Reinventing the (aero disc-specific) wheel with Hunt - Pt I

g'man wrote:

@HuntBikeWheels

How do you measure the performance of your wheels in a cross wind?

I watched the live stream, I seem to recall they said they were measuring drag force (coefficient of drag, Cd) and the moment (coefficient of moment, Cm).

fukawitribe wrote:

@IanMunro

Interesting points, thank you, got me thinking - quick question, what makes the wind gradient in the wind tunnel* significantly different from that on flat ground ? The drag is measured on the wheel as a whole isn't it, including the variability due to gradient ? Obviously the wind tunnel is a more idealised environment but the actual drag measurements should still be indicative of what scale of resistance to expect and it's change with yaw angle.

 

* assuming you haven't got the test platform isn't raised into the flow off the floor.

The primary difference is that apparent wind component caused by the movement of a bike through air in the real world has no wind gradient, only the portion of the apparent wind caused by actual real wind will have wind gradient. So you can iether mount the wheel close to the ground in the tunnel and get a gradient in both components*, or mount it on a plate mid stream on a plate and get no gradient, but you can't mount into it any which way that will simulate the two seperate flow components correctly.
*Though I suspect the sidewall boundary layer will be markedly different to outside.

In other words, depending on setup, you can measure the equivalent of a stationary bike experience a wind hitting it at an angle or simulate bike moving though still air,  but not a combination of both, or at least not correctly.

In reality the differences make very little difference compared to the human sat on top, but nevertheless companies are making claims about the relative merits of wheels based on an incorrect premise of what wind yaw angle a wheel is experiencing in the real world.  So it's possible if you believe the wind-tunnel scenario to have a fat wheel X with a higher 0 yaw drag than a thin wheel Y, but with a lower 10 and 20 degree drag than the thin wheel Y, and thus a claimed lower aggreate drag. But in reality that aggregation is incorrect, because those yaw drag values are incorrect. 

It's a bit like using VO2max as a metric for athletic performance. It's used not because it's a accurate indicator of performance, but because it's easy to measure. You can be pretty confident that someone with a VO2max of 30 is never going to beat a VO2max of 80, but you can say with zero confidence that someone with a VO2max of 81 is faster than someone with one of 80. 

This is the problem with windtunnel yaw angles - nice and easy to measure, gives a rough indication,  doesn't give a believable answer if X and Y are closely matched.

There's probably a nice PhD to be done in trying to determine if tunnel yaw values can be scaled to produce a more accurate representation of the real world. Though going down the route you also open the horrble can of worms as to what sort of turbulent air flow typical occupies the last  60 cms down to the ground on various road types.

IanMunro wrote:

fukawitribe wrote:

@IanMunro

Interesting points, thank you, got me thinking - quick question, what makes the wind gradient in the wind tunnel* significantly different from that on flat ground ? The drag is measured on the wheel as a whole isn't it, including the variability due to gradient ? Obviously the wind tunnel is a more idealised environment but the actual drag measurements should still be indicative of what scale of resistance to expect and it's change with yaw angle.

 

* assuming you haven't got the test platform isn't raised into the flow off the floor.

The primary difference is that apparent wind component caused by the movement of a bike through air in the real world has no wind gradient, only the portion of the apparent wind caused by actual real wind will have wind gradient. So you can iether mount the wheel close to the ground in the tunnel and get a gradient in both components*, or mount it on a plate mid stream on a plate and get no gradient, but you can't mount into it any which way that will simulate the two seperate flow components correctly.
*Though I suspect the sidewall boundary layer will be markedly different to outside.

In other words, depending on setup, you can measure the equivalent of a stationary bike experience a wind hitting it at an angle or simulate bike moving though still air,  but not a combination of both, or at least not correctly.

In reality the differences make very little difference compared to the human sat on top, but nevertheless companies are making claims about the relative merits of wheels based on an incorrect premise of what wind yaw angle a wheel is experiencing in the real world.  So it's possible if you believe the wind-tunnel scenario to have a fat wheel X with a higher 0 yaw drag than a thin wheel Y, but with a lower 10 and 20 degree drag than the thin wheel Y, and thus a claimed lower aggreate drag. But in reality that aggregation is incorrect, because those yaw drag values are incorrect. 

