Torque: Outdoors and on the Trainer

[Bailsibub]

So I've been fiddling around with the ride files from my new Powertap. After my team ride yesterday, I noticed a pretty big difference in the torque requirements for riding up mountains and riding on my Cycleops fluid at a similar cadence and power.

Have a look at these two ride excerpts:

That ride was done indoors.

And then this one:

This was well into yesterday's team ride...going up a mountain here with 7-10% pitches. Notice that my cadence is exactly the same as the trainer ride. And my power is pretty close to the same, too. But the torque is very different...almost double on average...but the peaks are really high. And the torque only dipped down to trainer-like levels when I was going around a tight switchback (the dips).

Does this mean that my trainer doesn't have enough inertia? That trainer interval was done with a 53x17. Granted, going to a higher gear would give more resistance, but it would also decrease cadence at the same power. And that still wouldn't mimic the requirements of climbing up a mountain.

Also, I get comments all the time about what a smooth rider I am, but as you can see from that outdoor file, my power was kind of all over the place. I wasn't really paying attention to the powermeter going up. On this climb, I made a conscious effort to just stay smooth and ride at an effort that I felt I could hold for 20-40 minutes (I didn't know how long the climb was...nothing here takes longer than 40 min). Are your outdoor climbing files similar? Do they go up and down as much as this? The reason I'm asking is that I don't think it really simulates riding on a trainer at the same cadence/torque/power for 20-60 minutes. And wouldn't that be negative for training?

 

[frenchyge]

1) No, your trainer doesn't have anywhere near the resistance of riding up a 7-10% mountain grade. That's why your wheel speed on the trainer was 18mph vs. 10mph on the mountain for the same power.

2) The difference you're seeing in torque is because the software is showing the "hub torque" rather than the "crank torque." Even though your cadence is the same in both examples, your hub/wheel is spinning much slower outdoors, which means that the hub torque is much higher for a given power. This makes sense if you think about the lower gearing that you were using outdoors, and how that transmits torque to the rear wheel better.

For both rides, if power and cadence are equal then your "crank torque" will be equal as well. The new versions of WKO+ are set up to display crank torque, but it doesn't look like your software is. Maybe there's a setting to change somewhere.

 

[acoggan]

The difference you're seeing in torque is because the software is showing the "hub torque" rather than the "crank torque." Even though your cadence is the same in both examples, your hub/wheel is spinning much slower outdoors, which means that the hub torque is much higher for a given power. This makes sense if you think about the lower gearing that you were using outdoors, and how that transmits torque to the rear wheel better.

For both rides, if power and cadence are equal then your "crank torque" will be equal as well.

Spot on.

I've always found it ironic that the SRM software does not report crank torque even though their powermeter measures it directly, whereas the Saris software reports the hub torque even though it is essentially meaningless (unless you're worried about tearing up your freehub, I suppose).

 

[Steve_B]

The new versions of WKO+ are set up to display crank torque

But if you are using a Power Tap, the crank torque is extrapolated by WKO so it's not perfect but it's good enough, especially in a relative sense for looking at things like this.

 

[acoggan]

But if you are using a Power Tap, the crank torque is extrapolated by WKO so it's not perfect but it's good enough, especially in a relative sense for looking at things like this.

Since power = torque x angular velocity, torque = power/angular velocity. The crank torque calculated by WKO from PowerTap data will therefore differ significantly from the true value only if 1) you use cranks much longer or shorter than 170 mm and/or 2) your chain and/or cogs are quite worn.

 

[cbjesseeNH]

...whereas the Saris software reports the hub torque even though it is essentially meaningless (unless you're worried about tearing up your freehub, I suppose).

My morning chuckle My old PT Pro needs a rebuild - I should have been watching that torque closer!!

 

[Bailsibub]

OK, I got it. I wasn't really thinking about the gear combination being different. So the crank inertial load is roughly the same then, and my trainer will suffice. Whew! That's good news.

