XFlow 2022

VALIDATION GUIDE

©2021 Dassault Systèmes. All rights reserved. 3DEXPERIENCE®, the Compass icon, the 3DS logo, CATIA, BIOVIA, GEOVIA, SOLIDWORKS, 3DVIA, ENOVIA, EXALEAD, NETVIBES, MEDIDATA, CENTRIC PLM, 3DEXCITE, SIMULIA, DELMIA, and IFWE are commercial
trademarks or registered trademarks of Dassault Systèmes, a French “société européenne” (Versailles Commercial Register # B 322 306 440), or its subsidiaries in the United States and/or other countries. All other trademarks are owned by their respective owners. Use
of any Dassault Systèmes or its subsidiaries trademarks is subject to their express written approval.
SIMULIA XFlow is © 2011 - 2021 Dassault Systèmes España, SLU .

Trademarks
XFlow, 3DEXPERIENCE, the Compass logo and the 3DS logo, CATIA,
SOLIDWORKS, ENOVIA, DELMIA, SIMULIA, GEOVIA, EXALEAD, 3D VIA,
BIOVIA, NETVIBES, and 3DEXCITE are commercial trademarks or registered
trademarks of Dassault Systèmes, a French ”société européenne” (Versailles
Commercial Register # B 322 306 440), or its subsidiaries in the U. S. and/or
other countries. All other trademarks are owned by their respective owners. Use
of any Dassault Systèmes or its subsidiaries trademarks is subject to their express
written approval.
DS Offerings and services names may be trademarks or service marks of Dassault
Systèmes or its subsidiaries.
Legal Notices
XFlow and this documentation may be used or reproduced only in accordance
with the terms of the software license agreement signed by the customer, or,
absent such an agreement, the then current software license agreement to which
the documentation relates.
This documentation and the software described in this documentation are subject
to change without prior notice.
Dassault Systèmes and its subsidiaries shall not be responsible for the consequences
of any errors or omissions that may appear in this documentation.
SIMULIA XFlow is © 2011 - 2021 Dassault Systèmes España, SLU .
For additional information concerning trademarks, copyrights, and licenses, see the
Legal Notices in the XFlow Installation Guide.
XFlow 2022 Validation Guide
 Contents

Contents i

Validation guide 1

1 Fundamentals 3
1.1 Lid-driven cavity flow . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Taylor Green vortex . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 Aerodynamics 11
2.1 NACA-0012 airfoil at Re = 500 . . . . . . . . . . . . . . . . . . . . . 11
2.2 S825 airfoil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Vortex cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4 1st AIAA High Lift Prediction Workshop . . . . . . . . . . . . . . . 27

3 Automotive 33
3.1 ASMO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 Ahmed body for Re=4.29E+6 . . . . . . . . . . . . . . . . . . . . . . 39

4 Free surface 45
4.1 NACA-0024 hydrofoil . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2 Dam-Break 3D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3 SYSSER50 hydrodynamics . . . . . . . . . . . . . . . . . . . . . . . 53

5 Acoustics 57
5.1 Simple Expansion Chamber . . . . . . . . . . . . . . . . . . . . . . . 57
5.2 Helmholtz resonator: Flute . . . . . . . . . . . . . . . . . . . . . . . 61

6 Heat transfer 65
ii

6.1 Natural convection in a cavity . . . . . . . . . . . . . . . . . . . . . . 65

7 Multi-phase flows 71
7.1 Rayleigh-Taylor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

References 72
 Validation guide

The aim of this guide is to present a set of test cases which show the applicability
of XFlow in a wide range of applications:
1. Fluid mechanics fundamentals
2. Aerodynamics
3. Automotive
4. Free surface flows
5. Acoustics
6. Heat transfer
7. Multi-phase flows
2
 1
Fundamentals

Contents
1.1 Lid-driven cavity flow . . . . . . . . . . . . . . . . . . . . . 3

1.2 Taylor Green vortex . . . . . . . . . . . . . . . . . . . . . . 7

1.1 Lid-driven cavity flow

The lid-driven cavity is a classical benchmark problem for viscous incompressible

fluid flow. It consists in a cavity where the upper boundary moves to the right, and
causes a rotation in the cavity. Side and bottom walls of the cavity are considered
no-slip, while the velocity at the upper wall is imposed to vx = 1 m s−1 (see Figure
1.1). Although there is a discontinuity of the boundary conditions at the two top
corners where the side wall meet the lid, this corner singularity plays a minor role
in the overall solution field.

