversion of airflow. It would also be interesting to see how hub design will affect the aerodynamics of the wheels since there are a huge variety of hub designs out there. Another interesting point to think about is, how could you make the bottom section of the wheel as aero as possible and at the same time have no effect on the top section of the wheel. Some sort of hybrid wheel, in which the bottom section stays at the bottom and top stays at top while the wheel is in rotation. But this design is way too complex to engineer since there would need to be a collection of different spinning points on the wheel.
Would you be so kind to make a video explaining why and if It's true that larger tires have less resistance/friction to the ground at lower pressure? 30 < 28 < 25 < 23? Is it just industry marketing bullocks ir there is real physic behind?
Pardon my incompetence. How does drag and turbulence affect a wheel, tyre, and spoke at bottom dead center when it is stationary irrespective of forward speed. Also at top dead center: wheel, tyre spoke combo is moving at twice road speed. Will viscous drag equate to (m/s x 2)^2.
Always a pleasure. Is there any rule of thumb way to estimate what fraction of the total drag comes from the wheels? It seems like for us bigger riders who aren't particularly fast, wheel drag is not something to worry a lot about.
Thanks for takling the time to make this video!! Zipp Binned the 1080 after the 808 tested faster. Can you speculate why the 808 might be faster Even though it has a shorter chord length? Also, it’s my understanding that the advantage of the toroidal rim shape is for cross wind stability. I’ve only ridden toroidal rims myself so can’t directly compare to the older style V shaped rims which seem more closely related to the NACA rim shape you mention. But I’ve always read that the v shaped rims (and presumably NACA shaped rims) cause the cross winds to steer the front wheel resulting in stability issues while riding. Of course when it’s windy I can feel my whole bike gets blown sideways and more so with 898s fitted, but I never feel any steering effect through the bars and find the 808s totally predicatable energy on moderately windy days, with forecast winds up to say 12-15m/s
I wasn't buying the golf ball dimples on the rim from Zipp either. I wouldn't be surprised if tire manufacturers took up the torch and started putting dimples on the tires though. At 11:00 you said that the aero benefits come from pure geometry as opposed to the toroidal shape. For the most part, would it be correct to say that Campagnolo Bullet 105s are at least every bit as aero as the Zip 808s provided that the tires used are a very similar width as the leading edge of the rims on both? Edit: Campagnolo Bullet rims are very similar in shape to the rim shape you drew and showed at 15:10
Great stuff. I kinda regret turning down Aeronautical Engineering at Southampton for Mechanical at Kings! The lure of the big city..... Also great to see Zipp's golf ball dimples rubbished; as you say they merely serve to increase skin friction. I strongly suspect the new 454 'shark's fin' design is more of the same. With one 50mm wheel resembling every other, you have to try and stand out in some way. Interesting too to see the importance of tyre to wheel interface and hopefully a rubber stamp for my 'silicone your wheel gap' vid!! What is your take on 'power to spin'? Although the majority of drag comes from the wheel's forward motion, one of the benefits of a deep rim wheel is shorter (and thus slower) spokes. Therefore, spoke profile, length and number are (IMO) all also important.
Yes, would you mind if I linked your silicone gap video in the description? On your second point. The reason for golf balls having dimples is to make the boundary layer go turbulent and therefore keep flow attached longer. You don't get that benefit on a bike wheel because the rotational speed is too high and the net velocity vector is never zero (there is always a tangential component). The CFD also shows the same thing. With the advent of disc brakes, they can ditch the brake track and make it more teardrop shape. That should produce a The spokes are a necessary part of the wheel and all of them have similar flow characteristics. Unfortunately the spokes will all have the same linear velocity for a given length and you gain more by having the deep rim than by having a long spoke. The shallow wheel advantage kicks in when the yaw angle becomes more of an issue. And as for Aero engineering. Out of a bunch of 30 of us, 3 went on to become engineers (myself included). The other 27 went on to work for large financial institutions as traders!
No problem for my silicone vid, I'd be honoured! As for the spokes, I've found that some carbon spokes (like Mavic R-SYS, large and circular) are quite poor aerodynamically, and will create significant rotational drag. Not all spokes are equal - some are more bladed than others. It seems to be an area that is often overlooked. Funny for career path - when I graduated in 1991, there was no hope of an engineering job (recession) so I ended up in trading through desperation!
Idk all the physics of the subject, but real world feel is this; 2 very similar bikes same make, one has Zipp 303 and other Yoleo 60 mm. Both fast, responsive and very close in weight. The real diff is how they handle wind. Zipp stay put and the Yoleos are a handful at times. Had a decent recently with a side wind.... let's just say I'm glad I was able to stay up right, had to clean my bike shorts lol....now Yoleo only on calmer days.
Cool vid :) But your calcs on density (Aus v UK) are possibly a bit off (esp the 60w difference). If you assume: AUS: temp 35C, Alt 20m, HPa 1020, Rel H 20% Air density is 1.147. Assume Crr 0.0052, 300w @ 30kmh = CdA 0.768 (which is super high, typical road racing cyclist ~0.330, typical triathlete ~0.285) UK: temp 10C, Alt 20m, HPA 1020, Rel H 90%, Air density is 1.248. Assume CdA 0.768, Crr 0.0052 Speed = 8.12 m/s = 29.2 kmh instead 30kmh. Power to ride @ 30kmh = 323w or only 23w difference. Maybe you could get a 60w difference due to air density, but it would be very extreme, and perhaps not really typical riding conditions. Air density certainly has a big effect on the success or failure of hour TT records, where higher altitudes do allow slightly higher speed and make success more likely.
Hi Rob, Thanks for your question. I didn't calculate the air density, it has come from Jet Engine test conditions data from real places. For the UK, I took average density recorded over december at 1.29kg/m^3, for Australia I took the conditions of Cyprus in July 1.09kg/m^3. Additionally, the air becomes much more viscous as it approaches zero celsius and therefore the drag coefficient increases. In your estimation you've assumed it's constant. It becomes more like trying to go through treacle than going through air. I don't know if you have an Aerospace background or not but the governing factor is the Reynolds number, in practice it would define the detachment points of airflow. I think 10 degrees for a UK winter is charitable! Thanks Hambini
Nothing replaces simulation to actually identify the vortices which indicate high and low pressure zones. As it was explained to me, the dimples on the Zipp wheels are designed to mimic the effect of dimples on golf balls which have a profound effect on drag by reducing the low pressure zone (drag) behind the ball as it travels through the air. As I understand it, the dimples allow a golf ball to travel approx. twice as far as one w/o. I believe to be some merit in their design, but how much would actually require a lot of testing (ie. a lot of money and time) to debunk. The biggest problem that I see in Zipp's dimples is that air isn't able to travel over or under and they still expect a spherical shape of the dimples and the ball itself for air to detach and reattach especially on the top half of the wheel. I suppose the reason companies like Look and Cervelo have adopted NACA airfoil designs is that they're proven, and wheel patterns would not be so different aside from the fact that frames are measured in a static environment with only yaw variables. As always, amazing video!
