[charan3007]

Hi everyone,
I am trying to simulate NACA 0012 wing in m=0.03 and Reynold number =129000. i have been using ROE. below is cfg which i am using. sst is not at all showing stall but some what sa is showing stall at AOA=30 which quite unusually. please help me what solver type, liner solver, preconditioner should be used in this cases, where i am doing the mistake thank you in advance

SOLVER= RANS
KIND_TURB_MODEL= SST
INNER_ITER= 3000
CONV_FIELD= DRAG
CONV_RESIDUAL_MINVAL= -3
CONV_STARTITER= 10
CONV_CAUCHY_ELEMS= 100
CONV_CAUCHY_EPS= 1E-10
MACH_NUMBER= 0.03871
AOA= 10.0
INIT_OPTION= REYNOLDS
FREESTREAM_OPTION= TEMPERATURE_FS
FREESTREAM_TEMPERATURE= 284.9
REYNOLDS_NUMBER= 129000
REYNOLDS_LENGTH= 0.15
NUM_METHOD_GRAD= GREEN_GAUSS
CFL_NUMBER= 5.0
CFL_ADAPT= NO
MUSCL_FLOW= YES
SLOPE_LIMITER_FLOW= VENKATAKRISHNAN
MUSCL_TURB= YES
SLOPE_LIMITER_TURB= VENKATAKRISHNAN
VENKAT_LIMITER_COEFF= 0.05
LINEAR_SOLVER= FGMRES
LINEAR_SOLVER_PREC= ILU
LINEAR_SOLVER_ILU_FILL_IN= 0
LINEAR_SOLVER_ERROR= 1E-6
LINEAR_SOLVER_ITER= 5
LINEAR_SOLVER_SMOOTHER_RELAXATION= 1.0
MGLEVEL= 0
CONV_NUM_METHOD_FLOW= ROE
ROE_LOW_DISSIPATION= FD
ENTROPY_FIX_COEFF= 0.01
TIME_DISCRE_FLOW= EULER_IMPLICIT
CONV_NUM_METHOD_TURB= SCALAR_UPWIND
TIME_DISCRE_TURB= EULER_IMPLICIT
CFL_REDUCTION_TURB= 1.0

[aero_amit]

First thing - Can you really predict stall with RANS models?

Coming to numerics - Mach number is very low (M=0.03), Roe scheme may not be suitable, try SLAU2 or AUSMPLUSUP instead alongwith USE_ACCURATE_FLUX_JACOBIANS=YES
Try CFL up to 10 or so (do hit and trial +-), also try SLOPE_LIMITER_FLOW= VENKATAKRISHNAN_WANG

Above may be more suitable settings, but consider the problem you are trying to solve in the first place ..... (update the findings nevertheless)

[charan3007]

thank you Amit for your reply, i came across transition model that is available in SU2 under incompressible rans using SA with BC transition mode. this has predicted the case almost near accurately with wind tunnel results. But i ahve a dobt, in Incompressible RANS +SA+BC is using 2nd order in flow equation but could able to manage only 1st order in turbulence equation (MUSCL_TRUB= NO). how much i should believe in these result, i also solve using Fluent, SU2 is outperforming Fluent in this accept. I am very doubtful with these results, please help me how i can assess this and also in 2 nd order its getting diverged. i want to know whether 2nd order is compulsory for this case or not. thanks you in advance

[fd_flux]

Originally Posted by charan3007

thank you Amit for your reply, i came across transition model that is available in SU2 under incompressible rans using SA with BC transition mode. this has predicted the case almost near accurately with wind tunnel results. But i ahve a dobt, in Incompressible RANS +SA+BC is using 2nd order in flow equation but could able to manage only 1st order in turbulence equation (MUSCL_TRUB= NO). how much i should believe in these result, i also solve using Fluent, SU2 is outperforming Fluent in this accept. I am very doubtful with these results, please help me how i can assess this and also in 2 nd order its getting diverged. i want to know whether 2nd order is compulsory for this case or not. thanks you in advance

have you solved the problem? I am in the same situation

[bigfootedrockmidget]

Looks fine to me...



based on the UQ example:
https://su2code.github.io/tutorials/..._Opt_NACA0012/

[charan3007]

I have solved it using incompressible RANS in version 7.2 with BC transition model. the Y plus is close to 1 and i also did grid convergence study, by this procedure I could be able to get confidence on the results. For my problem, i prefer V7.2 than V7.5