Power Required

That is quite pessimistic!

Drag will be mostly frictional and an easy way to estimate the drag is to multiply the flat plate frictional drag with a form factor. For a keel bulb it is assumed to be 1+1.5*thickness ratio, thus for 20% thickness it would be 1.3.

A 1.25 m * 0.25 m cylinder has a surface area of 1.0 m2. A rotated airfoil has a bit less say 0.7 m2. The frictional drag coefficient at 8 kn is about 0.0034 for a smooth surface. At 4 kn it is 0.0038 so you can almost assume it to be constant.

Now the frictional drag F=0.5*rho*Cf*A*V^2 and total drag is about 30% more due to form drag.

Thus at 8 kn the drag is 20 N friction and 26 N friction + form and at 4 kn 5.6 N / 7.3 N.

The above assumes a turbulent conditions. If you are able to keep the front part laminar, the total drag will be less.

Surface roughness is at least as critical for the turbulent conditions as it is for laminar. In laminar conditions roughness does not increase drag, but it may induce turbulence. In turbulent conditions roughness increases drag, if it is not below limit. At 8 kn you need to have roughness below ~20 um, which means sanding with about P400 paper. At 100 um roughness the drag will be about double at 8 kn.

 

Mmmmm.....and when you have done all this fiddling, you add some kind of propulsion device aft; all the beautiful theoretical streamlines have changed again......!

The Lutz et al paper is fine for a towed vehicle or one moving due to gravity, but these laminar flow shapes are extremely sensitive to non-optimum operation, such as yawing or a propeller flow. As Jehardiman notes, the more classical shapes with turbulent flow are far more tolerant in this respect.

 

Hello,
I am for a classical approach - drag coefficients for slender body of revolution from the printed literature, plus appendage drag modelled as friction drag of a double-faced flat plates. Hoerner is ok, and is full of graphs, but there are other sources. NACA measurements for airplane nacelles or for airship hulls is a good source, for example, and many papers are available on-line.

Laminar flow at this model scale would be unstable and could lead to a course instability of the UAV, imho. So I believe you should go for a fully turbulent flow, or a laminar flow with a forced (and hence controlled) transition point.
Also, in order to simplify the construction (and also the modularity of the vehicle) you might consider a torpedo hull shape, rather than a more complex airfoil-like or laminar shape.

A clean streamlined body with length-to-diameter ratio of 5 (but it is not the optimum ratio - 6 would be better) should have a Cd of around 0.065. At a speed of 8 kts it gives a body drag of 28 N. To this you should add a drag due to appendages, as said before.

If your AUV will have to maneuver (rotate and/or accellerate) at a given rate, then you should also consider the added mass drag, which is an apparent mass increase due to unsteady inertial hydrodynamic forces. The added mass is somewhat more difficult to calculate, but there are techniques to do that (slender body theory).

 

As I stated at the beginning - 50N is a nice conservative figure to work with if you are deep enough not to worry about wave making!!!

Rick W

 

The REMUS is highly developed, having said that, it does not fulfill all current mission criteria.
And if you look back at the American auto industry of the 1970s you might have come to the conclusion that no one could penetrate their market.

 

I totally agree that REMUS is not a do all, but a PRIUS is not a duce-and-1/2, and who in the 70's would have seen that the pick-up/SUV would be 1/2 the market now because cars got too small to accomplish the mission (transport a well feed American or family) given the design breif (increase fuel economy). One of the important things in my job (US Navy RDT&E) is to tell the sponsor what is, and what is not, possible near term given the mission requirements....and when a contractor is promising something that can't be delivered.

In the end, cars, boats, and super-models are all the same....fashion dictates that the body styles and size change from year-to-year, but under the paint the running gear evolves very slowly.

To Joakim:
Here is a fun exercise...Consider your last sentence on surface roughness then do a survey of the size of suspended biomaterial/biota in the nearshore or benthic environment. Also look at the turbulence in the water mass itself from some of the circulation current studies (and we will ignore wave orbitals which is a whole other matter...). The water itself is against you. Laminar flow for vessels looks good on paper and in simulations, but as baeckmo and daiquiri (as well as Hoerner) point out, real life body and environmental conditions render it moot.

