Volume coupling module #183

prasadadhav · 2021-06-17T12:59:08Z

In this PR I add a class for the Volume coupling module.
This will form the basis for further development on the Volume Coupling module, it's data fields and test cases.

TODO list:

I updated the documentation in docs/
I added a changelog entry in changelog-entries/ (create directory if missing)

Co-authored-by: David Schneider <david.schneider@ipvs.uni-stuttgart.de>

…hard coding the dimensions

davidscn · 2021-06-18T05:31:47Z

Adapter.C

+            if (module == "Volume_Coupling")
+            {
+                Volume_Couplingenabled_  = true;


Style-wise, the adapter uses nearly exclusively CamelCase, VolumeCouplingEnabled_. I guess it would be good to stick to it.

davidscn · 2021-06-18T05:33:25Z

Adapter.C

@@ -186,6 +190,14 @@ bool preciceAdapter::Adapter::configFileRead()
            }
        }

+        // If the Volume_Coupling module is enabled, create it, read the
+        // Volume_Coupling-specific options and configure it.
+        if (Volume_Couplingenabled_)


May I ask you to add this option in the if block below in order to bail out in case no module was enabled?

davidscn · 2021-06-18T05:37:54Z

Volume_Coupling/Volume_Coupling.H

+
+
+    //- Name of the Porosity
+    std::string nameYour_Volume_Field_ = "Your_Volume_Field";


Do you want to read this name through the dictionary?

This was just a placeholder. I have added `Fluid_Velocity' and 'Temperature'

We should take this from the configuration file of the adapter.

davidscn · 2021-06-18T05:39:38Z

Volume_Coupling/Volume_Coupling.H

+#include "Volume_Coupling/Generic_volScalarField.H"
+#include "Volume_Coupling/Generic_volVectorField.H"
+


These includes are still missing.

Yes, will be pushing things soon.
I should have named the PR with Draft as I was still working on cleaning up few things.

I converted the PR to draft, mark as "ready for review" when you are ready. :-)

…scalar and vector fiels on, so as to reduce the efforts for doing the same coding again and again

prasadadhav · 2021-06-27T19:03:51Z

Regarding clang-format. The issue was that I was using Ubuntu 18.04. And clamg-format 11 is not supported on any OS below Ubuntu 20.04. I updated my OS to 20.04, and installed the correct clang-format version.

But I am unable to use the format-code.sh script. I get the following error (this error is repeated multiple times)

find: ‘clang-format’: No such file or directory

I tried changing the permissions for .clang-format file but still the same error.

MakisH · 2021-06-28T08:48:52Z

Regarding clang-format. The issue was that I was using Ubuntu 18.04. And clamg-format 11 is not supported on any OS below Ubuntu 20.04. I updated my OS to 20.04, and installed the correct clang-format version.

But I am unable to use the format-code.sh script. I get the following error (this error is repeated multiple times)
find: ‘clang-format’: No such file or directory
I tried changing the permissions for .clang-format file but still the same error.

Does just running clang-format work on your system? What does which clang-format return? If you start typing clang-format in your terminal and press Tab twice to auto-complete, do you see any version-specific alternatives?

prasadadhav · 2021-06-28T08:55:41Z

Regarding clang-format. The issue was that I was using Ubuntu 18.04. And clamg-format 11 is not supported on any OS below Ubuntu 20.04. I updated my OS to 20.04, and installed the correct clang-format version.
But I am unable to use the format-code.sh script. I get the following error (this error is repeated multiple times)
find: ‘clang-format’: No such file or directory
I tried changing the permissions for .clang-format file but still the same error.
Does just running clang-format work on your system? What does which clang-format return? If you start typing clang-format in your terminal and press Tab twice to auto-complete, do you see any version-specific alternatives?

Running just clang-format on my system does nothing, I did this in the OF adapter directory.
whcih clang-format returns nothing in the terminal
The alternatives I get after double tab are following:

clang-format-10        clang-format-6.0       clang-format-diff-11   
clang-format-11        clang-format-diff-10   clang-format-diff-6.0

MakisH · 2021-06-28T08:58:44Z

Then just change the format-code.sh script to point to your clang-format-11. This is probably not the default because of the upgrade process, and probably you can fix it somehow.

