Explanation of drag factor

[Allan Olesen]

Have you done a poll of every single person in order to come to this conclusion?

Did you understand the message I was trying to get through?

[ilan]

Have you done a poll of every single person in order to come to this conclusion?

Did you understand the message I was trying to get through?

Yes, that you polled every single person and that they all agree with your statement.

[jackarabit]

In SimCity, fizzix elaborates and also obscures metaphor. In the case of the rowing ergometer, the metaphor or analog is boat. The rowing ergometer is certainly not a boat but it is a vessel of the synthetic imagination.

http://eodg.atm.ox.ac.uk/user/dudhia/ro ... meter.html

[Citroen]

None of us know how fast the flywheel spun during our row, and we don't care.

Have you done a poll of every single person in order to come to this conclusion?

-ilan

There's a significant unknown in the maths.

I = ∑ r^2 δm
m / t = a ω
P = E / t = ( E / m ) ( m / t ) = k ω^3

Every single rower except the folks at Concept2 is unlikely to know the value for mass (m).

So it's impossible to integrate the watts value and get the flywheel speed. The performance monitor also keeps the tacho signal hidden.

So Allan's assertion is correct.

[Ombrax]

Every single rower except the folks at Concept2 is unlikely to know the value for mass (m).

Mass, or rotational inertia?

Which, btw, isn't that hard to measure:

https://youtu.be/B35V-UWE9I8

(apologies for the ads you may get before the video)

[Citroen]

Every single rower except the folks at Concept2 is unlikely to know the value for mass (m).

Mass, or rotational inertia?

Which, btw, isn't that hard to measure:

https://youtu.be/B35V-UWE9I8

(apologies for the ads you may get before the video)

No ads for me. Part 2 https://www.youtube.com/watch?v=lhRuOQQ9iks

So you've got to spend 20 minutes watching YT videos, then take your ergo to pieces to make measurements.
I'll say it again, every single rower (who isn't a physics major) is unlikely to know or care.

[Nomath]

There's a significant unknown in the maths.

I = ∑ r^2 δm
m / t = a ω
P = E / t = ( E / m ) ( m / t ) = k ω^3

Every single rower except the folks at Concept2 is unlikely to know the value for mass (m).
...

You are mixing up different 'm's here. In the equation for the moment of inertia I, δm is the radial distribution of mass in the flywheel.
In the second and third equation, m is the mass of the air pumped by the rotating flywheel.
Even the folks at Concept2 will not know the value for the mass of pumped air, as it varies continuously, but they do know the moment of inertia I.

[NavigationHazard]

Paul Smith's old ErgMonitor program actually allowed you to track flywheel rpm if you wanted. There's no particular value to the data, in my opinion, but it was there if you wanted it. This is a screenshot (selected more or less randomly) from a workout I did on ErgMonitor back in 2007. The graph here is set to emulate a PM force-curve screen (same handle-force scale vs time on the x axis).



Handle force here is in purple, and is what you'd see on a PM screen. I've added instantaneous watts in red and flywheel rpms in green. The scales for the latter two curves are on the right.

On this stroke, flywheel rpms at the catch were about 1200. That's a function of deceleration during the recovery phase of the previous stroke. They peak at about 1500 rpm, maybe 5/100ths of a second or so before the finish. Note that all three metrics graph out as parabolas. Note also that instantaneous watts and rpms would both continue off to the right if the time axis were extended to show the duration of the entire stroke cycle. Instantaneous watts peaked at about 650. Average for the entire stroke cycle was 437.3, taking into account both the rising curve during the drive and the longer but less steep "tail-end" falloff beginning just before the recovery phase. Flywheel rpms were back down to 1200 or so at the catch for the next stroke.

Values like watts (derived from flywheel data) that are averaged over the entire stroke cycle don't tell you a whole lot in themselves. What's missing from pace/watts is what the rower did biomechanically to produce those average values, along with how sustainable the effort might be.

[Ombrax]

Paul Smith's old ErgMonitor program actually allowed you to track flywheel rpm if you wanted. There's no particular value to the data, in my opinion, but it was there if you wanted it. This is a screenshot (selected more or less randomly) from a workout I did on ErgMonitor back in 2007. The graph here is set to emulate a PM force-curve screen (same handle-force scale vs time on the x axis).

Neat stuff. Thanks for posting it.

As an engineer I'm a bit of a geek when it comes to this sort of thing, but I try not to let my "day job" stuff intrude too much into the rest of my life - that tends to take the fun out of it. So I find things like this interesting, but stay away from heavy-duty number crunching for my cycling or rowing.

[msongy]

“Paul Smith's old ErgMonitor program...” Is this program available anywhere, and does it require surgery to interface with the PM? Or does it even interface with the PM?

I, too, spend down time figuring out not only the physics, but the usefulness of display data, especially in ErgData. Most recently, I’ve been trying to figure out how to potentially display (drive time:stroke time) ratios, stroke by stroke. Should be between 20 and 50% (or 1:1 thru 1:4 if drive:recovery). I realize the stroke data isn’t super accurate, but it’s accurate enough to spit out spreadsheet results - I had to video the displays to collect data. I wish an ErgData window had the option to display this ratio stroke by stroke to help “teach” the rhythm you THINK you’re maintaining. (If it could teach me to dance, that would be good too.)

Alternatively, I played with the PM force curve enough to know the x axis (accelerating time) attempts to auto-adjust for best display, and that recovery time isn’t part of the time displayed, nor can it be visually interpolated. But the curve function does do what it’s supposed to.