This dimple business will be covered in the aerodynamic masterclass video. You are right in the sense you are trying to reduce the pressure drag by keeping the boundary layer attached more. It does not work on bike wheels because they have a fairly smooth tyre so the layer never develops.
I saw a vid where ex F1 Sauber old mate from Suicide sorry, Swiss Side, talked at length about sailing effect. I understand the principle where windsurfers can travel faster than the wind but is this a factor in cycling or simply BS? Thanks, really good video.
My great doubt is about what is better, aero profile carbon rims or low weight aluminium rims (I mean 100g less per wheel in aluminium wheel) to ride in velocities like 35km/h in a flat course..
i really like how you make it more simpler to understand. thanks a lot Mr.Hambini. i got a question tho.... talking about drag is there a certain speed at which it will start to take place if we are to consider deep section wheels?
Very interesting. Just wondering, if the top of the front wheel heading towards the front of the front wheel is creating drag, if you where to put a full coverage fender from the top of the wheel towards the front of the wheel, would that have less drag going over the fender vs hitting the tire rotation going forward? Great video, thanks for taking the time to show this.
Interesting video, thanks. What are your thoughts on the spokes? 16 or so spokes all with a leading and trailing edge as the wheel spins? Would deeper rims reduce this drag as the spokes are shorter? Where would a tri spoke wheel fit into all of this? Thanks.
Aerodynamically, without any doubt a flush tyre will be faster. The other bit of your question which is regarding rolling resistance is speed related. There is a tipping point which i estimate is about 26-30km/h. Above 30km/h the drag reduction from a flush tyre wheel combo outweighs the rolling resistance difference offered by a wider tyre/narrow wheel combo. This speed range is wide because there is a large variation in wheel widths and tyre widths
Makes sense to me, I go faster on my own where speeds are slower on my 24mm wheels and faster in a group on the 50mm. In a TT the faster you go the faster you go! By that I mean you hold onto the slip stream of vehicles longer so get that extra tow. Keep up the good work, engineering is natures way of saving you money ;-)
Not sure is this has been discussed in the comments below but I must admit that I haven't read the entire thread. Question is this. Do the spokes (whether bladed or not make a difference on the upper part of the wheel (ie) (in the nastiest part of the vortex). Do the spokes smooth that out a little? Just a thought... And BTW.. Nice job of the video!! I am not an aerospace engineer but I am smarter than the average bear.. So I enjoyed watching and made perfect logical since. Thanks again! Sergio
Hi Hambini, cheers for the no BS and putting the engineering facts out there. Cycling is full of marketing hype. On another note. If the wind is coming from a yaw angle behind you. (Pushing you along) will the wheels being deeper give you a sail effect, and if so, does the shape effect this. Or is it still just down to the chord length. Cheers
When travelling at Thirty Mph, if the bottom of the wheel is moving at Zero, the bike moving at Thirty Mph, wouldn't the top of the wheel be travelling in a forward motion at Sixty Mph. A thirty Mph headwind would contact the Top of the wheel travelling at Sixty already meaning the Top of the wheel would be going through the air at a relative Ninety Mph. Or am I missing something.
In some testing I made on another unrelated field but using a spinning disc, the air tended to be pulled around the circumference, which I don't think is modeled in any of these simulations.
Bit late to the party here but curious to hear your thoughts. With all this data you have meticulously researched . For an accomplished cyclist who is proficient in most cycling disciplines ...fast group rides....including all aspects of cycling. What do you think is the ideal rim depth. Reason I ask I run 35mm Bora Ultra and these are rarely found wanting during pacy group rides with accomplished cyclists at speeds over 40 km/h . Other guys have deeper rims and I am wondering if I am missing something. Many thanks
What's the best mental model for considering pressure drag assuming no ambient wind? Do I consider the top of the wheel to be moving through the airstream at 2x bike speed and the bottom to be stationary? Or do I consider a uniform airstream from top to bottom? Would a wheel sliding on ice at 30 km/h have the same pressure drag as one rolling on a flat surface at the same speed. If skin friction doesn't exist and the wheel is perfectly radially symmetric, I don't see how the airstream "knows" if the wheel is spinning or not.
Well, the video was uploaded a while ago, but I have to ask this anyway: According to what we see here, wouldn't a smaller wheel have a better aerodynamic? Plus, lighter and considerably stronger...
I like this video very much. Please do the same with forks and dropbars and frames and disk brakes and front mech's, Basically everything..If you could manage that before I buy my next bike, I would really appreciate it:)
I don't think I'd have enough content to do a video on that. However to give you an explanation: Drag is composed of two things, skin friction and pressure drag. Skin friction drag is just that, the drag caused by air flowing over a surface. Pressure drag is caused by the back draft that an object leaves behind it as it travels forwards. The dimples increase the skin friction drag and reduce the pressure drag. This is why golf balls are dimpled. The only issue on a bike wheel is this does not work because the relative speeds of the wheel around it's circumference vary.
Great video. So in your opinion, what's the best valued aero wheel for around USD 1k? I was looking at Chinese carbon (56mm rim depth at 1550 grams) for around that price but wasn't sure about quality. Alternatively I could look at more main stream brands. Thought?
I'd go for Yoeleo wheels. They aren't the cheapest "chinese" wheels but they are well regarded on multiple sites. However, if you want wheels that are essentially indestructible. I'd get Shimano C60's. They do weigh more and cost more but they have good bearings and an aluminium brake track. So in the rain they are miles better.
curious to hear #hambini's thoughts about really old Zipp wheels, I have some 440's from the late 90's, hope titanium hubs, dt bladed spokes and 22mm continental tubular tyres
Thanks for your explanation about aero wheels. Out of curiosity, have you done any simulation in connection with valve length? For example, say you have a 38mm deep wheel and try with 48,60 and 80mm valve length. There would be more drag but I suspect it's negligable because the valve sits in the same turbulence that is created from the trailing edge of the wheel.
I have a rim brake bike - Giant TCR and was thinking of getting some great all round depth say approx 40mm carbon wheels that are great in the wet & for steep Alpine descents & big climbs. Can you recommend the best top 3 wheels. I am concerned with heat build up on the steep descents & possible blow out of tyres/rims etc Cheers
Very informative. I have a basic understanding of what makes a wheel set "aero"; however, like others I wonder at the industry's move to toroidal shaped rims. Why if it isn't better? Also, there seems to be a lot made of "better handling" in crosswinds with current wider rims compared to tear drop shaped ones. Any thoughts on why the toroidal shape has been adopted by the industry? Also, does the wide toroidal shape offer better handling compared to the thin deep dish rim?
