Fluent Input Rationale

1. Solver (Pressure based Solver)

Pressure Based Solver: Recommended for incompressible flows. [1]
Density based solver: recommended for compressible flows, used to study shocks

Additional info
A lot of the answers have correctly pointed out that pressure based solvers are only useful
for low Mach number or essentially incompressible flow calculations. However, to get a
deeper insight, it is also important to ask the following questions:

1. Why does a density based solver lead to higher computational cost for
incompressible flow simulations?
2. Why is pressure based solver not preferred for compressible flow simulations?
Essentially, everything boils down to deeper understanding of fluid physics. It turns out that
density based solvers calculate their timestep based on the acoustic timescale. In case of
low Mach number flows (M<0.3), speed of sound is very high (almost approaching infinity).
Therefore, in order to ensure stability of the solver, a very low timestep will be needed
leading to extremely high computational cost.

A pressure based solver solves this problem in an indirect manner. For strictly
incompressible flow calculations (M=0), a Poisson’s equation for pressure can be obtained
by manipulating the continuity and momentum equations. The pressure based solver solves
this equation in an implicit manner. It turns out that while performing this manipulation, we
have completely neglected the acoustic timescales (due to M=0 approximation).

Therefore, the pressure based solver now calculates the timestep based on flow velocity
which is much smaller than the acoustic velocity. This reduces the computational cost for
incompressible flow simulations. To get a more intuitive insight into the physics, my above
statement can also be interpreted as saying that the pressure based solver resolves all
features based on the flow timescale and neglects features emerging from acoustic
timescale.

It turns out that this advantage of calculating timestep based on flow timescales only works
well until moderate Mach numbers. As the Mach number increases, flow velocity and
acoustic velocities become comparable. Therefore, the computational cost becomes
comparable for both types of solvers. Furthermore, the incompressible flow approximation
no longer holds at moderate Mach numbers. Therefore, in addition to added computational
cost, pressure based solver will also start giving out wrong results at higher Mach numbers.
Hence, pressure based solvers cannot be useful for high Mach number flows.

In summary, fluid physics and numerical methods decide the validity of any solver. There
have been some efforts made in the past to develop numerical methods which work for all
Mach numbers. However, these efforts have not been very successful. For greater detail on
the theory behind pressure based solver, I recommend you to read Chapter 2 of Pope’s
book on Turbulent Flows. There is an article entitled “The role of Pressure” which provides a
detailed derivation behind the Poisson’s equation for pressure.
https://www.quora.com/What-is-the-difference-between-pressure-based-solver-and-density-
based-solver-in-Ansys-Fluent-What-is-the-criteria-while-selecting-one-of-them#:~:text=Pressure
%20based%20solver%20is%20generally,is%20for%20compressible%20flow
%20simulations.&text=with%20no%20difficulty.-,For%20supresonic%20flows%20it%20is
%20advisable%20to%20go%20with%20density,the%20energy%20and%20momentum
%20equations.

2. Energy Equation (Off)

The energy equation is the mathematical formulation of the law of conservation of energy
For isothermal flows and incompressible flows, there is no change in internal energy and
the pressure term in NS equation becomes independent of the thermodynamic
quantities therefore the energy equation should be turned off.

3. Model (Spalart Almaras)

A single transport equation model solving directly for a modified turbulent viscosity. Designed
specifically for aerospace applications involving wall-bounded flows on a fine near-wall mesh.
Predicts the lift coefficient most well when compared to other models for subsonic flows (K
epsilon, , K-w). [2]

4. Boundary Condition (Pressure far field inlet, far side and outlet)
Pressure far field boundary condition is used to model free stream condition of air flow. When
the flow over the wing is turbulent, a wake region is developed also referred to as trailing edge
noise. This region diminishes as we move away from the trailing edge and converts back to free
stream condition. The outlet boundary must be placed far enough until the flow characteristics
become that of free stream condition. [3, 4] hence the far field is placed at 20 times the chord
length, taken from the leading edge of the wing, which is the minimum limitation [5]

5. Meshing Approach (viscous sublayer resolution)

Wall Function Approach: fluid mixing etc. it is not valid where airflow separation is taking place

Viscous Sub layer Resolution: used to study forces and heat transfer near the wall. Used for flow
separation problems and Involves determining y+ for the model to accurately capture the
viscous sublayer by setting the element height near the wall. [6]

No of Inflation layers
Inflation layers should be sufficient enough to capture the boundary layer development. A case
study indicates that a minimum of 20 layers are required with y+ of 1 to efficiently capture the whole
boundary layer.
https://www.linkedin.com/pulse/what-y-should-i-run-part-3-how-many-inflation-layers-required-
mendis/

Note: value of y+ after meshing should be check around the wing profile to match the calculated value
of y+. Along with the velocity profile and turbulent viscosity ratio at the chord to check if both the
profiles are being captured smoothly or not.