 

I was pointing out the importance of surface roughness to drag caused by turbulent flow. I was not saying that smooth surface would keep the boundary layer laminar. Or are you saying that surface roughness is not important in real life? I calculated almost identical drag to daiquiri and that is based on a smooth surface without laminar flow.

 

I realize that this thread was intended to be a simple question on power consumption, but it has broadened in scope, to include hull shapes as well.
So, on that expanded topic, I will bring up a interesting observation as well.
It turns out that a sailfish is the fastest marine animal with a recorded top speed of 68 mph. For the fastest animals, the speed reduction from the first one in the list, to the eleventh one, is nearly a factor of 2; however they all have some things in common.

All 10 have a high length to width ratio in their tail fins.
All of them resemble airplanes (no coincidence, I imagine).

None of them are completely revolved around an axis.

Obviously there are many other factors to consider in a fish or a sub, besides flow resistance.

I find it interesting though, that the “thinnest” of the bunch, is also the fastest.


Sailfish (Istiophorus platypterus), leaping 68
Swordfish (Xiphias gladius), leaping 60
Marlin (Makaira), leaping 50
Wahoo (Acanthocybium solandri), leaping 47.88
Yellowfin Tuna (Thunnus albacares), leaping 46.35
Blue-fin Tuna (Thunnus thynnus), leaping 43.4
Albacore (Thunnus alalunga), leaping 40
Bonito (Sarda), leaping 40
Mahi Mahi (Coryphaena hippurus), leaping 37
Flying Fish (Exocoetidae), gliding 35
Killer Whale (Orcinus orca) 34.5
Dall's Porpoise (Phocaenoides dalli), leaping 34.5
Shortfin Mako (Isurus oxyrinchus) 31

They have a lot in common and a lot of differences.

Notice the top three all have bills, and flatter than the rest. Of course a flounder is flat, a ray is flat, and neither of them is very fast.

So obviously the propulsion is more important than nuances in hull flow dynamics.

Take a look here to see the different body types.

http://www.einfopedia.com/fastest-fishes-of-the-world-top-ten-sharpest-marine-animals.php

 

It seemed to me, he was referring to factors other than laminar flow would have an overall larger impact on performance.

 

About your question relative to fish swimming speed...

There are many things about the fish hydrodynamics that our science is trying to understand, but we are still far from the goal.
There is one thing that all the members on your list share in common: they are living creatures made of flesh, bones, nerves etc., perfectioned by the nature through millions of years of evolution. This brings them an distinctive advantage we are still far from reaching - the possibility to actively modify the waterflow around their bodies, down to the microscopic level.
Their skin is full of receptors (nerve terminals) which sense the waterflow from big vortex and streamlines, down to a tiniest turbulence. They act as a feedback for the rest of the mechanism, which automatically adaptst to the flow field around it. The outer skin is made of small vortex generators (scales in case of fish and dermal denticles in case of sharks - which are not fish, ok) tightly intertwined with lubricating devices (mucous glands). Acting together they can modify the boundary layer condition in any single point of their body surface, through mechanisms we are still struggling to grasp. The skin itself lays upon an elastic layer of muscles, storing and releasing the elastic energy on demand during strokes.

It actually appears that the whole fish body, apart from obviously being a hull, is one single propulsive device which is instantaneously adapting to the unsteady flowfield around it, squeezing everything that can be squeezed out of it in therms of propulsive efficiency. So the hull and the propulsion are very efficiently blended to form one single device, when considering fish and other marine fast-swimmers.

On the other hand, we are still stuck with rigid inanimate hulls and comparably simple propulsive devices which, in most cases, can be optimized for only one single flow regime... That is basically why we still have an enormously long way to go until we reach the nature, imho. It is basically our ignorance in the field of hydrodynamics and our inability to make efficient morphing mechanical devices that makes us separate the propulsion device from the hull device and from all the the rest of devices on the ship.

But I am sure that sooner or later we will arrive to emulate nature much better, thus saving God knows how much energy which we are wasting right now. I am actually pretty much convinced (though we all will hardly arrive to see it...) that we will one day arrive to grow our ships and airplanes, like we do with trees.

