Simulation of hypersonic flow: NaN in the residual

smanist · January 26, 2015, 01:43

Hi guys,

I have just begun using SU2, which I hope I can use to simulate hypersonic flow with real gas effects. I assumed that is possible, because I noticed the existence of TNE2 module in SU2. I have tried the RAMC-II test case and ran a Mach 12 double-wedge airfoil case for Euler solution. They worked well.

However, when I try to get RANS solution of the double-wedge case, I get this error

Code:

------------------------------ Begin Solver -----------------------------

!!! Error: There is a NaN in the residual. Now exiting... !!!

In my case, I configured flow condition to be H=12km in atmosphere,

Code:

FREESTREAM_TEMPERATURE= 220.61
REYNOLDS_NUMBER= 8.07145E7
REYNOLDS_LENGTH= 1.0

The output of SU2 is correct, as it shows

Code:

Specific gas constant: 287.058 N.m/kg.K.
Free-stream pressure: 20646.1 Pa.
Free-stream temperature: 220.61 K.
Free-stream density: 0.326019 kg/m^3.
Free-stream velocity: (3573.08, 0) m/s. Magnitude: 3573.08 m/s.
Free-stream total energy per unit mass: 6.54177e+06 m^2/s^2.
Free-stream viscosity: 1.44323e-05 N.s/m^2.

For the rest of the configure file, the key settings are, (the whole file is just too long to be posted here)

Code:

PHYSICAL_PROBLEM= NAVIER_STOKES
KIND_TURB_MODEL= SA
VISCOSITY_MODEL= SUTHERLAND
MARKER_HEATFLUX= ( UPPER, 0.0, LOWER, 0.0 ) % Wall BC
MARKER_FAR= ( INLET, OUTLET ) % Farfield
CONV_NUM_METHOD_FLOW= ROE
SPATIAL_ORDER_FLOW= 2ND_ORDER
SLOPE_LIMITER_FLOW= VENKATAKRISHNAN
TIME_DISCRE_FLOW= EULER_IMPLICIT
CONV_NUM_METHOD_TURB= SCALAR_UPWIND
SPATIAL_ORDER_TURB= 2ND_ORDER
SLOPE_LIMITER_TURB= VENKATAKRISHNAN
TIME_DISCRE_TURB= EULER_IMPLICIT

I also tried Mach 2 (Re=1.34524E7), after several iterations, same issue happens. For Mach 3 (Re=2.01786E7), same thing, except that there is also a message,

Code:

There are 4 non-physical states in the upwind reconstruction.

I am still playing with the setting to see what happens. Any idea what might be wrong here? Any help will be appreciated.

Thank you in advance!
Danny

hlk · January 26, 2015, 19:05

Thank you for your question

For changing the flight conditions, please note that the solution will be sensitive to the quality of the mesh, and that a mesh that is well tuned for a particular Reynolds number and Mach number may not perform well at other conditions due to not having enough refinement in the boundary layer or near shocks. Similarly, when switching between models a different mesh may be needed - especially when switching between an inviscid (euler) and viscous model (rans).

If you have not done so already, I recommend making a new mesh for the wedge which has y+ <= 1 at the wall. (cfd-online has a calculator that can help with this at http://www.cfd-online.com/Tools/yplus.php).

Quote:

Originally Posted by smanist

Hi guys,

I have just begun using SU2, which I hope I can use to simulate hypersonic flow with real gas effects. I assumed that is possible, because I noticed the existence of TNE2 module in SU2. I have tried the RAMC-II test case and ran a Mach 12 double-wedge airfoil case for Euler solution. They worked well.

However, when I try to get RANS solution of the double-wedge case, I get this error

Code:

------------------------------ Begin Solver -----------------------------

!!! Error: There is a NaN in the residual. Now exiting... !!!

In my case, I configured flow condition to be H=12km in atmosphere,

Code:

FREESTREAM_TEMPERATURE= 220.61
REYNOLDS_NUMBER= 8.07145E7
REYNOLDS_LENGTH= 1.0

The output of SU2 is correct, as it shows

Code:

Specific gas constant: 287.058 N.m/kg.K.
Free-stream pressure: 20646.1 Pa.
Free-stream temperature: 220.61 K.
Free-stream density: 0.326019 kg/m^3.
Free-stream velocity: (3573.08, 0) m/s. Magnitude: 3573.08 m/s.
Free-stream total energy per unit mass: 6.54177e+06 m^2/s^2.
Free-stream viscosity: 1.44323e-05 N.s/m^2.