Thankyou for your questions. I don't work in the bike industry so some of this is my opinion more than anything else. 1. The true aerofoil shape is difficult to implement on a wheel because the holes where the spokes come through require a minimum thickness. It's easier to do this on a flat edge or slightly curved edge but not on an aerofoil trailing edge. A toroidal shape is therefore cheaper to make. Also the influence of the brake track and the tyre need to be included. 2. Aerodynamically, if the wheels have the same depth and you adopt a NACA profile or a toroidal profile, the difference between the two will be negligible. There is a particular angle call the stall angle where the aerofoil no longer generates lift. In the case of a wheel which has rotation this angle will not be constant and will be more like a pulsation effect. If I go away from the technical theory and look at practice, I have a set of Shimano C60 wheels and a set of Zipp 808's. The firecrests are allegedly immune to crosswinds but this is total rubbish, the C60's kill them every time - that is not to do with any profiles, it's purely based on the rim depth. Hope that helps Hambini
Mavics don't seem to do well in independent wind tunnel tests and reviews I notice, even though it has as you say the preferable NACA profile. Why do you think that might be?
That's a very good question.
The most important consideration are the tyres, mounting a tyre that protrudes wider than the edge of the mavic rim will cause significant turbulence at the back of the wheel. It will also cause turbulence where the clincher meets the tyre bead.
Speed tests are often conducted at 40km/h (50km/h in the latest bikeradar test), they do this because if they did this at a lower speed, there would be insufficient difference between the best wheel and the worst wheel to justify the horrendous price difference. The average "good" rider will probably be going at 35km/h (air density will affect this), most club riders are at 28km/h. At these "mortal" speeds a NACA aerofoil is quite effective because it's side to side vortex shedding is limited. A toroidal aerofoil is really only good in completely steady state conditions such as a wind tunnel. Out on the road which is where it matters, there are various bits of road furniture that upset the airflow coming onto the rider - technically this is called transient analysis.
Items such as lamp posts, a passing car, a hedge all affect the airflow. Another factor which nobody seems to have commented on is when a rider is going full gas, their bike tends to rock slightly from left to right, only by a few degrees but this upsets the oncoming airflow. An aerofoil which sheds fewer vortices during this process will be more aerodynamic.
Measuring the drag of a car in a wind tunnel is not a perfect science either but the car's velocity relative to the effects caused by street furniture is much greater. A car can be doing 120km/h and the street furniture is maybe pulsing the airflow by 2-3km/h hence the delta is massive. On a bike which is doing 35km/h and the furniture is providing 2-3km/h of pulsing the difference is much smaller.
So to summarize, the rim depth has the most effect, mounting a wide(r) tyre onto a narrow rim is a big no no and a wind tunnel is a poor measure of real world (pulsing) conditions because wind is not constant in speed or direction and the speed magazine's use for their analysis is a speed that most riders do not or cannot ride at.
I want to emphasise that at around 35km/h there would be little to separate any of the wheels provided the section depth was the same and the tyres were appropriately sized. The choice of aerofoil section is really only a marginal gain. My choice for the NACA aerofoil is the best choice of those marginal gains.
So does that mean an open mold rim would be fine? There is a big price difference between that and a zipp or enve
Aerodynamically yes, Enve and Zipp may or may not have better hubs but the aerodynamics of the wheel will show little difference below 35km/h
All good points there. Wheel makers are designing wheels that beat each other in the wind tunnel without consideration for real world conditions.
As you say, a side gust of 20km/h is inconsequential to a 1500kg car travelling at 100km/h, but very significant to a 70kg rider at 30km/h.
The 'rocking' bike is interesting too and will create extra turbulence. Do wheel makers test for this? No!
My V-profile Lightweights are BETTER in sidewinds than the more toroidal Corima 47 S+ (same depth). Neither is a problem in winds up to 70km/h or so.
Showing the difference between a 50mm and a 30mm wheel at 35km/h is hard enough, let alone the difference between 2 x 50mm wheels.
I have tried to model the rocking motion but it is extremely difficult. You basically generate a vortex to the right and then to the left but it's not consistent at all. This is coupled with micro steering corrections/movements. Your point is correct though, the wind tunnel versus real life is probably 70% accurate at best. Accounting for road furniture is difficult to do in a wind tunnel
Further, the guy at bikeradar described shimano and campag wheels as "canon fodder". I'm sure he has done his credibility wonders.
A rare video of Hambini without any swearing!
Considering the complexity of the subject, you presented it very well. There is a teacher in you!
Yeah :)
I agree wholeheartedly! Very well done!
I live in the US and I’m a member of a Cycling Club with 600 plus members, I’m glad I found this channel and can share all this knowledge with my Cycling buddies.
Congratulations and keep up your outstanding work.
How did the members take this?
I have bike raced since the early 80's. One of my friends who bike raced had a dad who in my opinion is the founder of modern fluid dynamics, and has authored a number of college texts on fluid dynamics. I remember one day years ago running into him and we talked about aero bicycle wheels, namely Zipps. They had just come out with the dimpling. When I asked him what he thought, he replied, "Figuring out drag coefficients is a black art, even with the best computer programs. I don't know where Zipp gets their numbers from but they are really creative." Thank you for this.
I had no idea about aero in wheels (or anything else) until I watched this video. Very informative and clear. Thanks so much.
This was an incredibly well done presentation Mr. Hambini. I think my students will very much enjoy watching this. So, I am going to show this to them tomorrow. (I am a physics instructor.) Thank you and kind regards.
holy nuts as a high school physic student, this stuff consolidates my learning!
Thank you for taking the time to make a clear explanation of a complex topic.
Awesome video. Just discovered your channel a few weeks ago and its content is gold.
Hey amazing video and I really enjoyed it! Just a disclaimer, I have no degree in engineering but as I understand some of the research out there is that the reason the wheel industry moved away from the NACA profile and into toroidal shapes is due to stability considerations. Apparently, the average yaw angles that cyclists encounter are between 5-10 degrees and this is where the toroidal shape performs better than NACA profiles (and some claim to have a "sailing" effect" akin to sailing into the wind). Also, I read somewhere that the discontinuity between the tire and the rim actually creates a beneficial vortex that causes airflow to reattach to the rim (if the tire is narrower than the wheel). And one last thing about toroidal shapes is that it creates a similar profile between the leading edge and the trailing edge of the wheel-tire system. For sure some of these is marketing as some of the research where than by the wheel manufacturers but I would like to hear your thoughts on this.
Thanks again for the great video!
Hello Franz, thankyou for the comments.