You can google out the terms "fish hydrodynamics" and will find out many interesting facts about it. Some of them are here:
http://www.people.fas.harvard.edu/~glauder/reprints_unzipped/LauderMadden2006.pdf
http://jeb.biologists.org/cgi/reprint/199/10/2139.pdf
http://jeb.biologists.org/cgi/reprint/203/2/193.pdf
http://www.ias.ac.in/currsci/sep252000/carpenter.PDF
etc...

 

Given all that; I was not comparing animals to rigid hulls. I was comparing one species to another, and trying to distinguish obvious differences in shapes to see if there is a patter to faster species relative to slower ones.
As I said earlier, all of the fastest animals have some distinctive similarities.
1. High aspect ratio tails
2. Shapes generally thought of as streamlined.
3. Within the category of fast fish, bills are fastest.
Are bills fastest because they’re bills or is it because they’re thin? Dolphins (Mahi) are thin but not as fast as bill fish…etc.
Aside from all of the biological magic; what makes one fish faster than another.

One more thing, when did sharks get reclassified? I always thought they were boneless, and bladderless, fish.

 

True, surface roughness is important...up to a (seperation) point.

Model testing and the Schoenherr line, have shown that there is a problem with predicting laminar flow attachment. Seperation in clean undisturbed water begins about Rn 1x10^5 and flow is fully turbulent by Rn 2x10^6. Hoerner in his discussion of laminar flow goes on to point out that laminar flow is never observed above Rn ~ 1x10^5 if cross flow above ~10% Va is observed. So for bodies of moderate size,( say 4m long by 1m dia @ 5 knots, i.e. a human powered submarine because I know this off the top of my head) laminar flow is only possible in the first 10-15 cm of the body. After that it is turbulent. Once the turbulent boundary has developed, absolute surface roughness is not nearly as important as boundary layer attachment. Like the dimples in a golf ball, the bound layer can significantly reduce pressure drag, sometimes more than the surface roughness increases apparent skin friction due to it size. Aspect ratio and forebody/afterbody shape is very important in this so there are no hard and fast rules.

Additionally, there is the whole matter of propulsion and control. The adaptation of a propeller to the wake behind a body of revolution is a significant matter to the overall efficency of propulsion. Bodies that have minimum self drag may not have minimum propulsive power requirements. It is a demonstrated fact the laminar shaped bodies have less control stability which require larger control surfaces exiting the layer. Again, this is a manipulation of the boundary layer thickness and seperation point to get a desired effect.

Finally, fishes are the expert in boundary layer manipulation for reduced drag and propulsion. They are a whole different topic altogether because as daiquiri points they aren't rigid and can actively control thier skin roughness depending on need.

 

So back to the original.
We have not agreed on which design philosophy is more efficient.
We don’t agree what method to use to create a design and test it.
We do agree that other design criterion needs to be considered.
Do we agree that somewhere in the middle a hull of this size will consume 230 watts at 8 knots?\
Do we agree that power consumption increases roughly at the cube of velocity?
I’m lending towards following Rick on this last part.
I see I have a long ways to go.
Should we start a new thread on actual hull design or do you guys prefer to continue on the same thread even if it gets off topic.

 

Bare with me on this fish thing for one second.

All things being equal. What makes one fish faster than another.

I get it that they are more effecient than rigid hulls, but that does not preclude them from being part of a design evolution.

We're starting from scratch here. why not try to achive the most effecient form posible?

 

Yes, I agree, that the laminar region is quite small. For a sailboat it could be up to 0.5 m or so.

I don't think separation is an issue for a well designed profile at small angles of attack. A golf ball is a totally different case, since it has a huge form drag compared to friction drag. For a airfoil profile (rotated or not) the friction is much bigger than form drag, thus it is (almost) all about reducing friction. For that surface roughness is most important.

I once compared the roughness needed to induce a turbulent boundary layer to the roughness needed to increase the drag of a turbulent boundary layer and found out that the latter was smaller for a flat plate at sailboat size and speed. Thus surface roughness requirement for minimal drag can be more important for the turbulent region than for the laminar region.

I was involved in a national project which studied the use of drag reducing agents (DRA) to reduce the pressure losses (caused by turbulent fricton) in pipes (e.g oil pipes). Even at few ppm level the friction was lowered 10-50% by this polymer. It consists of long molecular chains that are assumed to dampen turbulence in the boundary layer.