CFD Online Discussion Forums - pimpleDyMfoam simulation keeps blowing up

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM (https://www.cfd-online.com/Forums/openfoam/)

- - pimpleDyMfoam simulation keeps blowing up (https://www.cfd-online.com/Forums/openfoam/99920-pimpledymfoam-simulation-keeps-blowing-up.html)

Thanks for that! I have executed that command and noted that I have two regions in my constant/polyMesh/cellZones file.

With AMI I have noticed that the dynamicMeshdict requires a faceZone. I tried to create this using setSet with the following sequence:

faceSet innerFace new patchToFace AMIMoving <that's my moving zone's face>
faceZoneSet innerFace new setsToFaceZone innerFace region1 <I think this is my inner region>

Out of interest is there a way of finding out which region relates to which of my two merged meshes? I did try region0 too!!

rerunning pimpleDymFOAM, now, just reproduces the rotation of the whole model that I have been fighting with for so long.

Any other suggestions? I can zip up my model if that might help? It's a bit convoluted though since I am running snappyHexMesh(Castellated) / flattehMesh / Extrude / snappyHexMesh(Snap) on each mesh before merging them.

Kindest Regards

Andrew

Cant you just zip the final case where it rotates the whole mesh.

Use dropbox, Gdrive or something else if its too big for CFD-online

Hi.... Here is the dropbox link. I've tried to tidy up the directory structure a bit and put notes in the shell script files.
https://dl.dropboxusercontent.com/u/...eTestAMI2D.zip

Best Regards
Andrew

Hi cant we just get the final case.

The shell scripts does not work properly.

Best

Duh, sorry, didn't read your last entry fully..... this link gives the model at the last step after the merge and splitMeshRegions.

https://dl.dropboxusercontent.com/u/...innemanAMI.zip

The Master directory is AMI

Regards

Andrew

Oh... The shell scripts probably didn't work right because I'm running OF on Mac OS X.... I think the grep statements are formed differently to linux.

Andrew

Hi

Just change this in the dynamicMeshDict

Code:

solidBodyMotionFvMeshCoeffs

{

    cellZone        region0;



    solidBodyMotionFunction  rotatingMotion;

    rotatingMotionCoeffs

    {

        CofG        (0 0 0);

        radialVelocity (0 0 360); // deg/s

    }

}

Everything works fine

https://docs.google.com/file/d/0Bxal...it?usp=sharing

Thank you so much for your help....... although it does make me feel incredibly thick! I think I got fixated on the faceZone entry in dynamicMeshDict and didn't look into what the possibilities were!

Out of interest though why was my faceZone method not working? In my setSet step (not in the model I shared late yesterday but in the setBatch file in the original share) I thought I was making a faceZone which incorporated the cellZone region0 and the AMIMoving face.... obviously this was wrong but I'd like to try to understand what was wrong about it? Is this something you could advise me on?

Once again, thanks you so much for your help and patience!

Kindest Regards

Andrew

I'm trying to take a slightly different track with my model now and I'm wondering how best to achieve the result. I've been experimenting with a basic mesh (essentially cavity) and creating an inner and outer mesh using setSet. I have been able to create cellSets for the inner and outer regions but if I want to rotate the inner cellset what would be the best way to do this?

The basic mesh has NO internal feature, like in the cavity tutorial, so I am building internal features like the cellSets. I'm trying to rotate the cells in an AMI / dynamicMesh fashion like my last experimentation so I need to create two patches for the AMI faces. should this be based on createBaffles / mergeOrSplitMeshes -split or it there a more appropriate method like splitMeshRegions?

I'm doing this in order to reduce the amount of time spent meshing for future models since sHM/flatten/extrude/sHM for each region then merging and splitting the mesh seems quite involved if I can sHM/flatten/Extrude/sHM once and then introduce the internal AMI features I require. Also I don't seem to be able to get the much closer distances between the two meshes that I would like using the merge method.

