[zhengjg]

Dear SU2 developers and users,
I am using SU2 eagle in my research.
But I encountered a few problems which have not been resolved.

1. How can I add an constant source term to energy equation in SU2 3.0 eagle? Is there any module in the software suit to implement this source term?

2. Is there module to add user-defined boundary conditions?

3. Is there any tutorial or example of demonstrating the setup of an transient flow simulation using explicit Runge-Kutta or dual-time-stepping methodologies, say the flow past a cylinder at higher Reynolds number with periodic vortex shedding as shown in flyer of SU2 eagle and earlier version?
I am interested in parameters configuration and setup involved in the two time discretization for unsteady flow.

Thank you very much
JG

[fpalacios]

Originally Posted by zhengjg

Dear SU2 developers and users,
I am using SU2 eagle in my research.
But I encountered a few problems which have not been resolved.

1. How can I add an constant source term to energy equation in SU2 3.0 eagle? Is there any module in the software suit to implement this source term?

2. Is there module to add user-defined boundary conditions?

3. Is there any tutorial or example of demonstrating the setup of an transient flow simulation using explicit Runge-Kutta or dual-time-stepping methodologies, say the flow past a cylinder at higher Reynolds number with periodic vortex shedding as shown in flyer of SU2 eagle and earlier version?
I am interested in parameters configuration and setup involved in the two time discretization for unsteady flow.

Thank you very much
JG

Dear JG,

Thanks a lot for your interest in SU2.
To add a source term in SU2 is easy but you should be familiar with C++: There is an elegant way based on creating a class in Numerics (files numerics_direct_mean.cpp, and numerics_structure.hpp). The class CSourceGravity is a good example. Once you have created the class, you should go to Numerics_Preprocessing subroutine (definition_structure.cpp) and allocate the class
numerics_container[iMGlevel][FLOW_SOL][SOURCE_FIRST_TERM] = new CSourceGravity(nDim, nVar_Flow, config);

There is a quick way to add a source term just modifying the residual in the subroutine
void CEulerSolver::Source_Residual(CGeometry *geometry, CSolver **solver_container, CNumerics *numerics, CNumerics *second_numerics, CConfig *config, unsigned short iMesh)

To add a new boundary condition, my recommendation is to copy one of the typical ones. Like
void CEulerSolver::BC_Far_Field(CGeometry *geometry, CSolver **solver_container, CNumerics *conv_numerics,
CNumerics *visc_numerics, CConfig *config, unsigned short val_marker).

And add the new boundary condition at the end of
void CIntegration::Space_Integration(CGeometry *geometry, CSolver **solver_container, CNumerics **numerics, CConfig *config, unsigned short iMesh, unsigned short iRKStep, unsigned short RunTime_EqSystem).

There are some unsteady problems in the Testcases folder, for example in the TestCases/optimization_euler/pitching_naca64a010/ folder. But as far I remember in this forum (maybe an old post) we also provided the config case for an unsteady problem. Anyway to run an unsteady problem, at least you should add the following to a typical config file.

% ------------------------- UNSTEADY SIMULATION -------------------------------%
%
% Unsteady simulation (NO, TIME_STEPPING, DUAL_TIME_STEPPING-1ST_ORDER,
% DUAL_TIME_STEPPING-2ND_ORDER, TIME_SPECTRAL)
UNSTEADY_SIMULATION= DUAL_TIME_STEPPING-2ND_ORDER
%
% Time Step for dual time stepping simulations (s)
UNST_TIMESTEP= 0.001
%
% Total Physical Time for dual time stepping simulations (s)
UNST_TIME= 1.0
%
% Number of internal iterations (dual time method)
UNST_INT_ITER= 200


Best,
Francisco