Effects of the drag factor on performance and on physiological parameters

[Nomath]

I have taken a look at that paper. Nowhere do the authors claim this 25 W difference is due to chain losses that I saw. It could be shock cord losses but to see 25W it would require a shock cord force of 10 lbs being pulled 1 m 30 times a minute (22W by my calc). Shock cord pull doesn't seem that high to me but I haven't measured it. If it were shock cord losses I would expect the expert losses to be greater than the novice losses (they are pulling the chain further in the same time) but they are not. This difference is hard to explain.

I simply don't understand why Concept2 didn't simply measure the strain on the support where the cog is. It would be easy enough to then know the horizontal component add that to the speed of the chain (which they already measure) and they have actual power. It is what it is I suppose.

I used the expression 'drivetrain losses' for the 25W discrepancy found by Boyas, which obviously includes shock cord losses. Advised shock cord tension is 2-3 kg, i.e. 20-30N. Let's assume a handle displacement of 1.3 m, then the work in pulling the shock cord over that distance is 25-40J. At a stroke rate of 25 spm, this amounts to 10-17W, which explains a substantial part of the 25W.

Concept2 uses a very clever way to determine power from the angular speed of the flywheel, at least since their model C. I don't know exactly when they introduced it, but it could be around the year 2000. At that time power sensors on bikes (SRM) were very expensive and almost uniquely used by pro's. This has changed. Although force sensors and displacement sensors are much more common now, they are rarely wireless. Adding them to a C2 erg would probably make the erg a lot less robust and add significant cost. For what benefit? But it can be done.

[Nomath]

...
Please note that a 57% increase in DF resulted in a 13% drop in drive speed and an almost 10% increase in power.

Comments?

I think this is roughly what you would expect from the physics of the C2 ergometer. This is most easily seen by looking at equation (3) in the paper Behind the Ergometer Display written by Prof. Marinus van Holst and accessible on the internet (link). Because my typeboard doesn't provide symbols for 'omega' = angular velocity, I am not able to reproduce that equation here.

Suppose we are able to deliver just the right power to the flywheel to keep its velocity constant. Then the last term d(omega)/dt is zero and the equation reduces to P = constant * omega³. The constant is proportional to the drag factor. The speed of the handle (drive speed) will be proportional to omega. Hence drive speed should be inversely proportional (P/DF)^0.33. With P=1.1*Po and DF=1.57*DFo, we obtain V=0.89*Vo. At constant power, the drop in drive speed would be roughly 14%.

I am delighted that you do not cite anymore long sequences of previous comments. Nobody reads those again. It helps to focus on the lines that you want to discuss and explain more carefully what you add. I didn't get the meaning of "RPE was tough on 204df and this felt really heavy.....'

[Dangerscouse]

I didn't get the meaning of "RPE was tough on 204df and this felt really heavy.....'

I did the test, and what I was trying to say was that when it went over circa 170 it felt notably heavier in the stroke, especially at a drag factor of 204. Just in case you don't know what RPE is, it's Rate of Perceived Exertion, so, not surprisingly it felt like a 9, or 9.5/10, compared to the lower drag factors.

I set it to undefined rests and only rowed again when it had dropped to 100, to try and mitigate against fatigue. I'm not going to pretend I understand the physics / mathematics of it all, so I'll leave that well alone, but my power / pace increased over the higher drags, rather than stayed constant. I'm not sure if this makes any difference, but I thought I'd mention it

[frankencrank]

Isn't that the same as saying Rowing is less efficient? And, that, I presume, doesn't include the losses from the oar handle to the water.

Yes, a recent study by Lindenthaler ('Differences in Physiological Responses During Rowing and Cycling Ergometry in Elite Male Rowers', in Frontiers in Physiology, 2018) estimated 18.1% Gross Mechanical Efficiency at 50% of VO2max on a Concept2 and 22.1% on a WattBike. At 75% of VO2max, the GME on a Concept2 was 18.6% and on a WattBike 22.6%. So there is about 4% difference in GME between the two activities. Note that the participants were elite rowers who regularly used a cycle ergonometer, not elite cyclists. Part of the difference should be attributed to the way power is measured on a C2, about which the authors express some reservations.

