Coast Downs and Wind Direction

[82zman]

Strangely my best coast downs (I define best by getting realistic/expected Cdas and Crrs) seem to come when I do the coast downs with a slight tailwind, yep a tail wind. The trick seems to be that the tail wind is light or less than the 8 mph coast down completion speed. This is the second time I have run across this pattern and quite honestly do not know what to make of it. My worst data comes from coast downs in cross winds I get very high friction numbers and Crrs in the .012 range which I know is garbage data.

Given a choice what do the experts say about coast downs in relative to wind direction?

So you know I am very meticulous about the CDs and calibration rides. Before the CDs and cal rides I let the unit stabilize at riding temp and then do a tilt cal and wind offset, then after the CDs and cal ride I recheck the offset and tilt to make sure they did not wander too much.

Any thoughts????

[racerfern]

I used to do CDs with little or no wind and maybe even a tailwind, until Coachboyd got me to try the bowl. He showed me that the more time spent rolling along just allowed more time for errors to take place. I found a spot that's a bowl, down 5%, up5%. In a matter of minutes I get to do CDs in both directions (I go very early before the cars affect the wind).

Then I do my cal ride on a road/bike path that is very lightly traveled and so far so good.

FYI, ideally the bowl would be less than my 5% each direction. But it is what it is. You would really like to go from 20 to 8 in 15-30 seconds.

[rruff]

Strangely my best coast downs (I define best by getting realistic/expected Cdas and Crrs) seem to come when I do the coast downs with a slight tailwind, yep a tail wind.

What is your slope like?

Having a slight tailwind should help the resolution on Crr. Crr becomes a bigger factor in the overall drag during the low-speed part of the CD, and this is especially true if you have a tailwind. So, it sounds like a good idea.

[82zman]

I am using a slopes from 1-3%, just enough to slow me down, I have already tried using the "bowl" but even on a "bowl" I got crappy results and the only common thread was the cross winds.

Also does anyone know why 4 miles was chosen as a cal ride? The reason why I ask is this. Why not apply the same logic that I have seen in many places on this forum, that is to keep the CDs as short as possible (use a "bowl") in time in order to reduce the effects of errors being introduced. For instance it is not uncommon that after returning after a ride my tilt calibration is a bit off. If this happened during a cal ride it would be interpreted incorrectly as rider tilt. Additionally a shorted cal ride would reduce the amount of wind gusts and traffic. Lastly It would be much easier to find an ideal one mile stretch than a two mile stretch. Just some thoughts.............

[rruff]

For determining your riding tilt, the longer the ride the better... but 4 mi should be long enough. For wind offset the only issue is that the wind can shift or change by the time you get back to where you started. It would be more accurate to do many out-back laps over a shorter course, but this isn't good for riding tilt since turning effects the accelerometer. At any rate you can check these things as many times as you like if you are concerned that your values might be off.

[stevevarnum]

Okay, the $64 question: How do you KNOW your coast downs, cal ride, Crr, Cda, etc. are good...what constitutes a good reading? I did what I thought was a good set of coast downs and cal ride using the bowl principle. In my case the bowl was only down at the start and then leveled off. I got the "good ride" "seal of approval" so what else is there? My power readings are low (about 185-195), but at 71+ years old and only a year riding, I would expect them to be low so I have no idea what is good or bad...

Thanks
Steve

[racerfern]

The key is consistency. If you have what appears to be a good set of data points that make a good profile, then get on and ride. As long as you do the basics of wind offset and tilt check before every ride, you'll know how much you're improving.

Once in a while (every 30 days at first) do the fitness test. Find a place to do the test that has relatively consistent conditions. You'll be comparing apples with apples. One more thing, increasing from 185w to 200w averages is a significant accomplishment. Don't look for huge jumps, you won't hit 250watts this year. Then again maybe you will!

Good luck.

[coachboyd]

Okay, the $64 question: How do you KNOW your coast downs, cal ride, Crr, Cda, etc. are good...what constitutes a good reading? I did what I thought was a good set of coast downs and cal ride using the bowl principle. In my case the bowl was only down at the start and then leveled off. I got the "good ride" "seal of approval" so what else is there? My power readings are low (about 185-195), but at 71+ years old and only a year riding, I would expect them to be low so I have no idea what is good or bad...

Thanks
Steve

The most important thing is that afterwards, the software KNOWS which coastdowns are the best ones. There are a lot of determining factors for looking at a good coastdowns. If you have a couple bad coastdowns the software will put very little weight on those ones. If you have some great coastdowns, the software will use those equations with more weight on the numbers from them. This is why it's important to do a bunch of coastdowns when doing a profile. The software will take care of the rest.

And make sure you do a tilt, wind offset, and find a nice road on a day with very little wind. It makes the process much easier and more reliable. If you are doing a coastdown and get passed by a car (from either direction), hit the center button to abort that coast.

[lorduintah]

Interesting effect I noticed - doing CDs with 5 sec. or 1 sec. data intervals - FRIC values were quite different by about a factor 3 higher with 5 sec recording inervals than with the 1 sec data.

I also calibrated tilt carefully and my cal ride gave me a 0.6 tilt correction - but also a very high Crr came out of this (0.012) - edited the tilt to zero and ended up with a Crr of 0.0049 - much more in line with previous results for me and the wheels I use - on the same stretch of paved road for many CDs and a portion used for the Cal Ride too. I have also noticed that the tail wind (light) leads to more consistent values than a headwind or cross wind.

Lately it has been way too windy in Dallas to really go back and try again (calibration, that is) - so I wait to see if I can get the Cal Ride to give me some values - the CDs gave very consistent result, so I think I have learned how to coast, at least.

Tom

[bex]

Hello everyone,
I did my coast downs and cal ride a couple of days ago
It was too windy I think.
All my cd's were good. I didn't know they could be bad. What causes a coast down to be bad?

Because of my fit test results I decided to re do my coast downs and cal ride today.
It was a much less windy day but 6 out of 10 coast downs were bad. Thats when I found out you can have a bad one.

When I downloaded into Ibike3 I was unable to save the profile because of a bad tilt. Why? I did a tilt and wind just prior and they wee both good.
These are my numbers, aero bars by the way.
Weight Bike and rider 181Lbs
Aero 0.218
wind scaling 1.265
CDa 0.173
Fric 10.345
riding tilt 0.2
Crr 0.0148
This is (of course) my original profile from a couple of days ago.
These are the results from today that I cannot save as a profile
Aero 0.226 Wind scaling 0.690 CdA 0.327
Fric 3.296 Rid tilt 1.9 Crr 0.02309
Weight 181 Wheel Circ 1927


Also, under the 4-mile ride section is Cm 1.020. What does this mean?

And on a lighter note, so Steve wants to do a Century on his 100th birthday, I want to be shot by a jealous husband on mine.

Cheers

Bex

[Morocco Mole]

cm relates to drivetrain efficiency, check out the following thread

http://www.ibikeforum.com/viewtopic.php ... 204&p=1086

[rruff]

Okay, the $64 question: How do you KNOW your coast downs, cal ride, Crr, Cda, etc. are good...what constitutes a good reading?

One simple way to check is to zero the offset, then put your computer on the "coast" screen and then coast in a variety of situations. You can expect the reading to fluctuate quite a bit at higher speeds, but it will average around zero if your calibration is good. If it tends to always be either positive or negative, then it would be a good idea to redo your CDs and cal ride.

[stevevarnum]

Bex: I am sorry for not reading this thread in a long time...hence the lateness in recognizing true genius!

I am seriously considering changing my goal!!! I had to laugh outloud at your post!!!

Thanks for a great laugh!!

Steve