Innovative Turbulence Modeling:

SST Model in ANSYS CFX ®

Technical Brief Most industrial flows include turbulent structures, which

cannot be resolved numerically on currently available
computers. To overcome these limitations, CFD methods
solve the Reynolds-averaged Navier-Stokes equations,
Accurate and using turbulence models to compute the averaged
effective turbulent turbulent stresses. These models often limit the accuracy
flow simulations
of CFD simulations.
ANSYS CFX-5™ offers a wide
range of turbulence models, k-ε model SST model
from the standard k-ε model
to Second Moment Closure
(SMC) and Large-Eddy-
Simulation (LES). SST, one of
the most effective models in Figure 1: Streamlines for plane diffuser flow for both models. The SST model predicts the
CFX-5 has been well tested separation zone in close agreement with data, whereas the k-ε model fails to capture the
for accuracy. In addition a physics of this flow entirely.
laminar to turbulant flow
transition model is now
available from ANSYS.

Turbulence Models The Shear Stress Transport

The standard k-ε model is used in the Model
prediction of most turbulent flow To solve this problem, new models have
calculations because of its robustness, been developed. One of the most
economy, and reasonable accuracy for a effective is the Shear Stress Transport
wide range of flows. However, the (SST) model of Menter, [1]. The model
model performs poorly when faced with works by solving a
non-equilibrium boundary layers. It turbulence/frequency-based model
tends to predict the onset of separation (k–ω) at the wall and k-ε in the bulk
too late and to under-predict the amount flow. A blending function ensures a
of separation. Separation influences the smooth transition between the two
Figure 2: Comparison of the axial
velocity profiles with experimental overall performance of many devices, models. The SST model performance has
data. such as diffusers, turbine blades and been studied in a large number of cases.
aerodynamic bodies. It also has a strong In a NASA Technical Memorandum, [2],
influence on other effects, such as wall SST was rated the most accurate model
heat transfer and multi-phase for aerodynamic applications.
phenomena. Predicting reduced
separation usually results in an Dr. Menter heads the turbulence
optimistic prediction of machine modeling program at ANSYS CFX,
performance. In some applications, this resulting in a robust accurate

CFX ® ™
can have dangerous consequences, a
notable example being the prediction of
wing stall on airplanes.
implementation of SST in CFX-5.
Advanced near-wall treatments have
been developed, allowing users to
benefit from SST, even for grids with
reduced near wall spacing.
www.ansys.com
 Innovative Turbulence Modeling:
SST Model in ANSYS CFX ®

Technical Brief A typical application of SST is the flow Reynolds number formulation to a wall
separation in a plane diffuser, shown in function treatment based on grid density.
Figures 1 and 2. The model predicts The user can then make optimal use of the
correctly the separation, which has been advanced performance of the turbulence
missed by the k-ε model. More examples model for any given grid.
can be found in [1,2] where a wide range
of Reynolds and Mach number cases have Figure 4 shows the heat transfer
been computed. prediction for a flat plate boundary layer
on a number of grids with different near-
When the separation point is fixed by the wall spacing. The usual near-wall
geometry, such as in sudden pipe treatment of the w-equation, available in
expansions, k-ε usually over predicts the standard CFD codes, leads to an
Figure 3: Non-dimensional wall heat turbulence length scale, leading to over- immediate deterioration of the results for
transfer coefficient for an axisymmetric prediction of the heat transfer at grids with reduced near wall resolution,
sudden pipe-expansion. [3]
reattachment. SST avoids this deficiency but CFX’s automatic wall treatment of
and Figure 3 shows that it is in closer the SST model gives accurate results for a
agreement with experimental data. The wide range of grid densities.
graph also shows simulations with a
2-layer k-ε model, which significantly
under predicts the Nusselt number. Conclusion
The SST turbulence model available in
CFX-5 offers significant advantages for
CFX’s Near-Wall Treatment
non-equilibrium turbulent boundary
An important issue in turbulence
layer flows, including heat transfer
modeling is the numerical treatment of
the equations in regions close to walls. predictions. CFX’s new approach to the
The near-wall formulation determines the near-wall treatment of the equations
accuracy of the wall shear stress and heat frees the user from the stringent grid
transfer predictions. CFX has recently resolution requirements of other CFD
introduced a new formulation for wall codes.
function-based models, called scalable
wall functions. This is the only available References
formulation that allows users to apply [1] Menter, F.R., “Zonal Two Equation k-ω Turbulence
Models for Aerodynamic Flows”, AIAA Paper 93-2906,
arbitrarily fine grids without violating the 1993
underlying logarithmic profile [2] Bardina, J.E., Huang, P.G. and Coakley, T.J., “Turbulence
Modeling, Validation, Testing and Development,” NASA
Figure 4: Grid sensitivity for wall heat transfer
assumptions. Technical Memorandum 110446, 1997. (see also Bardina,
predictions J.E., Huang, P.G. and Coakley, T., “Turbulence Modeling
Validation”, AIAA Paper 97-2121)
For SST, the new wall boundary [3] Baughn, J.W., et al., “Local Heat Transfer Downstream of
an Abrupt Expansion in a Circular Channel With Constant

CFX ® ™ treatment exploits the simple and robust

near-wall formulation of the k-ω model
and switches automatically from a low-
Wall Heat Flux”, Journal of Heat Transfer, Vol. 106, 1984,
pp. 789-79

ANSYS, Inc. Toll Free USA/Canada: Publication subject to change without prior notice.
www.ansys.com Southpointe
275 Technology Drive
1.866.297.9724
Toll Free Mexico:
ANSYS, ANSYS Workbench, CFX, and any and all ANSYS, Inc. product and service names are registered
trademarks or trademarks of ANSYS, Inc. or its subsidiaries located in the United States or other countries.
Canonsburg, PA 15317 001.866.297.9724 ICEM CFD is a trademark licensed by ANSYS, Inc. All other trademarks or registered trademarks are the
USA Europe: property of their respective owners.
724.746.3304 44.870.010.4456
ansysinfo@ansys.com eu.sales@ansys.com ©2004 ANSYS, Inc. All Rights Reserved. PDF0072 11/04