Aero Testing Discussion

Ptrckzq

Alright I've been interested in doing aerodynamic testing, ie, velodrome testing to measure CdA. However I'm not a fan of the price tag! My understanding is that they take speed, power, temperature, pressure, humidity and knowledge of the tyres Crr to calculate drag? Firstly does anyone know if this is even remotely correct?

I do study physics and did some reading into relevant areas of fluid dynamics and the ideal gas laws and I've made a working excel spreadsheet and based of some very rough trials from past rides I'm getting CdA numbers that make sense, so that's encouraging. But does anyone have experience working with this? it would be great to be able to go out and regularly do my own aero testing with the only requirements being a way to measure air pressure, humidity and temperature(and potentially wind speed if a closed velodrome or such isn't an option) any help developing this idea would be greatly appreciated

Thanks in advance!

cougie

What about Boardman and his wind tunnel ? I'm sure they'd be all over this kind of thing.

imposter2.0

Ptrckzq wrote:

My understanding is that they take speed, power, temperature, pressure, humidity and knowledge of the tyres Crr to calculate drag? Firstly does anyone know if this is even remotely correct?

I do study physics

You're the one that studies physics - you tell us?

58585

Google Search for golden cheetah aerolab Chung analysis. If you have a power meter you can do your own field testing.
Consensus seems to be that wind tunnel is not really the way to go - aero specialists will now have you on a velodrome and collect live data as you are actually riding. There is a lot of progress on new products soon which aim to give you “live” CdA figures. Within the next 5 years I expect one of these products will materialise, and it could move the goalposts so to speak.

joey54321

Yup, Chung analysis in Golden cheetah is the way to go. Ride laps of a shortish loop and use virtual elevation to see crr/cda.

A couple of points though:
- You will need a speed sensor, GPS is too inaccurate
- You can buy a portable weather station for pressure, temp, etc...
- You have to throw out any run in which a car passes you in either direction (or do enough runs that you can account for this noise)
- You need to have a loop in which you don't touch the brakes at all and can maintain your aero position 100% of the way round. This is normally the sticky point, I couldn't find a suitable venue near me.
- Keep tyre pressure the same for each session. Temperature affects Crr as well as Cda so you'll need to account for this. You need to do further roll down tests to extract cda from crr, but unless you do those on the same course as the cda testing its mostly guess work (though the important thing is to keep it all consistent, if you change tyres you'd need to re-do the cda-crr testing).

Ptrckzq

You're the one that studies physics - you tell us?[/quote]

I didn't say whether I was any good!

Ptrckzq

joey54321 wrote:

Yup, Chung analysis in Golden cheetah is the way to go. Ride laps of a shortish loop and use virtual elevation to see crr/cda.

A couple of points though:
- You will need a speed sensor, GPS is too inaccurate
- You can buy a portable weather station for pressure, temp, etc...
- You have to throw out any run in which a car passes you in either direction (or do enough runs that you can account for this noise)
- You need to have a loop in which you don't touch the brakes at all and can maintain your aero position 100% of the way round. This is normally the sticky point, I couldn't find a suitable venue near me.
- Keep tyre pressure the same for each session. Temperature affects Crr as well as Cda so you'll need to account for this. You need to do further roll down tests to extract cda from crr, but unless you do those on the same course as the cda testing its mostly guess work (though the important thing is to keep it all consistent, if you change tyres you'd need to re-do the cda-crr testing).

Thanks for the response.
So you're saying an indoor velodrome would be the sort of gold standard for this DIY aero testing? And If the passing of cars is going to be substantial issue without a large enough n, then I imagine you'd want very still wind conditions? Or say I were to perform the testing on an outdoor velodrome, would both the fact that you change direction and so regularly reduce the "noise" from wind? Also how do you go about using a rolldown to seperate CdA from Crr?

joey54321

I cant remember the exact process for separating the crr from the cda, try google (also have a look at the Timetrialforum, lots of people there that do this sort of stuff).

Indoor velodrome is king for this sort of testing, but I believe outdoor velodromes are very good as well. They are short enough that any consistent wind is sort of evened out and you can do a large number of laps to get a high 'n' (as you put it).

