-o0| At the end of having browsed in this site |0o-
Cast a vote here if you think it should move up in the Top-100 ranking: Top-100 Rowing Sites
You will be transferred to the Top-100 site, but may then return here by going "BACK"

Vignette Rowing Computer Research Vignette

(Or: How Rowing Really Works)

© 2001 Atkinsopht (03/12/13)
Tweet #RowingPhysics
Tweet to @Atkinsopht

With Results from the ROWING Computer Model
Make Your Boat Run Faster

Oar vignette


In regard to the use of the content of these pages I ask only that:
-o0| Credit be given where credit is due |0o-

This website and the research it represents is a work in progress.
Return here from time to time for corrections, additions, and refinements.
Occasionally you will find contradiction with previously posted material and opinion--thus advanceth Science.

Note: This page has been translated into:
Beylorussian by Olga Skachko. Into
German by Marina Dmitrieva. And into
French by Kate Bondareva.

In effect what I have done here is to have conjured up an infinity of shells, oars, and blades; of crews of all sizes, capabilities, and rowing styles; and of lakes, rivers, and oceans for them to row upon. This infinity of ghosts haunts an infinity of space in which they may materialize in almost any form to experiment in their world: a comprehensive computer program.

The utility of a reasonably good computational model of any system is not that it can produce perfect absolute results but that it can produce good relative results. In no other way can changes in a variable of interest be evaluated while all others are held constant. With ROWING I would be unwilling to predict individual race results for any collection of boats and crews. But, because ROWING considers every significant variable, I have confidence in ROWING's ability to predict an advantage for a difference between one mode of rowing and another.

Atkinsopht will share information with any rower, coach, or crew wishing to understand the possible benefit of a rational approach to improvement in oar blade efficiency and shell speed. Get in touch with me.


From the computer model, "ROWING"--based on its physics, mechanics, and hydrodynamics--I have found some interesting things to share with the rowing community. Extensive use of this model has convinced me that there are subtle ways to increase shell speed without increase in total rower power.

A computer model treats only of the mechanical aspects of rowing and rowers. Crews win races through their capacity to add to the base mechanical skill of body and equipment those crucial elements which one can hope will never lend themselves to modeling: teamwork, resolve, spirit, etc. Beneath these intangibles, however, it behooves any winning crew to tune its mechanical base to the highest practical degree.

Note: I have done what I can--with a model written to compute in English units--to present results in metric units.

News & Comments on Applications

Findings from the exercise of the ROWING model offer the possibility of increase in the efficiency of rowing in the following areas:
1. Blade surface area (and shape)
2. Peak force management
3. Rower strength and rigging geometry
4. Peak force vs. rating
5. Catch bow angle (for scullers)
6. Blade cant angle (for sweepers)
7. Peak oarhandle pull (timing of)
8. Blade immersion depth

A Mathematical Model Comparison

Here is a comparison of results from two independently developed and comprehensive mathematical models of rowing- one, ROWING, a FORTRAN model and the basis of this website, and another, a MatLab model, by Marinus van Holst.
We consider the independent concurrence of results from these two codes as lending strong credence to the validity of the individual models.

Note: See also a new third (2010) rowing computer modeling site by Roosendaal

Rowing Faster

Here is a commentary on Volker Nolte's exhaustive compendium of rowing arcana: Rowing Faster, 2nd. Ed., Human Kinetics, 2011. I invite the various authors to post responses to the commentary which, per force, is limited to those technical aspects of rowing which are subject to computer modeling.

Indications

The several variables modeled and investigated so far by ROWING are illustrated and discussed in detail in the various application notes that follow.
Coaches paying close attention simultaneously to Indications 1, 2, 3, and 7 might hope for an improvement in average speed on the order of 2%.

1. Blade Surface Area-
The ROWING model indicates that blade surface area should be as large as a rower can easily manage. This is the first and easiest thing a rower or crew should work on in order to increase efficiency and speed.

2. Peak Force Management-
The ROWING model indicates an optimum point and timing for the peak of the oarhandle force profile in the course of the drive. Intuition might tell one that the peak force should coincide with the oarshaft ninety-degree point- a correct hunch as it turns out. This technique will be hard to learn without direct oarhandle force feedback available on the boat itself- the next wave for N-K?

3. Rower Strength & Rigging Geometry-
The ROWING model shows that there is, for every rower, a best oar length and lever ratio depending upon the rower's peak oarhandle pull (his strength) thus providing a rational approach to the optimization of rigging arrangements. An example is given for coxless fours. This, too, will be hard to implement without the same force feedback instrumentation suggested above.

4. Managing the Free Return-
The ROWING model indicates that there is no possible way to coach the free return for improvement in shell speed, except by increasing the "float".

5. Peak Force vs. Rating-
The ROWING model shows that it seems better to pull hard at a lower rating than to ease up at a higher rating; all while doing equal rower total work.

6. Catch Bow Angle-
The ROWING model indicates the optimum catch bow angle and allays fear of the worrisome "pinch-point" for scullers.

7. Blade Cant Angle-
The ROWING model indicates an optimum cant angle between oarshaft & blade which may avoid current physical restrictions on smaller catch bow angles for sweeps.

