[lethu]

I am transferring a post I wrote in the wrong forum :

Hello everyone,

I am trying to simulate particles transport in a water channel.
So first, I have simulated water running in the channel with air on top of it using interFoam. I have unphysical results at the outlet such as in the topic
HTML Code:
Channel flow using InterFOAM
I extracted the interface of this case anyway. Now I am trying to run simulation with uncoupledKinematicParcelFoam.
I want to use the latest results of this interFoam simulation, only for the U file, as initial conditions.
Problem is the domain is now smaller than the first simulation (but more cells).

It seems that mapFields is not ok with that.

Does anyone know how to solve this problem ?

I have already studied the hopper tutorial.

I have put the name of the new patch that has been created and that is kind of slicing the old domain. (in cuttingPatches)

But I still can't map my U field...

If anyone has the solution, thanks !

Old domain :
https://drive.google.com/open?id=0B7...lJ0MG1nd1NFNEU

New domain :
https://drive.google.com/open?id=0B7...1VERS1wMVZ3ODg

Here the message I got from it :

Source: ".." "cadreValeursInitiales"
Target: "C:/PROGRA~1/BLUECF~1/ofuser-of4/16-EssaisCadre/particules" "05-cadreKs1mNut0EvenFinerMesh"

Create databases as time

Source processor 0

Source time: 60
Target time: 0
mesh size: 160610

Target processor 0
mesh size: 160610
--> FOAM Warning : 
    From function void Foam::meshToMesh0::calcAddressing()
    in file meshToMesh0/calculateMeshToMesh0Addressing.C at line 155
    Source patch conduite_inlet has no faces. Not performing mapping for it.
--> FOAM Warning : 
    From function void Foam::meshToMesh0::calcAddressing()
    in file meshToMesh0/calculateMeshToMesh0Addressing.C at line 155
    Source patch conduite_outlet has no faces. Not performing mapping for it.

Mapping fields for time 60

    interpolating U


--> FOAM FATAL IO ERROR: 
size 79900 is not equal to the given value of 160610

file: ../cadreValeursInitiales/processor0/60/U from line 18 to line 80816.

    From function Foam::Field<Type>::Field(const Foam::word&, const Foam::dictionary&, Foam::label) [with Type = Foam::Vector<double>; Foam::label = int]
    in file X:/OpenFOAM-4.x/src/OpenFOAM/fields/Fields/Field/Field.T.C at line 295.

FOAM exiting

[wyldckat]

Quick answer: I haven't had the time needed to look into your cases, but I suggest that you don't map between decomposed cases, specially if you've not yet very familiar with OpenFOAM.

I'm referring to the mapping being done to or from the folder "processor0". It's preferable that you use reconstructPar on the first case and use mapFields between the two reconstructed cases (the main folder of each one) and not the decomposed cases (don't use the "-parallel*" options in mapFields).

[lethu]

Thank you for your answer.
I did what you proposed and my mapping did work.

I don't understand why it didn't work in parallel cases.

If you have some time about the outlet BC issue, I would really appreciate your help !

Thanks !

[eb19]

Originally Posted by lethu

Thank you for your answer.
I did what you proposed and my mapping did work.

I don't understand why it didn't work in parallel cases.

If you have some time about the outlet BC issue, I would really appreciate your help !

Thanks !

Hi

I have been trying to map the field from a domain with coarse mesh and a domain with a finer mesh. The two domains are exactly the same.

Here, I name the folder containing a coarse mesh with the variable fields u, p_rgh and alpha.water as 'coarse', and the folder containing the fine mesh with the variable fields as 'a04f1v3t'. Both are located in the same folder.

I want to map the pre-calculated field at t=3.32334 from the 'coarse' to the 'a04f1v3t'. So, the time folder '3.32334' exist in the 'coarse' and also in 'a04f1v3t'. The time folder '3.32334' exists in the 'coarse' has the pre-calculated fields: U, p_rgh and so on.

Then, I typed the command line :
mapFields coarse -sourceTime 3.32334 -case a04f1v3t

But, I got this following error:

Build : 2.3.0-f5222ca19ce6
Exec : mapFields coarse -sourceTime 3.32334 -case a04f1v3t
Date : Nov 28 2017
Time : 16:46:01
Host : "cv-epb14"
PID : 22808
Case : ./a04f1v3t
nProcs : 1
sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Disallowing user-supplied system call operations

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Source: "." "coarse"
Target: "." "a04f1v3t"

Create databases as time
Case : ./coarse
nProcs : 1

Source time: 3.32334
Target time: 3.32334

Create meshes

#0 Foam::error:rintStack(Foam::Ostream&) at ??:?
#1 Foam::sigSegv::sigHandler(int) at ??:?
#2 in "/lib/x86_64-linux-gnu/libc.so.6"
#3 Foam:rimitiveMesh::calcPointOrder(int&, Foam::List<int>&, Foam::List<Foam::face> const&, int, int) at ??:?
#4 Foam:rimitiveMesh::reset(int, int, int, int) at ??:?
#5 Foam:olyMesh::initMesh() at ??:?
#6 Foam:olyMesh:olyMesh(Foam::IOobject const&) at ??:?
#7 Foam::fvMesh::fvMesh(Foam::IOobject const&) at ??:?
#8
at ??:?
#9 __libc_start_main in "/lib/x86_64-linux-gnu/libc.so.6"
#10
at ??:?
Segmentation fault (core dumped)

My mapFieldsDict is as follows:

patchMap
(
bottom bottom
inlet inlet
atmosphere atmosphere
frontAndBack frontAndBack
chamber chamber
sides sides
);

cuttingPatches
(
);


My question is:
1. Do I have to create the time folder '3.32334' with empty files in the target folder 'a04f1v3t' ?

2. I used refineMeshDict locally in the target folder. I mean I didn't refine all the mesh, but only in some part of the domain. Will this cause a problem when mapping the field?

3. Where should I put the source folder actually?

Please advise me or hint me things that you think is the cause... I searched around, and I dont find people having a similar error message.

Regards
eb19