Also interesting is the observation "it was noted how quickly rowers in this investigation recovered from maximal WattBike exercise when compared with maximal C2 exercise, despite similar VO2peak, Rating of Perceived Exertion and Blood lactate."

Now this is a very interesting paper. There are several problems that interfere with drawing some conclusions. (For instance, did not control for variables that can influence efficiency like stroke rate or pedal speed.) Of course, the biggest question is why did the rowers test better on the bike? I thought it was well established that rowers had higher VO2max than cyclists yet, here the rowers had higher VO2max on the cycle. Why???? This is not a calibration issue. They went anaerobic sooner on the rower, higher lactate on the rower. They tested better doing something they don't do everyday. Why???

And, while a 4% difference in efficiency doesn't sound like much when one is talking efficiencies of 20% a 4% difference is actually a 20% difference. HUGE!!! When a study on my product showed a 2% increase in efficiency in cyclists (from 20% to 22%) those "in the know" thought it impossible.

The authors really make some silly comments that make no physiological sense in trying to explain the conundrum of their results. For instance, "As such, while rowing maximally O2 extraction at the tissue level compensates to ensure sufficient aerobic metabolism." So, somehow, the tissues compensate to ensure sufficient aerobic metabolism when the rower is rowing anaerobically?

It is unclear what they meant that the rowers recovered more quickly after rowing when they actually produced more lactate when rowing. Not sure I would trust that observation as meaning much.

The other issue this paper points out is it is impossible to calibrate the CII? CII may have become the standard that everyone trusts, but from a research perspective, it is sort of worthless. If we were to add the 25 watts the CII misses (from that other study) then the difference between the two modalities becomes much less, bringing that 40 watt difference down to 15.

[frankencrank]

Here is the result of one reported experimental subject result today. I have asked my participants, before starting, to run through increasing drag factors looking specifically at how it changes drive speed and watts. After this they will go up more slowly to allow for some adaption. Here is what the report was. He did 1000 m intervals with about 1.5 min rest between. (blanks were not reported)

DF. Drive Speed. Power
131 --- 2.18 --- 307
142 ----------- 312
169 --- 2.14 --- 324
191 --- 1.99 --- 328
204 --- 1.92 --- 334
127 ----------- 318
Drive length didn't change (1.48 mostly)
"RPE was tough on 204df and this felt really heavy, but up to 169df it wasn't too bad. I struggled to get the same power when I dropped it down to 127df from 204df"

RPE was tough because he wasn't adapted to the stresses. Didn't keep him from improving though. Then when DF got "easier" he struggled. What would he do if he was adapted to higher DF? What DF would it take to cause his power to drop?

Please note that a 57% increase in DF resulted in a 13% drop in drive speed and an almost 10% increase in power.

Comments?

Not a single comment? Not.a single person has run off and repeated this themselves to see what they got? No need to rely on some study. Everyone here (or most people) have the same machine and can repeat the "experiment" so don't be afraid.

[frankencrank]

...
Please note that a 57% increase in DF resulted in a 13% drop in drive speed and an almost 10% increase in power.

Comments?

I think this is roughly what you would expect from the physics of the C2 ergometer. This is most easily seen by looking at equation (3) in the paper Behind the Ergometer Display written by Prof. Marinus van Holst and accessible on the internet (link). Because my typeboard doesn't provide symbols for 'omega' = angular velocity, I am not able to reproduce that equation here.

Suppose we are able to deliver just the right power to the flywheel to keep its velocity constant. Then the last term d(omega)/dt is zero and the equation reduces to P = constant * omega³. The constant is proportional to the drag factor. The speed of the handle (drive speed) will be proportional to omega. Hence drive speed should be inversely proportional (P/DF)^0.33. With P=1.1*Po and DF=1.57*DFo, we obtain V=0.89*Vo. At constant power, the drop in drive speed would be roughly 14%.

I am delighted that you do not cite anymore long sequences of previous comments. Nobody reads those again. It helps to focus on the lines that you want to discuss and explain more carefully what you add. I didn't get the meaning of "RPE was tough on 204df and this felt really heavy.....'

This isn't what you thought would have been expected in that initial post comparing DF 100 and 150. You would have liked to have seen an intermediate data point. Now you have more (plus drive speed).

You gave me a nice "this is what we would expect" drop in drive speed "with constant power" but let me point out again, power didn't stay constant, it increased. That is what I expected. It should happen to you also. Try it.