Figure 1.1: Geometry and boundary conditions.

4 Fundamentals

The solution field will depend on the Reynolds number

ρ vx L
Re = (1.1)
µ
This validation case computes the laminar incompressible flow for a 2D driven cavity
at Re = 1000 and compares XFlow’s results with the ones presented in [1]. Fluid
properties are ρ = 1 kg m−3 and µ = 0.001 Pa s, in a L = 1 m square cavity
discretized with 128x128, 256x256 and 512x512 lattice points.
The case should be set up as an external flow, using a Generic Rectangular Domain,
with initialized flow of vx = 1 m s−1 , vy = 0 m s−1 . The origin of the system
should match the center of the doamain. All walls should be resolved with no slip
condition, except the top wall which has vx = 1 m s−1 . Turbulence should be turned
off for this Re, and it is essential to activate the high order boundary conditions:
Environment > Engine > Advanced Options > High order boundary conditions.
This imposes a a high order scheme for the velocity at the lid, which gives a relative
error of O(10−3 ) whereas it is of O(10−2 ) with the default boundary conditions.
With these settings, the flow field will become stabilized after 50s of simulated time.
The velocity field computed with XFlow is plotted in Figure 1.2. The vx velocity
distribution along vertical centerline is shown in Figure 1.3 and Table 1.1. In order
to obtain the displayed results, it is suggested to create plot lines with extremes
located at the voxel of the lattice elements adjacent to the walls, and a number of
points equal to the lattice resolution (128, 256, 512). For example, for the 128x128
resolution, the plot line should have the following settings:
V ertex1 = (0, -0.4961, 0)
V ertex2 = (0, 0.4961, 0)
Number of points =128

y 128x128 256x256 512x512 Ref. [2] Ref. [1]

0.9688 0.57874 0.58261 0.58212 0.57492 0.58031
0.9531 0.47825 0.47344 0.47341 0.46604 0.47239
0.7344 0.19412 0.19016 0.19013 0.18719 0.18861
0.5 -0.06265 -0.06174 -0.06209 -0.0608 -0.06205
0.2813 -0.28269 -0.27995 -0.28073 -0.27805 -0.2804
0.1016 -0.30371 -0.30217 -0.30168 -0.2973 -0.30029
0.0625 -0.20525 -0.20415 -0.20344 -0.20196 -0.20227

Table 1.1: Horizontal velocity vx along the vertical centreline. Comparison of

XFlow’s results for different resolutions with the reference solutions.
1.1 Lid-driven cavity flow 5

Figure 1.2: Velocity field.

Figure 1.3: Horizontal velocity vx distribution along the vertical centreline.

6 Fundamentals

Additional quantitative comparisons of pressure and vorticity are presented in Tables

1.2 and 1.3.

y 128x128 256x256 512x512 Ref. [2]

0.9688 0.05388 0.05264 0.05207 0.051514
0.9531 0.05275 0.05150 0.05091 0.050329
0.7344 0.01396 0.01304 0.01259 0.012122
0.5 0.0 0.0 0.0 0.0
0.2813 0.04064 0.04043 0.04038 0.040377
0.1016 0.10581 0.10490 0.10452 0.104187
0.0625 0.11148 0.11026 0.10972 0.109200

Table 1.2: Pressure p along the vertical centreline for Re = 1000. Comparison of
XFlow’s results for different resolutions with a reference solution.

y 128x128 256x256 512x512 Ref. [1]

0.9688 11.907 11.26092 10.9246 9.47810
0.9531 6.66407 5.97346 5.68313 4.86280
0.7344 2.35950 2.34076 2.33212 2.09090
0.5 2.35591 2.32528 2.31102 2.06690
0.2813 2.49017 2.50176 2.50973 2.26780
0.1016 1.68321 1.74846 1.78189 -1.63520
0.0625 2.52116 2.54815 2.56153 -2.31740

Table 1.3: Vorticity ω along the vertical centreline. Comparison of XFlow’s results
for different resolutions with a reference solution.
1.2 Taylor Green vortex 7

1.2 Taylor Green vortex

Taylor Green vortex a is three dimensional periodic, transitional and incompressible

flow which one can study the generation of small scales by three-dimensional vortex
stretching and the resulting turbulence. This process controls the turbulent-energy
dynamics and therefore the global structure and evolution of the flow.

