It also suggests that wind tunnel measurements taken at a fixed, non-zero yaw don’t tell you everything either, huh? Hmm…who’s been telling us this all along??
Hmmm…It’s funny, I looked at the chart and had the exact opposite thought: it seems that the positions that are fastest @ 7.5 also seem to be the fastest overall (at least for road–probably not for an indoor track). Of course, that could just be happenstance for this particular rider and tests.
It also suggests that wind tunnel measurements taken at a fixed, non-zero yaw don’t tell you everything either, huh? Hmm…who’s been telling us this all along??
Hmmm…It’s funny, I looked at the chart and had the exact opposite thought: it seems that the positions that are fastest @ 7.5 also seem to be the fastest overall (at least for road–probably not for an indoor track). Of course, that could just be happenstance for this particular rider and tests.
Except, if for example, you had to choose between positions #7 and #8 based on JUST the 7.5 deg data and pick the lower drag at that yaw, you’d actually end up with an overall higher drag.
It’s sounding more and more to me like the most useful field test method is the “Chung method” where you use a loop type course and have some wind present and your speed is varied. That way, at least you get an “average” CdA over a range of yaws instead of just a single zero yaw number.
The problem with that approach is that wind is rarely constant, which makes it difficult to compare results obtained on different days. For example, I’ve been doing this for almost 10 y using my power and speed data from TTs, and I clearly get a lower overall average CdA when the wind is stronger, presumably due to the fact that this results in a higher overall average yaw angle that allows me to “sail” better.
True, dat. But, for those of us on the "Cheapass Aero"TM development plan, short of going into the wind tunnel, it’s going to get me a better answer than just posting a pic on Slowtwitch and asking for suggestions, right? 
Totally agree with you. It’s a real treat though when one of these folks really does open up. It doesn’t happen often though.
Then what is the point of your post?
I’m quite certain that some of the smart people here have learned something of value from my various posts about this trip.
Wow, take a humility pill, please!
if for example, you had to choose between positions #7 and #8 based on JUST the 7.5 deg data and pick the lower drag at that yaw, you’d actually end up with an overall higher drag.
Or (probably, since I’m just eyeballing it at the moment) #4 vs. #6.
well lots of dicussion and we not even do not know how the test was done
(meaby even the wind came from behind,we just do not know)
evry one has his on idea
as you know all wind is rarely always come from same angle
and if the test was done wrong in the tunnel you get lot of noise
example of noise
go from 0 to 30 degree in 2.5 degree steps in let say + x axel
go back to
30 to 0 degree in same 2.5 steps in + x axel your bound to have different readings
this due the problem how the windtunnel functions
but if you do the same in - x axel so 30 to 0 you should have the same readings (meaby tiny off)
if not then the problem is in the windtunnel not every point in the test section has the same reynolds number
this makes the windtunnel in netherlands so extremly good and stable
so again do not put any conclusion to the plot i put on
since there are some things on the plot people cannot see
also to make a good judgement you must isolate the rider from the bike
and test the bike alone
and look at their interaction at the problems they generate
the plot is from a top world athlete is you can calc you see how low his drag is at 50 km/h and the watts needed are very low
but then again there still problems to solve to optimize things
It’s sounding more and more to me like the most useful field test method is the “Chung method” where you use a loop type course and have some wind present and your speed is varied. That way, at least you get an “average” CdA over a range of yaws instead of just a single zero yaw number.
The problem with that approach is that wind is rarely constant, which makes it difficult to compare results obtained on different days. For example, I’ve been doing this for almost 10 y using my power and speed data from TTs, and I clearly get a lower overall average CdA when the wind is stronger, presumably due to the fact that this results in a higher overall average yaw angle that allows me to “sail” better.
I’ve given up on trying to generate absolute drag numbers in field testing. That’s partially because it’s never calm where I live, and partially because I don’t have access to a portable weather station. What is still of value I think is going out and testing 2-3 positions (or peices of equipment) against each other, alternating positions/equipment regularly to tease out any weather changes. Keep testing new changes against baseline, fastest on the day becomes the new baseline. Lather, rinse, repeat.
It’s sounding more and more to me like the most useful field test method is the “Chung method” where you use a loop type course and have some wind present and your speed is varied. That way, at least you get an “average” CdA over a range of yaws instead of just a single zero yaw number.
The problem with that approach is that wind is rarely constant, which makes it difficult to compare results obtained on different days. For example, I’ve been doing this for almost 10 y using my power and speed data from TTs, and I clearly get a lower overall average CdA when the wind is stronger, presumably due to the fact that this results in a higher overall average yaw angle that allows me to “sail” better.