[frankencrank]

I have taken a look at that paper. Nowhere do the authors claim this 25 W difference is due to chain losses that I saw. It could be shock cord losses but to see 25W it would require a shock cord force of 10 lbs being pulled 1 m 30 times a minute (22W by my calc). Shock cord pull doesn't seem that high to me but I haven't measured it. If it were shock cord losses I would expect the expert losses to be greater than the novice losses (they are pulling the chain further in the same time) but they are not. This difference is hard to explain.

I simply don't understand why Concept2 didn't simply measure the strain on the support where the cog is. It would be easy enough to then know the horizontal component add that to the speed of the chain (which they already measure) and they have actual power. It is what it is I suppose.

I used the expression 'drivetrain losses' for the 25W discrepancy found by Boyas, which obviously includes shock cord losses.

You did use the term drivetrain losses and then, immediately, added (chain + sprocket wheel). It was not obvious that you were including the shock cord when you mentioned drivetrain losses.

[hjs]

Here is the result of one reported experimental subject result today. I have asked my participants, before starting, to run through increasing drag factors looking specifically at how it changes drive speed and watts. After this they will go up more slowly to allow for some adaption. Here is what the report was. He did 1000 m intervals with about 1.5 min rest between. (blanks were not reported)

DF. Drive Speed. Power
131 --- 2.18 --- 307
142 ----------- 312
169 --- 2.14 --- 324
191 --- 1.99 --- 328
204 --- 1.92 --- 334
127 ----------- 318
Drive length didn't change (1.48 mostly)
"RPE was tough on 204df and this felt really heavy, but up to 169df it wasn't too bad. I struggled to get the same power when I dropped it down to 127df from 204df"

RPE was tough because he wasn't adapted to the stresses. Didn't keep him from improving though. Then when DF got "easier" he struggled. What would he do if he was adapted to higher DF? What DF would it take to cause his power to drop?

Please note that a 57% increase in DF resulted in a 13% drop in drive speed and an almost 10% increase in power.

Comments?

Not a single comment? Not.a single person has run off and repeated this themselves to see what they got? No need to rely on some study. Everyone here (or most people) have the same machine and can repeat the "experiment" so don't be afraid.

What did you expect? Ofcourse we know this, almost every newby starts out at a high drag and finds out that lower is better and is more comfortable and gives less injury risk. So its the opposite, everybody has done this. I personally can’t row slower than 1.45 on drag 200. Which is fast for longer work. At shorter work, that drag, gives so much time to pull against, that you more or less snap in two. Sprinting at high drag is the biggest risk on injury. See this happening time after time.

The best example why high drag is flawed is look at the 500m top guys. Those use high drag a lot, the train themselves to be very strong, but yet they all find out that drag 150 ish, gives them the best result and thats what is often used.

Otw, there is a different point, there nomatter what, the biggest limitation is speed of motion. A boat goes with 20 km, so to put energy in the boat, the oars need to travel at least the speed of boat to catch the water. A 125/35 drag, for men, comes out as best one to mimic this condition.

[frankencrank]

I didn't get the meaning of "RPE was tough on 204df and this felt really heavy.....'

I did the test, and what I was trying to say was that when it went over circa 170 it felt notably heavier in the stroke, especially at a drag factor of 204. Just in case you don't know what RPE is, it's Rate of Perceived Exertion, so, not surprisingly it felt like a 9, or 9.5/10, compared to the lower drag factors.

I set it to undefined rests and only rowed again when it had dropped to 100, to try and mitigate against fatigue. I'm not going to pretend I understand the physics / mathematics of it all, so I'll leave that well alone, but my power / pace increased over the higher drags, rather than stayed constant. I'm not sure if this makes any difference, but I thought I'd mention it

Thanks for coming and clarifying. My guess is this isn't what you expected plus I'll bet you couldn't imagine how hard it was after you went back to the "easier" drag factor. That was a great touch.

This data pretty much confirms my hypothesis (yours is the second of the group to do so but your data included drive speed). The question to be answered now is can long term adaption to these higher drag factors (slower drive speeds) result in long-term improvement in performance. My guess is yes.