The computation domain is a periodic cube defined as −πL ≤ x, y, z ≤ πL, with

a reference length L = 1 m. The fluid properties are ρ0 = 1 kg m−3 , p0 = 0 Pa,
V0 = 1 m/s, µ = 0.000625 Pa.s and the initial conditions (see Figure 1.4) for this
flow evolution is given by:

x y z
ux = V0 sin( )cos( )cos( ) (1.2a)
L L L
x y z
uy = −V0 cos( )sin( )cos( ) (1.2b)
L L L
uz = 0 (1.2c)
ρ0 v02 2x 2y 2z
p = p0 + (cos( ) + cos( ))(cos( ) + 2) (1.2d)
16 L L L

Figure 1.4: Taylor Green vortex initial condition: isosurface of vorticity coloured by
the velocity field

The solution field will depend on the Reynolds number:

ρ0 V0 L
Re = (1.3)
µ

This validation case computes the incompressible flow at Re = 1600 and compares
XFlow’s results with the ones provided by the Université Catholique de Louvain [3]
using a dealiased pseudo-spectral code for which neither numerical dissipation nor
numerical dispersion errors occur. The domain has discretized with 64 × 64 × 64,
8 Fundamentals

128 × 128 × 128 and 256 × 256 × 256 lattice elements in order to check the grid
convergence.
The goal of this validation case is to compare the kinetic energy dissipation rate
 = dE
dt and the energy spectrum function E(k) with the spectral data .
k

The kinetic energy dissipation rate is the energy dissipated due to the work done
by the fluctuating viscous stresses in resisting deformation of the fluid material by
the fluctuating strain rates. XFlow allows to export directly the kinetic energy (Ek )
integrated over the whole volume (Ω) from the Function viewer (Other integrals >
Overall kinetic energy):

v·v
Z
1
Ek = ρ dΩ (1.4)
ρ0 , Ω Ω 2
Then, the kinetic energy dissipation rate is calculated by the time derivation of this
kinetic energy provided by XFlow.

All the grid configurations have been simulated with the same time step dt = 0.001
s. In the Figure 1.5, it can be seen that, increasing the refinement, the turbulence
dissipation gets closer to the spectral data. It has been calculated with a central
finite difference as numerical approximation of the derivate ( = dE
dt ) .
k

Figure 1.5: Energy dissipation rate in function of time at different resolutions

1.2 Taylor Green vortex 9

The energy spectrum function E(k) is a characterization of the turbulent flow in a

case of homogeneous turbulence. The magnitude k is the modulus of the wavevector
which corresponds to some harmonics in a Fourier representation of the flow velocity
field u(x). This u(x) is defined as it follows:

ZZZ
u(x) = û(k)eik̄x̄ d3 k (1.5)

where û(k) is the Fourier transform of the flow velocity field. The total kinetic
energy is calculated as Z ∞
T KE = E(k)dk (1.6)
0

Using dimensional analysis, the energy spectrum function according with the third
Kolmogorov’s hypothesis is:

E(k) = Cε2/3 k −5/3 (1.7)

In order to compare the theoretical result with the decay of the energy in the
simulations, it is necessary to compute the Fast Fourier Transform and consequently,
the vector k̄ = (kx , ky , kz ) is calculated in the three directions of the space. The plot
is carried out taking into account the contribution to the kinetic energy q of the three
components of the velocity and the wavelength is calculated as: k = kx2 + ky2 + kz2 .
These results are plotted in the following figure:

Figure 1.6: Total Kinetic Energy spectra

10 Fundamentals

As it can be appreciated in the Figure 1.6, the decay of the kinetic energy is produced
for higher wavelengths when the lattice length unit is reduced. Indeed, at larges
wavenumber or smaller sizes, as the lattice length is reduced, the mesh is able to
capture smaller eddies and, consequently, there are more eddies . Besides this, the
Figure 1.6 suggests that the decay of the kinetic energy in the short wavelength is
similar to the slope k 5/3 , but, instead, in the smallest scales the simulations have
bigger decay, which relates to the LES filter.
Finally, the volumetric isosurface computed at t = 20 s with XFlow is plotted in
Figure 1.7.

