Negative crr

[fmullin]

I ride a trike. I had done a profile calibrations in September but wanted to get updated values.
I reran the calibrartion this morning -- did the 2 mile ride and then a lot of coast downs (15+).
When I got home, I found that the iBike S/W determined my calculated crr was a negative value.
What can cause that?
One of the challenges with a trike is trying to do the tilt calibrarion. Unless the road surface is perfectly flat (which doesn't exist), then it is a bit of a crapshoot trying to get a consistent value. Is it possible that a poor tilt calibration could cause this result?
Reading your documentation, it appears that the crr is determined largely by the road surface (0.0055 for a smooth road, 0.008 for a rough road and 0.01 for dirt -- or something like that). A trike has three wheels. Do you think it is a reasonable estimate of crr to be 1-1/2 times the normal road bike value? So would 0.008 be a reasonable value for normal roads?

[lorduintah]

Do you have the iBike mounted securely? What kind of road surface and was it all downhill coast downs?

A poor tilt calibration shows up in two ways first. The actual calibration process - with the two 180 degree turns - you get bad tilt correction on the screen. The second will show up on cal ride as a bad cal ride after the calc cal is completed.

You should probably do a longer cal ride (4 or more total miles). The CDs could be done in both directions where you are running them.

What does your wind scaling value look like?

For now you could re-run the analysis with a fixed crr of around 0.065 and see if everything else makes sense - wind scaling, tilt correction... or use the crr you got back in September and check the other values (crr should not have changed much - unless the tire pressure is radically different, the tires or road are different - high winds? - wheels or brakes rubbing a little?

Tom

[racerfern]

Just a couple of thoughts here:

I helped out a Cat 1 rider that keeps a tire within a tire on his training bike. He cut the bead off the inner tire and laid it inside the outer tire solely for the purpose of avoiding flats. So far mission accomplished. We came up with a Crr of .0098 and then wondered if it could be true. So, we did a coastdown side by side as equally as possible. I rolled 15 seconds further than him before I got down to the 8mph cutoff. That's a huge amount of time and distance. The first part of the CD wasn't that pronounced, but below 12mph it was amazing how he just ground to a stop.

So try doing a CD exactly beside someone else and see if you slow down substantially quicker. It might give you some guidance. Or take out a road bike (if possible) and mark the start/finish spots then go back with the trike.

[travispape]

I ride a trike. I had done a profile calibrations in September but wanted to get updated values.
I reran the calibrartion this morning -- did the 2 mile ride and then a lot of coast downs (15+).
When I got home, I found that the iBike S/W determined my calculated crr was a negative value.
What can cause that?

What version of FW do you have installed? In general, a negative Crr means that the algorithm attributed too much of the resistance during the coast-down to aerodynamic resistance.

One of the challenges with a trike is trying to do the tilt calibrarion. Unless the road surface is perfectly flat (which doesn't exist), then it is a bit of a crapshoot trying to get a consistent value. Is it possible that a poor tilt calibration could cause this result?

Tilt cal can cause this, but so can a tailwind when doing the coast-downs. Make sure that if there is any wind to do your coast-downs into the wind. The best thing to do is to take advantage of any calm wind conditions to do your coast-downs.

Reading your documentation, it appears that the crr is determined largely by the road surface (0.0055 for a smooth road, 0.008 for a rough road and 0.01 for dirt -- or something like that). A trike has three wheels. Do you think it is a reasonable estimate of crr to be 1-1/2 times the normal road bike value? So would 0.008 be a reasonable value for normal roads?

I think Crr might scale less than 1.5 time because the rolling frictional force scales with mass on the wheels. Since you distributing approximately the same mass over more wheels, I think the net result should be in the same ballpark as for two-wheelers. If you are using 700x23c tires, I would think it would still be closer to the 0.0055 value.

[travispape]

For now you could re-run the analysis with a fixed crr of around 0.065 and see...

If you have a GenIII unit, I think a fixed Crr in the 0.0065 ballpark (I think Tom accidentally omitted a zero) would be a good way to go.

[fmullin]

Thanks to you all for the responses and suggestions. Here is additional informations
1. I am completely up-to-date on firmware (4.17).
2. The iBike is securely mounted and does not wobble.
3. I did a successful tilt calibration in my garage before beginning, drove to my calibration ride location, and then began the (3+ mile) out-and-back calibration ride.
4. When I went to start the coast-downs, I was told I needed to do an new tilt calibration (evidently 30 minutes had passed since I did the tilt calibration in my garage). While I was finally able to get an acceptable value, I was suspicious that the result was truly good (more on this below).
5. I then did my 15+ coast downs. The road surface was pretty smooth. It was gently U shaped so that I was able to rapidly get to 20+ MPH (on a slight downhill part) then coast to 8-mph (on a slight uphill part). I would then turn and do the same going the other way.

Now back to the tilt calibrataion issue. With a trike it is a challenge because you can't put your wheels in the same place when you do a 180 degree turn. You can return the wheels to the same point, however, when you do the second 180 degree turn.
I have marked a "calibration area" in my garage in which I have two rows of three points.
1 2
2 1
1 2
The floor is FLAT though not necessarily LEVEL. But I believe the 6 points are very close to being in one plane.
Starting in position 1, I initiate the tilt calibration. Then I rotate and use position 2. Finally I rotate again to position 1. But the iBike is consistently saying the tilt calibration fails. It seems that there may be a problem with my iBike in that it is not able to get a consistent tilt calculation. Just to make sure the battery wasn't the problem, I put a new battery in the iBike first.

[racerfern]

If the floor is flat even though it's not perfectly level then don't swap back and forth. Put the bike in your designated spot and don't move it for all three parts of the calibration. This might yield better results than fighting the trike out on bad pavement.

I do that on my floor which has a tilt of .2% and get very acceptable results. I'm not advocating this, I'm just saying for a quick and dirty tilt cal if you know the floor is just about level then do them all in one direction. Of course you still need to get to your starting spot within 30 minutes. Or find a flat spot closer to the start of your cal rides.