Understanding Geometry's Role in Causing Cavitation | Designing an Impeller

[Spimon]

So I'm designing an impeller and rather than how to prevent cavitation, I'd like to know how to cause cavitation.

I've read up on fluid shear stress and Couette flow, but this all appears to be based on relative velocity dV, of parallel discs, separated by a small distance dX, and fluid of viscosty μ. Shear Stress, τ = μ dV/dx.

I also read that cavitation occus when pressure exceeds vapour pressure for the given fluid a the given temperature (3.2 kN/m2 for water @ 25 oC). I'm happy to accept all this.

What I don't think I understand is how the geometry of the impeller affects cavitation.

Videos of a cavitating marine propeller show the cavitation bubbles all the way along the length of the leading edge, right from the hub to the tip, but not on the hub (despite the hub having the same relative velocity as the leading edge, close to the hub). This suggests to me that geometry does play a role in caviaton, not just relative velocity. I'd exect an impeller with a very square leading edge to cavitate at a lower relative veloity than a very sharp leading edge. Is this a reasonable thought?In water at 25 oC at 1500rpm, my 80mm impeller seems to produce a shear stress of 22.4 Pa
Nowhere near the 3'200 Pa required for cavitation. Further calculations seem to indicate that to get an 80mm impeller to cavitate would require it to rotate at 215'000 rpm! Surely cavitation is dependant on more than just diameter?

I have Solidworks Simulation and plan on verifying results with this, but I'd like to have some idea if what Solidworks tells me is correct.

Thanks for ANY help anyone can provide!

 

[256bits]

I also read that cavitation occus when pressure exceeds vapour pressure for the given fluid a the given temperature (3.2 kN/m2 for water @ 25 oC). I'm happy to accept all this.

The pressure has to be less than vapour pressure for a bubble to appear, and the bubble collapses when encountering an increase in pressure.

Videos of a cavitating marine propeller show the cavitation bubbles all the way along the length of the leading edge, right from the hub to the tip, but not on the hub (despite the hub having the same relative velocity as the leading edge, close to the hub). This suggests to me that geometry does play a role in caviaton, not just relative velocity. I'd exect an impeller with a very square leading edge to cavitate at a lower relative veloity than a very sharp leading edge. Is this a reasonable thought?

The pressure behind the leading edge of the propeller is below vapour pressure, so you get cavitation. Where on the hub do you expect the pressure to be below vapour pressure?
Square edge, round edge, knife edge - do not make assumptions that you cannot back up without some explanation or mathematics, as gut feeling can be most incorrect a lot of times.

You should maybe investigate and try to understand cavitaton by reading up a bit more on it.

 

[Pkruse]

This is weird. I've always worked hard to eliminate cavitation, or if I can't to mitigate the damages. Never gave a thought to causing it.

But it has been my experience that after you identify all the causes, the real world will show you some that you never thought of. My favorite is cavitation caused by vibrational modes in the fluid, or in the structure it flows through.

 

[Bobbywhy]

After spending over 20 years trying to avoid cavitation near the face of high power sonar transducers I developed a great curiosity about the phenomenon. High intensity acoustic pressure waves cause it, especially at shallow depths. Interestingly, we can detect the onset of cavitation in this "near field" because the bubbles cause a drastic change in the acoustic impedance and this "pulls" the output amplifier driving the transducer. Just by monitoring that current we can cut back on the driving power if cavitaion begins.

Here are a few bits and pieces for anyone who is interested:

"The snapping shrimp competes with much larger animals such as the Sperm Whale and Beluga Whale for the title of 'loudest animal in the sea'. The animal snaps a specialized claw shut to create a cavitation bubble that generates acoustic pressures of up to 80 kPa at a distance of 4 cm from the claw. As it extends out from the claw, the bubble reaches speeds of 60 miles per hour (97 km/h) and releases a sound reaching 218 decibels.[11] The pressure is strong enough to kill small fish.[12] It corresponds to a zero to peak pressure level of 218 decibels relative to one micropascal (dB re 1 μPa), equivalent to a zero to peak source level of 190 dB re 1 μPa at the standard reference distance of 1 m. Au and Banks measured peak to peak source levels between 185 and 190 dB re 1 μPa at 1 m, depending on the size of the claw.[13] Similar values are reported by Ferguson and Cleary.[14] The duration of the click is less than 1 millisecond.