Figure 1.7: Isosurface of vorticity with 2563 elements and dt = 0.001 s at t = 20 s

 2
Aerodynamics

Contents
2.1 NACA-0012 airfoil at Re = 500 . . . . . . . . . . . . . . . 11
2.2 S825 airfoil . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Vortex cell . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
st
2.4 1 AIAA High Lift Prediction Workshop . . . . . . . . . 27

2.1 NACA-0012 airfoil at Re = 500

The NACA-0012 airfoil is a widely-used wing section that has zero camber and a
maximum thickness to chord ratio of 12 percent. As it is shown in Figure 2.1, the
NACA-0012 is a symmetric two-dimensional profile with a very smooth aerodynamic
shape.
This validation case presents the XFlow results for the flow past a NACA-0012
at zero angle-of-attack and at a Reynolds number of 500. These results are further
compared with reference data; the comparison is based on the CFL3D code from the
National Aeronautics and Space Administration (NASA)1 [4, 5] due to the lack of
experimental data for such a low Reynolds number. A two-dimensional single phase

Figure 2.1: NACA-0012 profile.

1
Acknowledgements: Validation data have been kindly provided by courtesy of NASA Langley
Research Center and David P. Lockard.
12 Aerodynamics

external aerodynamics analysis has been performed using a virtual wind tunnel of
dimensions 60 × 40 m and a NACA-0012 profile of chord length L = 1 m. In order
to reach a Reynolds number based on the chord length equal to 500, the simulation
parameters have been set according to Table 2.1. As shown in Figure 2.2, the

Free-stream velocity vref 50 m s−1

Density ρ 1 kg m−3
Dynamic viscosity µ 0.1 Pa s
Chord length L 1m
Reynolds number Re 500
Angle-of-attack α 0 degree

Table 2.1: Simulation conditions.

Figure 2.2: Spatial resolution in XFlow.

spatial resolution is 2.56 m for the far field, and 0.005 m around the airfoil profile
and within the wake area. The spatial discretization has been defined through a
region of refinement instead of using an adaptive refinement in order to ensure that
the symmetry of the NACA-0012 is respected. The discretization ended up with 1.3
million elements in 9 levels of refinement. Since the flow is laminar, no wall functions
have been used to model the boundary layer.
Regarding temporal discretization, the time step was set to 0.004 s, which
corresponds to a Courant number of 1 with respect to the lattice size and the free-
stream velocity. Due to the fact that XFlow solver is inherently transient and this
case is steady, the analysis has been run until the aerodynamic coefficients stabilize
in time. The values of these coefficients are given in Table 2.2; the ones predicted by
XFlow are quite similar to those from CFL3D. The drag coefficient (Cd ) has a relative
error of -2.0678% with respect to CFL3D results, whereas the lift coefficient(Cl )
2.1 NACA-0012 airfoil at Re = 500 13

is actually even more accurate as it should theoretically be theoretically equal to

zero because of the NACA symmetry at zero angle-of-attack. XFlow profiles of

XFlow CFL3D Objective

Cd 0.173 0.1741 0.1741
Cl 10−15 −0.538 × 10−5 0

Table 2.2: Aerodynamic coefficients comparison.

normalized X and Y velocity components as well as pressure coefficients have been

compared with CFL3D data at five vertical sections: x/L = 0, 0.25, 0.5, 0.75 and 1
(see Figure 2.3). Figure 2.4 shows the X-component of the velocity field normalized

Figure 2.3: Data plot lines.

by the reference velocity vref = 50 m s−1 . The results of XFlow and CFL3D
are perfectly matching for the five sections. The profiles are as expected in both
codes: they tend to zero in the airfoil thickness and to one (or slightly more) on
the sides where the boundary layer is fully developed. For the Y-component of the
velocity field (normalized by vref ), again XFlow results are almost perfectly matching
with those of CFL3D, as shown in Figure 2.5. Nevertheless, one can observe some
differences close to the airfoil wall, specially for section x/L = 0.5. This might be due
to the size of the first element within the boundary layer which is not fine enough.
However, the differences between the two codes are small. The pressure coefficient
pstatic
Cp is defined as Cp = 1 2
, where pstatic is the gauge static pressure. Figure 2.6
2 ρ vref
shows the pressure coefficient distribution at the five sections. More differences are
now noticeable, especially at x/L = 0.5, 0.75 and 1.0. In general, the Cp tends to
be slightly over-estimated.
14 Aerodynamics

Figure 2.4: X-component of velocity u(x, y) at x/L = 0.0, 0.25, 0.50, 0.75 and 1.0.
2.1 NACA-0012 airfoil at Re = 500 15

Figure 2.5: Y-component of velocity v(x, y) at x/L = 0.0, 0.25, 0.50, 0.75 and 1.0.
16 Aerodynamics