The key thing to realise from this analysis is the type of aerofoil profile has very little effect in comparison to the length of the aerofoil chord (ie the depth). The difference between a toroidal and a naca profile will not be much but the difference between a box that is 80mm and a Zipp 404 will be massive - the box will win most of the time. When you get to a yaw angle where large degrees of separation take place, it is more advantageous to have a shallower section wheel (especially in the front).
In the range of 5-10 degrees, a NACA 0024 profile should be better than a toroidal shape because it would be no where near separating. I estimate at the Reynolds number for the top of the wheel, the stall angle is about 12 degrees. Link here airfoiltools.com/airfoil/details?airfoil=naca0024-il
Great work, well done. I'm both, bike mech and a pilot, and really enjoyed the video. Thanks.
Thanks for the feedback
Wow!!! I am impressed. Video was very professional without any usage of foul language. Thanks Hambini
Yes, not that it bothered me, but during the PPP he used the expression "Oh dear", unsub!
Ok, I love your enthusiasm toward this aspect of airflow. Obviously you're extremely bright and into bicycling. While I realize the complexity of including the hub and spokes would be astronomical...they represent a MASSIVE amount of area which cause a significant amount of turbulence ~ I didn't consider that road "furniture" would play into this equation either...that was shocking;-) Thanks for the upload!!!
There is going to be another video which deals with "advanced topics" stay tuned for it!
This was outstanding! And it help me decide on a getting an upgrade. Thank you so much!
Have been a fan since day one. This is to date, in my opinion your best video. Thank you for the masterclass Hambini I learned a lot from this video.
Years back, I had 30mm aluminum clincher and 44mm carbon tubular wheels in the classic "V" shape, which was popular back then (10-15 years ago). I don't really recall having to fight the wind very much and having my bike blown all over the place with these wheels. With my HED Belgium C2's (24mm) and Enve 3.4's (35mm front and 45mm rear), the wind is brutal and my front wheel gets incredibly sketchy with the "U" shape.....even my 24mm HED's get blown around in gusty crosswinds - and the tire profile fits much better/cleaner with my current wheels than my older (and narrower) wheels, which had the bulb shape going on back then.
the amount of twitchyness is related to how much separation you get. it might be a particular combination that sets it off. largely the separation is related to the Reynolds number of the wheel and fluid
Hambini: that's true to a degree but twitchy behaviour also is affected by centre of pressure both for the front wheel in isolation (with the imbalance causing a steering torque) plus the bike as a whole. E.g. If the entire bike centre of pressure was at the back axel (approx) the bike would not be turned.
Jof Arnold Also, the degree of head rake has a dramatic effect on the tire staying straight...less angle twitchier steering...too much angle and it "flops" into turns. In the quest for overall weight reduction those are negleted b/c of target riders, target weights, cost and deadlines~ It's a black art that many manufacturers overlook.
Fascinating insight Mr Hambini. The big take away for me was how quickly, given the number of major assumptions/simplifications made, you got into deep water! Despite this, there are still calls for you to delve deeper into the murky world beyond haha! Good luck to your PC with that! For the mortal rider constrained inside the rules of competition, the potential gains seem too small to bother about, outside, stream liner bodywork would seem less bother than individual attention to each and every component, if of course you're prepared to compromise the look of your bicycle in that way.
Found you through TheDarkInstall's 'dark' live sessions.
Insanely detailed videos, subbed. Awesome stuff, keep it coming!
Thankyou for the positive comments!
Bringing the science back and pushing the marketing bs to the back of the queue! Thank you!
Thank you so much! I have been an aircraft enthusiast all my life,and what your saying makes so much sense. There's so much marketing hype out there! Really grateful to you for blowing these myths apart!!! Subbed,
You are welcome!
I love this portal AC. We can move around the house. Definitely need it for the hot summer.
Since today I love physics again and I listen carefully :D
Awesome great vid as usual! You have my appreciation and thanks for busting these myths and marketing hype!
I've always been searching for unbiased independent test results on aero drag and I find comfort in knowing that the cord length is what makes it count mostly.
Would be great if moving forward there could be an analysis on the differences in drag between a disc and non-disc aero wheel. Looking forward to more of your word Hambini. Thank God for people like you =)
Jin KX thanks for the feedback it is appreciated
❤
I do agree with you about the dimples as well!-) Also the sheer PRICE difference between the top tier equipment and the intermediate range matched to "aero advantage". The KWs needed to achieve higher velocity are reduced over hours...BUT at what cost and how many non-pro riders ACTUALLY have the leg power/stamina to seriously take advantage of it? I understand the appeal (like super cars etc) but....😱💥😁
My UA-cam watching is now within 3 years of realtime!
As a casual observer, are you certain you modeled the 2nd wheel with rotation? The leading edge, especially on the top slice looked too nice to me.
As a TV consumer, MythBusters did a show with golf ball-like dimples on the surface of a car, and found a noticeable difference in gas mileage. The dimples in the first wheel may be more than snake oil if you have interest in following up!
Hambini Sir, I summise that, to use a 23mm tyre then a rim of x would be sufficient for minimum drag and a 25 tyre will have a deeper rim depth (y) because the teardrop/NACA shape of the rim&tyre system needs to be bigger to incorporate the bulge or balloon of the tyre. A 28mm tyre would require, an even deeper rim. Would that be right enough?
Yes that's correct and a very good summary
Great! Impressive review far more thorough than any other I have come across.
Thanks for the modelling. This draws a question (from my non-engineering profession background)... why would more people not incorporate a carbon fender with good aerodynamic shape characteristics to cover the top half or more of the front wheel (assuming you weren't worried about crosswinds as much and its a calm day)? Wouldn't that reduce a lot of the front trailing edge turbulence effects by creating a continuous surface (more or less) to flow into the frame better? Obviously the speeds aren't comparable, but you see now in some Moto2 bikes where the front fenders cover a large portion of the fork/brake surface now, which probably aides in not having wind hit all those surfaces in weird ways. Has anyone ever experimented with front carbon fenders/shapes to test the affects on this with good results?
It's a pity you had no answers, excellent questions. I noticed that Moto2 front fender. I guess it may work
This is the first of your videos I’ve watched and it was awesome, thank you so much! I studied engineering (a long time ago) and its great to have such a clear explanation of the fundamentals of drag forces on wheels.
Have you calculated the CdA of the deep section 808 versus say a 404? I’ve just invested in an 808 for my TT’s and keen to understand the CdA improvement likely from the deeper section.
Also, I saw in the comments you didn’t see much benefit in a rear disc, at least in side wind conditions. I’ve been riding a disc this season compared to a rear 404 last season. It is definitely faster. I’ve calculated my CdA using the formulae for cycling resistances (inputting other known parameters such as power, weight, temperature, humidity, air pressure, and reasonable estimates for mechanical efficiency and rolling resistance). Based on my calcs across many efforts this season and last, my CdA has been lowered by 0.014. The disc is the only difference between this season and last, so I had attributed that to the disc. Interested in your thoughts as to whether you agree or not.
