designing a fast rowboat

Then you need to consider the interior layout.
Where to put a slot for an umbrella?
Where's the optimal location of an esky?
Should you shift the esky to a better position when it's empty?

Forget hydrodynamics!

Cheers,
Leo.

 

Ah. Well in that case consider the following.

If one is to be carrying an esky full of cold beer for medicinal purposes, and if one is going to be using a sliding seat, one has the opportunity for a gain in efficiency. Given that speed variation during the stroke results in increased expenditure of energy for the same average speed, one should mount the esky full of cold beer on its own slide.

This would be hooked up to the sliding seat via basic ropes and blocks. So, when you shoot forwards on the sliding seat during the stroke the esky (with its associated payload of beer) would shoot aft on its slide. This makes sense in medicinal terms, because when you are leaning towards the bow you wont be able to reach the esky anyway.

When you shoot aft down your slide on the recovery the esky shoots forward on its slide. This will place it in a convenient location for grabbing another beer before the catch of the next stroke. Presumably most people would have to have a slight pause between strokes, but I'm sure a trained athlete could grab a beer without stopping.

Since a decent load of beer will probably end up weighing much the same as the rower the forces should all balance out rather well and the boatspeed should stay much closer to the mean speed.

I haven't patented this idea yet but it's probably worth looking into that.

 

After that much beer you might need to install a self-bailing system. You wouldn't want to pinch an important pipe in the sliding seat mechanisms.

 

Leo:

I took another look at the curves in your post #826; I assume they show results of towing tests at constant speed. I thought they looked close to square law so I calculated the sqrt(N)/S, for the left curve it ranged from 3.34 to 3.5 over the range 2-6k, for the right one the range was 3.42 to3.53, so drag is almost exactly proportional to the square of speed for these examples. This suggest that the drag is almost entirely due to skin drag, has little form drag - as it should for a well-designed and finished hull and also little wave drag - another design goal.

I was surprised by the speed variation in the curves for your post #833, about 31%, even more than the 26% I computed for my trivial example in my post #822 which I thought might be excessive, although I have seen smaller variations elsewhere.

To simplify my math I assumed a sinusoidal speed variation with time. I haven’t been able to find actual plots of short-term speed variation. On page 8 of this link http://www.scribd.com/doc/21985291/Valery-Kleshnev-Rowing-Bio-Mechanics a sinusoidal curve is shown but only as an example. Another study I found used a square wave as an example. Most studies that I have found which look at speed variation are more concerned with long term variations related to strategy, and also emphasize blades, technique, race tactics and strategy, muscle efficiency and biometrics. The “boat efficiency” measure seems to give an optimistic result compared to average power required to maintain average speed. In the attached I used my sinusoidal speed variation assumption and plotted the results. With the approach I used this time it would be possible to plug in the actual speed variation during the stroke cycle, if known. Incidentally, there was an error in my earlier numbers which were scratched up rather hastily; the average power values were too high.

 

But the thing Leo's earlier post showed (and it surprised me too) was that although the resistance may vary pretty much as the square of the speed when running at constant speeds, things go all wonky when the boat is being accelerated and decelerated through the stroke cycle by a couple of hefty gorillas. This means that in practice you can't work it out that simply, which is a nuisance but you get that.

PS: Just had my first play around with Michlet 807. Interesting, even if it doesn't calculate the effects of sliding eskys full of beer.

 

The drag loops for the rowing pair I showed earlier include the effect of dynamic sinkage and trim (aka "squat") as well as the effect of the rowers' masses.

At the catch, when the rowers are furthest sternwards, their weight tends to lift the bow. The boat is going slowest, so the pitching moment is not very large, and the boat tends to be bow-up at that instant.

When the rowers are furthest bow-wards, their weight tends to push the bow down, but the pitching moment tends to lift the bow, and the two moments act to cancel each other out, i.e. the boat stays fairly level.

I suspect that the stubbier boats you guys are considering, and the Froude number range you will be rowing at, means that dynamic forces and moments will be fairly low. Therefore you can assume that the hull be bow-up or bow-down purely because of the moments induced by the location of the rowers. That's good news because it means you can use (the free program) Michlet to give you reasonable approximation to the drag of the hulls you are considering.

Incidentally, I think that estimating the air drag is probably the toughest calculation, and that is where very complicated CFD programs should shine. Unfortunately, because of the very bluff irregular shapes of the rower, the hull and the oars, it is a very challenging computation.
I can only advise you to stay thin, and try to develop a hatchet-shaped face to reduce air drag.

Leo.

 

I think that, for the sorts of boats we're generally talking about in this thread, any hull that's close to optimal for level trim will be equally close for the sort of trim changes likely to occur in practice. IOW, I don't think I'd be bothering running calculations for varying trim.

I just ran a few things in Michlet and Godzilla to check a design I've been working on. I used the ITTC line and no form factors since that seemed a good basis for comparison of hulls even if the absolute values are not quite right. What I found was that my one hull was about as good as the multiple hulls Godzilla came up with (a bit better at some speeds and a bit worse at others) which is certainly not discouraging.

Furthermore, the Godzilla hulls wouldn't really work in practice because it always seems to generate semi-elliptical immersed midship sections and I need quite a bit of flare above the waterline. This means I require a different immersed shape anyway.

So on that basis I think it might be time for me to stop worrying about design and start thinking about building.

ETA: Attached a shot of the resistance curves scaled to match and overlaid into one image. The fainter curves are Godzilla hulls optimised for 5 knots and 6 knots. The brighter curves are my design. Speed range of the graphs is 1 knot to 6 knots.