It's a bit like using VO2max as a metric for athletic performance. It's used not because it's a accurate indicator of performance, but because it's easy to measure. You can be pretty confident that someone with a VO2max of 30 is never going to beat a VO2max of 80, but you can say with zero confidence that someone with a VO2max of 81 is faster than someone with one of 80. 

This is the problem with windtunnel yaw angles - nice and easy to measure, gives a rough indication,  doesn't give a believable answer if X and Y are closely matched.

There's probably a nice PhD to be done in trying to determine if tunnel yaw values can be scaled to produce a more accurate representation of the real world. Though going down the route you also open the horrble can of worms as to what sort of turbulent air flow typical occupies the last  60 cms down to the ground on various road types.

I've had problems measuring drag in a wind tunnel due to the velocity gradient being different in the wind tunnel to real life.

It seems to be messing up the boundary layer because the velocity gradient is opposite (I.e. moving air over stationary object rather than air with no velocity being accelerated around an object). EDIT: This isn't true as I realised the refernce frame is simply reversed (this does not affect boundary layers.

The really annoying thing about this is I'm most interested in the boundary layer because I'm trying to control (delay) flow separation. I can do this in CFD but can't in the wind tunnel due to the instability at transitional Reynolds numbers.

IanMunro wrote:

The problem with  yaw values in windtunnels, or outside for that matter is it's largely B.S

Is it BS or is it a method of evaluating the increasingly smaller variations in wheel aerodynamics? The whole point is comparing wheel x with wheel y as once you've bought the wheel it doesn't matter any more, you just ride the bike.

Everyone knows that a wind tunnel is not the real world. What we like is reliable comparisons so wind tunnel performance is probably as good as it gets.

I'm not sure that the difference in wind speed and turbulence between the ground and your headset is really all that important. In the end it's your body that causes more than 80% of the drag so the wind that hits YOU - or more specifically your aerodynamics - is what slows you down most.

The difference between wheel x and wheel y is pretty small in the scheme of things. Flo Wheels' figures for time saved with their 20-spoke 30mm wheel compares well with the baseline Mavic Open Pro with 32 round spokes but deeper rims do not have a proportional increase in time saved: http://www.flocycling.com/wheels_front_flo_30.php

IanMunro wrote:

The problem with  yaw values in windtunnels, or outside for that matter is it's largely B.S

If you are moving through still air at 20mph the apparent air speed at head height is 20mph, and the apparant air speed at 50mm off the tarmac is also 20mph.
However. If you are stationary and there happens to be a 20mph wind blowing at head height, there will not be a 20mph wind blowing at 50mm off the tarmac, due to a combination of wind gradient and outside wind turbulance. 
So what does this mean? 
People have measured apparent wind yaw angles with onboard bike instruments placed at height somewhere between that of the top of the wheel and handlebar height. They've then made the assumption that this is then a yaw angle that stays consistent all the way down to ground. Even if you make the very optimistic assumption that you've got non turbulent wind, then at ground level you will always have a yaw angle of zero, and then as you move vertically upwards the yaw angle will become closer to that at the measument height. Of course in reality you've got hedges, grass verges and all sorts of crap next to the road that means the apparent wind a wheel really experiences bears no reality to that of simple mathemtical models.

The upshot being comparing wheels at different yaw angles in a wintunnel, is lovely in that it gives you data sets that allow you to say that wheel x performs better than wheel y at a given yaw angle, but as wheels never experence the same yaw angle across their profile when actually used on a bike outside, it doesn't appear to give any true indication of performance.  It's great for adverts though

 

Interesting - and here's another way to look at it. The tyre at the road level is stationary in a linear horizontal direction, while the top of the wheel is going something like twice as fast as the bike.

If you can't get your head around that - think of this, if the tyre isn't linearly stationary at the bottom then you are either in a skid or wheelspin scenary.

I point this out because it's another difference between the wind tunnel and real world. In the wind tunnel, at least in the case of the one Hunt used, the bottom of the wheel isn't stationary linearly along the horizontal axis in relation to the floor, although by spinng the wheels up to the same speed as the wind is being blown through, the wheels are essentially passing through the airflow in a similar way to real world scenario (the bottom of the wheel will be travelling backwords at the same speed as the wind, while the top of the wheel will have effective airflow of double the wind speed in the tunnel.

