Hi fellow geeks (slow day at work)

I have heard about the "chung" method for using your bike on windless days as its own wind tunnel

Here is what little i know of the method:
-find a windless day
-ride a flat loop, trying to hold constant speed
-record average power and speed
-if you do this enough times, varying things like aero helmets, you can determine faster/slower helmets, etc etc

What i dont know
-i know there is some way to correct for minor variations in speed between runs to back to the CDA- how do you do this?
-what is the formula to run backwards from speed/watts to CDA? i have ron ruffs spreadsheet- but is that the correct formula?

Thanks

Noah

--------------------------
Proud member of the Guru Cartel, EH!
Just found it myself (great paper)

http://anonymous.coward.free.fr/...cda/indirect-cda.pdf

--------------------------
Proud member of the Guru Cartel, EH!
Hi fellow geeks (slow day at work)

I have heard about the "chung" method for using your bike on windless days as its own wind tunnel

Here is what little i know of the method:
-find a windless day
-ride a flat loop, trying to hold constant speed
-record average power and speed
-if you do this enough times, varying things like aero helmets, you can determine faster/slower helmets, etc etc

What i dont know
-i know there is some way to correct for minor variations in speed between runs to back to the CDA- how do you do this?
-what is the formula to run backwards from speed/watts to CDA? i have ron ruffs spreadsheet- but is that the correct formula?

Thanks

Noah

http://anonymous.coward.free.fr/...cda/indirect-cda.pdf

You're going to need to "code up" another spreadsheet to take the power data and calculate the "virtual elevation" as outlined in the .pdf.

BTW, I've found that an "out and back" course where you can make your turns on an uphill section (to avoid braking since braking "looks" like a steep climb to the method) works pretty well (possibly better than a loop). You still need to make multiple "laps".

Also, it doesn't need to be completely windless...low wind helps, but that's one of the nice things about the method is that it is less sensitive to this than the typical "regression method".

The major downside (as compared to a windtunnel) is that the evaluation is really only good for zero, or close to zero yaw conditions. Then again, it IS a lot less expensive ;-)

http://bikeblather.blogspot.com/
Has anyone coded up an excel sheet for this already? if so, anywhere i can download it?

Thanks muchly

Noah

--------------------------
Proud member of the Guru Cartel, EH!
have you considered using an iBike?
Hi.

I have uploaded my copy here: http://folk.ntnu.no/arntovf/chung.xls . I believe I once found it at biketechreview. There is data on it from a test I did last year.

I see in the OP that you mention a flat loop and constant speed as important for this method. This is true for 'normal' aero testing (curvefitting power to speed for a range of test runs at various speeds), but for the Chung method you fit the power and speed data to to the elevation changes on the loop you ride. Consequently, you should have some variation in speed (but no braking!) and elevation on the loop to ensure a robust data set.
Last edited by: walser: Mar 11, 08 13:39
Hi.

I have uploaded my copy here: http://folk.ntnu.no/arntovf/chung.xls . I believe I once found it at biketechreview. There is data on it from a test I did last year.

I see in the OP that you mention a flat loop and constant speed as important for this method. This is true for 'normal' aero testing (curvefitting power to speed for a range of test runs at various speeds), but for the Chung method you fit the power and speed data to to the elevation changes on the loop you ride. Consequently, you should have some variation in speed (but no braking!) and elevation on the loop to ensure a robust data set.
In your sheet you have the following constants:

crr 0.0053[/url] rho 1.22[/url] m 89[/url] cda 0.259[/url] g 9.81[/url] time interval 1.279[/url] V factor 3.6[/url]
I get all of them except the time interval - can you remind what that represents? I thought it might be the powertap's recording interval but that would be 1.26 seconds.
Thanks

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio
Hi.

I have uploaded my copy here: http://folk.ntnu.no/arntovf/chung.xls . I believe I once found it at biketechreview. There is data on it from a test I did last year.

I see in the OP that you mention a flat loop and constant speed as important for this method. This is true for 'normal' aero testing (curvefitting power to speed for a range of test runs at various speeds), but for the Chung method you fit the power and speed data to to the elevation changes on the loop you ride. Consequently, you should have some variation in speed (but no braking!) and elevation on the loop to ensure a robust data set.
In your sheet you have the following constants:

crr 0.0053 rho 1.22 m 89 cda 0.259 g 9.81 time interval 1.279 V factor 3.6
I get all of them except the time interval - can you remind what that represents? I thought it might be the powertap's recording interval but that would be 1.26 seconds.
Thanks
You're right, 'time interval' is the recording interval used on the power meter. I don't know/remember why I changed it from 1.26 in this case. I was probably checking something out.
Yeah - got it. Thanks. I also added in the distance covered as I think the chart is better shown vs distance than vs time and did a bit of a tidy up on the sheet so you could paste in any data pretty easily.

Here is an example of me doing a tempo workout at a local and popular training circuit in Sydney. It is remarkable in its accurate representation of the circuit used. I chose a tempo workout session of 90 minutes on the road training bike, probably riding up on the hoods. I checked against some old files where I used to have elevation data from a polar HRM many years ago and it said the elevation variance in the park was 44 metres. Winds were a light westerly on this day and much of the circuit is sheltered by trees, so it's a good candidate for this type of analysis. The circuit is mostly hotmix/asphalt is in reasonable condition.

What I did notice is that the derived elevation data seemed to be more sensitive to the Crr input than I thought would be the case.

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio
BTW - here's the power file showing power and speed for the data shown in the elevation profile:

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio
It is remarkable in its accurate representation of the circuit used.

Cool.

