CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   bubble dynamics (

ndabir August 16, 2011 16:32

bubble dynamics
Hello everyone,

I would be very happy if anyone can help me with this question. I would like to simulate a single bubble behavior in water under ultrasonic sound wave. This behavior consists of growth and collapse of bubble. Does anybody know cfx has this capability? anybody had similar experience with cfx?

I am not sure whether to use cfx or Fluent for this simulation? Does anybody know which one would be more powerful?

Thank you,


ghorrocks August 16, 2011 18:40

I suspect both CFX and Fluent can do it. I have done similar modelling and used fluent for it but I suspect either could do it. In my experience micro bubble collapse and expansion is a very tricky thing to model as you can get negative absolute pressures - and the problem is this is physically correct and I do not know of any CFD code which can handle negative absolute pressures.

ndabir August 17, 2011 11:00

Thanks for your reply. Do you know if I want to write some code in cfx, which language I should use? C or Fortran? because I think the problem discussed above cannot be modeled only by the use of software itself and I need to do some modifications in the software.

ghorrocks August 17, 2011 18:44

There should be no need for C or fortran. You should be able to set this up using the normal CEL and CCL.

ndabir November 17, 2014 18:44

Hi Glenn,

I started this post almost 3 years ago and now I was reviewing it. You mentioned that a negative pressure might happen during bubble growth and collapse simulation. I should say you were absolutely right! I am facing those negative values.

Do you know what is the reason for these negative values? Is it because of the process of vaporization of water?

Do you know how is it possible to overcome this issue?

ghorrocks November 17, 2014 20:06

The physical process which can lead to negative absolute pressures is the delay caused by phase change. When a very fast pressure transient drops the pressure in a liquid the phase change process needs to nucleate and draw in energy to undergo the phase change. This happens very fast, but not instantaneously. So for a small period of time a negative absolute pressure can exist.

Alternately, imagine the stresses in a beam under load. On the tensile side of the beam the stresses will be negative and probably enough to cause negative absolute pressure. The same thing can happen in liquids.

As it is a real physical process it cannot be "overcome". Ideally you can develop a CFD solver which can handle negative absolute pressures but the applications where it occurs are quite rare so few people are interested in it. Certainly all the commercial CFD codes I am aware of just assume if the pressure goes negative there must be an error. So you might have luck with writing your own solver which handles negative absolute pressures better. I have tried weird fluid EOS where I tried to allow the liquid to tend towards zero pressure gracefully and reduce density artificially to keep the pressure positive but I never got it working.

This is a very tricky area and I have not found a solution to it - and would be very interested to hear if you find a solution to it.

ndabir November 17, 2014 20:27

Thanks for your detail explanation. That makes sense.
This problem mainly arises after the bubble completely collapses and the bubble starts to rebound.

I use VOF model with very small time steps (less than 1 nanosecond) and try to use fine mesh but the problem still remains. I am not sure if accounting for mass transfer over the surface of the bubble will solve this issue or not. Do you think accounting for evaporation and condensation will help?

ghorrocks November 17, 2014 20:33

I too was using sub-nanosecond time steps and had the problem at the collapse point.

Fine mesh will not help, it is not a mesh resolution issue. I do not see how evaporation and condensation will help either as they are slow processes. The only way I can think of to get this working on a normal CFD solver is to bend the EOS of the fluid to add artificial compressibility as the pressure approaches zero such that the pressure does not go negative. But as I said I have played with this a reasonable amount and not got it working yet.

All times are GMT -4. The time now is 01:29.