The snap can also produce sonoluminescence from the collapsing cavitation bubble. As it collapses, the cavitation bubble reaches temperatures of over 5,000 K (4,700 °C).[15] In comparison, the surface temperature of the sun is estimated to be around 5,800 K (5,500 °C). The light is of lower intensity than the light produced by typical sonoluminescence and is not visible to the naked eye. It is most likely a by-product of the shock wave with no biological significance. However, it was the first known instance of an animal producing light by this effect. It has subsequently been discovered that another group of crustaceans, the mantis shrimp, contains species whose club-like forelimbs can strike so quickly and with such force as to induce sonoluminescent cavitation bubbles upon impact.[16]

The snapping is used for hunting (hence the alternative name "pistol shrimp"), as well as for communication. When feeding, the shrimp usually lies in an obscured spot, such as a burrow. The shrimp then extends its antennae outwards to determine if any fish are passing by. Once it feels movement, the shrimp inches out of its hiding place, pulls back its claw, and releases a "shot" which stuns the prey; the shrimp then pulls it to the burrow and feeds."
http://en.wikipedia.org/wiki/Alpheidae

Also, see: http://en.wikipedia.org/wiki/Cavitation

And this: http://en.wikipedia.org/wiki/Sonoluminescence

 

[Pkruse]

Wow. Thanks for posting that, Bobbywhy.

Sounds like how cavitation damage happens inside of a diesel engine, in the cooling jacket around the cylinder liners. Piston slap against the liner causes cavitation bubbles in the coolant, which eats holes though the liner. Also known as "liner rot."

They prevent that by adding chemicals to the coolant. I've always wondered how the chemicals could prevent the cavitation. He who does not maintain his antifreeze coolant gets water flowing into the combustion chamber and a very sweet smelling exhaust.

 

[Spimon]

Thanks for the replies everyone!
I'm certainly doing much more research (this is just a small part of it) and am currently setting up a test rig to try to get what I want happening.

Square edge, round edge, knife edge - do not make assumptions that you cannot back up without some explanation or mathematics, as gut feeling can be most incorrect a lot of times.

This is exactly what I'm trying to do :P Just the appropriate equations are hard to come by. I think my calculations are such a rough estimate that they're no longer worth considering. This is the tough bit :(

My best bet might be to find a CFD guru to assist with some modelling.

I realize the pressure has to be below vapour pressure (since corrected from my first post) for cavitation to occur, but I'm trying to determine the best 'shape' for the impeller to create these conditions, at the lowest possible angular velocity...without having a huge diameter...

The other main problem I'm having is that any impeller I try creates a vortex and introduces air this way. I think I need to eliminate this and reduce tank swirling before I can see what's going on.

 

[Bobbywhy]

Spimon,
Just one small point: You write that

"The other main problem I'm having is that any impeller I try creates a vortex and introduces air this way."

If I understand this correctly you are saying the cavitation bubbles contain air. That can only be true if air is dissolved in the water. When we "rip" a liquid apart to form a caviation bubble the gas inside it is whatever gas was dissolved in the water...not necessarily air.

 

[bigfooted]

You might be interested in the book of Brennen, which is available for free online at caltech
authors.library.caltech.edu/25017/1/cavbubdynam.pdf

 

[NeonfOx]

Simplest way, is run it far from BEP of the impeller (hp out/hp in)
or Drop the NPSHa (net positive suction head avaliable) below the NPSHr (net positive suction head required) of the impeller
by adding pressure reducer to suction, or pulling from negitive suction head

well that how to do it in Normal Pumps