If the Ocean St. climb is a little short in elevation gain (compared to that altimeter) and there was a light westerly, I'm guessing the climb goes from west to east.

I've been thinking of ways to separate the estimates of Crr and CdA.
It is remarkable in its accurate representation of the circuit used.

Cool.

If the Ocean St. climb is a little short in elevation gain (compared to that altimeter) and there was a light westerly, I'm guessing the climb goes from west to east.

I've been thinking of ways to separate the estimates of Crr and CdA.

Thanks. I need to work out how to address slight winds from a constant direction.

The climb goes north however, so a slight crosswind, if any. Although in the park the wind can feel like it changes depending on where you are.

Here is a map of the route snipped from Google Maps and roughly shown in red by my less than perfect mouse tracing!

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio
It is remarkable in its accurate representation of the circuit used.

Cool.

If the Ocean St. climb is a little short in elevation gain (compared to that altimeter) and there was a light westerly, I'm guessing the climb goes from west to east.

I've been thinking of ways to separate the estimates of Crr and CdA.

Yeah - that would be good. It was having prior knowledge of the elevation gain that enabled me to play with the Crr and CdA numbers - otherwise the elevation gains were insufficient. This was about the max I could get.

Also - I should note that the polar used to show quite a range of elevation gains on any one day's ride or for different days, with many metres variance.

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio

Also - I should note that the polar used to show quite a range of elevation gains on any one day's ride or for different days, with many metres variance.
That doesn't surprise me. They also appear to be slightly speed sensitive.
OK - I couldn't help myself, I put a write up in my blog....

http://alex-cycle.blogspot.com/...-power-meter-77.html

Now off to pimp it :)

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio
And for fun, here's one more, this time from a 1hr Masters road race, just to see how it copes with variability of bunch riding, race tactics and changing environmental conditions (it warmed up from about 1C to 12C during the race).

I note that CdA drops significantly due to riding in a bunch compared to the solo training ride shown before.

You can see how the change in temperature affects the chart, up to when the temperature stabilises, then the laps appear to plateau. I also happen to know the elevation variance on this circuit is approx 20m. It is a car and motor cycle road racing circuit and the road surface is excellent.

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio
And for fun, here's one more, this time from a 1hr Masters road race, just to see how it copes with variability of bunch riding, race tactics and changing environmental conditions (it warmed up from about 1C to 12C during the race).

Yikes. 11C increase in temp over about half an hour?

So, how much did you "save" by riding in the bunch compared to taking a solo flyer?
And for fun, here's one more, this time from a 1hr Masters road race, just to see how it copes with variability of bunch riding, race tactics and changing environmental conditions (it warmed up from about 1C to 12C during the race).

Yikes. 11C increase in temp over about half an hour?

So, how much did you "save" by riding in the bunch compared to taking a solo flyer?
Well it's a mid winter early morning race in Western Sydney, so the day just post dawn is really nippy but as soon as the sun gets going, it warms up pretty quickly. It was 15C by the time I left after presentation and got to about 20C.

Not sure about the solo flier CdA. Intersting how my CdA on the day I was training in the Park was 0.384 and in this race I was 0.250. That in itself is a huge difference.

As for the solo flier - it was a bit embarrassing. Thinking I had it won by miles, I sat up to enjoy crossing the line solo, not knowing I was being caught by a chaser who pipped me by a wheel! :o

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio

Also - I should note that the polar used to show quite a range of elevation gains on any one day's ride or for different days, with many metres variance.
That doesn't surprise me. They also appear to be slightly speed sensitive.

The plot shown is from a run I did with my Polar PM on my bike in tandem with my PT. This was on my road bike in the drops and the Polar watch was mounted on the bars.

The black line is the "virtual elevation" that's calculated from the power data using Robert's method. The purple line is the elevation data reported by the Polar, corrected to the starting elevation from the power data. As you can see, there's a pretty large discrepancy at the "low points" of the elevation plots (where the speed is highest) with not much difference at the "peaks". The red line is the "speed corrected" values calculated by assuming that the Polar pressure sensor was at a "stagnation point" where the kinetic energy of the moving air was converted to an increase in presssure. Actually, it was ~90% of the kinetic energy that was the best fit.

http://bikeblather.blogspot.com/

The plot shown is from a run I did with my Polar PM on my bike in tandem with my PT.

That seems like a fair amount of error. I can understand that there's not a lot of real estate to work with on the Polar watch but it seems like the static port isn't in the best spot (at least, from the point of view of bike applications).

Hmmm. So you're saying there are conditions under which some PMs do a better job at measuring elevation change than some altimeters. Cool.
That seems like a fair amount of error. I can understand that there's not a lot of real estate to work with on the Polar watch but it seems like the static port isn't in the best spot (at least, from the point of view of bike applications).

I don't think it's a function of where the port is on the watch, but rather that the watch itself is mounted at a "stagnation point" of the rider+bike system (i.e. on the handlebars.)

Hmmm. So you're saying there are conditions under which some PMs do a better job at measuring elevation change than some altimeters. Cool.

Yep...pretty ironic, huh? Especially considering you asked this question oh so many years ago...

"How good could the HAC4 be?"

and

"How to turn a \$1500 Powermeter into a Cheap Altimeter"

With the conclusion that: "The next time someone asks me how good the HAC4 could be at measuring power, I'll say it's about as good as using a powermeter to estimate altitude gain."

I guess you might have to slightly reword that conclusion, huh? ;-)

(BTW, I realize in the analysis above you used the Polar elevation data in your calculations to simulate the HAC4. I wonder if the results change appreciably if they are corrected for speed?)

http://bikeblather.blogspot.com/