[frankencrank]

Not a single comment? Not.a single person has run off and repeated this themselves to see what they got? No need to rely on some study. Everyone here (or most people) have the same machine and can repeat the "experiment" so don't be afraid.

What did you expect? Ofcourse we know this, almost every newby starts out at a high drag and finds out that lower is better and is more comfortable and gives less injury risk. So its the opposite, everybody has done this. I personally can’t row slower than 1.45 on drag 200. Which is fast for longer work. At shorter work, that drag, gives so much time to pull against, that you more or less snap in two. Sprinting at high drag is the biggest risk on injury. See this happening time after time.

The best example why high drag is flawed is look at the 500m top guys. Those use high drag a lot, the train themselves to be very strong, but yet they all find out that drag 150 ish, gives them the best result and thats what is often used.

Otw, there is a different point, there nomatter what, the biggest limitation is speed of motion. A boat goes with 20 km, so to put energy in the boat, the oars need to travel at least the speed of boat to catch the water. A 125/35 drag, for men, comes out as best one to mimic this condition.

Of course we know this? Really? Then why a thread entitled "Effects of the drag factor on performance and on physiological factors" as if everyone doesn't know? You know you can improve power by increasing drag factor and, yet, you choose not to? Or, you know you can increase power by increasing DF but choose not to because you are afraid of injury?

The fact that newbies learn that a lower drag factor is more in line with their capabilities is not evidence that a lower drag factor is better for the experienced. The fact that "everyone" did this little experiment when they were newbies doesn't mean the results would be the same as they improve.

Your data that, from a training and performance perspective, that "easier" is better?

Your data that injuries are reduced on lower drag factors? (sprinting or not) I'll admit injuries are increased when one does something new and or different but that is due to change, not drag factor necessarily.

We all like what we are used to. Change is hard. Change like this takes time to adapt. Cyclists have the same problem, change that can make them better usually slows them down before they see improvement (egos are a tough thing to overcome). Many cannot see the forest for the trees. Elites, as a group, (there are exceptions - the very best never stop asking how they might improve even more) are the hardest to change as they are used to winning and afraid of change from what works for them. Just below elites are the easiest to change as they are looking for something to make them better.

So, give me the data that suggests one DF is better than another. Where is the data that says the recommendation that DF should be between 100-115 is best.

What is wrong with rowers experimenting with DF to see if they can find a better DF for them? Do you recommend they not do this?

If everyone knows this, why does it seem to me nobody knows it? When I told those who joined my experiment group what I wanted to do not a single one came back and said, "we know that."

[frankencrank]

Otw, there is a different point, there nomatter what, the biggest limitation is speed of motion. A boat goes with 20 km, so to put energy in the boat, the oars need to travel at least the speed of boat to catch the water. A 125/35 drag, for men, comes out as best one to mimic this condition.

Your data to support this 125/35 ratio is best? This is a simple problem to solve. if one knows the optimal speed of the rower to produce the desired power and the speed of the boat (and "differential speed) of the oar to deliver the rowers power) it is a simple matter of adjusting the fulcrum point of the oar to achieve both.

A cyclist has the same problem. They cannot put a single erg of energy into the pedal until they have accelerated their foot up to the pedal speed. That takes energy.

[Nomath]

...
This isn't what you thought would have been expected in that initial post comparing DF 100 and 150. You would have liked to have seen an intermediate data point. Now you have more (plus drive speed).

In my inital post I took care to present the results of two separate studies by Daniel Kane in an objective way, by plotting his numbers hidden in a table in a couple of graphs that are probably easier to understand for anybody. Also, I included 'physiological parameters' in the title to underline that underlying effects are often more relevant than superficial effects.
I feel that these data should be related to the exercise for which they are collected, viz. a stepwise increase in power in blocks of 3 minutes. Whether the trends apply for,e.g. a 5K time trial remains open.

Secondly, a technical comment. I doubt whether the C2 can accurately measure drive speed. As far as I know, it only has 3 magnetic sensors near the flywheel. After the catch the handle speed must first match the speed of the flywheel before the PM 'knows' that the rower is in the drive phase. Since good rowing technique requires that at the catch arms are fully extented, handle speed = body speed. To speed up the body to the required 1.5-2 m/sec, takes some time and distance. Let's assume that the rower has a constant acceleration a from the catch. With Newton's formulas, we have speed v=a*t and distance s=½*a*t². Eliminating t, we get s = ½*v²/a. For v=1.5 m/sec and a=5 m/sec², we find s=0.23 m and t=0.3 sec. So at an initial body acceleration of half the earth' standard gravity, which is quite demanding for a cyclic effort, the C2 system probably misses some 23 cm of the movement. The initial force to accelerate a body of, say, 60 kg (i.e. legs discounted) at 5 m/sec² is 300N.

