Torque and effects on training, rollers vs stationary

NomadVW said:

I was reading through one of the e-books I've got and it talked about torque on rollers having negligible changes for changes in power output as compared to a stationary trainer (ala fluid, etc).

I decided to take a look back at power files of my rollers vs trainer time and compare torque requirements for power output.

I took a handful of files that show, at least anecdotally, that torque on my Kurt Kinetic is a significantly steeper slope as watts increase than it is on my Minoura rollers with a mag resistance unit at same cadence/power.

Do any of you take this into account when planning training on rollers?

I'm able to get enough resistance to do all the way through L5 work on the rollers. What physiological differences am I going to be experiencing when I am requiring less torque for equivalent power output at equal cadence? I understand that on the road if I want to work different sets of muscles I can shift up a few gears and push a bigger gear at equal power at lower cadence (that is, higher torque). But in this case it's an equal power at equal cadence at lower torque.

Thoughts?

Not sure I quite get it. Are you looking at hub torque or the back calculated crank torque?

NomadVW said:

it's an equal power at equal cadence at lower torque.

Thoughts?

If this is what you're observing, it means the torque you're looking at is hub torque. The crank torque should be the same -- if not, it means your power meter is off.

Yes, I'm looking at the PT hub torque. So then... when reading Arnie Baker's High Intensity Training and he says

Arnie Baker said:

Rollers have limited intensity-training applications, because they do not allow changing torque. In any given gear, even though cadence, and therefore speed may double, torque requirements change hardly at all....

He's talking specifically of wheel torque in which case really its a moot point. Not even moot, rather completely and blatantly wrong as the legs are trained on crank torque and not hub torque.

Looking back at the files now in Cpeaks, I see that the back-calculated crank torque ramps comparably with equal cadence, equal power though the wheel torque in the raw files does not.

NomadVW said:

He's talking specifically of wheel torque in which case really its a moot point. Not even moot, rather completely and blatantly wrong as the legs are trained on crank torque and not hub torque.

Maybe he rides a fixed-gear on his rollers. 😄

The torque requirements certainly change because of increased friction and air-disturbance on rollers, but unless you have a resistance unit the torque curve would be very flat compared to a fluid trainer. I would guess he's talking about rollers without a separate resistance unit, in which speed would have to increase a lot in order for torque to increase a little.

NomadVW said:

Do any of you take this into account when planning training on rollers?

No.

Rolllers on easy days, fluid trainer on harder days.

It sounds like with a flater power curve you would tend to drift into a higher than optimum cadence if you aren't paying attention. This isn't a bad thing it will just hide your power improvements for the given intensity until you shift up to a bigger gear.