 

[acoggan]

Since power = torque x angular velocity, torque = power/angular velocity. The crank torque calculated by WKO from PowerTap data will therefore differ significantly from the true value only if 1) you use cranks much longer or shorter than 170 mm and/or 2) your chain and/or cogs are quite worn.

Doh! Crank length has nothing to do with the conversion of cadence to angular velocity and vice-versa...I must have been thinking of quadrant analysis/circumferential pedal velocity.

I guess that's what I get for posting at 5:30 a.m....

 

[Steve_B]

I wasn't really thinking about the gear combination being different. So the crank inertial load is roughly the same then, and my trainer will suffice.

Yeah, IMO power across a reasonable range of cadence (~70-100 RPM) is pretty much the same, at least with me it is. So 90 RPM on the trainer at 300 W seems to prepare me for climbing a mountain at 70 RPM, also at 300 W.

I just went through an analysis of power vs. speed by cadence/gear combo for my trainer to try to understand why I hate the thing. It has power vs. speed curve that is very similar to the open road but because the of the curve, the gear combos I have been riding in to get the power I want restricts me to (what I feel is) a narrow range of gear combo/cadence and I think that this is what is boring me about riding it. Outdoors, terrain forces you to change gears to maintain power. Inside, you can obviously play games with gear combos, getting out of the saddle and shifting into a harder gear, etc. It's just that this is sort of artificial on top of the fact that the whole experience of riding a trainer is artificial therefore it's just one more thing to hate about the whole experience. You can tell that I love riding trainers so much, can't you?

I guess that the logical thing to do, aside from moving to a warmer climate, is to try a trainer with a flatter power vs. speed curve to expand the possible gear range. I tried my friend's E-motion rollers and it seemed to fit the bill. Their data supports this too.

 

[Steve_B]

Doh! Crank length has nothing to do with the conversion of cadence to angular velocity and vice-versa...I must have been thinking of quadrant analysis/circumferential pedal velocity.

I guess that's what I get for posting at 5:30 a.m....

Well, you got the formulas right anyway so partial credit is awarded.

I agree with you about worn cogs and chain. I can tell you that to the extent that I am able to find conditions in my ride files that generate nearly ~ the same power under different cadences, I don't get ~ the same crank torque. However, I have not gone through the analysis to see if the slight differences in power would account for those differences in crank torque or not.

 

[frenchyge]

I can tell you that to the extent that I am able to find conditions in my ride files that generate nearly ~ the same power under different cadences, I don't get ~ the same crank torque.

If the cadences are different, then why would you expect the crank torque to be the same? Did I misread what you meant?

 

[Steve_B]

If the cadences are different, then why would you expect the crank torque to be the same? Did I misread what you meant?

Good question. For the same power, if the angular velocity is different then the torque would have to be too. Similar to Andy, I think I had a brain fart. Maybe I was thinking of hub torque. I dunno.

 

[Bailsibub]

You can tell that I love riding trainers so much, can't you?

I guess that the logical thing to do, aside from moving to a warmer climate, is to try a trainer with a flatter power vs. speed curve to expand the possible gear range. I tried my friend's E-motion rollers and it seemed to fit the bill. Their data supports this too.

Steve, I know a lot of guys that can't take riding on the trainer. I guess I'm one of those who can bang out the kilojoules for hours no problem.

I know quite a few riders that love the e-motions. I remember the "inside ride." That thing is bigger than my apartment here in Seoul! The e-motions are a much better idea!

I've ridden Kreitlers in the past. I'd like to get some of the e-motions, but as I'm in Korea and the Won has been plummeting to the dollar (causing me to pay out the booty for my Powertap, even at a team price), I won't be getting them any time soon. Plus, this summer when I was doing VO2 intervals, I was just thinking how spent I was at the end of each one. It's hard enough just to keep going after those, not to mention focus on staying up and riding on rollers...even e-motions I imagine.

I'm fine doing my workouts on a trainer. And anyway, I get out on the weekends here with the team...hitting some pretty steep climbs and hairy descents. In fact, a teammate cracked a Pinarello on Saturday on a descent. Overcooked a turn at 40 mph and laid it down about 30 meters in front of me. Ouch.