Efficiency Improvements Using Force Measurement Part III

Efficiency Improvements Using Force Measurement Part III

with No Comments


The last two blogs focused first on the Recovery and then on the Drive.  Both used actual boat data-logger force measurement derived Impulse (F x t) curves plus rowing tank real-time impulse and handle speed curves as part of the discussions.  We will continue to refer to those and similar curves shown below to expand on this discussion.

While the data logger output of impulse, boat velocity, boat acceleration and the subsequent calculation of handle acceleration relative to the moving boat presents the purest rowing force measurement information, it suffers from information overload and not being able to be viewed in real-time.


Why Use Force Measurement In A Tank:

When used in a rowing tank a force measurement system suffers from the lack of boat acceleration and velocity, but benefits from being able to observe impulse and handle speed plus stroke length, gate angle and slip in real time. In addition, each force curve is unaffected by the mass movement of others rowing in the tank. In a boat the athletes are interconnected, so it is harder to discern who might be creating an anomaly in the curve.

The simulation match between boat and tank is FORCE according to Newton’s 2nd Law: (F=MA), where: M boat x  A boat = M tank water (impacting blade) x A blade.  (Note: That nowhere in the simulation equation for force does water speed appear, refuting any reason for a mechanically assisted moving water tank.


Force Measurement In The Boat:

To explain further:  Rowing in a moving boat involves the blade of the oar moving through still water. The blade (if correctly buried) only slips relative to the water along a defined path while translating in and out from the boat and generating a component of force that propels the boat forward.  The blade follows a curved trajectory first out from the boat and than back tracing a pattern similar in shape to a cursive small letter “e”.  A good analogy is a cam follower being forced to follow the profile of a cam.  The water viscosity working against the blade speed reacts like a solid surface and resists getting out of the way, so the blade being fixed to the shaft, which is fixed to the boat at the pin is forced to move relative to the water first tip to tail and then once past the orthogonal, tail to tip.

In a manner almost directly in contrast, the rowing tank blade is sized small enough so that it can rip through the surrounding water mass to generate its force following roughly a circular arc.  The trick to good tank design is to make the force magnitude represented by the impulse curve (force-curve) and handle speed  curve roughly equal to what is experienced in the boat by modulating blade size and oar mechanics.

Once the force simulation is satisfied between the boat and the tank to an acceptable degree than real-time feedback training for rowing can be done in the tank to achieve an optimal stroke that is both economical and efficient at near to racing cadence.

The displayed force and handle speed curves become essentially real-time feedback to the athlete that are similar enough to curves downloaded from a boat data logger so that an economic shape of the curve, area under the curve, and slope of the curve can be mimicked by all to achieve synergism.  Rowers training in this way will over time quickly converge on near identical force curve similarity.  In short they will approach a shared optimal rowing technique.

The question then becomes: What does an optimal and efficient technique look like in terms of shape, area under the curve, and slope of the curve?  To answer that question we look for examples of the curves from some of the fastest, most accomplished elite rowers and note what their curves have in common.  We then use basic physics mechanics to form a hypothesis that we can use to test less experienced rowers.

Below are four Curve Sets:  (Click To Enlarge the Force Curves With Handle Speed Curve Embedded and See Captions)

Starting fro the top left we have:  a top US National Team and Olympic lightweight man sculler, a top US multi National Team and Olympic HW man, a multiple Olympic gold medal women, and a US junior lightweight woman sculler (when tested), eventual 2 time under 23 US National team single sculler.


The Catch:

Just below is two sequential curve plus an imbedded screen grab video from the LW Male sculler taken just before the catch, and at the true catch, We think that his curves in general represents a good example of efficient, fast rowing. (Click to Enlarge and See Captions) If you click cycle back and forth you will see in the video how quickly he inserts the blade matching the relative speed of boat to water without any body movement other than his hands/arms.

The catch is identified on the graph by the long, vertical, thin black curser line bisecting all the curves.  Note the force curve (the  smaller, almost rectangular shaped mountain at the top, overlapping the larger sinusoidal oar angle curve). In #1 just before catch curve you can see that force has barely started on the footboard just as he is about to quickly initiate the catch, with the blade just about to touch the water surface.  Boat deceleration (red line) is at its maximum (off the chart!), and boat velocity (blue line) has decreased to minimum.

Now skip down the page to the four Handle speed curves. These are the same curves embedded in the four curve sets of the individual scullers.  Note the first curve of the Lt Men’s Sculler:  At the furthest extension of approximately -65 degrees the angular acceleration of the handle goes from 0 to 200 radians  (at least on the starboard {green} side).  In almost no time the sculler matches the velocity of boat relative to the water at about 5 m/s, its minimum.  That is a graphical description of a very quick catch!