 

I agree that you have pushed it as far as you need. And yes, start building!

The latest version of Michlet (9.10) allows you to use as many as 42 shape parameters and to include flare. It is also quite useful for approximating real hulls with mathematical functions which can be useful for some types of work. One day when I get some time I will update the manual and release the code.

Have fun!
Leo.

 

Hello guys,

I have read the last 3-4 pages (57 pages is just too much ) and I see that you are estimating the rowing boat resistance without taking into consideration the pitch, surge and heave added masses - or at least I failed to notice any.
We all know from experience, and Leo's post #833 shows it in a graphical way, that both boat's motion and crew's power input are time-variable. In the post #844 Ancient Kayaker has modeled the power input as sinusoidal curve. But then, how come you have decided to neglect the added masses?
I recall reading that the surge-motion added mass (for example) for a Series-60 ship at low speeds, for example, should be something like 3.0-5.0% of the ship mass (but can go up to 10-15% in shallow waters). Do you consider it important or is it neglectable for your power estimates? That's just surge, but heave and pitch added-masses would add much more to the picture, imho.

Cheers!

 

The speed variation already incorporates the impact of the moving masses that you mention, but perhaps you missed my point. In my calculations I start from the speed variation, calculate the hydrodynamic drag from that, compute the energy lost to drag and average that to get the average power input. Then I point out how much more that is than the power needed to maintain a steady speed.

It’s a simplification and there are many factors involved. I am focusing on this one issue because it seems to be getting less attention than it deserves. Various documents I can find on the Net address issues such as rowing technique in great detail, looking for fractions of a percent improvement.

The extra power required to maintain the average speed of the boat due to speed variations is considerable, we are speaking of double percentage figures here! With that much potential saving, there are dramatic improvements in speed to be obtained. The improvements may be even greater than I have suggested, as reducing the speed variation should also reduce pitching.

The speed variation is mostly caused by the moving masses of the rowers. This in turn results from the system of oars, outriggers, sliding seat etc., in which the rower is operating. A different arrangement should be able to reduce the surging and still retain the efficiency of the rowing system.

I am not involved in rowing (except as a means to get a becalmed sailboat back to the launching ramp) so perhaps there are people addressing this issue and devices already available. Of course, competition rules may also be preventing the development.

 

Ok, now I understand better the sense of your calculation in the post #844.
But I was not refering to "moving masses". It is not the same thing as "added mass". The latter exists even if there are no moving masses on board.

Consider a body accelerating through a fluid along a rectilinear path, with a constant acceleration "a". You would expect that a force necessary to mantain that acceleration is:
Ftot = Drag + M_b*a
where Drag is the drag of the body (function of speed) and M_b*a is the inertial force due to acceleration. M_b is the mass of the body, which can be either fixed or moving (like in case of a rowing shell),

But if you try to measure the actual force necessary to move the body, you'll discover that it is different from what it should be according to the above equation.
The measured force will be bigger, because you are accelerating not only the body, but also the water around it. In other words, the body behaves as if it's mass had been increased by a certain amount, called "added mass" (I'll indicate it with "M_added"). This fictious added body mass accounts for the inertial force due to the accelerated water.

The sum of the body mass plus the added mass is called "apparent mass" and the above equation becomes:
Ftot = Drag + M_app*a,
where M_app = M_b + M_added.
You can see that the total force has increased by a quantity M_added*a, so the required power will also increase by a quantity M_added*a*V.

But that's just the case of steady forward rectilinear acceleration. The rowing shell also heaves and pitches, all of which displaces and accelerates water - which means more added masses to take into consideration.

Hope that makes my previous post more clear.

Cheers!

 

Slavi,
You made some good observations.
To get a rough estimate of the surge added mass in deep water, you can use the simple results for a Rankine ovoid :

madded/D = B/(3L-B)

where

madded = added mass
D = displaced mass
B=beam
L=length

For Olympic rowing shells the fraction madded/D is very small.

Pitch and heave are more complicated, but they are also quite small. The differences between two different very slender hulls of similar length and beam are hardly worth considering.

Leo.

 

Sliding riggers. And yes, they are banned in competition.

When Leo posted his graphs I figured something like this must be happening, although I hadn't heard of it before. From our points of view the problem with it is that calculating it accurately is likely to be an intractable problem, and even if we could get a reasonably accurate solution I'm not sure how much help it would be. I think it's probably one of those things you can't do much about, so there would be little or no scope for optimisation.

 

Ahah! Typical! Competition is supposed to about "testing improvements in a competitive environment" until a real improvement shows up and suddenly it is undesirable and banned! I suppose the forward-facing rowing rig is also banned for being too safe!

A good shell will beat a good kayak and a hydroplaning kayak is already available. The pitching of a standard rowing shell would be enough to destabilize a hydrofoil setup but with the sliding rigger ...

Of course as the boat rode higher on the foils it would increase the oar angle to the horizontal, but the same happens to the paddle in the kayak. Problem should be soluble with a bit of thought and development - it would go like stink. If the public got to see one cruising past an olympic shell folks would want to see it in the Olympics, just like they wanted to see snowboarding and - no doubt -will soon get to see women's ski-jumping ...
but I guess
that I digress

 

The main reason for banning sliding riggers is that all rowing shells would have to be changed to remain competitive.

I agree with you about introducing faster boats and I think they will force their way in when younger people have more of a say. I thought the BMX was the best event to watch at the 2008 Olympics. I kept thinking that youngsters had to invent it as a way of showing old fogies that cycling could be fun. We need to introduce the same attitude into boat events.