Power Required

So you say don't worry about laminar flow and focus mostly on drag?

 

The fish leaping is quite a different situation to a deep water vehicle. Wave drag is a big factor with leaping. There will be less wave drag for a slender body. So be careful when comparing near surface operation with deep water operation- totally different situations.

The efficiency of a foil section is related to its aspect ratio of foil width to cord length. A high aspect foil is more efficient. Hence the reason you see wide but short tails on fast swimmers. You can use JavaFoil to compare the performance of same foils with different aspects. The options page enables you to set the aspect and it provides a correction for the induced drag.

Boat props are the way they are primarily for draft constraint and cavitation. They typically have much lower efficiency than aeroplane propellers. An efficient water propeller that is not concerned with cavitation and draft will look like an aeroplane propeller. You can use JavaProp to determine performance of such a propeller although it has a limited range of foils and Re# but enough to suit your requirements.

I personally think that laminar flow on a real hull is a pipe dream but it may be worth chasing. It can reduce drag markedly.

My actual experience with these hulls was tow testing small models and I built the boat pictured as an experiment. Wave drag dominated. It would have been impractical to set the hull deep enough to overcome wave drag. So my model testing was not highly relevant because I could easily set the hull deeper on the model. In fact all the tesing on the model was a diameter deeper than what I could practically use in my pedal boat scale.


Rick W

 

I assume you used local Rn calculations, but: Did you base your analysis on flat plate boundary layer growth? Did you consider velocity changes due to 3d form? Did you integrate drag by differential area along the length?

I'm curious because I have done the same analysis for submerged bodies of revolution in relationship to RANS analysis.

 

I think it is not diffucult to make a hull, that would have less than 30 N drag at 8 kn (not including appandages). Buying (or making) a propeller to reach the ~90% efficiency Rick used is much more difficult. But if you have a drag of 30 N, then 230 W is easily enough.

Yes power need inceases rather accurately V^3.

 

Everything is never equal...

One of big things in fish speed is power density and biomass. If you notice, the larger fish go faster, this is because they actually heat up which gets more power out of their muscles. On the flip side, they have to get rid of the heat buildup which can kill them, so they tend to need more surface area to volume. Hydrodynamics comes into play here, so you get long thin bodies on fish and shorter, broader bodies on marine mammals which are trying to hold heat in.

This is why nuclear/conventional submarines and fast long range torpedos use liquid fuels and look like they do...power density as well as lower drag at the speed they want to go for the volume. So depending on your power density, mission speed, range and loiter, and any mission sensors, your vehicle shape will change to optimize propulsion. Given that you may be size constrained, you may find yourself with an impossible task without a "Mr. Fusion" from the movie "Back to the Future".

 

The prop -

I have attached the original JavaProp design and one for an off-the-shelf prop that is pictured. It is difficult to buy a model plane prop that has a P/D >1. The APC pictured is 1 and is the highest I have seen with adequate strength for your application however the diameter is more than you specified. You can get 12X12 model plane props but they will struggle with the loading.

Bolly used to have a 15X25 prop but they will only make them in quantity lots now. I have one only. Again it is bigger diameter than you wanted.

It is not hard to make a prop that will get better than 80% efficiency for this application - about 4 hours of welding and shaping. I can also put you in contact with a firm that will mill to specific design.

Rick W

 

Here is some more information: http://www.boatdesign.net/forums/boat-design/golf-ball-boat-install-dimple-plate-20413-3.html

I'm not that deeply involved in the boundary layer physics, just enough to be able to use RANS CFD (mostly with wall functions and non-marine apliations) and some simpler calculations. A friend of mine developes RANS low Re turbulent models and I consulted him on how to estimate the roughness that would have an effect on the transition.

So what would be the difference between flat plate and a rotated airfoil? The velocity increases rapidly in the front part. That should stabilize the flow thus keeping it laminar longer, but does it also make the roughness needed for transition smaller due to higher velocity?

I did not do any drag comparisons. I just wanted to know the smoothness criteria for differen regions of a sailboat hull. What I did not find out was how the roughness criteria changes inside the turbulent region. Any reference on that?

 

Ahhm... Uuuups! Happens...

 

1) As about design philosophy, I think that we all here share the same thought about avoiding laminar hull in this case. Though it is also true that we don't really know what would the typical mission of you ROV look like... Apart that, it is up to you to decide will you go for a more hydrodynamically efficient 3D-airfoil shape, or for a simpler to make and more payload-flexible torpedo shape. It is hydrodynamic efficiency vs. constructive and operative practicity... A compromise, just like any design is.
2) Given an objective difficulty of translating data for 2D bodies obtained through Javafoil (or through some other airfoil database) into 3D flowfield around a body of revolution, and given the fact that literature is available with measured resistance curves of 3D bodies of shapes similar to yours, I suggest you the second approach. Again, it's up to you.
3) that is definitely an universal rule in any design process...
4) as a conservative preliminary value, I would say yes. Given the uncertainity at this stage, it could be anything between 150 and 250 W, but staying closer to the higher value can only make you good for now. You will know the final true only after a tank testing of your underwater ROV.
5) Yes, it is a straightforward math. Drag is equal to
D = rho/2 V^2 Af Cd (where Af is the frontal area of your ROV)
or also, Joakim's method:
D = rho/ V^2 Aw Cf (where Aw is wetted area of ROV's hull).
Since power is equal to D*V, it means that it grows proportional to V^3 , if fully turbulent flow is assumed.

 

That's nice, I didn't think of that...

 

Thanks Rick,

So your gut feeling is that a 15 X 24 would be better. It would need to be optimized for 2 to 4 knots. The 8 kt figure is for max speed and not often used.
When it gets down to it I have a CNC mill so that won't be much a problem.
My inclination would be to build it out of maple and sandwich it with carbon reinforced epoxy.

Next on the list, motors, fortunately I have a little knowledge of that.

 

I had considered elements like muscle mass and muscle flinch frequency.
However, and I apologies in advance for drawing out this fishy discussion.

BUT...
A Wahoo and a sailfish are very similar in both size and shape, but a sailfish is way faster.

Personally, I think it's the tail, but it might be that big ole pole coming out of his snout.

There has to be more to it than size that makes bill fish so much faster than the rest.

What do you think is the reason a sailfish is faster than a Wahoo?
I’ve caught a few of both.

Now I’ll eat a Wahoo, but I would prefer to catch a sailfish. One of these days maybe; on a fly rod.

 

The attached image is one I had milled. The P/D is about 2 - quite aggressive pitch for easily driven boat with similar drag/speed to what you are considering. This prop has efficiency around 85%.

It took a few attempts to add enough meat for tool passing from my design but it worked out well. This one is aluminium and was linished by hand to remove the tool marks rather than going to fine finish on the mill. That saved a lot of machine time.

Anyhow it shows you what a high efficiency marine prop looks like for these low Re# applications.

Rick W

 

Or Aluminum.
That looks good.
I dont think it would be practical to attemp to get it that smooth in a mill.

 

Is there any reason why people don’t use elliptical shapes in a cross section for underwater hulls?
Some of the design elements would lend themselves to fitting better in an oval space, rather than a round one.