Now look at #2 – the true catch.  Cycling back and forth between screen grab #1 and #2 notice that the scullers body does not move, only the hands raise quickly. Note how quickly the boat responds (red curve) by going from extreme deceleration to extreme acceleration.  Also note the extreme slope of the force curve, almost vertical terminating at about 75% of maximum.  That force is generated almost exclusively by the boat velocity and by the momentum of the boat at the most extreme oar angles.  It is referred to as lift. But that should be the subject of another Blog.

Again: Referring to the first of the four handle speed curves – the light weight men’s. You can interpolate the force change from #1before catch to #2 at catch as accruing on the handle speed graph during about 7 degrees of oar angular movement during the catch (due to boat speed).  Blade angular velocity cannot exceed boat velocity at the catch. Note how at the beginning of the catch angular velocity is 200 radians and a millisecond later the angular acceleration of the oar has been cut in half to 100 radians!  This illustrates a very key point.  At the catch, if the blade is inserted quick enough and deep enough it will initially only rotate as fast as the boat will allow.  At the catch any additional force applied over what is needed (about 75% max) to maintain the minimum boat speed will be internal energy expended with no work done, or an isometric exercise!

(Click To Enlarge the four Handle Speed Curves and See Captions)



Referring again to the Handle Speed Curves of the four examples of quite accomplished scullers.  You will see the same characteristic curve at the catch with all of them where the handle speed is constrained by boat speed.  The point that this makes is the futility and inefficiency of driving too hard with the legs initially.  You must limit force until the boat progresses enough to have a favorable mechanical advantage with the leg angle.  About 75% maximum pressure will be enough until the legs get to a bio-mechanically strong position to be able to drive down hard.


The Drive:

(Click to Enlarge and See Captions)

Notice that as the drive progresses the boat velocity stays at its minimum until well past the orthogonal position (when the oar is perpendicular to the boat).  From there until the end of the stroke the rowers mass (which is about 6 times the boat mass), is trying to move the boat in the wrong direction – toward the start rather than the finish line). This mass resistance becomes exponentially less as the drive progresses, allowing the boat to accelerate and the velocity to increase due to oar handle acceleration. (i.e., shown by an increase slope of the blue velocity line).  At the same time the handle speed curve has a noticeable upward slope from 100 radians toward 200 radians.  While the blade is in the water the steeper the slope of the handle speed curve the faster the boat.  Boat velocity and handle speed are of course directly related.

When rowing in the tank where there is no boat velocity.  Tank handle acceleration is shown by upward slope of the handle speed curve on the drive (i.e., top of the curve from left to right). This technique must be achieved without “washing out”.

The best shape of the impulse curve is a full curve slightly rounded on the top.  Trapezoidal is an even better, even our young scullers curve shows this, taking into account her inexperience at the time with much less power and a slower catch.  What you don’t want is a narrow high peak curve.

In genera, the larger the area under the impulse curve, the better. (i.e., with the area being the integral of the curve).


The Recovery:

The 2nd part of the stroke – the recovery, can best be thought of as the other side of the equal sign in Newton’s 2nd law.  It is Mass times Acceleration   A, whereas the Force part of F = MA can be considered to be the drive and the Mass x Acceleration the recovery.  You may ask what mass are we referring to?  It is primarily the rowers mass, since it is approximately 6 times that of the boat.  During the drive phase of the stroke we use our kinetic energy (i.e., 1/2 mass times velocity squared).  This energy is now available as potential energy of position (with no gravity to help us, except the slight incline of the slide).  The effect of this available energy is graphically shown by the large boat velocity increase during the recovery to maximum boat speed just before the catch solely due to the mass movement of the rower from finish to catch.  This increase in boat velocity on the recovery is at least equal to, or more than that seen generated from Catch to Finish during the drive.  This is despite boat frictional resistance increasing as the square of the speed.

Rewriting Newton’s 2nd law results in the Impulse/Momentum equation t = M V.  This should be familiar since Force and Time are the x, y coordinates for the Impulse (force) curves and F x t is the definition of impulse. Since Impulse equals Momentum we can define the rowing stroke (cycle) as Drive velocity is the result of Impulse and the recovery velocity the result of momentum.

