Moving boundary?
 

Gas with high volocity & pressure || lquid slug

The problem is to calculate the impact of slug to the pipe wall of a 90 degree bend. I set a simple model as follows:

For gas:

rho0_t + (rho0*v)_x = 0;

(rho0*v)_t + (rho0*v*v+p)_x = 0.

Let u be the mean velocity of the liquid slug and assume u equals to the velocity of gas at downsteam, so that, for liquid slug,

du/dt=p/(rho1*L), where L is the lengh of slug, p is the pressure of gas at downstream.

I will use ROE's scheme to solve the conservation laws, but I don't know how to deal with the moving boundary for gas. I know that there are various methods to tracking the interface such as VOF method, Level set method and moving mesh method. As I am a newbie in CFD, could anyone tell me which method is relatively easier to solve this problem and, suggest some reference. I really appreciate any comments on the problem.

regards,

Hao Meng

Re: Moving boundary?
 

For a start calculate approximately how far the gases might move and define a bigger domain with appropriate conditions

Re: Moving boundary?
 

Thanks for your response, But I am afraid i havn't catched your point, what do you mean defining a bigger domain?

Do you mean to treat both gas and liquid slug as one fluid and just set wall boundary? I really appreciate your more detail.

Re: Moving boundary?
 

You can start doing that for the beginning treating them as the same fluid with say some mean values.

As far a s domain is considered what i meant was adding an artificial domain outside your normal domain of calculation. This is referred in the literature as perfectly matched layer. You gradually disssipate your solution in the perfectly matched layer using some sort of filter and an exponentially decaying function. When the gas enters into that domain start altering the size of the perfectly matched layer .

Re: Moving boundary?
 

Thanx! Could you please recommend an article which uses the perfectly matched layer method?

Re: Moving boundary?
 

The references can be found in the journal of computational physics. You can find references for both finite difference and volume schemes