I have tried creating a faceZone based on the inner cellSet then creating baffles but this doesn't seem to work, if I try to use moveDynamicMesh it just fails with a segFault (I'm using MacOSX so it's perhaps not as well manner as in Linux!)

I would welcome anyone's suggestions and guidance on this.

regards

Andrew

Hi all,
I have a basic problem to setup a case for pimpleDyMFoam solver,
I got some idea from this post, still I am not clear.

In my case I have 3 domain Inlet Volume, fan Volume, outlet volume.
you can imagine the case is just a pipe, sub divided in to 3 volume (see attached Fig).
The middle one is suppose to rotate. So far I am using MRF its going fine.

Now I want to use pimpleDyMFoam, I dont know how to treat the in between faces.
This is going to be my first try please help and correct me,

steps what I understood from the previous post is,

1) I need to split the domain in to three.
As long as i am going to use sliding mesh, so the mesh no need to be conformal I think.

2) Then I will have three .msh files

I have few questions,

before exporting the mesh, while giving BC in Gambit what BC, should I use for the faces (interface? internal?)

where I need to place this three .msh file? all in one folder? or separately?

after converting this mesh, there will be three constant folder.

Do I need to edit anything before merging this mesh? if so where and what I need to edit.

Please help me to go further.

Thanks,
Sivakumar

Quote:

Originally Posted by sivakumar (Post 465937)

hi
first of all MRF is usually used in steadystate cases. Pimpledymfoam is for transient cases, or at least transient solution cases. For the dynamic case where the domain does indeed rotate, you have to set the region(s) which rotates. the interfaces you should set as AMI so that the solver can handle non-conformal patches that happen in rotating regions.

but if u have a steadystate case with a cyclic domain with a mrf region then you should use other solvers, like the MRFsth.

try to figure out what u need first, is your case transient or ss? and then u go from there..

gl

Hi Calim , thanks for your reply, I don't know which question is forced you to answer like this.

After checking all my needs and possibilities, I have decided to use that solves.

Siva

Hi There,
I tried and followed the steps mentioned in this thread, I dont know which step I am missing.

Please help me to sort out the problem.

Here is the step which I followed,

1) I have divided my domain in to 3 volume, each volume has its unique faces, then non conformal has been generated. (4 interface are defined AMI_1, AMI_2 ......)

2) fluent3DMeshToFoam

3) I have modified the AMI boundaries under case/constant/boundary ( as jiejie explained in his post)

I am not sure what are the steps I need to perform more.

While executing checkMesh I am getting the following error,

Code:

Create time



Create polyMesh for time = 0



Time = 0



Mesh stats

    points:           2959944

    faces:            8672812

    internal faces:   8468870

    cells:            2856947

    boundary patches: 20

    point zones:      0

    face zones:       1

    cell zones:       2



Overall number of cells of each type:

    hexahedra:     2856947

    prisms:        0

    wedges:        0

    pyramids:      0

    tet wedges:    0

    tetrahedra:    0

    polyhedra:     0



Checking topology...

 ****Problem with boundary patch 3 named top0 of type wall. The patch should start on face no 8547530 and the patch specifies 8554881.

Possibly consecutive patches have this same problem. Suppressing future warnings.

 ***Boundary definition is in error.

    Cell to face addressing OK.

    Point usage OK.

    Upper triangular ordering OK.

    Face vertices OK.

   *Number of regions: 3

    The mesh has multiple regions which are not connected by any face.

  <<Writing region information to "0/cellToRegion"



Checking patch topology for multiply connected surfaces ...

