[LauraSumner]

Hi!

I have created a solver based on pimpleDyMFoam which I can post process using paraview to find the pathlines of particles injected into the flow using particle Tracer etc.

However, this is incredibly slow and time consuming, and the main issue is the paraView just crashes.

I need to be able to find the paths of particles in my flow for long times and therefore large datasets. My paraView simply cannot handle this. Plus this is a pretty ugly way of finding such information.

My initial thought was to research how to post process this kind of thing using the command line instead of running paraView. However the most elegant solution would be some combination of solvers which have massless particles added to the flow, so that I can just output their coordinates at each timestep instead of post processing at all. The solution must run in parallel as well!

My questions are:
1) Is this possible?
2) If so, is my method below anywhere near the right kind of thing?
3) If it is, what am i doing wrong?!

METHOD:


I am trying to merge the icoUncoupledKinematicParcelFoam somehow with my solver, because I thought that because of this line:

U.correctBoundaryConditions();

it would read in the newly calculated velocity field (U) and evolve the cloud accordingly. I have looked at other people's solvers that I have found from researching this problem and it seems they do similar things, however the cases I have found have been for old versions of OpenFOAM (I use 3.0.1) so it's hard to compare/I cant compile their code.

So far I have added all the files I think I need to the options file of my new solver, and added these lines of code within my time loop:

Info<< "Evolving " << kinematicCloud.name() << endl;

U.correctBoundaryConditions();

laminarTransport.correct();

mu = laminarTransport.nu()*rhoInfValue;

kinematicCloud.evolve();


(i also added this at the beginning of my main:
argList::addOption
(
"cloudName",
"name",
"specify alternative cloud name. default is 'kinematicCloud'"
);
)



when i compile i get errors saying:

error: expected primary-expression before ‘.’ token
Info<< "Evolving " << kinematicCloud.name() << endl;


and that

error: ‘mu’ was not declared in this scope
mu = laminarTransport.nu()*rhoInfValue;


which is strange because I am sure all the header files called by the icoUncoupledKinematicParcelFoam are included and that solver returns no such errors.



Perhaps this is also a foolish idea to begin with, any insight would be appreciated!!

Thanks

Laura

[rasool_soofi]

Hi

did you ever find out the solution? Having the same issue myself!