\\wsl.localhost\Ubuntu\home\bananaboi i store my work in $FOAM_RUN, so should be fine?

should be yes, stuff in your home dir isn't supposed to disappear

Im screaming in despair. Is manually opening paraview the only way to view the mesh? Any scripts?

You can automate paraview to output images using python

From command line without having to open it

what do you mean by manually opening paraview?

Although like, it'll still need to load your data into ram, so it'll be as slow as opening paraview if your mesh is really big

You can also use tools like PyVista to view the mesh using Python

Cool, any examples please.

I mean, for now i click paraview, run a script in it (often requires processing one case by clicking things then write a script for the others)

While i can replace that initial setup with some config in a file

hopet this helps to see the mesh!

Thanks

Guys, do I use icoUncoupledKinematicParcelFoam after I run my pimpleFoam case with all the timesteps or instead of?

anyone got a solution to this

Is mapFields -consistent still robust, if the mesh remains the same but my patch type has changed? e.g. I have an inlet but now it's closed so it's a wall, but it's the same patch on the same mesh.

I would assume that mapField overwrites the boundary entry. But you can check it by using the mapFields and then go into the 0-directory and evaluate the files. In each file the patch entries are still given. If you search for it you will see what the entry is after mapFields

Has anyone dealt with block-coupled matrix systems where simple geometries converge fine, but more complex geometries lead to oscillatory residuals? I have a system of 8 strongly and weakly coupled equations with 8 unknowns. Mesh issues have largely been resolved. With weaker coupling, the solver converges, but stronger coupling causes oscillations. Currently using deflated PCG for weakly coupled and BiCGStab for strongly coupled block systems. I’m considering switching to PCG with a reference value for a full-Neumann equation to ensure uniqueness. Looking for insights on handling oscillations in block-coupled systems, similar to implicit pressure–velocity or conjugate heat transfer problems. Anyone interested in discussion on this with some know how, I can further explain the equations etc, I do not like to dump them here as it might be overwhelming for some.

Hello everyone. I’m applying for PhD programs with a focus on CFD and I’m looking for someone experienced in the field who could review my SOP. I will also be asking my supervisor to review it, but I’d like to get a more unbiased perspective from someone I don’t personally know. I understand this isn’t the kind of help people usually ask for here, so I would really appreciate it if anyone is willing to spare some time.

might be worthwhile to post it upfront so that someone can review it without having to ask you for it

can someone explain like every part of CFD, I am new to this I fear, and I know like 3 lines of C, but I can teach myself right?

you go to google and you find video, please refer to #810161483903401984

Im trying to get my own STL file into OpenFOAM, but when I try to insert said file into my ubuntu/openfoam files it says that I need permission to do that action

Where are you trying to put your file?

Guys, what do you think would be the best approach to simulate snow packing under a moving train? I'm talking about a train circulating while it's snowing, with a few cm of snow on the ground, or maybe a dozen of cm max. I'm interested in seeing how the snow might (or not) be blown away by the train and if snow might stick and pack under it.

My first thought is to go for one of the lagrangian solvers, maybe something like MPPICFoam, but I'm not sure how to simulate the snow on the ground as it would probably require a lot of particles. I'm also not sure to be able to properly model the relative motion between the train and the snow on the ground.

There are euler euler simulations for solid-gas systems, for instance in moving beds, which is something simular that you have, I would not know if that would be accurate but it could work, as there is no discrete boundary something like VOF would like, but euler-euler could work preseumably

Something like preonlab maybe?

Seems to be up their alley