hello just an small detail to add (someone commented too..) surfaceTransformPoints has almost (i say almost as it does not have the decompose option) that transformPoints but it is for the surface mesh (.stl) :) if anyone is looking for that is nice to know
thanks mr jozef for the video. How do you set for the base case i.e. getting 0, constant, and system folder and how to process the .obj file become the polymesh data so that we can open the .foam file?
Hi, is there a way to transform the points and keep it fixed and saved onto that location. There is an STL file, which, when I import, does not coincide with the origin of the blockMesh.
Hi Jozsef, thanks for the video. I work with OF for urban wind comfort. I was thinking if its possible to "cut out" a proposed building and insert another building proposal in it's place, without having to remesh the entire context. Thanks, Mathias
Nice and useful video! vielen Dank! I have a question for the sets after scaling: I want to run a simulation in micrometer scale. I did blockMesh and snappyHexmesh all in default(in meter scale) and had transferred all the points into um scale, these are all ok. unfortunately, after running I found an unacceptable result. I think the reason is I should also change the set on boundary conditions and time-steps or fvsolutions? If I have to do this, is there any rules? All things times 1e-6?
You have to enter everything in SI units. Maybe you added an inlet velocity of x um/s but instead you have to use x 1e-6 m/s. Possibly this could be it. But difficult to tell.
Good video. Is there any tool available you know about which can rotate fields which are located in the postProcessing folder as well? (e.g. fields from a specified surface written out at specific time steps)
Hi Jozsef, thank you for the interesting content. I have a question to ask: is it possible to translate mesh in between solution, e.g. I complete 0.8 s simulation, and then I translate the mesh, and re-run the simulation for another 0.8 s, with a total simulation of 1.6 s. Is is possible ?
Hello Jozsef, saw your video. I want to divide my inlet patch into 3 different patches to assign different BC. Is there any utility for this? Also I am using blockMesh for generating my mesh.
hello, thank you for the video, it made me realise that there might be a way of translating/rotating only the products of the snappyhexmesh without moving the boundary volume. i tried the surfaceTransformPoints command but it just moves my stl, which in retrospect i should have anticipated. Do you know of any ways you can achieve that?
József Nagy József Nagy yes, exactly. To give some context, we have a formula student car and at some point in our research we need to test the car in various heights from the road, so by just moving the inner boundaries we could save a lot of computational time. Otherwise we would have to crreate a mesh for the same geometry around 10 times ( i suppose the same principle will apply for the roll and pitch )
@@apotl8584 The question is, how big the movement is compared to the actual geometry. The mesh might be morphed in a strong way and then your cell quality will suffer.
@@OpenFOAMJozsefNagy the movement would be a maximum of 20 millimeters while the geometry is around (2.5 , 1.5 ,1.2 ) in meters in the x,y,z axis accordingly. I suppose that an evaluation of the results would be in order, assuming the mesh movement is feasible
I would never do that as STL consists of triangles and the quality of the cells would never be good for a CFD simulation. STL is only there as a guide for hex dominant meshers to do the job,