Figure 2.6: Pressure coefficient at x/L = 0.0, 0.25, 0.5, 0.75 and 1.0.
2.2 S825 airfoil 17

2.2 S825 airfoil

The S825 airfoil has been designed for horizontal-axis wind turbine applications by
the National Renewable Energy Laboratory (Colorado, USA). The report of the
design and experimentation of the S825 airfoil [6] exposes the different objectives
and constraints set for the design, as well as the methodology of measurements
which have been conducted in the NASA Langley Low-Turbulence Pressure Tunnel
(LTPT) [7]. As explained in [6], the main objectives were, first, to reach a maximum
lift coefficient of at least 1.40 at a Reynolds number of 2 × 106 . Second, a low profile-
drag coefficients should be obtained between 0.40 and 1.20 of the lift coefficient. Two
main constraints were to keep the zero-lift pitching-moment coefficient greater than
-0.15, and also to have an airfoil thickness equal to 17% of the chord. The final two-
dimensional design is as shown in Figure 2.7, with a chord length equal to 0.45715
m.

Figure 2.7: S825 airfoil shape.

Experiments have been conducted at different Reynolds numbers based on the chord
length, however this validation case will only treat the Reynolds number 2×106 since
it has been used for most of the data provided by [6]. The Mach number is 0.1 and
the experimentation has been done with transition free (smooth) and with transition
fixed by roughness at specific locations.
The objective of this case is to validate the pressure distribution and aerodynamic
forces predicted by XFlow at low Mach number and different angles of attack (AoA).
The calculations have been performed with XFlow for a range of angles of attack
between -4 and 10 degrees every two degrees. All the calculations are transient due
to the nature of the XFlow solver and use the Wall-Adapting Local-Eddy turbulence
model, which belongs to the Large Eddy Simulation (LES) approach. Wall function
models in XFlow assume that the boundary layer is fully turbulent, therefore it is
not possible to model transition or prescribe a transition location.
Two-dimensional single phase analyses have been performed using a virtual wind
tunnel of 60 m × 40 m and a velocity at the inlet of 43.7493 m s−1 . The angle of
attack is varying by rotating the geometry instead of projecting the inlet velocity
vector, since XFlow allows easy manipulation of the geometry.
The fluid has a density of 1 kg m−3 and a dynamic viscosity of 10−5 Pa s accordingly
18 Aerodynamics

to the Reynolds number based on the airfoil chord length (Re = 2 × 106 ). The
simulations have been run for 1 second of physical time, with a time step of 0.002
s. The resolution scale at the far field is 1.28 m, using the adaptive refinement
algorithm available in XFlow. The walls and the wake are resolved with a scale of
0.0025 m, as shown in Figure 2.8.

Figure 2.8: Resolution refinement near the airfoil and the wake.

The solution for the static pressure and velocity flow variables at final time for zero
angle of attack can be observed in Figure 2.9.
For each angle of attack, the curve of pressure coefficients (Cp ) has been extracted
in XFlow using a cutting plane field distribution which projects the selected field
on the upper and lower sides of the airfoil. The Cp has been computed as following,
being Vref equal to 43.7493 m s−1 :
pstatic
Cp = 1 2
(2.1)
2 ρ Vref

For the angles of attack -4, -2, 0, 2, 4, 6, 8 and 10 degrees, the pressure coefficient
distribution along the airfoil has been compared with the transition free experimental
data presented in [6]. The results for angle of attack between 0 and 6 degrees are
in good agreement with the experiments, as shows Figure 2.10. On the upper side
of the airfoil, the pressure coefficients are slightly under-estimated when the angle
increases but still match reasonably with the experimental data.
However, when angles are increased to 8 and 10 degrees of AoA then results are
getting less accurate, as shown in Figure 2.11. The pressure coefficient tends to
be more under-estimated near the leading edge of the upper part. This could be
2.2 S825 airfoil 19

(a) Static pressure

(b) Velocity

Figure 2.9: Static pressure and velocity flow fields at final time for AoA = 0 degrees.

explained by the lack of transition model or the LES model, which is not fully
consistent for 2D simulations.
Another last series of angles of attack have been studied, this time for negative
incidence. Again, XFlow predicts with accuracy the pressure coefficient distribution
for -2 and -4 degrees, as shown in Figure 2.12.
Finally, Figure 2.13 compares the angle of attack vs. lift coefficient for theoretical
[6], experimental [6] and XFlow results. For positive angles, the lift coefficient is
slightly over-predicted by XFlow but in good agreement with the theoretical results.
20 Aerodynamics

(a) (b)

(c) (d)

Figure 2.10: Airfoil pressure coefficient distribution for different AoA: a) 0 degrees,
b) 2 degrees, c) 4 degrees, d) 6 degrees.
2.2 S825 airfoil 21

(a) (b)

Figure 2.11: Airfoil pressure coefficient distribution for different AoA: a) 8 degrees,
b) 10 degrees.