Alternatively, you can set an alias clang-format="clang-format-11" in your ~/.bashrc.

prasadadhav · 2021-06-28T09:02:54Z

Then just change the format-code.sh script to point to your clang-format-11. This is probably not the default because of the upgrade process, and probably you can fix it somehow.

Alternatively, you can set an alias clang-format="clang-format-11" in your ~/.bashrc.

Thank you Makis. This worked properly. Pushed the changes for clang-format.

Edit:
Now just the build is failing due the GenericVolumeVectorField. But this fails for OF v2012. Works with OF v7. Will try and fix asap.

prasadadhav · 2021-07-26T13:26:42Z

@MakisH @davidscn I would like some hints/direction on this one.
Although I am able to compile (& use) the adapter for OF v7, we have compile errors for v2012.

I went through the OF API, OF source code for OF v7 and v2012 relating to motion solver, I couldn't find any difference between the two versions.

This is the error for quick reference.

Volume_Coupling/Generic_volVectorField.C: In member function ‘virtual void preciceAdapter::Volume_Coupling::Generic_volVectorField::read(double*, unsigned int)’:
Volume_Coupling/Generic_volVectorField.C:76:9: error: ‘fixedValuePointPatchVectorField’ was not declared in this scope
   76 |         fixedValuePointPatchVectorField& pointGeneric_volVectorPatch =
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Volume_Coupling/Generic_volVectorField.C:76:42: error: ‘pointGeneric_volVectorPatch’ was not declared in this scope
   76 |         fixedValuePointPatchVectorField& pointGeneric_volVectorPatch =
      |                                          ^~~~~~~~~~~~~~~~~~~~~~~~~~~

Volume_Coupling/Generic_volVectorField.H

davidscn · 2021-08-19T10:50:01Z

Interface.C

+        for (uint j = 0; j < patchIDs_.size(); j++)
+        {
+            // Get the face centers of the current patch
+            const vectorField faceCenters =
+                mesh.boundaryMesh()[patchIDs_.at(j)].faceCentres();
+
+            // Assign the (x,y,z) locations to the vertices
+            for (int i = 0; i < faceCenters.size(); i++)
+            {
+                vertices[verticesIndex++] = faceCenters[i].x();
+                vertices[verticesIndex++] = faceCenters[i].y();
+                if (dim_ == 3)
+                {
+                    vertices[verticesIndex++] = faceCenters[i].z();
+                }
+            }
+        }


How does the configuration for such a case look like? Do I need to specify all involved patches or are all patches automatically included. I just want to raise a warning that we might run into issues with empty patches in case we run 2D simulations.

This is unanswered question. Configuration actually ignores different type of mapping for volumes.

@engenegr what do you mean by "ignores different type of mapping for volumes"?

@engenegr what do you mean by "ignores different type of mapping for volumes"?

I guess it means that there is no configuration option to set particular boundaries empty.

MakisH · 2022-05-13T07:45:40Z

Some news: we are picking up this together with a student, to integrate it in the main branch of the adapter, write documentation, and add some tutorials on the website. Expect updates over the next months / by the end of the year.

yinjun5200 · 2022-05-30T08:15:01Z

Hi, I used your volume coupling adaptor. I slightly modified it, it works fine with compilation. But I have an error which states that "there were undefined symbols". This is because I added some additional variables "alpha (volume fraction)" and "AF(acoustic force)" to the adapter. I know I have to include the required libraries to Make/Options. But those variables are added in "CreateFields.H" file in my modified solver, which do not have a library. Any suggestions to add those symbols?

prasadadhav · 2022-05-30T08:48:46Z

Hi, I used your volume coupling adaptor. I slightly modified it, it works fine with compilation. But I have an error which states that "there were undefined symbols". This is because I added some additional variables "alpha (volume fraction)" and "AF(acoustic force)" to the adapter. I know I have to include the required libraries to Make/Options. But those variables are added in "CreateFields.H" file in my modified solver, which do not have a library. Any suggestions to add those symbols?

