Sub physics - experts needed
 

I'm tuning my first simple movement model for my subsim. I'm trying to set it to some realistic behavior. There is however too many things I don't know. So I have few questions.

I found on internet some data about subs, for example good old 688. I have weight, size, engine power, maximum speed. Now this alone is almost enough for much of the simulation. But not all, and I also need as much real data as possible to see if my model indeed is realistic. I'd like to know:

1) Acceleration at different speeds, especially at 0.
2) How long the subs takes to stop after engines are cut off.
3) Does sub have screw with variable pitch ? How is it used exactly ?
4) What is the turning rate at different speeds ? How much rudder is needed to achieve this ?
5) Why do subs have lower max speed on the surface ?
6) How is the depth controlled ? How much does the tanks, how much the planes ? How strong effect trim tanks has ? How are they used ? How are the common surface and dive procedures from these aspects ?

Please don't wild guess. I can do that too. I want some real info, with references, if possible.

Wouldn't regard myself as an expert by any means, but I do have some info and knowledge that is relevant, so here is that. But first, some pictures you might find useful:

http://i78.photobucket.com/albums/j1...adbury/LA2.jpg

http://i78.photobucket.com/albums/j1...adbury/LA3.jpg

http://i78.photobucket.com/albums/j1...adbury/LA1.jpg

Next some info on why subs are slower on the surface:

Three main types of drag act on a body going through a fluid. The most obvious one is drag caused by the actual shape, ie. Structural Drag, which is why subs, aeroplanes, cars etc are streamlined in an effort to reduce this. Next you have Form Drag, which is caused by low pressure from the fluid passing over the surface of the vessel, this creates low pressure over the surface of the sub as the molecules speed up and are strung out further apart, thus reducing pressure (this is what an aeroplane's aerofoil exploits to create lift, the technical name being the bernoulli principle, you can demonstrate this yourself by holding the back of a spoon in running tapwater, the curved back of the spoon will be pulled into the flow). Last you have skin drag, this is where the boundary layer (i.e the thin stratum of water close to the sub's surface) can erupt into tiny eddies and bubbles; these disturb the laminar flow, which slows things down. There have been experiments to try and inject polymers into the water which hug the skin of the sub and make it more slippery by improving the laminar flow, but there's not much data on the success or failure of this. Western analysts thought that was what the twin lines down the surface of the 'Akula' were when they first saw them, of course they aren't, they are simply safety line attachment rails for crew members on the deck.

Anyway, what all that technobabble above basically means is, that the flow over the sub is hugely disturbed when it is on the surface, and that makes it prone to massive amounts of drag in the disturbed flow of the water. Movement through this turbulent water probably also affects the efficiency of the screw too, as that would not be getting a clean flow either. Also interesting to note is that a submarine actually gets smaller (and therefore less prone to drag, though not by much) when it dives deep, in the case of a LA Class sub, it apparently shrinks by as much as 750 cubic feet when going deep as the water pressure compresses it.

Some other stuff on the LA Class sub:

The LA class sub has five main ballast tanks, when all those are flooded, they add 800 tons to the boat's weight of 6,100 tons, so of course it sinks to a certain depth based on how much water is in the tanks, but typically, the boat is set for negative buoyancy and held at a specific depth by the planes (i.e. dynamically), this is so that it will not accidentally rise if it slows to a speed where the planes become ineffective. Typical dive angles incidentally, are between 3 and 5 degrees. As far as I'm aware the screw's pitch is fixed (but someone may know better, as they are generally a closely guarded secret), one interesting fact that I do know though is that screws with odd numbers of blades are quieter than those with even numbers of blades, as they are less likely to set up harmonic sound waves, which is why you most often see seven blades on a sub screw.

:D Chock

Thanks .. many new things for me. Like the screw blades count. I guess it should be prime number .. so there are not subharmonics between main shaft and blades frequency.

So main tanks are completely filled with water, trim tanks are used to set buoancy to slightly negative and to keep sub level at low speeds. Main ballast tanks are blown from shallow depth.

And sub on the surface is slower because of turbulent flow and lower screw efficiency, both pretty good reasons (btw. I love the wake LA makes on the surface, still thinking about how to make it :rotfl:, so far no working idea)

In which book ya got these cool pictures? :hmm:

They are taken from Hunters of The Deep (ISBN 1-84447-222-1)

:D Chock

Here is some good technical material that may help you in your project:

http://ocw.mit.edu/OcwWeb/Mechanical...otes/index.htm

and

http://web.nps.navy.mil/~me/tsse/NavArchWeb/1/toc.htm


I'd be interested in sharing notes with you, as I started a similar project a couple of years ago. For example, what value are you assuming for drag coefficient? I calculate that a sub with 15,000 shaft horsepower, 25-kts max speed and 10-m hull diameter should have an estimated drag coeff of 0.15. Given that, you should be able to answer your question, "How long the subs takes to stop after engines are cut off?"

Thanks for links ! Tons of information :up:

Yes, I can compute drag from power and max speed, my results for LA is 0.23. This does not include the frontal area of fins or sail, so in the end it will be little smaller. And that is at max speed with some possible turbulence, so at lower speed it can be less. Anyway some real life data would be helpfull to know 'it fits'.

For now the most 'blank areas' in my math are: acceleration at zero speed and turn rate.
For acceleration at zero speed I now use simple estimate from max speed. The force needed to keep sub at the max speed is simple: power/speed. I use this force for all speeds. It gives acceleration of 0.017g. It takes about 25 seconds to reach 32 kts.

Horizontal movement is quite alright, I mean simmilar to DW. I'm tuning diving now. I'd like to add some simple particle syystem so sub movement is more obvious and then I'll make this public.

My drag estimate was based on Sturgeon-class SSN for size, power and speed. I would expect the Sturgeon-class to have a lower drag coefficient than the LA, because it has a lower length-to-beam ratio.

Max acceleration, of course, would mean cavitating at shallow depths. If you limit prop speed to below the cavitation threshold, acceleration to flank speed will take longer, of course. Somewhere I have notes on that and I even wrote a program. I no longer have the latter, but I seem to recall acceleration to flank from a dead stop took around 60 seconds without cavitating.

Sounds like we're both in the same ballpark.

Hey Bill (I mean hail you O big sub guru) you SHOULD know about the screw. Is it variable pitch ? Will you tell ?

Nope, its not variable.

Good .. that simplifies some things.

Yet another problem. How fast does the screw rotate ? In DW there are numbers like 7 for TPK, with 30 knots it gives 210 revolutions per second. Seems too much to me.

It could be true if "Turns" referred to turns per seconds.

TPK is the propeller efficiency, I don't have the real algorithm, but in game it seems to be calculated as follow :

Speed = ( Freq * 15 ) / TPK
or
Freq = ( Speed * TPK ) / 15
or
TPK = ( Freq * 15 ) / Speed

So TPK may be "Turns(per 1/4 minute = 15*Freq = rpm/4) Per Knot"

Hm ! Good point .. I always thought that you obtain speed as demon frequency*TPK, but that is not true. Your formulas seems to be correct.

So in DW, 688 with TPK=7 at 30 KTS would have 14 turns per second, at 5 KTS it would be 2.33 turns per second, which sounds much better.

See link below. This book does a nice job of explaining much of the complexity associated with trim systems and control surfaces.

Complete Idiots Guide to Submarines: http://www.amazon.com/Complete-Idiot...6082917&sr=8-1

^^^^
That's a hard to find book. I can't even find it on eBay.