Again, awesome video, I’ll now be looking at your other content.
Hi Briony, It's a bit difficult to say for certain that it's the disc that is causing the difference. I have never found these generic formulas to be particularly accurate - they get you in the ball park but to get to 0.014 of precision would be pushing it.
However, parking that to one side. The disc will give you an Aero improvement, the impact would be a function of your body and bike in relation to it.
Thanks Hambini. When you say you don't find the generic formulas accurate, what generic formulas are you referring to? I mean the physic equations that allow you to calculate speed based on gravitational resistance, air resistance, rolling resistance, and mechanical resistance. While there is some variability day to day (probably because of differing wind conditions), you can minimise these errors by choosing a loop course. And by measuring CdA numerous times for each configuration, I think you can generate reasonable average CdA figures for each configuration over time. The chart below is an example of CdA with different clothing, keeping all other parameters constant. If I was proficient in CFD I'd try that also. But without CFD, the field test is the best I've got. Hey, does CFD allow you to quantify the CdA benefit of a deep dish Zipp 808 versus say a Zipp 404 or 202? I'd be very interested. Thanks.
The generic formula used for drag is CDA * 1/2 rho V^2 to give the drag force and then apply a sin or cos correction for angle. In my view this is not accurate and especially bad for a wheel, Wheels are rotating so the CD value changes as the wheel angle (yaw) changes. The formula you use above assumes a Fixed CDA. I guess at the end of the day, it's how much of an accurate answer you want. Hope that helps Hambini
OK, I understand, thanks. Yes, the calculated CdA for a field test is an average of various yaw angles, and is also for the entire and unique bike + rider system. By isolating one change in equipment and doing numeous field tests I’ve been able to determine statistically significant differences in the CdA for different configurations. But there is variability day to day and I see that the CFD method has some major advantages. Thanks again for the insights Hambini.
If you're chasing that little of a gain it might be better to understand what your true intention is behind it. Max speed? Overall avg speed? What chassis is best suited for that? The number 1 most important question of all...how do you train? Do you have the leg power/endurance to get the full advantage out of your mathematical endeavors? Math is a great start...but out on the road and in competition are the TRUE tests of what works. I love your attention to detail! I'm sure your one hell of a competitor!!! 😁😀👍
Good Morning Sir, As i understand it, since the wheel is at any one time 50% aerodynamic to the direction of travel [please correct me if im wrong] Would there be any benefit to attaching a LMP1 style wing or bodywork type idea to the fork, to mask/protect the trailing edge (6 o clock to 12 o clock) of the wheel to prevent those vortices creating drag? (I understand that the first thing F1 designers would do in regulation-free rules, is encase the wheels). Cheers Hambini!
You have truly enlightened my inner geek, I thank you young Sir.
I'm not that young! but I'll take the compliment
@@Hambini Age is relative and I'm a relatively old git.
Awesome software and great explanation, thanks! Mavics are certainly great in a perfect world but in reality you rarely hit the wind head on. Also the rough surface is used in many disciplines including aviation to reduce early separation and the results are spectacular.
Hambini, your videos keep restoring my faith in humanity. I have a question, though. I understand that spinning up wheels require power. But once wheels are in motion, their kinetic energy is constant and there should be no need to use power (i.e. add kinetic energy to them) -- unless they somehow interact with the air and drag it along. Is that what you had in mind -- that the air resistance of spinning wheels is different from the resistance of wheels at rest? Thank you for your answer.
Great video, very well explained! Considering the Zipp logo actually covers some of the dimples up, they must have almost no effect on the airflow.
Thanks for the feedback. I think it's more of a marketing gimmick to differentiate
Also, vortex isn’t drag but indicative of drag.
Thrust energy is charged into volume of air molecules by compression during the frontal section that is followed by energy return from air molecules in the expansion phase by the trailing section, provided that we have lamina flow during expansion.
Energy in vortex is harnessed from energy in the expansion flow. Suppress vortex generating can conserve the expansion energy back on forward thrust. The end effect is what we call low drag.
The dimples: Schuberth helmets used these to reduce airsound. The theory behind this is a controlled airstream without random stalls. The efficiency of the helmet got worse but the difference between silence and tornado feelings became less.
And this is why Zipp put the dimples on the rim. The efficiency got worse but the crossover performance became more repeatable.
BTW: NACA was the predecessor of the NASA and developed wing profiles for airplanes. You referred to the fully symmetric ones (0009 or 0012), but they made a lot more. However for bikes we need the fully symmetric ones only. Nobody came to the idea to use a RG 15 profile for crosswind situations (would be funny if the wind direction changes).
Right, finished it. Jeez, that was good mate. You seriously need to do more stuff like this. Phenomenal.
Question; the fork, hub and spokes will obviously affect air flow, so does that improve things for the top half of the wheel or add to the messy flow?
It makes it more messy, One of the things that you can do is when you disturb the airflow you can put a structural member immediately behind it to effectively have it operating in a drafted section.
For example My cervelo S5 has a "dropped" downtube, whoever designed the bike realised they were going to get turbulence generated by the front wheel which they could do little to avoid and put the downtube as close to the wheel because it has to go somewhere so lets put it in an area where there is little velocity.
This is similar to a bike rider drafting a car, they are sitting behind the car, the car has disturbed the air and they are riding along in an area with no appreciable wind. If they slip slightly back then the air has a chance to straighten itself and they get the full brunt again.
The type of spoke has less effect as the rim depth gets bigger because the spoke is shorter and the relative velocity reduces.
Great video! Curious to see more about the effects when the air hitting the wheel at different yaw angles.
That's a waste of time, real air outside of a windtunnel is never laminar, the various yaw angles are htting at the same time. Pick any angle and you will get numbers, all of which are meaningless.
Hi, great video. I’m wondering if there’s any new learnings since you made this video 3 years ago.
Specifically, I’m wondering if you still think the NACA profile is optimal in the real world where there is constant fluctuations in yaw and wind velocity.
Thanks!
I can definitely understand a leading edge / leading edge discussion, however, can a discussion of the trailing edge of the front edge really take place without the spoke, hub, and fork structures in place?
A fucking BMC time machine. I appreciate the Hambini reviews but would love to see the tests and data points collected on more down to earth frames and platforms.
Love your videos. This is the best one of them. Thank you for taking the time to make it. But I'm still confused
I get it that
1. Wheel depth is the overwhelming determinant of how aero the wheels are.
2. But steep yaw angles cause stall, and stall makes the wheel less stable and causes you to lose the aero benefits of the deep wheel.
2. A toroidal wheel is *supposed* to help solve that problem by deferring the stall to steeper angles, right.
Is that all correct? If so, well how often does a toroidal make a difference? How often does stall matter? Because in your example, you only used a head-on wind model.