8. Peak Oarhandle Pull-
The ROWING model indicates that a ten percent increase in rower peak oarhandle pull can be expected to yield a 3-1/2 percent increase in average shell speed at equal stroke rating. For an eight this translates into about three lengths in 2000m. Total rower power would increase about 7-3/4 percent; just about what would be expected from theory: P2/P1 = (V2/V1)3. Of course, for equal rower power, pull force can be traded off against rating; it seeming better to pull hard at low rating.

9. Oarshaft Flexibility-
The ROWING model indicates that oarshaft flexibility has virtually no affect on shell speed. Flexibility modifies oar handle (torso) speed and momentum but in such a way as to produce no change in the net work transferred to the oarlock or footboard. Flexing a well made shaft is efficient (little lost heat generated); however, opting for stiffness can conserve a bit of energy, but may be hard on the rower's back.

10. Blade Immersion Depth/Puddles-
It seems reasonable to expect that a well buried blade will be more efficient than one limited to the water's surface.

11. Oarblade Efficiency-
Instantaneous or position dependent oarblade efficiency is not a very useful concept.


Validating the ROWING Model

Until it has been demonstrated that a computer model has a base in reality it has neither credibility nor value as a performance predictor. I have confirmed the ability of ROWING to model the reality of singles, fours, and eights through comparisons between its output and the extensive on-the-water data of V. Kleshnev of the Australian Institute of Sport (AIS) [see Links]. Using input to the model supplied by Kleshnev (shell type, rower weight and height, rower peak force, oar geometry, stroke rate, etc.) ROWING accurately predicted virtually the same speeds as found on the water by the AIS. Moreover, ROWING pretty well matched the shape of the shell speed profiles of which the following is typical:

Speed profile

The speed profile is a function of the movements of the rower's mass on the slide and the speed and forces at the oarhandle.


The ROWING Model

The accompanying abstract of the paper "Modeling the Dynamics of Rowing" describes in general terms the nature and the operation of the model. The complete paper (ca.50 pages) may be had upon request for $75 (add $30 for shipping outside the US).

Download and Run the ROWING Model

If you have Windows 95, 98, or XP you may run ROWING yourself by downloading the ROWING program.
ROWING runs in MS-DOS mode under Windows 95, 98, and XP.


Rowing Ergometers

(Ergos)

Some Rowing Topics

Here follow some topics of interest to--and often misunderstood by-- rowers:

1. Oarblade Path Geometry- The absolute slip of the oarblade is large in spite of convincing appearances to the contrary.

2. Propulsive Force- The propulsive force exerted "on the water" by the oarblade has little to do with blade design or hydrodynamics.

3. Oarblade Vectors- In the literarure I have found no consistent exposition of the forces and velocities pertaining to an oarblade in the water slipping under load. This new presentation is a simplified one; for more detail see Oarblade Lift and Drag.

4. Oarblade and Wing Compared- The oar blade is not exactly an airplane wing.

5. Oarblade Lift and Drag- How lift and drag affect oarblade slip.

6. Shell Work- A significant portion of the useful effort required to advance a shell is transferred through the footboard.

7. Rower Power- Total rower power output includes the internal mechanical work done on the rower's own body, and which is there lost internally in friction and heat. To date this internal work has been almost universally ignored in rowing research.

8. Drive/Return Ratio- There is confusion in understanding the relation between stroke rate and drive/recovery (return) period ratio.

9. The Moving Water Fallacy- Oarblades do not "move" masses of water around while sweeping the stroke.

10. Shell Hull Friction- Data are sought for shell hull block and prismatic coefficients.

11. Soccer Balls for Oar Blades?- Soccer balls work well enough as oar blades but are not recommended.

12. Water Temperature- Shells go faster in warm water.

13. Asynchronous Rowing- Can the elimination of cyclical speed variations increase shell speed? The answer is no.

14. Coxswain on Sliding Seat- Perhaps tomorrow's cox can be enjoined to help out.

15. Counting Strokes- Racers would do well to pay attention to course stroke count.

16. Some Observations- Further research, propulsive efficiency, on-the-water testing, design, etc.

17. The Effect of Deadweight The effect of dead (displacement) weight on shell speed.

18. Shell Hydrodynamics- How ROWING makes shell resistance estimates.

19. Rigging Notes- How span and spread changes affect oarshaft angles: not by very much. And how span and spread changes affect "gearing": virtually not at all.

20. Upstream/Downstream- How rowing river currents upstream or down affects the power required for rowing.

21. Shell Air Resistance- It's time for someone to measure shell air resistance; here's how.

22. A Useful Empirical Equation- Quickly make estimates for new shell speed for changes in stroke rate, pull force, and fluid resistance.

23. Race Start- Fifty percent of race speed is achieved by the end of the first stroke.

24. Big Blades and Back Injury- Oar shaft flexibility may be able to compensate for increased blade surface area.

25. Sliding Riggers- Sliding riggers are fast but banned in competition.


A Rowing Physics Bibliography


Rowing Links


Send me your comment and criticism:
Email: Atkinsopht

Site Map Home Page