How to add new nodes in a sliding mesh simulation

Simo2001 · June 22, 2024, 13:01

Hello everyone,

I'm currently working with sliding meshes in su2 and I'm trying to modify the code to perform a new interface treatment.

The idea I developed requires to add some nodes in the grid in specific positions.
I know that, for each new point, I will have to define the coordinates of the new point but also, the neighbouring points, the linked edges, the connectivity, whether it belongs to a marker or not and possibly other things.
I assume I will have to work with the CGeometry class and its child classes but I don't know how exactly.

Does anyone know how to do that?

Thanks
S

bigfootedrockmidget · June 26, 2024, 14:43

This is very challenging.
Does it really need to be performed in real-time? If it is not needed every iteration, you can create a python function for it and just create an entire new instance of basically everything.

Simo2001 · July 17, 2024, 08:03

Actually, it's not needed every iteration. It has to be done only once at the start of the simulation.

bigfootedrockmidget · July 17, 2024, 17:48

Then I would just read the su2 mesh in python or something similar and manipulate the mesh. Just make sure that the mesh is conformal, no hanging nodes allowed.
Or if you have created the geometry in gmsh, you can rebuild it from there maybe...

Simo2001 · October 10, 2024, 05:03

Hello, some time has passed and I managed to update all the geometry classes as I wished.

On the other hand, now I have problems during the inizialization of the solver, in particular when the code loads the restart file. Indeed, I've been restarting the sliding mesh simulation with a previous steady multizone simulation. The problem is that the restart file has obviously less points than my current geometry. In theory, there's the appropriate function InterpolateRestartData() of the CSolver class which should perform the required interpolation but it doesn't work for my case and the points are being moved in a wrong way.

The problem is with how InterpolateRestartData() builds the variable Restart_Data which is the structure which contains the solution of each point of the restart file and the starting data for the following interpolation.

I tried to modify this function but I'm not able to manage how Restart_Data works. In particular, I noticed that it's ordered by global index and, since I had problems, I printed it out on the console and I noticed that the first half of the points contains the correct data whereas the second half it's all null.

Can someone explain me why? And how do I get the whole Restart_Data for all the points of the zone?

Thank you

June 22, 2024, 13:01	How to add new nodes in a sliding mesh simulation	#1
Simo2001 New Member Monza e Brianza Join Date: Jun 2024 Posts: 3 Rep Power: 2	Hello everyone, I'm currently working with sliding meshes in su2 and I'm trying to modify the code to perform a new interface treatment. The idea I developed requires to add some nodes in the grid in specific positions. I know that, for each new point, I will have to define the coordinates of the new point but also, the neighbouring points, the linked edges, the connectivity, whether it belongs to a marker or not and possibly other things. I assume I will have to work with the CGeometry class and its child classes but I don't know how exactly. Does anyone know how to do that? Thanks S

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Sliding mesh simulation Rotor	tomasOrt	Main CFD Forum	0	March 21, 2024 07:01
[snappyHexMesh] SnappyHexMesh/splitMeshRegion : region1 in zone "-1"	GuiMagyar	OpenFOAM Meshing & Mesh Conversion	3	August 4, 2023 13:38
2D Single Bladed VAWT Simulation with Sliding Mesh Problem	peter go	FLUENT	8	September 8, 2015 11:39
Moving mesh	Niklas Wikstrom (Wikstrom)	OpenFOAM Running, Solving & CFD	122	June 15, 2014 07:20
Icemcfd 11: Loss of mesh from surface mesh option?	Joe	CFX	2	March 26, 2007 19:10

June 26, 2024, 14:43		#2
bigfootedrockmidget Senior Member bigfoot Join Date: Dec 2011 Location: Netherlands Posts: 676 Rep Power: 21	This is very challenging. Does it really need to be performed in real-time? If it is not needed every iteration, you can create a python function for it and just create an entire new instance of basically everything.

July 17, 2024, 08:03		#3
Simo2001 New Member Monza e Brianza Join Date: Jun 2024 Posts: 3 Rep Power: 2	Actually, it's not needed every iteration. It has to be done only once at the start of the simulation.

July 17, 2024, 17:48		#4
bigfootedrockmidget Senior Member bigfoot Join Date: Dec 2011 Location: Netherlands Posts: 676 Rep Power: 21	Then I would just read the su2 mesh in python or something similar and manipulate the mesh. Just make sure that the mesh is conformal, no hanging nodes allowed. Or if you have created the geometry in gmsh, you can rebuild it from there maybe...

October 10, 2024, 05:03		#5
Simo2001 New Member Monza e Brianza Join Date: Jun 2024 Posts: 3 Rep Power: 2	Hello, some time has passed and I managed to update all the geometry classes as I wished. On the other hand, now I have problems during the inizialization of the solver, in particular when the code loads the restart file. Indeed, I've been restarting the sliding mesh simulation with a previous steady multizone simulation. The problem is that the restart file has obviously less points than my current geometry. In theory, there's the appropriate function InterpolateRestartData() of the CSolver class which should perform the required interpolation but it doesn't work for my case and the points are being moved in a wrong way. The problem is with how InterpolateRestartData() builds the variable Restart_Data which is the structure which contains the solution of each point of the restart file and the starting data for the following interpolation. I tried to modify this function but I'm not able to manage how Restart_Data works. In particular, I noticed that it's ordered by global index and, since I had problems, I printed it out on the console and I noticed that the first half of the points contains the correct data whereas the second half it's all null. Can someone explain me why? And how do I get the whole Restart_Data for all the points of the zone? Thank you