Any one could summarize the procedure of implementation of arbitrary lagrangian-eulerian method? Many thanks.

-Hao

The Computer Methods in Applied Mechanics and Engineering Journal has released 3 issues in volume 193 totally dedicated to ALE formulation.

Thanks, but which pages? Could you specify more detail.

Sorry, here it goes: Volume 193, Issues 39-41, Pages 4073-4456 (1 October 2004) The Arbitrary Lagrangian-Eulerian Formulation Edited by A. Huerta, A. Rodriguez-Ferran

and the link is: http://www.sciencedirect.com/science...sueURL&_tockey=%23TOC%235697%232004%23998069960%23518995%23FLA%2 3Volume_193,_Issues_39-41,_Pages_4073-4456_(1_October_2004)%2BMThe_Arbitrary_Lagrangian-Eulerian_Formulation%2BMEdited_by_A._Huerta,_A._Ro driguez-Ferran&_auth=y&view=c&_acct=C000037900&_version=1& _urlVersion=0&_userid=687360&md5=2ebc519bc093deeae 2e6995f0dd045be

if it doesn't work, enter www.sciencedirect.com, click on journals and then on Computer Methods .... Then you can choose the volume which interests you. Anyway, I've donwloaded some papers. If you want, you may send me an email address and I send you the papers.

thanks, i can access to them

I am doing a computation about gas-liquid in a pipeline, where a liquid slug, treated as incompressible, is driven by gas with high pressure and volocity.

What i am interesting is the pressure and velocity of the gas-liquid interface.

Is this problem a fluid-structure interaction problem?

Could you give some comments or give some reference to this kind of problems?

Thanks in advance.

i've read some papers, but i am still not very clear how to apply ALE in this case:

1. to tracking the interface, it can be done by setting the interface velocity equal to the fluid velocity at the interface, however, the interface velocity is also an unknown. so how to deal with it?

2. it is only need to take account of gas domain, isn't it? the up streamboundary is quite straightforward as inflow B.C, but how about the downstream?

many thanks in advance for your comments!

There is no need for moving mesh (ALE). Have a look at VOF (volume of fluid). The downstream boundary is indeed a little problem, where you have to find a smart solution. Otherwise your pipe will be empty after some time.

Have a look at OpenFOAM. It has probably the best VOF implementation at all and it is open source.

hi, Jorn

Could you explain a bit more detail why the pipe will be empty after some time.

Can you suggest something "smart" to the downstream boundary. I deduce the boundary by intergrating along characteristic line, so that i can approximate the location, pressure and velocity of interface at next time step, however, these quantities seem to depend on the start point i choose, and they are difficult to control, for example, the location of interface may be out of pipe.

Many thanks.

Hao

I tried to run a simulation of fluid waves in a half filled pipe some time ago.

It is easy do define an inlet velocity at one side of the pipe but since the wave should pass through the other end I was not sure what boundary condition to use there.