Sunday, November 8, 2009

Computrainer - Track simulation

It was a Computrainer weekend. The Soloist is still at the shop with the busted derailleur. And I didn't feel like sorting out some training wheels for the TT bike just to go out in the cold wet rain...

Yesterday I did an hour plus L2/L3 ride. At least according to the CT numbers. But really it felt a lot harder. Only 73TSS points for an hour and fifteen minutes. But I could barely walk after..

Today for fun I built a Track simulation 3d course. 200m, flat about the right dimensions....

This was fairly interesting. I did my normal "paced" type ride. I.e. set the CT pacer going with its constant wattage and then just try and stay in its draft. Again this felt harder and for what should have been equivalent to the track, a lot slower. Normally an easy pace spinning around would be about 35-38kph. The CT with about 200w was only getting me about 29-30kph.

I then tried to do the equivalent of a Take-a-lap drill. Normally at the track the group goes around at about 33-35kph and when the coach whistles the front person jumps off the front and rides hard until he catches up to the back of the group. Usually, depending on how fast you go, about 3-4 laps, 50-70 seconds.

When I tried this on the CT I basically failed miserably. Very hard to moderate the power to something I could hold and then just wasn't going fast enough. Even though the starting speed was only 28 I only was able to push and hold about 38 and just never made it all the way around.

So then I changed the drag to 50% AND lowered the pacer wattage to about 125 Watts... So it was only going about 25kph. Even then I was only JUST able to get around and then die...

Not quite sure what this says about the CT.. We know that the power numbers "seem" low. But here what we see is the speed for both power AND Perceived Effort being to low. I have never failed to catch the group in a take-a-lap drill on the track. Here I failed miserably with lower emulated speed and watts and higher perceived efforts.

A very good workout. I just would like to know if I can believe the numbers or should bias them up. Before it was easy to say that perhaps the Ergomo was too high... But here is some empirical, although still somewhat subjective, evidence that the CT "game physics" is not quite right.


Dave said...


I've had a similar experience doing short high speed intervals on the CT - it seems like my speed is lower on the CT that it would be outside.

However, I've actually found that at lower speeds, I get similar results to what I would do on the road (assuming on the hoods riding). For example, a 40 k tt or a long ironman ride.

I think this is because the CT assumes you are on the hoods (as opposed to an aerodynamic position in the drops). This would have a larger effect at higher speed. I would actually love to find out what CDA the CT assumes.

As another reference point, my CT reads about 7 watts higher than my Quarq on average, but if I use erg mode, the RPE seems higher than the CT power output.

Ryan said...

Also, I think rolling resistance differences (board track being a near-ideal surface) may be accounting for some of the missing speed. RR is a small component of effective drag at these speeds, but on the margin, that's all you have.

You've got real-world wattage files from the track, right? So it shouldn't be too hard to determine your hr-at-speed in a take-a-lap, and then torque the CT numbers until your hr matches up with real-world experiences.

Also, have you taken the baroque step of making your own "Rosetta Stone" by riding your Ergomo-equipped bike on the CT?