Body sizing, element size over BOI defines the average element size inside the BOI(body of influence).

6. Y+ for Spalart Allmaras

Recommended y+ for near wall treatments is y+ < 5.

https://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node410.htm
−13
+¿ √ 74 . ℜ 14 ¿
∆ y =L× y
+¿ v
y
y= T ¿
u

T
u =
√ Tw
ρ
2
T w =0.5 C f ρ U ∞
1
C f =0.0576 ℜd 5

Skin friction coefficient represents the fraction of total dynamic pressure that is felt as shear
stress on the surface

Where L= 0.5 m

S no Y+ Velocity Re First layer Height

1 5 154 ft/s 1.564 E 6 .039 mm
2 4 154 ft/s 1.564 E 6 .031 mm
3 3 154 ft/s 1.564 E 6 .0232 mm
4 2 154 ft/s 1.564 E 6 .0155 mm
5 1 154 ft/s 1.564 E 6 .0077 mm
Calculated through https://www.pointwise.com/yplus/

7. Reference Values
These values are used as a reference by fluent to derive physical quantities and non-dimensional
quantities. [7] In our case,

L = chord length

A= projected area of wing

 Density, viscosity = air density & viscosity

Pressure = operating pressure (if specified in operating pressure then input 0)

https://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node971.htm

Additional info

Some examples of the use of reference values include the following:

 Force coefficients use the reference area, density, and velocity. In addition, the
pressure force calculation uses the reference pressure.
 Moment coefficients use the reference length, area, density and velocity. In
addition, the pressure force calculation uses the reference pressure.
 Reynolds number uses the reference length, density, and viscosity.
 Pressure and total pressure coefficients use the reference pressure, density, and
velocity.
 Entropy uses the reference density, pressure, and temperature.
 Skin friction coefficient uses the reference density and velocity.
 Heat transfer coefficient uses the reference temperature.
 Turbomachinery efficiency calculations use the ratio of specific heats.

8. Coupling Scheme (SIMPLE)

The SIMPLE algorithm uses a relationship between velocity and pressure corrections to enforce
mass conservation and to obtain the pressure field.

SIMPLEC is employed in cases that are relatively uncomplicated problems (laminar flows with
no additional models activated) in which convergence is limited by the pressure-velocity
coupling, you can often obtain a converged solution more quickly.

For complicated flows involving turbulence and/or additional physical models, SIMPLEC will
improve convergence only if it is being limited by the pressure-velocity coupling. Often it will be
one of the additional modeling parameters that limits convergence; in this case, SIMPLE and
SIMPLEC will give similar convergence rates.

Hence SIMPLE coupling scheme is preferred. [8], [5]

References
1. F. R. Menter, P.F.G., T. Esch, M. Kuntz, C. Berner, Cfd Simulations Of Aerodynamic Flows With A
Pressure-Based Method. 24TH INTERNATIONAL CONGRESS OF THE AERONAUTICAL SCIENCES,
2004.
2. Siddiqi, Z. and J.W. Lee, A computational fluid dynamics investigation of subsonic wing designs
for unmanned aerial vehicle application. Proceedings of the Institution of Mechanical Engineers,
Part G: Journal of Aerospace Engineering, 2019. 233(15): p. 5543-5552.
3. Linyue Gao*, H.Z., Yongqian Liu, Shuang Han, Effects of vortex generators on a blunt trailing-
edge airfoil for wind turbines. Elsevier Renewable Energy, 2015. 76: p. 8.
4. Muhammad Saif Ullah Khalid, I.A., Nonlinear Reduced-Order Models for Aerodynamic Lift of
Oscillating Airfoils. Journal of Computational and Nonlinear Dynamics, 2017: p. 15.
5. ANSYS FLUENT 12.0/12.1 Documentation. 2009 [cited 2012; Available from:
https://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node245.htm
https://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node785.htm#sec-uns-simple-vs-
simplec.
6. Salim .M. Salim, a.S.C.C. Wall y+ Strategy for Dealing with Wall-bounded Turbulent Flows. in
nternational MultiConference of Engineers and Computer Scientists. 2009.
7. John D. Anderson, J., Fundamentals of Aerodynamics, ed. 5th. 2010.
8. BEZAVADA, S.P.K.a.S., Numerical simulation of air flow over a passenger car and the Influence of
rear spoiler using CFD International Journal of Advanced Transport Phenomena, 2012. 01.