@yinjun5200 To summarize, your adapter compiles fine. But when you try to run the coupled simulation, you get this error, correct?
Could be helpful if you share the error log. Could you also clarify where you add these fields? alpha & AF?
Also, the moderators can say if this deserves an issue on its own (GitHub or Discourse) or if it's ok to continue discussing here.

Hi, I used your volume coupling adaptor. I slightly modified it, it works fine with compilation.

Also could you please tell me if you solved the compilation errors for OF v2022?

yinjun5200 · 2022-05-30T09:12:11Z

@alphaoo1 Thanks for you kind reply. Yes, it compiles fine. But at the end of the compilation, it print out "Everything looks fine in wmake.log === ERROR: Building completed with linking problems: there were undefined symbol. I attached my ldd.log and wmake.log files there.

I added these fields in modified interFoam solver in "createFeilds.H" file, as seen in the attached file. I think I have to provide the library to link those variables in Make/options to solve this problem? But I don't have any developed library for those variables. I need your guide !

I used volume coupling, I think it should be fine to discuss there? But I don't mind if we need to create a new issue. :)

NO, I used OpenFOAM6, not OF 2022.

ldd.log
ols ===" I also attahched the log file there.
wmake.log

tirgendetwas · 2022-07-20T16:00:06Z

Volume_Coupling/Generic_volVectorField.C

+        int patchID = patchIDs_.at(j);
+
+        // Get the field on the patch
+        fixedValuePointPatchVectorField& pointGeneric_volVectorPatch =


Regarding the compilation issues with OFv2012: did you mean to make this a fvPatchVectorField? Since you are making a refCast to fvPatchVectorField anyway. I can't find information on fixedValuePointPatchVectorField in the OF v7 documentation (however, I am finding fixedValuePointPatchField)

MakisH · 2022-08-01T12:21:07Z

Adapter.C

+            //            if ( module.empty )
+            //            {
+            //                return false;
+            //            }


@alphaoo1 What was the intention here? Why is this commented out?

MakisH · 2022-08-01T12:21:59Z

Adapter.C

@@ -277,6 +293,12 @@ void preciceAdapter::Adapter::configure()
                    FF_->addWriters(dataName, interface);
                }

+                // Add Momentum-related coupling data writers


Why "momentum-related"?

MakisH · 2022-08-01T12:24:19Z

Adapter.C

@@ -186,6 +194,14 @@ bool preciceAdapter::Adapter::configFileRead()
            }
        }

+        // If the Volume_Coupling module is enabled, create it, read the
+        // Volume_Coupling-specific options and configure it.
+        if (Volume_Couplingenabled_)


(long term / document)

With this approach, we have the choice between {CHT, FSI, FF, Volume Coupling}. In the long run, we want volume coupling to be possible for all (?) of {CHT, FSI, FF}.

MakisH · 2022-08-01T12:26:24Z

CHT/SinkTemperature.C

@@ -32,15 +32,15 @@ void preciceAdapter::CHT::SinkTemperature::write(double* buffer, bool meshConnec
        tmp<scalarField> patchInternalFieldTmp = TPatch.patchInternalField();
        const scalarField& patchInternalField = patchInternalFieldTmp();

-        //If we use the mesh connectivity, we interpolate from the centres to the nodes
+        // If we use the mesh connectivity, we interpolate from the centres to the nodes


Let's merge develop to the volume_coupling branch first, then the volume_coupling branch to this branch, and then check the formatting again. Maybe some of these changes have already been introduced before.

MakisH · 2022-08-01T12:31:15Z

Interface.C

+        for (uint j = 0; j < patchIDs_.size(); j++)
+        {
+            // Get the face centers of the current patch
+            const vectorField faceCenters =
+                mesh.boundaryMesh()[patchIDs_.at(j)].faceCentres();
+
+            // Assign the (x,y,z) locations to the vertices
+            for (int i = 0; i < faceCenters.size(); i++)
+            {
+                vertices[verticesIndex++] = faceCenters[i].x();
+                vertices[verticesIndex++] = faceCenters[i].y();
+                if (dim_ == 3)
+                {
+                    vertices[verticesIndex++] = faceCenters[i].z();
+                }
+            }
+        }


@engenegr what do you mean by "ignores different type of mapping for volumes"?