For 3 months I've been trying to decide on a pair of aero wheels for my road bike and even after months of googling every aero wheel in existsnce, I'm still confused. Ignoring weight and price, what really matters for real world aerodynamics? Should one just go deep and ignore the shape? How often am I going to hit wind angles where that toroidal shape matters in the real world?
Perfect example I was looking at today. Deep-but-flat Fast Forward F6R (60mm/60mm depth) and shallower-but-toroidal Enve SES 4.5s (45/55 depth) . I've been through many such comparisons with many different wheels but after 3 months I have no idea what really matters between them in real-world winds.
go for the shallower but toroidal. Weight loss is great for uphill
I love your videos. I am confused though. You said that at the bottom of the wheel the drag is much less important because the wind is moving in the same direction as the wheel. But then when you started to talk about chord length, you demo'd it at the bottom of the wheel?
Great vid, clearly explained how the deeper section is more aerodynamic. I've had a (quick) look to see if anyone has asked this question already. Simply which wheel is going to be faster up a climb, a wheel that is heavier with a larger section or lighter wheel but has a shallower section.
If the lightweight wheel has a section of 33mm and weight of 1250g and the more aero wheel is 55 mm at 1400g, climbing at a speed of 14kmh. Climb has gradient of 7% (I don't know if that's relevant). Which wheel would be quicker on the climb? Thanks
Really good job at tackling a very complex subject. Thanks. The one question/surprise is why you didn't comment much on the stability effect of the toroidal shape at the average yaw angle. I bought into that argument in my latest wheelset purchase and I do think that i notice the benefit, versus my other bike which still has naca profile rims/tyres on it. Do you think that there is a stability benefit and, for you, might that outweigh pure aero cut through out on the open road?
Hi Hambini and thanks for sharing all this amazing knowledge.
Aren't you doing the air simulation on the wheel you design with the airflow perpendicular to the wheel instead of inline with iit as you do with the shimano one?
Thanks again for all the knowledge
Polite Hambini is a bit of a shock if you've only watched his later videos before this
I'd love to see how the airflow is different w/ the inclusion of a bottom tube (esp tight clearance back there), fork and esp the crown area. Most of the disrupted flow was on the top half of the wheel which is where those items would have a great effect. Not to mention spokes, but I'm sure those would be very difficult to model.
Going down the road the top of the wheel is double the speed and the bottom of the wheel is standing still. Freewheeling in a bicycle stand yes it would be going backwards. Not sure if this would change the airflow?? Thanks for your channel, I’m going to be purchasing some ntn bearings for my bicycle
Excellent rim review however I do have a couple conversation points. I think the spokes and hubs would have a significant impact on the airflow of the wheels, specifically the rim to spoke conversion of airflow. It would also be interesting to see how hub design will affect the aerodynamics of the wheels since there are a huge variety of hub designs out there. Another interesting point to think about is, how could you make the bottom section of the wheel as aero as possible and at the same time have no effect on the top section of the wheel. Some sort of hybrid wheel, in which the bottom section stays at the bottom and top stays at top while the wheel is in rotation. But this design is way too complex to engineer since there would need to be a collection of different spinning points on the wheel.
Would you be so kind to make a video explaining why and if It's true that larger tires have less resistance/friction to the ground at lower pressure? 30 < 28 < 25 < 23?
Is it just industry marketing bullocks ir there is real physic behind?
tip... use carbide spiral flush trim bits. And white glue for nut... makes it easier to remove if needed.
Pardon my incompetence. How does drag and turbulence affect a wheel, tyre, and spoke at bottom dead center when it is stationary irrespective of forward speed.
Also at top dead center: wheel, tyre spoke combo is moving at twice road speed. Will viscous drag equate to (m/s x 2)^2.
Always a pleasure. Is there any rule of thumb way to estimate what fraction of the total drag comes from the wheels? It seems like for us bigger riders who aren't particularly fast, wheel drag is not something to worry a lot about.
Thanks for takling the time to make this video!!
Zipp Binned the 1080 after the 808 tested faster. Can you speculate why the 808 might be faster Even though it has a shorter chord length?
Also, it’s my understanding that the advantage of the toroidal rim shape is for cross wind stability. I’ve only ridden toroidal rims myself so can’t directly compare to the older style V shaped rims which seem more closely related to the NACA rim shape you mention. But I’ve always read that the v shaped rims (and presumably NACA shaped rims) cause the cross winds to steer the front wheel resulting in stability issues while riding. Of course when it’s windy I can feel my whole bike gets blown sideways and more so with 898s fitted, but I never feel any steering effect through the bars and find the 808s totally predicatable energy on moderately windy days, with forecast winds up to say 12-15m/s
Zipp binned the 1080 after suckers were saturated and they needed to resell them new wheels.
Best video’s on you tube, thank you Sir
I wasn't buying the golf ball dimples on the rim from Zipp either. I wouldn't be surprised if tire manufacturers took up the torch and started putting dimples on the tires though.
At 11:00 you said that the aero benefits come from pure geometry as opposed to the toroidal shape. For the most part, would it be correct to say that Campagnolo Bullet 105s are at least every bit as aero as the Zip 808s provided that the tires used are a very similar width as the leading edge of the rims on both?
Edit: Campagnolo Bullet rims are very similar in shape to the rim shape you drew and showed at 15:10
Great stuff. I kinda regret turning down Aeronautical Engineering at Southampton for Mechanical at Kings! The lure of the big city.....
Also great to see Zipp's golf ball dimples rubbished; as you say they merely serve to increase skin friction. I strongly suspect the new 454 'shark's fin' design is more of the same. With one 50mm wheel resembling every other, you have to try and stand out in some way.
Interesting too to see the importance of tyre to wheel interface and hopefully a rubber stamp for my 'silicone your wheel gap' vid!!
What is your take on 'power to spin'? Although the majority of drag comes from the wheel's forward motion, one of the benefits of a deep rim wheel is shorter (and thus slower) spokes. Therefore, spoke profile, length and number are (IMO) all also important.
Yes, would you mind if I linked your silicone gap video in the description?
On your second point. The reason for golf balls having dimples is to make the boundary layer go turbulent and therefore keep flow attached longer. You don't get that benefit on a bike wheel because the rotational speed is too high and the net velocity vector is never zero (there is always a tangential component). The CFD also shows the same thing. With the advent of disc brakes, they can ditch the brake track and make it more teardrop shape. That should produce a
The spokes are a necessary part of the wheel and all of them have similar flow characteristics. Unfortunately the spokes will all have the same linear velocity for a given length and you gain more by having the deep rim than by having a long spoke. The shallow wheel advantage kicks in when the yaw angle becomes more of an issue.