    Patch               Faces    Points   Surface topology                  

    top2                37088    37653    ok (non-closed singly connected)  

    outlet              4484     4620     ok (non-closed singly connected)  

    center1             37088    37653    ok (non-closed singly connected)  

    top0                3965     4092     ok (non-closed singly connected)  

    center0             3965     4092     ok (non-closed singly connected)  

    inlet               2867     2976     ok (non-closed singly connected)  

    top1                8040     8268     ok (non-closed singly connected)  

    fan                 15720    15948    ok (non-closed singly connected)  

    ILR0                3055     3168     ok (non-closed singly connected)  

    ILR1                3055     3168     ok (non-closed singly connected)  

    OLR0                28792    29340    ok (non-closed singly connected)  

    OLR1                28792    29340    ok (non-closed singly connected)  

    CLR0                1800     1891     ok (non-closed singly connected)  

    CLR1                1800     1891     ok (non-closed singly connected)  

    FCLR0               1800     1891     ok (non-closed singly connected)  

    FCLR1               1800     1891     ok (non-closed singly connected)  

    AMI_1               2867     2976     ok (non-closed singly connected)  

    AMI_2               6000     6161     ok (non-closed singly connected)  

    AMI_3               6480     6649     ok (non-closed singly connected)  

    AMI_4               4484     4620     ok (non-closed singly connected)  



Checking geometry...

    Overall domain bounding box (0.239348 -0.371369 -0.75) (0.771515 0.354013 4)

    Mesh (non-empty, non-wedge) directions (1 1 1)

    Mesh (non-empty) directions (1 1 1)

    Boundary openness (3.92584e-14 6.56918e-14 -3.07888e-17) OK.

    Max cell openness = 3.36913e-16 OK.

    Max aspect ratio = 26.3734 OK.

    Minumum face area = 4.20635e-07. Maximum face area = 0.000104506.  Face area magnitudes OK.

    Min volume = 3.36453e-09. Max volume = 1.02511e-06.  Total volume = 0.931774.  Cell volumes OK.

    Mesh non-orthogonality Max: 59.4945 average: 11.6305

    Non-orthogonality check OK.

    Face pyramids OK.

    Max skewness = 1.5012 OK.

    Coupled point location match (average 5.3341e-12) OK.



Mesh OK.



End

Please give me some suggestions.

Thanks,
Siva

Greetings to all!

@Siva: I see that you have gotten some answers to your questions here: http://www.cfd-online.com/Forums/ope...tml#post466156

So I have no idea if you still are having problems with this.
If you are still having problems with this, I suggest that you create a simplified version of a case conceptually similar to yours, so that you can share it with us. That way it'll be easier to help you.

Because from the error message given by checkMesh, all I can figure out is that something went wrong in your editing of the file "boundary".

Best regards,
Bruno

hi linnemann
well I am also trying for propeller case, I used FV schemes, FV solution given in the propeller case. I meshed my propeller using Ansa.

Thing I want to ask you is how to select a fv scheme, Fv solution?, what should be the value of non orthgonality, max skewness?.

Hi.

try this thesis: Error Analysis and Estimation for the Finite Volume Method with Applications to Fluid Flows

http://www.h.jasak.dsl.pipex.com/HrvojeJasakPhD.pdf

cheers! ;)

Hi,

based on pimpleDyMfoam's wingMotion example,
I managed to mesh a new moving geometry.
Though, pimpleFoam crashes so I lowered time step to 1e-4.
Now pimpleFoam finishes in 3 iterations with poor results.

Details, Files & pictures attached in this thread, post #3 and #4
http://www.cfd-online.com/Forums/ope...tructures.html

Any clue how to make this work correctly ?
iy-a

PS : I just had to run pimpleDyMFoam in a loop using a bash script !

Quote:

Originally Posted by A.Wendy (Post 426084)

Hi

here you find the "cleaned" case. just run the Allrun file. The boundary ist changed by hand maybe you can automatize it.

if you have quest just send a massage

http://ubuntuone.com/3ZKPM8nv9xDgZaSkhsVcoO

best wishes

andy

Hey Andy,
I was trying to setup a similar case and stumbled upon this thread.
First of all thanks for the detailed explanation - it puts the problem in perspective.
I tried to download the case file but seems like the file is not present. Can you please upload to Dropbox or drive?
I think this is a starting point for anyone trying to understand Ami so I will put up the case on a permanent FTP for everyone.
Thanks again.
Sangram.