MakisH · 2022-08-01T12:40:27Z

Volume_Coupling/Volume_Coupling.C

+    // Read the name of the field Fluid_Velocity (if different)
+    nameFluid_Velocity_ = Volume_Couplingdict.lookupOrDefault<word>("nameFluid_Velocity", "Fluid_Velocity");
+    DEBUG(adapterInfo("    Fluid_Velocity field name : " + nameFluid_Velocity_));
+
+    // Read the name of the field Volume_Porosity (if different)
+    nameT_ = Volume_Couplingdict.lookupOrDefault<word>("nameT", "T");
+    DEBUG(adapterInfo("    Temperature field name : " + nameT_));


Application-specfiic

MakisH · 2022-08-01T12:40:45Z

Volume_Coupling/Volume_Coupling.C

+    if (dataName.find("Fluid_Velocity") == 0)
+    {
+        interface->addCouplingDataWriter(
+            dataName,
+            new Generic_volVectorField(mesh_, nameFluid_Velocity_));
+        DEBUG(adapterInfo("Added writer: Fluid_Velocity."));
+    }
+
+    else if (dataName.find("T") == 0)
+    {
+        interface->addCouplingDataWriter(
+            dataName,
+            new Generic_volScalarField(mesh_, nameT_));
+        DEBUG(adapterInfo("Added writer: T."));
+    }


Application-specific, check

MakisH · 2022-08-01T12:41:04Z

Volume_Coupling/Volume_Coupling.C

+    if (dataName.find("Fluid_Velocity") == 0)
+    {
+        interface->addCouplingDataReader(
+            dataName,
+            new Generic_volVectorField(mesh_, nameFluid_Velocity_));
+        DEBUG(adapterInfo("Added reader: Fluid_Velocity."));
+    }
+
+    else if (dataName.find("T") == 0)
+    {
+        interface->addCouplingDataReader(
+            dataName,
+            new Generic_volScalarField(mesh_, nameT_));
+        DEBUG(adapterInfo("Added reader: T."));
+    }


Application-specific, check.

MakisH · 2022-08-01T12:41:41Z

Volume_Coupling/Volume_Coupling.H

+
+
+    //- Name of the Porosity
+    std::string nameYour_Volume_Field_ = "Your_Volume_Field";


We should take this from the configuration file of the adapter.

MakisH · 2022-08-01T12:42:11Z

tools/format-code.sh

Just use whatever develop is using (after merging develop here).

tirgendetwas · 2022-10-04T21:51:54Z

We will continue our work on the volume-coupling branch of precice. Follow PR#255 for updates!

padhav and others added 19 commits May 20, 2021 17:19

added volume coupling option

36fe22e

added file for changelog

b3bf01f

correct spelling error

9d8c532

Co-authored-by: David Schneider <david.schneider@ipvs.uni-stuttgart.de>

Update in documentation: Specify which patches to consider

228cdf1

Co-authored-by: David Schneider <david.schneider@ipvs.uni-stuttgart.de>

Debug message fix

1167e04

Co-authored-by: David Schneider <david.schneider@ipvs.uni-stuttgart.de>

Update Interface.C general else condition continuing the if from above

8630530

Co-authored-by: David Schneider <david.schneider@ipvs.uni-stuttgart.de>

Made changes according to comments by David to add dim_ instead of …

b9138ac

…hard coding the dimensions

fixed dim typo

7442a96

added 3D differentiate as per suggestion by @makis precice#182 (comment)

f95830d

markdown fix

2e107f2

markdown fix

7ee6da2

Fixed clang format violations

1558794

Fixed clang format violations

ca49aed

Format code

5d32bde

Added Volume Coupling Module

56daa9e

Added volume coupling module

916eece

added file path to Make

10b79f9

Changelog for PR#183

ff9e0a9

fixed typo in file name

07bd8a5

davidscn reviewed Jun 18, 2021

View reviewed changes

MakisH marked this pull request as draft June 23, 2021 16:01

padhav added 4 commits June 26, 2021 10:59

added Generic fields. These are template classes to base your Volume …

1e4263b

…scalar and vector fiels on, so as to reduce the efforts for doing the same coding again and again

