[jf_vt]

Hi
I am trying to extract and write into a file the positions (x,y,z) and ideally closest cell volume but I still has to figure out how to to that, where the velocity magnitude exceed a threshold using swak4foam writeIf feature.

After reading some examples, I came with the following:

 {
      type writeIfSwakExpression;
      verbose yes;
      outputControlMode timeStep;
      outputInterval 1;
      valueType internalField;
      writeCondition "mag(U)>90";
      writeControlMode always;
      cooldownMode no;
      storeAndWritePreviousState no;
      writeConditionAccumulation or;
      variables (
         "X=pos().x;"
         "Y=pos().y;"
         "Z=pos().z;"
);
      storedVariables
        (

          {
            name X;
            name Y;
            name Z;
            initialValue "0";

          }
        );

    }

When the case run the log file provides the following:

swak4Foam: Allocating new repository for sampledMeshes
swak4Foam: Allocating new repository for sampledGlobalVariables
filterTest: "mag(U)>90" evaluated to true
filterTest starts writing
filterTest triggered writing of t=1
filterTest: Writing ....

which let think that all goes well ( I know the max velocity is larger than 300m/s in this case)

But when the case ends I cannot find the information I need.

Obviously I am doing something wrong but I dont know what
If someone could point me into the right direction, it will be much appreciated

Thanks
JF
ps: the motivation is to build a script that refine a mesh based on some physical criteria

[paoching]

is swak4foam compatible with current version of OpenFOAM v6 or OpenFOAM v1806?
I want to used groovyBC, which is a sub function in swak4foam and found this page:
https://openfoamwiki.net/index.php/Contrib/swak4Foam
but cannot find a version supporting v6 or v1806. Do I have to install a legacy OpenFoam?
what version of swak4Foam are you using with what version of OpenFOAM? Can you post me a link?

[jf_vt]

Originally Posted by paoching

is swak4foam compatible with current version of OpenFOAM v6 or OpenFOAM v1806?
I want to used groovyBC, which is a sub function in swak4foam and found this page:
https://openfoamwiki.net/index.php/Contrib/swak4Foam
but cannot find a version supporting v6 or v1806. Do I have to install a legacy OpenFoam?
what version of swak4Foam are you using with what version of OpenFOAM? Can you post me a link?

you already have the link
(almost) Everythng is explained there (with some news on the tweet feed as well).
TO answer your question
I am using swak 0.4.2 with foan-extend 4.1,but I read that swak should work with OF6 and 18.06
Just download the dev version on the mercurial website as explained on the wiki
JF

[gschaider]

Originally Posted by jf_vt

Hi
I am trying to extract and write into a file the positions (x,y,z) and ideally closest cell volume but I still has to figure out how to to that, where the velocity magnitude exceed a threshold using swak4foam writeIf feature.

After reading some examples, I came with the following:

 {
      type writeIfSwakExpression;
      verbose yes;
      outputControlMode timeStep;
      outputInterval 1;
      valueType internalField;
      writeCondition "mag(U)>90";
      writeControlMode always;
      cooldownMode no;
      storeAndWritePreviousState no;
      writeConditionAccumulation or;
      variables (
         "X=pos().x;"
         "Y=pos().y;"
         "Z=pos().z;"
);
      storedVariables
        (

          {
            name X;
            name Y;
            name Z;
            initialValue "0";

          }
        );

    }

When the case run the log file provides the following:

swak4Foam: Allocating new repository for sampledMeshes
swak4Foam: Allocating new repository for sampledGlobalVariables
filterTest: "mag(U)>90" evaluated to true
filterTest starts writing
filterTest triggered writing of t=1
filterTest: Writing ....

which let think that all goes well ( I know the max velocity is larger than 300m/s in this case)

But when the case ends I cannot find the information I need.

Obviously I am doing something wrong but I dont know what
If someone could point me into the right direction, it will be much appreciated

Thanks
JF
ps: the motivation is to build a script that refine a mesh based on some physical criteria

writeIfSwakExpression only triggers the writing of a complete timestep


storedVariable is "only" a special time of variable that keeps its value between timesteps (applications would be finding a maximum over time, summing up mass-flows etc)


What kind of data would you need? Only one position or all the cells "violating" the condition?

[jf_vt]

All the cell violating a condition ( that can be possibly complex ) ==>
It generate a cloud of points ( a file of x,y,z possibly with vol(cell) s an added information that can then be used to refine the mesh automatically in a scriptable mesher.
I currently do it through paraview threshold but this is not practical for large mesh size ( but sufficient to test the workflow).
JF

[gschaider]

Originally Posted by jf_vt

All the cell violating a condition ( that can be possibly complex ) ==>
It generate a cloud of points ( a file of x,y,z possibly with vol(cell) s an added information that can then be used to refine the mesh automatically in a scriptable mesher.
I currently do it through paraview threshold but this is not practical for large mesh size ( but sufficient to test the workflow).
JF

Have a look at manipulateCellSet in swakFunctionObjects (example in Examples/other/topoSetDamBreak). That creates a cellSet depending on a condition. cellSets are "natural" for a lot of OpenFOAM-tools


If you still need to output all cell centers then you can use dumpSwakExpression from simpleSwakFunctionObjects (which I hardly use because of its tendency to clutter the hard-disk)

[jf_vt]

Thank you

a quick look at the exemple indicates that manipulateCellSet is the function I need for my current endeavor.
I will play a bit with it to get acquainted first

As for dumpSwakExpression, you clearly documented the disk cluttering issue in the relevant example! I will nevertheless take the time to understand what it does.. just in case.

JF

[gschaider]

Originally Posted by jf_vt

Thank you

a quick look at the exemple indicates that manipulateCellSet is the function I need for my current endeavor.
I will play a bit with it to get acquainted first

As for dumpSwakExpression, you clearly documented the disk cluttering issue in the relevant example! I will nevertheless take the time to understand what it does.. just in case.

JF

Stupid question: have you considered using dynamicRefineFvMesh with a custom field that you calculate with an expressionField functionObject (instead of the alpha.water that is used in the damBreakWithObstacle-tutorial)?

[jf_vt]

Yes, briefly but I did not explore it.
Indeed, and to the best of my understanding of the documentation, dynamicRefineMesh will cut a cell in 8 or not based on a predefined criteria.

That does not work well for complex unstructured mesh such as the ones I am working with, as it increases contrast between cell that are cut and the ones that are not.
I cannot speak in general, but in most situation I can think of ( read: useful to me), this decrease the overall mesh quality.

On the other hand a scriptable mesher allows for more complex remeshing strategies in order to maintain mesh quality around the "bad" cell that will be reduced.