FEM_EULER vs. EULER Setup?

CSMDakota · October 4, 2019, 18:53

Hello,

I have a simple converging diverging nozzle case with a hexahedral mesh that converges OK to 12 orders residual reduction using traditional FVM "EULER" solver with SU2 (v6.2). I'd like to experiment with the FEM_EULER DG solver in SU2, but haven't found much documentation or examples. When I try to run with FEM_EULER, the solution throws an early error during the Geometry Preprocessing:

Error in "void CMeshFEM_DG::CreateFaces(CConfig*))":
Boundary face not found in localFaces. This should not happen.

Any suggestions on this error?

Thanks!
Brandon

pcg · October 16, 2019, 14:13

You need a high order mesh to run the high order (FEM) solver, as with all SU2 features the best line of action is to start from a working example and go from there, making incremental changes.
Clone the TestCase repo and have a look at the ones in folders with prefix "hom_" (high order methods).
There is only one developer for those solvers (and I'm not him) so I cannot help much more.

CSMDakota · October 16, 2019, 14:58

Thanks, Pedro - this helps, I wasn't using a higher order mesh, assumed the quadrature points would be generated by the solver on the base grid. I will have a look at the examples in the repo.

October 4, 2019, 18:53	FEM_EULER vs. EULER Setup?	#1
CSMDakota New Member Brandon Gleeson Join Date: Apr 2018 Posts: 26 Rep Power: 8	Hello, I have a simple converging diverging nozzle case with a hexahedral mesh that converges OK to 12 orders residual reduction using traditional FVM "EULER" solver with SU2 (v6.2). I'd like to experiment with the FEM_EULER DG solver in SU2, but haven't found much documentation or examples. When I try to run with FEM_EULER, the solution throws an early error during the Geometry Preprocessing: Error in "void CMeshFEM_DG::CreateFaces(CConfig*))": Boundary face not found in localFaces. This should not happen. Any suggestions on this error? Thanks! Brandon

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Euler unsteady with moving boundary conditions - problem setup	rocco.bom	SU2	0	January 24, 2017 10:11
2D Glass Melt Simulation Setup	marmz	FLUENT	5	October 9, 2016 16:25
hard to convergent with SU2 4.0 for Euler adjoint	Xianguu	SU2	1	July 14, 2015 18:09
[ICEM] surface/curve mesh setup	Studi	ANSYS Meshing & Geometry	15	November 12, 2014 01:32
[ICEM] Hexa mesh, curve mesh setup, bunching law	Anorky	ANSYS Meshing & Geometry	4	November 12, 2014 01:27

October 16, 2019, 14:13		#2
pcg Senior Member Pedro Gomes Join Date: Dec 2017 Posts: 466 Rep Power: 14	You need a high order mesh to run the high order (FEM) solver, as with all SU2 features the best line of action is to start from a working example and go from there, making incremental changes. Clone the TestCase repo and have a look at the ones in folders with prefix "hom_" (high order methods). There is only one developer for those solvers (and I'm not him) so I cannot help much more.

October 16, 2019, 14:58		#3
CSMDakota New Member Brandon Gleeson Join Date: Apr 2018 Posts: 26 Rep Power: 8	Thanks, Pedro - this helps, I wasn't using a higher order mesh, assumed the quadrature points would be generated by the solver on the base grid. I will have a look at the examples in the repo.