And as for Aero engineering. Out of a bunch of 30 of us, 3 went on to become engineers (myself included). The other 27 went on to work for large financial institutions as traders!
No problem for my silicone vid, I'd be honoured!
As for the spokes, I've found that some carbon spokes (like Mavic R-SYS, large and circular) are quite poor aerodynamically, and will create significant rotational drag. Not all spokes are equal - some are more bladed than others. It seems to be an area that is often overlooked.
Funny for career path - when I graduated in 1991, there was no hope of an engineering job (recession) so I ended up in trading through desperation!
HAHA kinda figured you were a mechanical engineer watching your videos
Is it that obvious!??!
same with the guy from cyclespeed tours :D takes one to know one I guess...
Idk all the physics of the subject, but real world feel is this; 2 very similar bikes same make, one has Zipp 303 and other Yoleo 60 mm. Both fast, responsive and very close in weight. The real diff is how they handle wind. Zipp stay put and the Yoleos are a handful at times. Had a decent recently with a side wind.... let's just say I'm glad I was able to stay up right, had to clean my bike shorts lol....now Yoleo only on calmer days.
Cool vid :)
But your calcs on density (Aus v UK) are possibly a bit off (esp the 60w difference). If you assume:
AUS: temp 35C, Alt 20m, HPa 1020, Rel H 20% Air density is 1.147. Assume Crr 0.0052, 300w @ 30kmh = CdA 0.768 (which is super high, typical road racing cyclist ~0.330, typical triathlete ~0.285)
UK: temp 10C, Alt 20m, HPA 1020, Rel H 90%, Air density is 1.248. Assume CdA 0.768, Crr 0.0052 Speed = 8.12 m/s = 29.2 kmh instead 30kmh. Power to ride @ 30kmh = 323w or only 23w difference.
Maybe you could get a 60w difference due to air density, but it would be very extreme, and perhaps not really typical riding conditions.
Air density certainly has a big effect on the success or failure of hour TT records, where higher altitudes do allow slightly higher speed and make success more likely.
Hi Rob, Thanks for your question. I didn't calculate the air density, it has come from Jet Engine test conditions data from real places. For the UK, I took average density recorded over december at 1.29kg/m^3, for Australia I took the conditions of Cyprus in July 1.09kg/m^3.
Additionally, the air becomes much more viscous as it approaches zero celsius and therefore the drag coefficient increases. In your estimation you've assumed it's constant. It becomes more like trying to go through treacle than going through air. I don't know if you have an Aerospace background or not but the governing factor is the Reynolds number, in practice it would define the detachment points of airflow.
I think 10 degrees for a UK winter is charitable! Thanks Hambini
To what degree (if any) would the various profiles of spoke guide or clean up those vortices?
Nothing replaces simulation to actually identify the vortices which indicate high and low pressure zones.
As it was explained to me, the dimples on the Zipp wheels are designed to mimic the effect of dimples on golf balls which have a profound effect on drag by reducing the low pressure zone (drag) behind the ball as it travels through the air. As I understand it, the dimples allow a golf ball to travel approx. twice as far as one w/o.
I believe to be some merit in their design, but how much would actually require a lot of testing (ie. a lot of money and time) to debunk. The biggest problem that I see in Zipp's dimples is that air isn't able to travel over or under and they still expect a spherical shape of the dimples and the ball itself for air to detach and reattach especially on the top half of the wheel.
I suppose the reason companies like Look and Cervelo have adopted NACA airfoil designs is that they're proven, and wheel patterns would not be so different aside from the fact that frames are measured in a static environment with only yaw variables.
As always, amazing video!
This dimple business will be covered in the aerodynamic masterclass video. You are right in the sense you are trying to reduce the pressure drag by keeping the boundary layer attached more. It does not work on bike wheels because they have a fairly smooth tyre so the layer never develops.
I saw a vid where ex F1 Sauber old mate from Suicide sorry, Swiss Side, talked at length about sailing effect. I understand the principle where windsurfers can travel faster than the wind but is this a factor in cycling or simply BS? Thanks, really good video.
One word. WOW!
My great doubt is about what is better, aero profile carbon rims or low weight aluminium rims (I mean 100g less per wheel in aluminium wheel) to ride in velocities like 35km/h in a flat course..
i really like how you make it more simpler to understand. thanks a lot Mr.Hambini. i got a question tho.... talking about drag is there a certain speed at which it will start to take place if we are to consider deep section wheels?
Hi Hambini can you do this simulation again with spokes :) cheers
It's in progress.
Very interesting. Just wondering, if the top of the front wheel heading towards the front of the front wheel is creating drag, if you where to put a full coverage fender from the top of the wheel towards the front of the wheel, would that have less drag going over the fender vs hitting the tire rotation going forward? Great video, thanks for taking the time to show this.
At 25mph, approx 250W approx. how much faster is a rear disk (e.g. Zipp sub 9) compared to a deep section (e.g. Zipp 808), roughly?
Very thorough, great video. I will go and look up Mavic now, I always thought Zipp were marketing clowns
Thanks for the feedback!
Interesting video, thanks. What are your thoughts on the spokes? 16 or so spokes all with a leading and trailing edge as the wheel spins? Would deeper rims reduce this drag as the spokes are shorter? Where would a tri spoke wheel fit into all of this? Thanks.
Cool calculations. I also use SW, but unfortunately without calculations.
can you do similar calculations but in the wind at an angle of 20 degrees?
I have watched this video 6 or 7 time now it's just so interesting.
So a flush 22 tub on my old Zip 808 is better than a 25 that has "better" rolling resistance ?
Aerodynamically, without any doubt a flush tyre will be faster.
The other bit of your question which is regarding rolling resistance is speed related. There is a tipping point which i estimate is about 26-30km/h. Above 30km/h the drag reduction from a flush tyre wheel combo outweighs the rolling resistance difference offered by a wider tyre/narrow wheel combo.
This speed range is wide because there is a large variation in wheel widths and tyre widths
Makes sense to me, I go faster on my own where speeds are slower on my 24mm wheels and faster in a group on the 50mm. In a TT the faster you go the faster you go! By that I mean you hold onto the slip stream of vehicles longer so get that extra tow. Keep up the good work, engineering is natures way of saving you money ;-)
So switching from 25 to 22 mm clinchers on my Mavic Cosmic naca profile is better? Even though I will loose out on gains from the wider tire.
Would it be worth it to run 23 front for aero and 25 rear for comfort? Internal width 19, 24 outer
Yes there is no harm in doing that.
ENVE SES Rims have wider tire in front, skinny one in back.
Not sure is this has been discussed in the comments below but I must admit that I haven't read the entire thread.
Question is this. Do the spokes (whether bladed or not make a difference on the upper part of the wheel (ie) (in the nastiest part of the vortex). Do the spokes smooth that out a little? Just a thought... And BTW.. Nice job of the video!!
