interFoam - stratified flow - problem with shear stress at interface
4 Attachment(s)
Dear Foamers,
I have a problem when simulating stratified flow of transformer oil and water with interFoam. I simulate in a 2D channel of width 0.1 m and 6 m length. The interface shall be at 0.05 m, so at 50 %. Water is the denser fluid, so at the bottom of the channel, and the transformer oil the lighter fluid. I start the simulation with a wall of zero thickness (splitMesh) at the interface to ensure the Poiseuille profile (see Folder 01_Poiseuille in tar). The analytic solution is matched perfectly (see Poiseuille_TransformerOil_Water.pdf). Then I take the last step as the starting point for the simulation without a wall. The resulting profile is in general not bad, but it does not match. Without any change in the profile I tried the following: run a longer simulation time (*1.5); enlarge the length of the channel (+ 2 m); enlarge number of cells in length (*1.5); enlarge number of cells in cross section (*1.5); schemes of 2nd order; surface correction factor from 1 to 0 (see TowPhaseFlow.pdf). :confused: The problem seems to be the shear stress compatibility condition at the interface (continuity of shear stress at the interface) (see ShearStress.pdf). I am waiting and hoping for your advice. Anja |
Another Try for stratified flow
3 Attachment(s)
Hello everybody,
I tried the two phase stratified, laminar flow for water treated as two different phases, a water-similar oil (HT350) and water and for water and air. Using water for both phases works perfectly, HT350 and water too, as their properties are very similar. Air and water is even worse than the transformer oil from above. I am not sure if that is a problem with the simulation or with the coding of the interface treatment. If anybody has an idea, it is very welcome. Thanks, Anja Attachment 1693 Attachment 1694 Attachment 1695 |
Hi, i m trying to simulate a two-phases Poiseuille flow like you but i have some problems. The water-air interface is not stable: the flow is not stratified. I setted up a case with: inlet / outlet boundary with a pressure drop and top / bottom wall (there's no gravity). How can i solve this problem and simulate a stratified flow?
Thanks in advance Emanuele EDIT: i have seen your attached file right now. Thanks RE-EDIT : i dont understand this part of how to - change /constant/polyMesh/boundary between faceSet command and splitMesh. I n what way i have to modify the boundary file?? RE-RE-EDIT: solved reading this post http://www.cfd-online.com/Forums/ope...thickness.html |
How can you obtained a no-zero value of U along the interface?? I setted U at fixedValue (0 0 0) and obviously at the interface U values are 0.
|
1 Attachment(s)
I made a new mesh but i have a problem with splitMesh: it returns this error
Code:
--> FOAM FATAL ERROR: I attach my case |
Hey Emanuele,
thanks for your interest in my work. I would like to ask you to do some steps the next time you post your case. Every command you enter in OpenFOAM will lead to a message on the terminal, if you do the commands as, for example,
In general, I did two different simulations. First, I simulate the velocity profile with a baffle (wall of zero thickness in OpenFOAM) seperating the two fluids. I obtained the two Poiseuille-Profiles. I did this to avoid the development of a wavy interface. With these results I deleted the interface (that is the entry for the two faces in the file of the time step to continue with) and continued to obtain the multiphase flow velocity profile. Additionally, I did not start with air right away because of some problems in setting up the case.. Therefore I started with water-water, water-waterLikeOil, water-HeavyOil and at last water-air as the difference in properties increases this way. I also did not use gravity, but surface tension. Please find attached my water-air case. In the post, probably in the afternoon, I will also write the commands and files together for the mesh with internal face walls (baffles). Due to a securty error, I cannot upload the tar.gz of my case. Please write me a mail to get the files. Hope it does help. Best wishes Anja |
How I did my mesh
My constant/polyMesh/blockMeshDict is:
Code:
The log.blockMesh file does look like: Code:
/*---------------------------------------------------------------------------*\ Code:
/*--------------------------------*- C++ -*----------------------------------*\ The log.faceSet file does look like: Code:
/*---------------------------------------------------------------------------*\ Code:
/*--------------------------------*- C++ -*----------------------------------*\ Code:
/*--------------------------------*- C++ -*----------------------------------*\ The newface1/2 are form the boundary file and will be the new boundaries. The command here is therefore: splitMesh interface interface_up interface_down |tee log.splitMesh and the log.splitMesh file does look like: Code:
/*---------------------------------------------------------------------------*\ last: do a checkMesh |tee log.checkMesh to be sure that everything went well |
The wrong part in my case is the faceSet command: i obtain a size (122) not equal to x size (16). At this resolution checkMesh fail but if i increase (eg 600) it gives a good output but faceSet still says a wrong size. Perhaps there is something wrong on my blockMesh file. I changed it according to your own and now it works.
Thanks PS It's important to delete *Zones files in constant/polyMesh before splitMesh command |
I am happy that your mesh works now. Hope, that the simulation will too.
|
All times are GMT -4. The time now is 08:20. |