The temperature is not constant at an isothermal wall BC

[LDK29]

Hello
I am doing simulations of a hypersonic cylinder flow. I defined an isothermal BC for the cylinder. But the results show that the temperature at the cylinder is not constant. The picture below shows the temperature distribution of the cylinder.

Another question is the EULER BC. The Mach number near the EULER BC is not reasonable. You could find it in the picture below. I have seen some similar issues posted before, it seems still remain. So is this problem fixed in develop branch?

My configuration file is shown below. Not just for the problems I post, but any advice about my configurations is useful for me.

`_**% ------------- DIRECT, ADJOINT, AND LINEARIZED PROBLEM DEFINITION ------------%
%
SOLVER= NEMO_NAVIER_STOKES
GAS_MODEL= AIR-5
GAS_COMPOSITION= (0.7543, 0.00713, 0.01026, 1e-5, 0.2283)
MATH_PROBLEM= DIRECT
AXISYMMETRIC= NO
RESTART_SOL= _NO

% ----------- COMPRESSIBLE AND INCOMPRESSIBLE FREE-STREAM DEFINITION ----------%
%
MACH_NUMBER= 8.98
AOA= 0.0
SIDESLIP_ANGLE= 0.0
INIT_OPTION= TD_CONDITIONS
FREESTREAM_PRESSURE= 476.0
FREESTREAM_TEMPERATURE= 901
FREESTREAM_TEMPERATURE_VE= 901
REYNOLDS_NUMBER= 10000
KIND_TURB_MODEL= NONE

% ---- NONEQUILIBRIUM GAS, IDEAL GAS, POLYTROPIC, VAN DER WAALS AND PENG ROBINSON CONSTANTS -------%
%
FLUID_MODEL= SU2_NONEQ

% -------------------- BOUNDARY CONDITION DEFINITION --------------------------%
%
MARKER_EULER= ( SYM)
MARKER_ISOTHERMAL= (CYL, 300)
MARKER_OUTLET= ( OULET, 5 )
MARKER_FAR= ( INLET )
MARKER_PLOTTING= (CYL )

%CATALYTIC_WALL = (CYL)
%SUPERCATALYTIC_WALL = YES
%SUPERCATALYTIC_WALL_COMPOSITION=(0.77,0.23,0,0,0)

MARKER_MONITORING= ( CYL )

% ------------- COMMON PARAMETERS DEFINING THE NUMERICAL METHOD ---------------i%
%
TIME_DOMAIN = NO
NUM_METHOD_GRAD= WEIGHTED_LEAST_SQUARES
%(WEIGHTED_LEAST_SQUARES,GREEN_GAUSS,LEAST_SQUARES)
%
CFL_NUMBER= 0.25
CFL_ADAPT=NO
CFL_ADAPT_PARAM=(0.2,1.2,0.25,0.8,1e-6)
ITER= 100000
%LINEAR_SOLVER= BCGSTAB
%(FGMRES,BCGSTAB,...)
%LINEAR_SOLVER_ERROR= 1E-6
%LINEAR_SOLVER_ITER= 10
%LINEAR_SOLVER_PREC =LU_SGS

% -----------------------------------------------------------------------%
%
CONV_NUM_METHOD_FLOW= MSW
%(AUSM,MSW,....)
ENTROPY_FIX_COEFF= 1e-6
MUSCL_FLOW= YES
SLOPE_LIMITER_FLOW= VENKATAKRISHNAN
VENKAT_LIMITER_COEFF= 0.05
TIME_DISCRE_FLOW= EULER_EXPLICIT
%(RUNGE_KUTTA_EXPLICIT, EULER_EXPLICIT, EULER IMPLICIT)

% --------------------------- CONVERGENCE PARAMETERS --------------------------%
%
CONV_RESIDUAL_MINVAL= -25
CONV_STARTITER= 10000

% ------------------------- INPUT/OUTPUT INFORMATION --------------------------%
%
% Mesh input file
MESH_FILENAME= project3.cgns
%
% Mesh input file format (SU2, CGNS)
MESH_FORMAT= CGNS

%MESH_FILENAME= viscwedge.su2
%MESH_FORMAT= SU2
HISTORY_OUTPUT= ITER, RMS_RES
SOLUTION_FILENAME= restart_flow.dat
TABULAR_FORMAT= TECPLOT
CONV_FILENAME= convergence
RESTART_FILENAME= restart_flow.dat
VOLUME_OUTPUT=SOLUTION,AUXILIARY, PRIMITIVE,VORTEX_IDENTIFICATION
VOLUME_FILENAME= soln_volume
SURFACE_FILENAME= soln_surface
OUTPUT_WRT_FREQ= 2500
WRT_RESTART_OVERWRITE= YES
OUTPUT_FILES= (RESTART, PARAVIEW, SURFACE_PARAVIEW)`**_

[LDK29]

Hi
About the second question I post, I have found that is my fault for a sloping symmetry. There is a 1e-9 m difference between the two endpoints of the symmetry in y direction.
For the first question, I would be grateful if someone can help me.

[WallyMaier]

Hi,
Regarding the symmetry boundary, this is a known issue in the SU2 solver, and we often avoid this by simulating a full domain....I know it isn't as computationally efficient.

The Isothermal wall condition behavior seems to be OK. The isothermal wall is a weakly enforced boundary condition, so we don't set the fluid temperature at the wall to be a certain temperature. The saw-tooth pattern is odd, and may be due to the numerical inaccuracies caused by the symmetry plane.

[LDK29]

Hi
Many thanks for the reply. I will have a try for the full domain later.