Coupling and mesh properties

[Halvor Aslaksen]

Hi!

My lab is preparing to make a coupling between Project Chrono and a CFD program (REEF3D). Our plan is to construct a mesh with the CFD program (.stl format), and communicate between the programs similar to the DualSPHysics coupling (we'll naturaly have to do this a bit different as REEF3D is not particle-based). REEF3D is a rigid mesh-based CFD program (mesh doesn't move), where information like pressure is saved in the mesh center and velocities at the mesh edges. 

We have a rough outline of what we need, but I just wanted to comfirm a couple of aspects.

When importing the mesh into Project Chrono, together with the live transfer of forces and moments calculated in REEF3D from a wave/waves, will the forces have to be applied on all mesh cells individually in Project Chrono or on a broader area? And will Project Chrono have to calculate these forces on the nodes itself,  and store information (like velocities and positions) on each individual node in Project Chrono before sending updated positions back to REEF3D? 

Thanks for any answers!

Halvor

[Bonaventura TAGLIAFIERRO]

Hi Halvor,

Thank you very much for posting this conversation!

Very nice to know that REEF3D is considering a coupling with Chrono.

In the coupling between DualSPHysics and the FEM module from Chrono, information is gathered from the fluid-to-structure interaction (this example is for a beam element), and then forces are applied to the nodes of the mesh formulated in Chrono. This is what the authors of the coupling decided to do to make it more efficient, also because it reuses functions already developed for evaluating the accelerations on solid geometries (floatings in DualSPHysics). In fact, each node's beam in Chrono has a corresponding rigid body in DualSPHysics. Forces on nodes worked just fine, so the coupling was finalized using the former approach. Note that for beams it is easier to built a very easy interface because around 15-20 nodes are sufficient to capture the dynamics of a flexible beam!

In general, you should first identify the element from Chrono that you want to couple, then it can be possible to define a strategy for it. Maybe you could create your sandbox considering a beam elements (there are many), and from there you can take it further. The structure of the coupling shouldn't change much when considering more complex structural models.

Please, feel free to ask for more details on this.

Regards,

Bonaventura

[Halvor Aslaksen]

Hi Bonaventura

Thank you for the quick and helpful answer. We will discuss what you've explained, and might come back with more detailed questions.

Best regards

Halvor