My son’s team parents got me into erging and competing, and I became fascinated with many benefits of rowing over 50. But competing really gets me focused on form and energy management. People often ask about drag factor, drive length and rating, but perfecting form and energy expenditure requires the mental discipline of timing force application and recovery within the target stroke rate. The oversimplification that Drag is just “feel” drives me nuts! There are so many related concepts whose displayed data seem bizarre unless you can conceptualize that Drag at 1 is like rowing with a spoon, whereas Drag at 10 is like rowing with a snow-shovel. Adjusting drag without watching what it does to these drive metrics is foolish - especially for competitive nerd-rookies like me!

Or it could be that I need one more thing to watch to help me ignore the pain!!! After all, very few metrics can be monitored during a networked 2k race.

[johnlvs2run]

On this stroke, flywheel rpms at the catch were about 1200. That's a function of deceleration during the recovery phase of the previous stroke. They peak at about 1500 rpm

Would you happen to have flywheel RPM data for a drag factor of 78 (half the 155.7) at the same stroke rate and pace?
If so, I'd be interested to see it.

[Allan Olesen]

Most recently, I’ve been trying to figure out how to potentially display (drive time:stroke time) ratios, stroke by stroke. Should be between 20 and 50% (or 1:1 thru 1:4 if drive:recovery). I realize the stroke data isn’t super accurate, but it’s accurate enough to spit out spreadsheet results - I had to video the displays to collect data. I wish an ErgData window had the option to display this ratio stroke by stroke to help “teach” the rhythm you THINK you’re maintaining. (If it could teach me to dance, that would be good too.)

Which PM do you use?

There is an Excel add-in named PMI, which will read the stroke data from a PM3 over a USB connection. I used that and wrote my own monitoring program in VBA, showing drive/recovery ratio, stroke energy, etc. And most importantly: Guiding me through Wolverine L4 sessions, which can be a bit hard to keep mental track of.

Unfortunately, that add-in does not want to talk my new PM5, so I am again rowing blindfolded.

[NavigationHazard]

“Paul Smith's old ErgMonitor program...” Is this program available anywhere, and does it require surgery to interface with the PM? Or does it even interface with the PM?

....

Paul and his development partner stopped upgrading ErgMonitor at least a decade ago and probably more like 15 years ago. The last version of it I'm aware of ran on Windows XP, which was superseded by Windows Vista in 2006. I am under the impression that a crucial dll file didn't/doesn't exist in later versions of Windows and that they didn't feel like reverse-engineering one. I kept it going for ages on an old desktop running Win 98, but eventually that platform died on me.

As for interfacing with the PM, the ErgMonitor screen was meant to replace a PM display. Normally the input cable from the machine connected to whatever computer you were running it on; you had to arrange the monitor yourself so that you could see it. It was possible to split the signal and run both the PM and ErgMonitor simultaneously, but there was little point since the latter could give you the same data and the same realtime visual display, plus a lot more. I now can't remember whether I had to rig a signal booster to run both simultaneously, but suspect I may have had to.

Would you happen to have flywheel RPM data for a drag factor of 78 (half the 155.7) at the same stroke rate and pace?
If so, I'd be interested to see it.

I doubt that I do, as I don't think my erg [the housing of which is usually clean] returns a value that low with the damper all the way closed. I'm not going to take the cover off and check, but it's somewhere around 90-92 units. FWIW, ErgMonitor allows you to track DF on any individual stroke. It turns out that ambient air currents -- to some extent affected by what you're doing vis-a-vis the fan -- will cause it to vary a few units one way or the other.

I can show you what happens to instantaneous watts at different drag factors. Here are two strokes at the same rate and pace, with the basic contours of the force curve as similar as is humanly possible. The first is at DF 144.1:



The second is at DF 191.9:



I played around with the graphic display on these to show the entire duration of the strokes. The purple curve is the force curve you'd see on a PM. The horizontal line is average watts: 336.6 in the first; 336.5 in the second. The red curve is instantaneous watts. Note that the amplitude is lower in the first screenshot, but that the curve takes longer to 'tail off' or 'decay' during the recovery. The amplitude is somewhat higher in the second, but the curve tails off faster. This difference is a function of flywheel drag. Other things equal, the higher the drag the higher the amplitude of the instantaneous watts curve, but the swifter the decay. The problem is, other things usually aren't equal biomechanically.

Bottom line: there is no one best solution to achieving a given pace at a given rating. Whether for a single stroke or (more realistically) for the succession of strokes that comprise a session piece or a test/race. From which follows a strong argument for doing enough of your training with the same parameters you intend to use for racing. How much is enough? That depends on the individual and the task at hand....

[Nomath]

I am very much puzzled by the description of the red curve in the graphs that you call "instantaneous watts".
Whose watts? The human rower delivers power to the rotating flywheel during the pull. That ends when the force curve (purple) drops to zero. During the recovery the rower delivers zero 'instantaneous watts', but the flywheel loses energy because it pumps air, so it delivers power onto the ambient.
If we look to the profile of the red curve, it gets a boost during most of the pull, peaks slightly before the end of the pull and continuously drops during the recovery. To me it looks like the red curve describes the kinetic energy of the rotating flywheel (units: Joule).

[NavigationHazard]

Which is expressed by convention by the monitor as watts, not joules per second. ErgMonitor calls them instantaneous watts to differentiate the calculated value at any given point from what you'd see on the monitor as a number averaged over the duration of the stroke cycle.