Added Volume Vector fields Fluid Velocity

67de185

Added the Field temperature.

b8d5701

included new files in the include

19425ad

padhav added 2 commits June 27, 2021 21:06

updated the Generic volume class name

0e72c1a

fixed md format errors

16eb84e

Code format using clang-format

579c5e4

davidscn reviewed Jul 26, 2021

View reviewed changes

Volume_Coupling/Generic_volVectorField.H Show resolved Hide resolved

davidscn reviewed Aug 19, 2021

View reviewed changes

prasadadhav mentioned this pull request Oct 7, 2021

Support multiple OpenFOAM versions with the same source #32

Open

MakisH mentioned this pull request May 4, 2022

Add a Fluid-Fluid module #67

Merged

MakisH mentioned this pull request May 13, 2022

WIP: Add volume coupling of temperature #97

Closed

MakisH mentioned this pull request Jun 26, 2022

Missing support for volume coupling #229

Closed

tirgendetwas reviewed Jul 20, 2022

View reviewed changes

ge72fah and others added 3 commits July 31, 2022 15:46

replace patchVectorField type

e702a42

fix other compilation errors (OFv2112)

240e8d4

formatting attempt

bdc1ada

MakisH reviewed Aug 1, 2022

View reviewed changes

tirgendetwas added 2 commits September 9, 2022 12:47

making internal fields iterations consistent

9e79ee6

replace const-reference with normal reference

c338a28

tirgendetwas merged commit d453b8c into precice:volume-coupling Sep 11, 2022

MakisH mentioned this pull request Oct 25, 2022

Volume coupling #255

Merged

2 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Volume coupling module #183

Volume coupling module #183

prasadadhav commented Jun 17, 2021 •

edited

Loading

davidscn Jun 18, 2021

davidscn Jun 18, 2021

davidscn Jun 18, 2021

prasadadhav Jun 27, 2021

MakisH Aug 1, 2022

davidscn Jun 18, 2021

prasadadhav Jun 22, 2021

MakisH Jun 23, 2021

prasadadhav commented Jun 27, 2021

MakisH commented Jun 28, 2021

prasadadhav commented Jun 28, 2021

MakisH commented Jun 28, 2021

prasadadhav commented Jun 28, 2021 •

edited

Loading

prasadadhav commented Jul 26, 2021

davidscn Aug 19, 2021

evd0kim Oct 8, 2021

MakisH Aug 1, 2022

davidscn Aug 1, 2022

MakisH commented May 13, 2022

yinjun5200 commented May 30, 2022

prasadadhav commented May 30, 2022

yinjun5200 commented May 30, 2022

tirgendetwas Jul 20, 2022 •

edited

Loading

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

MakisH Aug 1, 2022

tirgendetwas commented Oct 4, 2022



		//- Name of the Porosity
		std::string nameYour_Volume_Field_ = "Your_Volume_Field";

		#include "Volume_Coupling/Generic_volScalarField.H"
		#include "Volume_Coupling/Generic_volVectorField.H"

Volume coupling module #183

Volume coupling module #183

Conversation

prasadadhav commented Jun 17, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

prasadadhav commented Jun 27, 2021

MakisH commented Jun 28, 2021

prasadadhav commented Jun 28, 2021

MakisH commented Jun 28, 2021

prasadadhav commented Jun 28, 2021 • edited Loading

prasadadhav commented Jul 26, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

MakisH commented May 13, 2022

yinjun5200 commented May 30, 2022

prasadadhav commented May 30, 2022

yinjun5200 commented May 30, 2022

tirgendetwas Jul 20, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tirgendetwas commented Oct 4, 2022

prasadadhav commented Jun 17, 2021 •

edited

Loading

prasadadhav commented Jun 28, 2021 •

edited

Loading

tirgendetwas Jul 20, 2022 •

edited

Loading