[REQ] New Cavitation...

[Capt Jack Harkness]

Can anybody out there rig up some more realistic cavitation? First off, the bubbles should really originate from one of the blade tips, not the hub. Second, cavitation should be a stream of bubbles, not a milky cloud. For excellent examples of what I'm talking about, check out the movies K-19: The Widowmaker and U-571.

[swdw]

whether it's a stream of bubbles or a milky cloud depends on depth, amount of cavitation, speed of prop in relation to current speed, etc.

Just cuz hollywood chooses to show a stream of bubbles doesn't mean that's how it is. You can hear explosions in space in movies too . . . .

You are correct though in the fact that the bubbles appear near the tip of the prop.

[Capt Jack Harkness]

Well, since a picture's worth a thousand words, here's some real cavitation for ya.









So to get realistic cavitation, we need to have really, really tiny bubbles generated really fast (maybe as low as 30 bubbles per second, but ideally 200+ per second) near the prop tips, hopefully synchronized with prop rotation.

[castorp345]

Originally Posted by Capt Jack Harkness

So to get realistic cavitation, we need to have really, really tiny bubbles generated really fast (maybe as low as 30 bubbles per second, but ideally 200+ per second) near the prop tips, hopefully synchronized with prop rotation.

'sounds like an fps-killer to me...

[Capt Jack Harkness]

Yeah... I can't think of any better way to do it, though. Unless we could make 3D cavitation???

[andycaccia]

Maybe we could re-design the actual textures and make them look like the photos Capt Jack has posted...if possible, of course.:hmm:

[Capt Jack Harkness]

Right, well the only way to do that is to make it 3D. Take two copies of textures of a strand of bubbles arrange them in a "+" (making the trails visible from all sides) and spiral this around making one helix per blade and fix there rotation with prop speed (so that they stay connected to the blade tips). Then you could chop it up into a handfull of sections that appear in sequence as speed increases, thus lenthening the trails. It would probably use a lot less processing power to do it this way instead of making millions of individual bubbles.

Of course this is all academic unless someone can actually get it to work...

[vanjast]

AFAIK, propellor cavitation is the collapsing of a vacuum formed around the surface/trailing edges of a fast moving prop. Not really bubbles in the true sense, but looks like it.

This cavitation, is destructive for the prop and it's associated mechanisms.

[Capt Jack Harkness]

Well, technically cavitation is the formation of the vapor cavities in the first place as the pressure on the backside of a blade drops below the vapor point of water.

Root cavitation causes relatively large vapor cavities near the hub that collapse roughly once per revolution, making a beating sound. Root cavitation causes significant losses in efficiency, makes a lot of noise, and seriously erodes the prop over time (sometimes causing the loss of a blade).

Submarine propellers in the 1940s did not suffer from root cavitation as it was a well known phenomenon at that time. They did, however, suffer from tip-vortex cavitation, in which long thin vapor trails form in vortices at the blade tips (as seen in the pictures); this is the same effect you occasionally see at the tips of aircraft wings. Tip cavitation creates noise and erodes the prop, but far less than root cavitation. It also creates a distinctive swishing sound, once per blade per revolution.

Modern submarine propellers suffer from neither form of cavitation unless using full power at or near the surface...

[andycaccia]

Moreover, in K-19 the cavitation effect is somewhat "glossy" or polished...as it was chromed, if you understand my meaning.:hmm:

[AVGWarhawk]

If anything, the cavitation noise should be added to the game IMHO. I can live without the visuals. I often play 100% and would like the actual cavitation noise off the screws of vessels.

[swdw]

Ok, nice pics, but there is an issue, this is tip cavitation created in a cavitation tunnel with no flow disturbance due to current, flow distortion from a hull, etc. Plus, these tests were designed to produce stream cavitation at the tip- so they got what they wanted. These are not tests to see what happens under any conditions, but carefully controlled experiments.

First go here:
http://metocph.nmci.navy.mil/KBay/backgroundnoise.htm

Lookat the diagram of sheet cavitation and also remember the bubbles are disturbed from the perfect streams in your picture by the flow of water being affected by a number of variables.

Also, pressure changes the characteristics of cavitation and where it appears on the blade-i.e. depth.

You can have tip cavitation, sheet cavitation, or cloud cavitation occuring on a propeller.
Here's a basic reference:
https://www2.hcmut.edu.vn/~dmthien/course/lythuyettau2/w11(suc%20ben%20xamthuc)/cavitation.pdf

What the pictures you posted and most things you find on the web do not deal with is the effect of shear forces created by unsteady flow disturbing the path of the cavitating buubles. Such forces create local changes in pressure that can cause the collapse or expansion of cavitation bubbles and will also mix the bubble streams, which, especially in the case of cloud cavitation, causes the effect you say is not realistic.

Here's a pic from a CFD analysis around a nuke submarine which has less flow disturbance created by the hull than a WW2 sub. Even this pic does not illustrate the disturbance from ocean current, wave action, etc.

[swdw]

Originally Posted by Capt Jack Harkness

Modern submarine propellers suffer from neither form of cavitation unless using full power at or near the surface...

Really? You might get a little disagreement from some sub sailors as there are other circumstances when you can cause cavitation with the new props even at depth . Course, having been on 2 fast boats and one boomer, what would I know.

[Capt Jack Harkness]

Well, how about this: "submarine cavitation, which is relatively slight in shallow water, can be almost completely eliminated in deeper water." - Naval Maritime Forecast Center/Joint Typhoon Warning Center

Anyways, getting back to the topic, cavitation in-game looks bad. I think something should be done about it. There are numerous ways to make it prettier, none of which are entirely accurate (SH4 is not a fluid dynamics program), but they would nonetheless add more eye-candy to an already pretty game.

By the way, does anyone have an idea what the cavitation really looked like on both the three- and four-bladed props on our fleet boats during normal cruising both surfaced and submerged?

[swdw]

Originally Posted by Capt Jack Harkness

Well, how about this: "submarine cavitation, which is relatively slight in shallow water, can be almost completely eliminated in deeper water." - Naval Maritime Forecast Center/Joint Typhoon Warning Center

Sigh, this is true under normal operating conditions but does not apply to large changes in prop speed in relation to the hull speed when doing things like taking evasive action. A knuckle in the water is created by a combination of prop cavitation and cavitation off the control surfaces due to a high speed underwater turn. It can even be heard on passive sonar because of the noise being created by the bubbles collapsing.

By the way, does anyone have an idea what the cavitation really looked like on both the three- and four-bladed props on our fleet boats during normal cruising both surfaced and submerged?

There shouldn't be any visible when "normal" cruising submerged. Cavitation creates noise as you mentioned so submarines try to keep away from it. Controlled changes in bell speed help reduce the chance of cavitation.