Checking your calibration

[rruff]

I don't see how the "Dynamic Pressure" with units of Force/Area can be equated to Power with units of Force*Velocity by multiplying by a constant with units of Area.

It is multiplied by CdA *and* velocity. The power to overcome aero drag is .5*Da*Va^2*CdA*Vb

[daidnik]

Okay, thank you mds & rruff. Now I am beginning to understand. I see how the units of the product of CdA*Velocity provide the proper units for the Aero Frictional Power term to come out with the proper units.

I wonder now how the unit is assigning values to the terms CdA & Ws in the calibration process.

All this dialog may seem too academic at times, but my real interest is more to understand the relative sensitivity of the total power to the frictional terms assigned in the calibration process to the Rolling friction & Aero friction terms. I gather from reading the available iBike literature that separating these two components in direct measurement in the calibration process is not really possible, or I should say that is what I inferred from what I read. I wish I had a better quantitative sense of the interplay of the frictional constants assigned to these two terms in the calc of total power. This is really what I would like to have a better grasp on.

If one knew the algebraic expression that equates to the total power calc, then the relative sensitivity to 'error' in these constants is easy to analyze, but we don't quite have that.

I just got the iBike Sport several days ago, and I went thru the standard cal process in the 'Cal Ride' & CD's where both the Aero Frictional Constant & Rolling Frictional Constant are assigned (that's my understanding).

I then read a bit more of the literature and got a sense of the dilemma of separating these two frictional components in direct measurement. I also got the sense that the Rolling Frictional term makes a much smaller contribution to the overall power, i.e. the work done by the rider in heating up the bearings, tires & road is a lot less than the work done in changing his/her potential energy & pushing the air out of the way.

However, getting the relative values of the constants in these different terms has some importance because it will affect the power calc 'tracking' over the range of speed. The Aero Frictional contribution is proportional to the cube of the relative velocity of the air mass we 'cut thru', while the Rolling Frictional contribution is directly proportional to the speed of the bike.

Now, maybe a person could just say: 'Look, the rolling frictional constant is small, AND it scales linearly with Vb; ... don't worry about it.'

This is, in a sense what I did in my second cal, where I looked at the range of Crr values and the conditions and thought 'okay, I ride a 25mm front tire & a 28mm rear, and I don't like the feel of 120 psi and instead ride about 105 psi, so I think I'll assign a Crr of 0.0060 (the display doesn't show the decimal point, only '60' but I think this is right). Then I did the cal where CdA is measured using the Est Crr that I assigned. I did 5 CD's and this is what I think happened.

The values of the constants that I find in the unit are as follows:

Cst Aero: 0.245; Cst Fric: 5.897; Crr: use Est @ '60'

Now, I should say that thru all this process, I have not noted any real notable changes in the displayed power in a variety of riding conditions, so maybe I shouldn't worry about it. I just can't help wanting to understand the inner workings better. I LOVE THIS UNIT!!

Now I'm just trying to figure when I upgrade to get the software, etc. added into the picture, but I'd better wait a bit on that.

[rruff]

All this dialog may seem too academic at times, but my real interest is more to understand the relative sensitivity of the total power to the frictional terms assigned in the calibration process to the Rolling friction & Aero friction terms. I gather from reading the available iBike literature that separating these two components in direct measurement in the calibration process is not really possible, or I should say that is what I inferred from what I read. I wish I had a better quantitative sense of the interplay of the frictional constants assigned to these two terms in the calc of total power. This is really what I would like to have a better grasp on.

I can't say for certain how they do it, but I've written a spreadsheet to do my own coastdowns (for greater flexibility and precision), and I simply do a curvefit of speed vs time, using the equations of motion with Crr and CdA as variables. CdA dominates at higher speeds, and Crr becomes a more significant percentage at low speeds.

Now that VC has allowed us to peg the Crr (a very good thing IMO), it would be nice if we could do coast downs at higher speeds to make CdA easier to determine.

[Velocomp]

Now that VC has allowed us to peg the Crr (a very good thing IMO), it would be nice if we could do coast downs at higher speeds to make CdA easier to determine.

Some people can't ride their bikes 20 mph!

We'll look in to it...

[lorduintah]

I have one set up for my bike - CORIMA Aero (front) and a tri-spoke on the rear. The issue here is that < 20 mph the tri-spoke will give high Crr values with the CDs from 18 to 8 mph. [These wheels are most effective > 20 mph when the aero characteristics come into play.] (After a lengthy discussion with some of the LBS guys). I have a pair of Mavic Carbones that yield a very reasonable Crr of 0.004 - all have Conti 4000S tires as well as a set of Ksyrium.

I can live with that - I get reproducible watts and that is what I want for training. Just letting you know that some conditions may be constrained by how the calibration parameters are obtained and under what conditions other values might be found. A large part of the population is not going to do all their riding > 25 mph 90% of the ride time - if they are, then they should be over in France right now.

So I am sure someone will find a way to complain (I AM NOT!) and in this case one might argue that there should be two Crr values depending on speed... indeed the entire gambit of coefficients are not constants, but are fluid in nature over the range of operational conditions an that is just too complex a problem to solve and not worth the cost in my opinion.

Tom

[rruff]

Some people can't ride their bikes 20 mph!

I know this would be a bit of extra progamming, but I was thinking of having a user selectable starting speed, and also the ability to select the ending speed, time, or distance for the run. For CdA determination (Crr pegged) coasting downhill should give the best results... no need to ever go slow.

EDIT: Also, the ability to peg the CdA would be helpful for doing runs specifically to determine Crr (ie low speed).

[rruff]

So I am sure someone will find a way to complain (I AM NOT!) and in this case one might argue that there should be two Crr values depending on speed... indeed the entire gambit of coefficients are not constants, but are fluid in nature over the range of operational conditions an that is just too complex a problem to solve and not worth the cost in my opinion.

Crr and CdA do not vary much at all in the range of speeds that we ride. AFAIK, everybody who uses the equations of motion in cycling (and writes calculators for cycling power and speed) considers them constants. Aero benefits do not "kick in" at >20mph. I'd check your wheel bearings if you are always getting a high Crr... or possibly your tilt.

[lorduintah]

rruf -

Worth investigating, true.

Tom

[mds]

Regarding how important an accurate Crr is for correct displayed power, it usually isn't that important.

At my weight, a 20% Crr change from 0.005 to 0.006 adds 3 watts when climbing at 8mph and adds 9 watts on the flats at 24mph.

Getting CdA and Ws correct is much more important for moderate and higher speeds.

On my iSport, I use the Est CdA screen to tweak Aero to compensate for CdA and Ws estimation errors that result from calibration rides and CD's.

My goal has been to get reasonable displayed power when pedaling at FTP downhill at 35mph. I tweak Aero via Est CdA until I see something reasonable, for a given body position and set of clothing.

[rruff]

My goal has been to get reasonable displayed power when pedaling at FTP downhill at 35mph. I tweak Aero via Est CdA until I see something reasonable, for a given body position and set of clothing.

Just some thoughts on that... it might be better to put your unit on the "coast" screen and actually coast and look for ~0, rather than guess how much power you are putting out.

[mds]

Yes, I use the coast screen also for this purpose. But when I'm putting out fair bit of power at 35mph and see the display obviously 100 watts too high or too low, its easy to conclude that Aero needs a tweak. Maybe when I'm using the coast screen my position is not quite the same or maybe not pedaling while coasting versus pedaling with power at 35mph, these differences sometimes have an impact on power. But it is helpful to double check the coast screen too. I think you need to have reasonable expectations - the unit by nature is going to be influenced by CdA changes - body position or whatever - at high speeds.