Recovery in the tank: We don’t have boat velocity or acceleration curves in the rowing tank, and there is no useful driving force on the recovery in the tank from momentum because of the literally astronomical reversal in mass ratios from our predominant 6 to 1 when considering our mass to the boats mass which is supported by an almost frictionless surface of water, as compared to the mass of the earth which we are affixed to!

Recovery on the water: However, the handle speed curve clearly shows that the Recovery (bottom horizontal) is almost a direct mirror of the Drive (top horizontal) .  It shows what our mass is doing because it is connected to the handle.  Two of our model male scullers show a very nice mirroring of the handle speed on the recovery.  When done correctly the complete handle speed curve shape resembles that of a snack food “little fish”.

What the recovery part of the curve shows is a measured speed  after the hands come away with the  upper body swinging forward and the rowers mass moving toward the Catch (left vertical side of the curve) at a velocity about equal to that of the drive, or about -200 radians at racing rates. The rowers mass then speeds up in conjunction with the square up just before the catch in an effort to continue momentum and make de-acceleration of mass as short as possible.(Click to Enlarge and See Captions)


Instantaneous Tank Display:

Below is data curves from a display from an indoor tank equipped with force measurement instrumentation. Shown are instantaneous Force Curves (Impulse Curves) and Handle Speed Curves with a red “+” indicating the position of the port hand and a green “+” for starboard hand, which also indicates the relative positioning of the hands on both sets of curves. Note the bar graphs on the right side are refreshed once per stroke. The curves are from two 16 year old scullers so we try not to be judgmental. The more accomplished sculler is on the right (stroke) side.  The less accomplished with regard to technique is on the left (bow).

Reading the display going clockwise: Force curves are in the upper quadrant for both rowers with a red dot indicating port and the green dot starboard.  The next quadrant is handle speed represented as a bar graph with the length of the bar representing stroke length and the relative starting point of each bar indicating the furthest catch angle for each side of each rower.  Numerical catch angles are to the right of each bar.  There is a threshold value of ten newtons that has to be exceeded before the bar is shaded and that visually represents the true catch.  The length of shading represents total effective stroke length until the 10 newton threshold is crossed again.  The goal here is to have no white showing at the start of the bar, indicating a quick catch. It is almost impossible to eliminate all the white at the end of the bar due to momentum. However, an excessive length of “white” indicate premature disengagement from the water (i.e., washing-out or pulling down into ones lap).

The 3rd quadrant shows power per side for each sculler in watts and stroke rate.  This is not absolute power but relative power because the mechanics of the tank are so much different from the boat that it is beyond the calibration range of the software.

The last quadrant shows the handle speed for each sculler. The shape and slope is similar to that seen in the boat, with the exception of immediately after the catch being more continuous mainly because the blade is being pulled through the water rather than being supported by the water and forced to match boat speed, as in the boat.  The shape of the handle speed curve ends up looking like a “peanut”, curved downward at both ends with top and bottom mirroring one another.


(Click To Enlarge of Sample FM Data Screen)








Click on Video Below or Refresh Internet Browser First:



The above video shows the same two young scullers using the tank.  You can hear me coaching them by questioning them: “When you accelerate your hands at the end of the stroke, what does that do to the shape of the (handle speed) curve?”.  In this way they learn by doing, assisted by real-time feedback. This is a much quicker, and more long lasting coaching method, that works so much better than hoping that they will interpret your words correctly.

From experience we have learned that if the rower is self-coaching in the tank, he or she can only effectively look at one quadrant at a time.  By playing the video over-and-over you should see most of the following takeaways:  The bow person (left side) has trouble with the side-to-side synergism as compared to the stroke person (right side). The slope of the curve at the Catch is different between the bow and the stroke, and this is also reflected in the slip at the catch shown by the bar chart.  What you don’t see or know is that the person in stroke seat has primarily sculled, whereas the bow person has primarily rowed sweep,  The bow person is quite a bit taller while being short on each end of the stroke. You can see that there is less handle speed acceleration during the stroke with the bow person then the stroke. There is a different hand speed away from the release on the recovery by the two athletes, which does not help synergism.  The bow being a bit excessive with hands out of bow.

Problems like these can be easily seen and identified by the rowers in real time as well as by the coach. From our experience, identical problems occur on the water and in the tank. Once pointed out, the problems can usually be quickly fixed.

Also during the practice a video can be made of the output data along with a video of what the problems looks like from the stern or side-view and sent via cell phone immediately to the athlete. This a very powerful tool. When environmental conditions of wind and weather, or school constraints impact training time, then real-time tank sessions are a much better use of coaching time then on-the-water sessions.

By: Jim Dreher and Coleen Fuerst

Leave a Reply