The closed I suspect you could get to the real world scenario in a wind tunnel would be to have a true rolling road (giant treadmill), although the holding of the bike would become trickier.

matthewn5 wrote:

I've been impressed with my Hunt 4Season Aeros, solid and bearings feeling perfect half way through winter with daily riding.

But they don't weight the claimed 1579g, they're 1690g on my kitchen scales, without skewers. Does the rim tape weigh 105g per wheel? Be careful of claims about wheels

 

Where do you get 105g per wheel from? Even with my limited intellect I can calculate that's only 111g across both wheels!

Allowing for two layers of tape that's gotta use up a chunk of the 55.g per wheel

Even without tape, and allowing for manufacturing tolerances (of hubs, rims, spokes, bearings), extra grease, accuracy of your scales etc. that's still only about 7% difference

mikeymustard wrote:

matthewn5 wrote:

I've been impressed with my Hunt 4Season Aeros, solid and bearings feeling perfect half way through winter with daily riding.

But they don't weight the claimed 1579g, they're 1690g on my kitchen scales, without skewers. Does the rim tape weigh 105g per wheel? Be careful of claims about wheels

 

Where do you get 105g per wheel from? Even with my limited intellect I can calculate that's only 111g across both wheels!

Allowing for two layers of tape that's gotta use up a chunk of the 55.g per wheel

Even without tape, and allowing for manufacturing tolerances (of hubs, rims, spokes, bearings), extra grease, accuracy of your scales etc. that's still only about 7% difference

 

Touché! My bad maths. Thank you for the correction. 55g heavier per wheel it is.

 But isn't it surprising that the pair of wheels one gets to take home is always a bit heavier than the quoted value? Where are all the wheels that were lighter than the quoted figures?

HuntBikeWheels wrote:

CXR94Di2 thank you ever so much for your comment and feedback. We used 2 layers of the bronze tape up until Spring 2017 as it was pretty much an industry standard for resisting higher pressures, however as you correctlly point out it is not as good at sealing as other types of tape used on lower pressure MTB wheels. We now use a proprietory 2 layer 2 tape system, with a highly felxible well sealing cream tape under the bronze tape. Please note that the bronze tape has a very high tensile strength so does not creep over time with high pressures, that is why we still use it for the top layer. as an existing HUNT wheels customer we would be delighted to send you a full set of these tapes free of charge please just email us on thechase [at] huntbikewheels.com and we will send it out straight away. The single layer bronze tape was only used for our wind tunnel testing in the images in this article as we were obviously not riding these wheels and fitting tyres to the rapid protoypes was tighter than real rims as the inner surface finish was not smoothed like the outer surface.

paulrattew, usedtobefaster you have posted very well made points. We are definately considering how the yaw angles alter at different riding speeds, we are also preparing some research calculations on rolling resistance vs aero effect at different speed ranges.

The 45-48kmhr traditionally used for wind tunnel testing is, as you can probably work out, the kind of speed a TT pro would ride at, as you correctly state slower speeds around 22-25mph on the flat are much more realistic for a drop bar road bike and for real riders more like us, obviously these speeds alter over distance and terrain. You're abolsutely right that the yaw angle would significantly widen at these lower speeds and Pete and Luisa here are currently doing a lot of calculations (they used the word integration which scares me from my a-level physics!) to work out the time likely to be spent at each wind yaw angle. They are using the wind data other maufacturers have used in their wind averaged drag models to give us the best picture of which yaw angles drag is most important for realistic riding speeds. Also, we already have plans and capabilities to record wind yaw angles on bike, but of course we want to use other companies data and WAD models to show we are not just creating wind data that suits our product. We'll release a lot more data and information soon.

variflex, thank you ever so much for your feedback and I feel we owe you a better level of service here. We are all riders and never want a rider to feel the way you do about HUNT wheels, of course it is in our best interests and the riders we serve that you have a set of HUNT wheels you are happy with. We have a policy of never leaving riders unhappy if at all possible, so we have clearly let you down in this case. We will be in contact shortly to find the best path possible for you. I would like to assure you that the current model of Aero Light Disc wheels, whilst being designed as a fast lightweight performance wheelset with superlight hubs, has larger bearings and more extensive bearing shielding. I suspect the hubs you are using are an earlier model and no longer used on any of our wheels, they were specced as the offered the ultimate in fair weather performance for this fast/light wheelset. For our 4Season range we put a special emphasis on bearing longevity, and we have applied similar thoughts to the full range now and even to the extreme of speccing our own waterproof grease (which costs $1500 for 10kg) and have special operating procedures for grease application. We go to these lengths as British all weather riders like yourself we want bearings to last as long as possible. As I say, we will be in contact shortly and I'm very hopeful we can find a resolution that will work well for you.

