OpenFOAM construct for conditional field modification

jherb · May 23, 2019, 12:12

I would like to ask if there is a standard OpenFOAM construction/macro/??? to write the following algorithm in a more concise way:

Code:

forAll(TPatch, faceI)
{
    if (TPatch[faceI] > whatever)
    {
         xPatch[faceI] = get_something(TPatch[faceI] );
    }
     else
    {
         xPatch[faceI] = get_something_else(TPatch[faceI]);
    }
}

something like:

Code:

xPatch = (TPach > watever ? get_something(TPatch) : get_something_else(TPatch))

mAlletto · May 25, 2019, 13:34

I do not know precisely, but when construction the pressure contribution in the energy equation of of compressible flow it is done something similar:

Code:

    fvScalarMatrix EEqn     (         fvm::ddt(rho, he) + fvm::div(phi, he)   

    + fvc::ddt(rho, K) + fvc::div(phi, K)     

  + (             he.name() == "e"           ? fvc::div             (                 fvc::absolute(phi/fvc::interpolate(rho), U),                 p,                 "div(phiv,p)"             )           : -dpdt         )     

  - fvm::laplacian(turbulence.alphaEff(), he)  

   ==         rho*(U&g)       + rad.Sh(thermo, he)       + Qdot       + fvOptions(rho, he)     );

jherb · May 25, 2019, 18:04

Thank you for your answer. But this is only testing the name of the energy field and selected on of two possible ways for the calculation.

What I look for is a way to calcuate a field differently for each cell based on a condition.

alexeym · May 27, 2019, 08:06

Hi,

@jherb

There are neg and pos functions for fields, so your code could look like:

Code:

xPatch = pos(TPach - watever)*get_something(TPatch) + neg(TPach - watever)*get_something_else(TPatch))

jherb · May 30, 2019, 08:38

@alexeym: Thank you for your answer. This will work (if using neg0 to also catch all cells with value 0). But it would require to two both calculations/function calls for each cell. Also I am not sure it this would not require some temporary fields. So for large meshes this might be significantly slower than a cell-by-cell decision.

alexeym · May 30, 2019, 18:43

@jherb, well, surely it won't be faster (for any size of the mesh), since it requires creation of temporary fields and (at least) double evaluation of TPach - watever. Yet there is "vectorisation" mantra. Which is usually supported (or imposed) by certain prof.

Also your code is a candidate for either transforming into macro or a template.

May 23, 2019, 12:12	OpenFOAM construct for conditional field modification	#1
jherb Senior Member Joachim Herb Join Date: Sep 2010 Posts: 650 Rep Power: 22	I would like to ask if there is a standard OpenFOAM construction/macro/??? to write the following algorithm in a more concise way: Code: forAll(TPatch, faceI) { if (TPatch[faceI] > whatever) { xPatch[faceI] = get_something(TPatch[faceI] ); } else { xPatch[faceI] = get_something_else(TPatch[faceI]); } } something like: Code: xPatch = (TPach > watever ? get_something(TPatch) : get_something_else(TPatch))

May 27, 2019, 08:06		#4
alexeym Senior Member Alexey Matveichev Join Date: Aug 2011 Location: Nancy, France Posts: 1,938 Rep Power: 39	Hi, @jherb There are neg and pos functions for fields, so your code could look like: Code: xPatch = pos(TPach - watever)get_something(TPatch) + neg(TPach - watever)get_something_else(TPatch))

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Other] dynamicTopoFVMesh and pointDisplacement	RandomUser	OpenFOAM Meshing & Mesh Conversion	6	April 26, 2018 08:30
How to contribute to the community of OpenFOAM users and to the OpenFOAM technology	wyldckat	OpenFOAM	17	November 10, 2017 16:54
Suggestion for a new sub-forum at OpenFOAM's Forum	wyldckat	Site Help, Feedback & Discussions	20	October 28, 2014 10:04
Moving mesh	Niklas Wikstrom (Wikstrom)	OpenFOAM Running, Solving & CFD	122	June 15, 2014 07:20
Problem with rhoSimpleFoam	matteo_gautero	OpenFOAM Running, Solving & CFD	0	February 28, 2008 07:51

May 25, 2019, 18:04		#3
jherb Senior Member Joachim Herb Join Date: Sep 2010 Posts: 650 Rep Power: 22	Thank you for your answer. But this is only testing the name of the energy field and selected on of two possible ways for the calculation. What I look for is a way to calcuate a field differently for each cell based on a condition.

May 30, 2019, 08:38		#5
jherb Senior Member Joachim Herb Join Date: Sep 2010 Posts: 650 Rep Power: 22	@alexeym: Thank you for your answer. This will work (if using neg0 to also catch all cells with value 0). But it would require to two both calculations/function calls for each cell. Also I am not sure it this would not require some temporary fields. So for large meshes this might be significantly slower than a cell-by-cell decision.

May 30, 2019, 18:43		#6
alexeym Senior Member Alexey Matveichev Join Date: Aug 2011 Location: Nancy, France Posts: 1,938 Rep Power: 39	@jherb, well, surely it won't be faster (for any size of the mesh), since it requires creation of temporary fields and (at least) double evaluation of TPach - watever. Yet there is "vectorisation" mantra. Which is usually supported (or imposed) by certain prof. Also your code is a candidate for either transforming into macro or a template.