[jackarabit]

9x1’/R indeterminate (45~60”)

Does anyone here believe n=3~4 subjects constitutes a study? Absolute BS.

[frankencrank]

...
This isn't what you thought would have been expected in that initial post comparing DF 100 and 150. You would have liked to have seen an intermediate data point. Now you have more (plus drive speed).

In my inital post I took care to present the results of two separate studies by Daniel Kane in an objective way, by plotting his numbers hidden in a table in a couple of graphs that are probably easier to understand for anybody. Also, I included 'physiological parameters' in the title to underline that underlying effects are often more relevant than superficial effects.
I feel that these data should be related to the exercise for which they are collected, viz. a stepwise increase in power in blocks of 3 minutes. Whether the trends apply for,e.g. a 5K time trial remains open.

Secondly, a technical comment. I doubt whether the C2 can accurately measure drive speed. As far as I know, it only has 3 magnetic sensors near the flywheel. After the catch the handle speed must first match the speed of the flywheel before the PM 'knows' that the rower is in the drive phase. Since good rowing technique requires that at the catch arms are fully extented, handle speed = body speed. To speed up the body to the required 1.5-2 m/sec, takes some time and distance. Let's assume that the rower has a constant acceleration a from the catch. With Newton's formulas, we have speed v=a*t and distance s=½*a*t². Eliminating t, we get s = ½*v²/a. For v=1.5 m/sec and a=5 m/sec², we find s=0.23 m and t=0.3 sec. So at an initial body acceleration of half the earth' standard gravity, which is quite demanding for a cyclic effort, the C2 system probably misses some 23 cm of the movement. The initial force to accelerate a body of, say, 60 kg (i.e. legs discounted) at 5 m/sec² is 300N.

It is a poor assumption that the rower accelerates uniformly. Everyone knows that the catch occurs shortly after applying power, seems instantaneous even though it isn't. So, there is a quick acceleration up to flywheel speed. Then, there is a slow acceleration to the end of the leg extension, then probably a slower acceleration as the body leans back and the arms pull in to the end of the power phase. The drive speed that is given is the average drive speed from the catch to the start of the recovery. If I were a coach I would want to know the slide time (easy enough to do using your cell phone and a slow motion video) and compare that to the slide time on the water (easy enough to do using the same technology). I would want all of the members of the shell to have similar slide time. I would use the ergometer to see if I could optimize the rower then see if I could transfer that optimization to the water. I think most use it just as more volume hoping to see training effect improvement.

Is it better to train more or to train smart? How about both?

I am a little confused. Are you saying the average rower on the Concept2 doesn't feel a catch until the body has moved 23 cm (9 inches)? My guess is there is more to it than that. Perhaps the catch occurs sooner by simply bringing the shoulders back (and the oar handle with it). I think your number do not reflect reality.

[frankencrank]

9x1’/R indeterminate (45~60”)

Does anyone here believe n=3~4 subjects constitutes a study? Absolute BS.

Isn't the plural of anecdote, data?
Your result is more in keeping with what I expected. Too much will slow you down and too little will slow you down. You have a "best" DF for this power and your current training level. Everyone will have one. How should you know which one is best when you have 4 that are very similar. I would use HR as the tie breaker (lowest wins but you didn't record that). Anyhow, just as in cycling that optimal pedal speed isn't exact but there is a range. If you were racing perhaps you would choose the lower drive speed for the bulk of the race so you have room to increase drive speed for the sprint at the end of the race.

Edit: Let me add one more thing, If you wanted to use this test result to guide your future training I would choose that DF of 5, maybe even 4, as your starting point (I assume those are higher than what you are used to). The reason is the increase stress won't be to large from what you are probably used to as we don't want to risk injury by trying to do too much too fast. Once that level feels comfortable and normal test again and see if you need change again.

The previous result has occurred twice so far but in much more elite rowers.