 

Thank you all again for your comments and replies and I did just say to Mat at road.cc "these are very well informed comments...I have to say the readers have answered some of the comments so well and have covered quite a few of the points we would make in replies, pretty impressive. There is definitely another level of reader on road.cc compared to any other site I've been on."

Your help and advice really does go into helping us make better wheels for all of us riders so thank you again ever so much,

Tom Marchment HUNT | TheRiderFirm

 

 

There is no necessity to send me some rim tape, but thankyou for the gesture.

usedtobefaster wrote:

Interesting reading.  So the drag at certain yaw angle is different depending on the speed and from what I've observed the marketing of aero bikes wheels etc mostly refers to riding at 40km/h e.g. a 10sec saving at 40km/h,  so we can assume that products are designed to be optimal at and above this speed.   Sounds like there's a space in the market for products optimised for lower speeds where I suspect a lot of riders spend a majority of time riding at, say 30km/h, and so could benefit from reduced drag, particularly looking at the benefit in longer events and endurance riding - after all how many riders can ride a 100 mile sportive at 25mph that's got to be a small market surely?

 

Part of the reason that testing tends to use 40 or 50pkh is because the higher speeds make it easier to measure differences reliably. Lower speeds to not really change the way in which the aerodynamics work, the impact though would be smaller. Generally spaeking if the results show that one wheel is more aero than another at 40kph across the range of yaw angles tested, then that same wheel will be more aerodynamic at 30kph (the impact of this though will be proportionally smaller).

One of the big differences at lower speeds is that relative yaw angles are higher, which is why it's important to understand performance across a wide range of yaw angles. 

variflex wrote:

Love my Aerolight wheels apart from the bearings.. suposedly upgraded but have lasted less than 600kms on the rear wheel. Main hub and free hub knackered. Hunt wherent too interested and suggested buying more of their bearings. No thanks, have upgraded to Enduro for the same cost.

None of my numerous other wheelsets have ever needed bearing changes before several years use and some still going strong after 4 years.

Good luck with the wind tunnel work and development... but pls get the bearings right first!

Hi Variflex,

I am sorry to hear you're not satisfied with your bearings. If you could email us at thechase [at] huntbikewheels.com and we will do everything we can to make you happy!

I look forward to hearing from you,

Jack @ HUNT Bike Wheels.

CXR94Di2 wrote:

Dear Hunt Wheels, please stop using that transparent rim tape, it is terrible. I kept losing pressure when I bought wheels with it installed. Please use Stans tape or something equally good. Since renewing with Stans tape, the wheels have kept pressure. I must say I do like your wheels

God really? I avoid Stans, I find it too thick and stiff to work into the 'tyre well'...  I find the 3M made high tensile tape (as above) superb. Been using just one layer of it my front road wheel now for well over 12months, with no issues.  I suspect that your particular bad experience was down to it being fitted badly, either A. it wasn't worked into the 'well' enough or B. the rim wasn't cleaned enough before fitting. 

@Hunt - curious why you've changed to two types of tape per wheel? 

Mat Brett wrote:

paulrattew wrote:

If I'm running tubeless tyres then I don't want wheels with internal nipples, even if they are more aerodynamic. Making any adjustments to the wheels to keep them in true is a complete faff if the nipples are internal and you're running tubeless - so internal nipples fail the 'liveability' test for me. 

The Hunt wheels will have external nipples for those reasons. Ernst – who runs the wind tunnel – reckons the aero advantage of internal nipples is minimal.

RobD wrote:

good to see the bike the right way round in the tunnel (except one of the first pictures)

It's the right way around in all the pictures. Think you're mistaking the air intake for the fan.

You're totally right, shows I should pay more attention early in the morning, one large blue shape on the wall isn't the same as another similar looking large blue shape on the wall!