The Ergometer and Rowing Compared

(at Equal Stroke Rating)

© 2001 Atkinsopht (07/03/10)


It is not easy to forge an ergometer vs. rowing comparison owing to the difficulty of deciding upon what basis to make it. One could maintain equal rower power; invent a time or energy cost for the ergometer to cover some fictitious course distance; compare stroke rate or chain-pull; etc.

I have chosen here to maintain equal rating, equal drive/recovery period ratio, and equal chain-pull force and profile--the comparison being made between handle power and total rower power output.

The rowing calculations are made by the model ROWING and the ergometer calculations by the model ERGMOM. The ergometer fanwheel torque drag factor, K, will be adjusted (the inlet damper set) to produce equal drive/recovery period ratios--an adjustment not possible in the shell.


Rower and system:

  Rower weight: 90.0 kg;  Height: 1.95 m;  Peak pull capacity; 650 N
  Stroke rating: 30.0 1/min

The shell:

  Single scull- Length: 8.0 m; Beam: 0.28 m;  Weight: 15.0 kg
  Hull drag factor, Kw: 3.19 N-(sec/m)2 
  Air resistance,   Ka: 0.32 N-(sec/m)2

The ergometer:

  Fanwheel mass moment, Im:       0.1001   kg-m2
  Fanwheel torque drag factor, K: 0.000199 N-m(sec/rad)2
                                           (damper setting)

The Result

  Revised 10/17/02: Owing to an improved method of summimg the internal

  Constants & variables              Sculler          Exerciser
  ---------------------------------  ---------------  ----------------
  Weight, kg; Height, m              90.0;   1.95     90.0;   1.95
  Peak chain-pull, N                 650              650
  Stroke rate, 1/min                 30.0             30.0
  Drive/recovery period ratio        0.851            0.851 *
    * matched to rower's ratio by adjusting fanwheel damper setting

  Hull drag factor                   0.073 (&. air)   n.a.
  Fanwheel torque drag factor        n.a.             0.000199 damper
  Shell speed (avg.), m/s            4.23             n.a.   [setting
  Fanwheel speed (avg.), rad/sec     n.a.             101.8

  Oarlock/Fanwheel shaft power, W    229              272
  Captured body momentum power, W    +48               +0
  Shell/Fanwheel friction, W         ===> 277         ===> 272
  Oarblade/Ret.spg., brg. losses, W        81               27
  Lost body momentum power, W            +114             +121
  Total rower power, W                    ===> 472         ===> 420
  Recorded rower power (PM2), W                n.a.             272

  Oar/Pull-handle power, W           310 <===229 +81  299 <===272 +27
  System efficiency                  0.587 = 277/472  0.648 = 272/420


1. For comparable effort put in to overcome the friction of the device (shell or fanwheel) the rower expends 12 percent more total power than the exerciser. This because the rower is able to add 10 percent of his total effort through the footboard toward overcoming the friction of the shell--an addition denied the exerciser whose footboard is fixed (and, in any case, not mechanically connected to the fanwheel friction effort).

The only true way of devising an ergometer more realistically to model a shell is to "float" the device (as has been done by some) and to arrange that the footboard reaction be connected to the fanwheel so as to speed or to slow it as the force is positive or negative--just as the speed of a shell varies on the water.

2. The rower does more "handle" work than the exerciser because his blade losses are greater than are the exerciser's return spring and bearing losses.

3. Unless some amount is arbitrarily added to it, the recorded work done by the exerciser is under-reported at the PM-2--by as much as thirty-five percent. The body slide momentum work is omitted as well as the work done on the handle return spring. In this case one would have to add 148W (91kgCal/hr) to the recorded output in order to make up the unrecorded difference. Adding too much skews the values in the other direction. The amount to be added varies significantly with the weight and the size of the exerciser thus raising the possibility of some fairness issues in ergometer competitions. Concept-II adds 300kgCal/hr to the fanwheel work to account for the unrecorded effort on the slide. If what the PM-2 thus attempts to estimate is the net work done only at the ergometer, then 300kgCal/hr for the internal work is an overestimate: [300kgCal/hr vs. 121W *1.16 =141kgCal/hr actual internal work]. If, on the otherhand, the PM-2 total represents an estimated metabolic total, then the result is an underestimate [272W *1.16 *4 +300 =1,388kgCal/hr] vs. [420W *4 *1.16 =1,949kgCal/hr actual total output]--assuming a metabolic efficiency of 25%.

4. The sculler putting out 472 J/sec (Watts) takes 473 seconds at 4.23 m/sec to cover a 2,000 meter course; a total work output of 223,000 Joules.
In order to expend 223,000 Joules (i.e., "row" 2,000 m @ 4.23 m/sec) the exerciser (at only 420 J/sec) must work for 532 seconds; twelve percent longer than the rower.

5. The sculler, at the exerciser's total power output of 420 watts, would travel at 4.09 m/sec (at 27.8 1/min) requiring 489 seconds for 2,000 meters. (Because the oarhandle pull is unchanged the rating must decline.)

Home Page Site Map Rowing Page Ergometer Page