I am not an aerospace engineer but I am smarter than the average bear.. So I enjoyed watching and made perfect logical since.
Thanks again! Sergio
Nice video, but i think with a yaw angle the NACA profile will create more pressure drag then the toroidal profile! Have you tested?
Hi, below 12 degrees, the NACA is more areodynamic between 12 and 13 degrees the toroidal is more aerodynamic. Above 13, both have separated.
Hi Hambini, cheers for the no BS and putting the engineering facts out there. Cycling is full of marketing hype. On another note. If the wind is coming from a yaw angle behind you. (Pushing you along) will the wheels being deeper give you a sail effect, and if so, does the shape effect this. Or is it still just down to the chord length. Cheers
The hissing sound is absolutely annoying. Video is really good
When travelling at Thirty Mph, if the bottom of the wheel is moving at Zero, the bike moving at Thirty Mph, wouldn't the top of the wheel be travelling in a forward motion at Sixty Mph. A thirty Mph headwind would contact the Top of the wheel travelling at Sixty already meaning the Top of the wheel would be going through the air at a relative Ninety Mph.
Or am I missing something.
I had the same thought. The bottom of the wheel wouldn't be at minus 30, but zero and then it rotates upwards and forwards from there. Hambini?
Thank you for your honesty sir 👍🏁
What do you think of Hunt wheels?
What an education. Brilliant.
Hello Hambini. What do you think of the Fast Fwd wheels design with the “hour glass” profile, they call the DARC profile? 60mm depth specifically...
In some testing I made on another unrelated field but using a spinning disc, the air tended to be pulled around the circumference, which I don't think is modeled in any of these simulations.
Bit late to the party here but curious to hear your thoughts. With all this data you have meticulously researched . For an accomplished cyclist who is proficient in most cycling disciplines ...fast group rides....including all aspects of cycling. What do you think is the ideal rim depth. Reason I ask I run 35mm Bora Ultra and these are rarely found wanting during pacy group rides with accomplished cyclists at speeds over 40 km/h . Other guys have deeper rims and I am wondering if I am missing something. Many thanks
Campagnolo vento wheels. They look just like the wheel you create yourself.
What's the best mental model for considering pressure drag assuming no ambient wind? Do I consider the top of the wheel to be moving through the airstream at 2x bike speed and the bottom to be stationary? Or do I consider a uniform airstream from top to bottom? Would a wheel sliding on ice at 30 km/h have the same pressure drag as one rolling on a flat surface at the same speed. If skin friction doesn't exist and the wheel is perfectly radially symmetric, I don't see how the airstream "knows" if the wheel is spinning or not.
Well, the video was uploaded a while ago, but I have to ask this anyway: According to what we see here, wouldn't a smaller wheel have a better aerodynamic? Plus, lighter and considerably stronger...
Hi great channel 👍 what do you think of the 2021 Campagnolo Bora 45mm front and 60mm rear with Ceramic bearings
I like this video very much. Please do the same with forks and dropbars and frames and disk brakes and front mech's, Basically everything..If you could manage that before I buy my next bike, I would really appreciate it:)
Great way to get into cycling!
As an electrical engineer, thank you for being technical enough for me to keep up :)
Also, you made a big claim saying the dimpling is rubbish; I don't intend to argue, but can you make a video on that?
So when you did the computation, you model the wheel spin while modeling the incident air flow? If so, well done! Good attention to detail.
I don't think I'd have enough content to do a video on that. However to give you an explanation:
Drag is composed of two things, skin friction and pressure drag. Skin friction drag is just that, the drag caused by air flowing over a surface. Pressure drag is caused by the back draft that an object leaves behind it as it travels forwards. The dimples increase the skin friction drag and reduce the pressure drag. This is why golf balls are dimpled.
The only issue on a bike wheel is this does not work because the relative speeds of the wheel around it's circumference vary.
Great video. So in your opinion, what's the best valued aero wheel for around USD 1k? I was looking at Chinese carbon (56mm rim depth at 1550 grams) for around that price but wasn't sure about quality. Alternatively I could look at more main stream brands. Thought?
I'd go for Yoeleo wheels. They aren't the cheapest "chinese" wheels but they are well regarded on multiple sites.
However, if you want wheels that are essentially indestructible. I'd get Shimano C60's. They do weigh more and cost more but they have good bearings and an aluminium brake track. So in the rain they are miles better.
Many thanks for your engineering explanations very useful and helps counteract marketing bs, only one question why you wear a prisoner jumpsuit?
I like orange overalls!
curious to hear #hambini's thoughts about really old Zipp wheels, I have some 440's from the late 90's, hope titanium hubs, dt bladed spokes and 22mm continental tubular tyres
I've got the same, I was just about to try and sell them, might keep them now.
Thanks for your explanation about aero wheels. Out of curiosity, have you done any simulation in connection with valve length? For example, say you have a 38mm deep wheel and try with 48,60 and 80mm valve length. There would be more drag but I suspect it's negligable because the valve sits in the same turbulence that is created from the trailing edge of the wheel.
So, all things being equal, a 650b wheel should be more aero than a 700?
I have a rim brake bike - Giant TCR and was thinking of getting some great all round depth say approx 40mm carbon wheels that are great in the wet & for steep Alpine descents & big climbs. Can you recommend the best top 3 wheels. I am concerned with heat build up on the steep descents & possible blow out of tyres/rims etc Cheers
Very informative. I have a basic understanding of what makes a wheel set "aero"; however, like others I wonder at the industry's move to toroidal shaped rims. Why if it isn't better? Also, there seems to be a lot made of "better handling" in crosswinds with current wider rims compared to tear drop shaped ones. Any thoughts on why the toroidal shape has been adopted by the industry? Also, does the wide toroidal shape offer better handling compared to the thin deep dish rim?
Thankyou for your questions. I don't work in the bike industry so some of this is my opinion more than anything else.
1. The true aerofoil shape is difficult to implement on a wheel because the holes where the spokes come through require a minimum thickness. It's easier to do this on a flat edge or slightly curved edge but not on an aerofoil trailing edge. A toroidal shape is therefore cheaper to make. Also the influence of the brake track and the tyre need to be included.
2. Aerodynamically, if the wheels have the same depth and you adopt a NACA profile or a toroidal profile, the difference between the two will be negligible. There is a particular angle call the stall angle where the aerofoil no longer generates lift. In the case of a wheel which has rotation this angle will not be constant and will be more like a pulsation effect.
If I go away from the technical theory and look at practice, I have a set of Shimano C60 wheels and a set of Zipp 808's. The firecrests are allegedly immune to crosswinds but this is total rubbish, the C60's kill them every time - that is not to do with any profiles, it's purely based on the rim depth.
Hope that helps Hambini