Still wind is preferable, but then again it goes back to accuracy and precision. There is an argument that since you race in 'regular' winds then you should be testing in regular winds and just do enough runs to get a good confidence, however i'd say you are getting in to a level of precision there which is beyond what is really required.

svetty

58585 wrote:

Consensus seems to be that wind tunnel is not really the way to go

Yeah, that Boardman bloke knows f**k all about the aero aspects of bike riding. No idea why he's wasted his time setting up a cycling wind-tunnel facility as its 'not the way to go' :roll: :roll:

58585

Svetty wrote:

Yeah, that Boardman bloke knows f**k all about the aero aspects of bike riding. No idea why he's wasted his time setting up a cycling wind-tunnel facility as its 'not the way to go' :roll: :roll:

Haha, nice one. I take it you are booked in for an appointment already? I have a feeling you could be just the type of customer he is after ;-)

Alex_Simmons/RST

Ptrckzq wrote:

Alright I've been interested in doing aerodynamic testing, ie, velodrome testing to measure CdA. However I'm not a fan of the price tag! My understanding is that they take speed, power, temperature, pressure, humidity and knowledge of the tyres Crr to calculate drag? Firstly does anyone know if this is even remotely correct?

It's remotely correct.

In essence aerodynamics testing is balancing the energy equation books. There are various methods of doing this and yes they all require measurement and/or control of all the key factors in play.

Ptrckzq wrote:

I do study physics and did some reading into relevant areas of fluid dynamics and the ideal gas laws and I've made a working excel spreadsheet and based of some very rough trials from past rides I'm getting CdA numbers that make sense, so that's encouraging. But does anyone have experience working with this?

I've been field testing cycling aerodynamics with various protocols and analysis techniques for more than a decade. I also provide the services in a professional capacity. There are many who do their own testing - it can be done with enough time, patience and good process.

The mathematics of the physics has been researched, published and validated. See this reference for example:
http://www.aerocoach.com.au/math-model/

Ptrckzq wrote:

it would be great to be able to go out and regularly do my own aero testing with the only requirements being a way to measure air pressure, humidity and temperature(and potentially wind speed if a closed velodrome or such isn't an option) any help developing this idea would be greatly appreciated

Thanks in advance!

Typically the biggest variable to manage is wind. Even the smallest non noticeable variation will impact results and so a global wind measurement, while it can help, is typically not sufficient to deal with moment by moment variations, and so a calm day is preferred.

Outdoor testing is certainly a feasible option, the reality of uncontrolled air movement though means the resolution of such tests reduces, so it then becomes a matter of whether the limits of measurement is sufficient to reliably assess whether any given change is reliably detectable. There are devices emerging that purport to provide moment by moment wind data at the rider, however their validity is yet to be assessed.

There are various places which discuss self testing methods. There is a thread on Slowtwitch for instance labelled The Platypus thread which is devoted to people discussing one method known as virtual elevation:
https://forum.slowtwitch.com/forum/?pos ... 5#p3536905

There is the more classic regression analysis methodology as described by Jim Martin/Andy Coggan, which I briefly wrote about in 2006 and created a testing protocol spreadsheet. With this method you can parse out the CdA and Crr.
https://wattmatters.blog/home/2006/10/a ... power.html

There are others, e.g. Adam Haile's short track regression testing method.

Keep in mind that the professionals who provide this service enable a few things that home brew testing often does not:
- reliable exclusive access to venues (even just one person on riding on the opposite side of the track will affect results)
- real time data so that impacts of changes to equipment and/or position are immediately apparent and guide the decisions on what changes to consider next
- more rapid assessment of any one set up, it's time efficient
- and probably the most important factor - the experience and knowledge of the person conducting the tests to guide the process, what testing options are most suitable for your individual circumstance. This sort of understanding and experience takes years to accumulate.

Alex_Simmons/RST

58585 wrote:

Google Search for golden cheetah aerolab Chung analysis. If you have a power meter you can do your own field testing.
Consensus seems to be that wind tunnel is not really the way to go - aero specialists will now have you on a velodrome and collect live data as you are actually riding. There is a lot of progress on new products soon which aim to give you “live” CdA figures. Within the next 5 years I expect one of these products will materialise, and it could move the goalposts so to speak.

Wind tunnels, field testing and computational fluid dynamics are all complementary methods of testing cycling aerodynamics and each has particular advantages which make them suitable for different purposes.

Viewing live CdA data for a rider on a track has been available for many years. I've been using it since 2013. Of course it's been available in wind tunnels for decades.

Alex_Simmons/RST

Ptrckzq wrote:

Also how do you go about using a rolldown to seperate CdA from Crr?

With regression analysis with suitable test run data where you control a key variable, usually via doing test runs at different velocities (but you could also vary mass for instance).

In the Power-Speed equation, rolling resistance is a linear factor, aerodynamics is cubic, so it is possible to pry them apart with sufficient data as described above.

But for most home brew tests, by using the same tyres, tyre temps and pressures and conducted on same track, you can reasonably assume a fixed Crr value and then ascribe changes to CdA. In this manner you can still assess what the relative difference in CdA between two set ups is, even if the absolute CdA values have some inherit bias. Naturally it does require careful protocol so that such assumptions are valid.