SU2-NEMO sphere simulation diverges
 

Hello

I am trying to simulate a sphere at Mach 15 using SU2-NEMO. So far when I run my simulation I get expected results for the first 1200-1300 iteration however afterwards the results seem to diverge (see the attached image).
I tried to reduce the Mach number to 10 where the same problems came, however at Mach 5 no problems were present.

What I can see from the results, the stagnation region does not grow as expected (the shock does not stay normal to the symmetry axis at higher iterations) which I expect to be the problem.

Has anyone had similar problems and if so, how did you solve them?

best.
Peter


Here is my config file
---------------------------------------------------------------------------------------------------------------------------------
MACH_NUMBER=15
AOA= 0.0
SIDESLIP_ANGLE= 0.0
FREESTREAM_PRESSURE= 21.96
FREESTREAM_TEMPERATURE= 246.87
FREESTREAM_TEMPERATURE_VE= 246.87
INIT_OPTION=TD_CONDITIONS

% ---- NONEQUILIBRIUM GAS, IDEAL GAS, POLYTROPIC, VAN DER WAALS AND PENG ROBINSON CONSTANTS -------%
%
FLUID_MODEL= SU2_NONEQ
GAS_MODEL= AIR-7
GAS_COMPOSITION= (5E-13, 0.767, 0.23, 0.001, 0.001, 0.000999999999, 5E-13)
TRANSPORT_COEFF_MODEL= GUPTA-YOS

% -------------------- BOUNDARY CONDITION DEFINITION --------------------------%
%
MARKER_ISOTHERMAL=(Wall, 300.0)
MARKER_FAR=(Inlet)
MARKER_OUTLET= (Exit, 21.96)%, symmetry, 21.96)
MARKER_SYM= (symmetry)

%MARKER_PLOTTING= (Wall)
%MARKER_MONITORING= (Wall)

% ------------- COMMON PARAMETERS DEFINING THE NUMERICAL METHOD ---------------%
%
NUM_METHOD_GRAD= WEIGHTED_LEAST_SQUARES
CFL_NUMBER= 3.0
ITER= 2500
LINEAR_SOLVER= BCGSTAB
LINEAR_SOLVER_ERROR= 1E-6
LINEAR_SOLVER_ITER= 5

% -----------------------------------------------------------------------%
%
CONV_NUM_METHOD_FLOW= AUSM
MUSCL_FLOW= YES
SLOPE_LIMITER_FLOW= VENKATAKRISHNAN
VENKAT_LIMITER_COEFF= 0.05
TIME_DISCRE_FLOW= EULER_IMPLICIT

% --------------------------- CONVERGENCE PARAMETERS --------------------------%
%
CONV_RESIDUAL_MINVAL= -20
CONV_STARTITER= 10

% ------------------------- INPUT/OUTPUT INFORMATION --------------------------%
%
MESH_FILENAME= SphereMesh.su2
MESH_FORMAT= SU2
SOLUTION_FILENAME= restart_flow_gy.dat
RESTART_FILENAME= restart_flow_gy.dat
TABULAR_FORMAT= TECPLOT
CONV_FILENAME= convergence
VOLUME_FILENAME= soln_volume
SURFACE_FILENAME= soln_surface
OUTPUT_WRT_FREQ= 100
SCREEN_OUTPUT= (INNER_ITER, RMS_DENSITY_0, RMS_DENSITY_1, RMS_DENSITY_2, RMS_DENSITY_3, RMS_DENSITY_4, RMS_DENSITY_5, RMS_DENSITY_6, RMS_ENERGY, RMS_ENERGY_VE, LIFT, DRAG, TOTAL_HEATFLUX)

Hi,
Is this the Ma=15 flow over a cylinder from the testcases repository? If yes, was anything changed?

Hello
Yes, I used the testcase repository where I could not find the proper mesh which lead me to create my own.
The only parts of the cfg file that should differ would be the CONV_RESIDUAL_MINVAL and OUTPUT_WRT_FREQ

Hi,
The meshes for the testcases are in a separate repository because of their size:

https://github.com/su2code/TestCases


Please try with the original mesh and see if that helps. If it does, you can compare the meshes and see where the differences are.

I have calculated this example before, but can you tell me how the boundary is defined. Where is the symmetric boundary and where is the exit?

Hi,

In the attached image, there is only a wall and farfield boundary condition.

However, in the provided cfg file, there are symmetric boundaries that are not represented in the grid.