For the rest of the configure file, the key settings are, (the whole file is just too long to be posted here)

Code:

PHYSICAL_PROBLEM= NAVIER_STOKES
KIND_TURB_MODEL= SA
VISCOSITY_MODEL= SUTHERLAND
MARKER_HEATFLUX= ( UPPER, 0.0, LOWER, 0.0 ) % Wall BC
MARKER_FAR= ( INLET, OUTLET ) % Farfield
CONV_NUM_METHOD_FLOW= ROE
SPATIAL_ORDER_FLOW= 2ND_ORDER
SLOPE_LIMITER_FLOW= VENKATAKRISHNAN
TIME_DISCRE_FLOW= EULER_IMPLICIT
CONV_NUM_METHOD_TURB= SCALAR_UPWIND
SPATIAL_ORDER_TURB= 2ND_ORDER
SLOPE_LIMITER_TURB= VENKATAKRISHNAN
TIME_DISCRE_TURB= EULER_IMPLICIT

I also tried Mach 2 (Re=1.34524E7), after several iterations, same issue happens. For Mach 3 (Re=2.01786E7), same thing, except that there is also a message,

Code:

There are 4 non-physical states in the upwind reconstruction.

I am still playing with the setting to see what happens. Any idea what might be wrong here? Any help will be appreciated.

Thank you in advance!
Danny

smanist · January 26, 2015, 22:03

I see! Thanks for your advice. Let me refine my mesh.

January 26, 2015, 01:43	Simulation of hypersonic flow: NaN in the residual	#1
smanist New Member David Hwang Join Date: Feb 2014 Posts: 5 Rep Power: 12	Hi guys, I have just begun using SU2, which I hope I can use to simulate hypersonic flow with real gas effects. I assumed that is possible, because I noticed the existence of TNE2 module in SU2. I have tried the RAMC-II test case and ran a Mach 12 double-wedge airfoil case for Euler solution. They worked well. However, when I try to get RANS solution of the double-wedge case, I get this error Code: ------------------------------ Begin Solver ----------------------------- !!! Error: There is a NaN in the residual. Now exiting... !!! In my case, I configured flow condition to be H=12km in atmosphere, Code: FREESTREAM_TEMPERATURE= 220.61 REYNOLDS_NUMBER= 8.07145E7 REYNOLDS_LENGTH= 1.0 The output of SU2 is correct, as it shows Code: Specific gas constant: 287.058 N.m/kg.K. Free-stream pressure: 20646.1 Pa. Free-stream temperature: 220.61 K. Free-stream density: 0.326019 kg/m^3. Free-stream velocity: (3573.08, 0) m/s. Magnitude: 3573.08 m/s. Free-stream total energy per unit mass: 6.54177e+06 m^2/s^2. Free-stream viscosity: 1.44323e-05 N.s/m^2. For the rest of the configure file, the key settings are, (the whole file is just too long to be posted here) Code: PHYSICAL_PROBLEM= NAVIER_STOKES KIND_TURB_MODEL= SA VISCOSITY_MODEL= SUTHERLAND MARKER_HEATFLUX= ( UPPER, 0.0, LOWER, 0.0 ) % Wall BC MARKER_FAR= ( INLET, OUTLET ) % Farfield CONV_NUM_METHOD_FLOW= ROE SPATIAL_ORDER_FLOW= 2ND_ORDER SLOPE_LIMITER_FLOW= VENKATAKRISHNAN TIME_DISCRE_FLOW= EULER_IMPLICIT CONV_NUM_METHOD_TURB= SCALAR_UPWIND SPATIAL_ORDER_TURB= 2ND_ORDER SLOPE_LIMITER_TURB= VENKATAKRISHNAN TIME_DISCRE_TURB= EULER_IMPLICIT I also tried Mach 2 (Re=1.34524E7), after several iterations, same issue happens. For Mach 3 (Re=2.01786E7), same thing, except that there is also a message, Code: There are 4 non-physical states in the upwind reconstruction. I am still playing with the setting to see what happens. Any idea what might be wrong here? Any help will be appreciated. Thank you in advance! Danny Last edited by smanist; January 26, 2015 at 01:45. Reason: Typos

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
buoyantSimpleFoam and watertank	Tobi	OpenFOAM Running, Solving & CFD	100	December 18, 2022 09:15
Cannot run the code properly: very large time step continuity error	crst15	OpenFOAM Running, Solving & CFD	9	December 14, 2014 19:17
SixDoFRigidBodyMotion under OF2.3 ( self oscillating cylinder)	Scabbard	OpenFOAM Running, Solving & CFD	1	July 22, 2014 05:50
pisoFoam with k-epsilon turb blows up - Some questions	Heroic	OpenFOAM Running, Solving & CFD	26	December 17, 2012 04:34
Could anybody help me see this error and give help	liugx212	OpenFOAM Running, Solving & CFD	3	January 4, 2006 19:07

January 26, 2015, 22:03		#3
smanist New Member David Hwang Join Date: Feb 2014 Posts: 5 Rep Power: 12	I see! Thanks for your advice. Let me refine my mesh.