I’ve given up on trying to generate absolute drag numbers in field testing. That’s partially because it’s never calm where I live, and partially because I don’t have access to a portable weather station. What is still of value I think is going out and testing 2-3 positions (or peices of equipment) against each other, alternating positions/equipment regularly to tease out any weather changes. Keep testing new changes against baseline, fastest on the day becomes the new baseline. Lather, rinse, repeat.
You should look at using the “Chung Method” if you haven’t already:
http://anonymous.coward.free.fr/...cda/indirect-cda.pdf
No need for a portable weather station, a flat road, or calm conditions…just a power meter.
http://lswt.tamu.edu/info.htm#Wind_Tunnel_Performance_Characteristics
well link does not work
i have some plots of differant reynolds numbers of the sections of differant tunnels
so i tought check them or they still the same
but i couldn’t
or the meaby even not update it because it a complicated measurement of a tunnel
.
You should look at using the “Chung Method” if you haven’t already:
http://anonymous.coward.free.fr/...cda/indirect-cda.pdf
The problem I have with the “Chung method” is that all you really get are (Crr,Cda) pairs. That is, if you assume on value of rolling resistance, you get one CdA. Assume another, you get a completely different one. It can be useful if you know your Crr relatively well, but otherwise it’s no better (IMO) than the standard F=drag*v^2+roll type regression. I personally prefer to perform kinematic modeling whereby I can account for variable power (i.e., downloaded power data) and wind conditions.
Josh at Zipp (I wonder if anyone ever calls him “Joshatzipp” in real life?) would disagree with you as to why this occurs - and since Zipp has tested their wheels in various wind tunnels, you’re essentially claiming that this is an extremely common problem.
my toughts are they never been in dutch aerospace windtunnel
Y
The problem I have with the “Chung method” is that all you really get are (Crr,Cda) pairs.
the problem also is that its assumed that over the whole course the crr is the same
and i can tell you its not
then you even do not calc the corners and curves in a test
as with corners the tyres have a greater CRR
also with high wind the tyre have a bigger deformation (higher CRR) then you world expect
and this is not calc in the test
Remind me again: how many Dutchman have stood on the moon?
(Just teasing!)
well it must be that his ancestors came from holland
well after all there was no native american who stood on the moon
or…
just teasing
o yeah if my ancestors did not build new amsterdam and (yes my family build the house there )
what would have become of the great USA
again just teasing
Josh at Zipp (I wonder if anyone ever calls him “Joshatzipp” in real life?) would disagree with you as to why this occurs - and since Zipp has tested their wheels in various wind tunnels, you’re essentially claiming that this is an extremely common problem.
my toughts are they never been in dutch aerospace windtunnel
Remind me again: how many Dutchman have stood on the moon?
(Just teasing!)
No one’s ever really been to the moon. It’s a well known fact.
Since these are changes, any of them could be the fastest. Heck 2 and 8 could be bolt upright and thus have less change with wind direction.
a total guess is that 6 is the fastest. the CdA decreases until the bottle come out of the wind shadow, then the disc rear starts showing it’s stuff and overcomes even that???
You should look at using the “Chung Method” if you haven’t already:
http://anonymous.coward.free.fr/...cda/indirect-cda.pdf
The problem I have with the “Chung method” is that all you really get are (Crr,Cda) pairs. That is, if you assume on value of rolling resistance, you get one CdA. Assume another, you get a completely different one. It can be useful if you know your Crr relatively well, but otherwise it’s no better (IMO) than the standard F=drag*v^2+roll type regression. I personally prefer to perform kinematic modeling whereby I can account for variable power (i.e., downloaded power data) and wind conditions.
Believe me, it’s much better than the regression method, if only from the standpoint of how much easier it is to perform the testing.
I understand about the assumption of Crr. But, if one is using the same tires, the same pressures, on the same course, and at relatively the same temperatures, a reasonable assumption is that the Crr is constant enough.
Isn’t there a 1/2 mile car race track anywhere around where you live? A lot of times places like that will let you use the facilities either for free or cheaply, since you can’t really do any damage to the road surface. From what I have found, it’s significantly more powerful than “windless day” testing or garbage like that. You just record the wind speed/direction during your runs and use that as part of your data set when plotting out position or equipment differences.
Chris
The point I was making was exactly what you finished with. You get 0 deg. yaw data only.
Which, while highly accurate, is not indicative of any race I have been in.
I think generally we see that good positions as 0 yaw are good positions at other angles, but the same is not always true with equipment. Primary example being the Kestrel KM40, which performs stellarly at 10 deg and greater yaw, and only pretty good at 0.
(That is from my data, no a tunnel, so accuracy is not guarenteed)
Chris
. But, if one is using the same tires, the same pressures, on the same course, and at relatively the same temperatures, a reasonable assumption is that the Crr is constant enough.
so what you saying that you riding on differant surfaces the CRR is the same
well i can tell you its not
I think generally we see that good positions as 0 yaw are good positions at other angles,
no that is not true for positions
.