(a) (b)

Figure 2.12: Airfoil pressure coefficient distribution for different AoA: a) -2 degrees,
b) -4 degrees.
22 Aerodynamics

Figure 2.13: AoA (α) vs. lift coefficient (Cl ) for theoretical, experimental and
XFlow results, adapted from [6]
2.3 Vortex cell 23

2.3 Vortex cell

Trapping vortices is a technique that prevents vortex shedding in flows past bluff
bodies. Vortices forming near bluff bodies tend to be shed downstream but, if the
vortex is kept near the body at all times, it is called trapped.
This validation case compares XFlow results with experimental data for the flow
inside a vortex trapping cavity (vortex cell). The geometry consists in a rectangular
channel of section 520 × 52 mm with a spherical vortex cell of 45 mm depth located
at mid-length as shown in Figure 2.14. The boundary condition at the inlet is set
to a constant fluid velocity of UL = 36 m/s and the gauge pressure to 0 Pa at the
outlet. The fluid has been initialized to UL in the whole domain except inside the
vortex cell where it is 0 m/s in order to reach quicker the pseudo-steady state.

Figure 2.14: Vortex cell geometry.

The experimental data from [8, 9] provide the normalized X-component of the
velocity measured along a vertical line going from the bottom of the sphere up
to the upper wall of the channel (see Figure 2.15). The vertical coordinate along the
line is normalized by the line length L = 52 mm.

Figure 2.15: Line for measurement.

The two-dimensional studies led by [10] based on steady RANS turbulence models
show how sensitive are the numerical results for the vortex cell flow depending
24 Aerodynamics

on the turbulence model and the choice of the numerical scheme. XFlow uses
Large Eddy Simulation turbulence models, which are inherently three-dimensional.
Furthermore, although the vortex cell seems a two-dimensional flow, turbulence
effects (important near and inside the cavity) need a three-dimensional analysis to
be accurately modeled.
Unfortunately three-dimensional analyses may involve a large number of elements
and long simulation times. The refinement algorithms available in XFlow (near the
walls and adaptive wake) allow to minimize the number of elements, but tend to
introduce numerical dissipation when passing from one element size to another and
has been found to be inaccurate especially in the boundary layer that detaches from
the leading edge of the vortex cell.
The following results were obtained using a uniform resolution of 1 mm in the
whole domain and a time step of 10−6 s. This resolution leads to a total of 1.1
million elements. The total simulation time solved is 0.9 s at a frequency of 500
Hz. Averaged results are required in order to analyse the pseudo-steady state of the
solution.
Figure 2.16 shows that XFlow 3D results are globally in good agreement with the
experimental data. The areas of less accuracy are at y/L around -0.7 and the peak
around 0.1. Nevertheless XFlow is able to predict the experimental velocity profile at
the cavity entry (−0.4 < y/L < 0) better than the RANS calculations and correctly
predicts the vortex speed at y/L = −0.8.
Finally, Figure 2.17 shows the averaged velocity field computed by XFlow. It is
possible to observe the creation of vortices at the leading edge of the cavity, what is
not evident in steady calculations.
2.3 Vortex cell 25

Figure 2.16: Comparison of XFlow 3D results with experimental data and RANS
results from [10].

Figure 2.17: XFlow averaged velocity field.

26 Aerodynamics
2.4 1st AIAA High Lift Prediction Workshop 27

2.4 1st AIAA High Lift Prediction Workshop

The 1st AIAA CFD High Lift Prediction Workshop (HiLiftPW-1), took place in June
2010 in Chicago (USA). The challenge was to simulate a half aircraft configuration
composed of a body and a 3-element airfoil with a plane of symmetry (as shown in
Figure 2.18) for a range of high angles of attack.

Figure 2.18: Trapezoidal wing, configuration 1.

This validation case reports the results obtained with XFlow for the Test Case 1 of
the High Lift Prediction Workshop [11]:
ˆ Trapezoidal wing Config 1 (slat at 30 degrees, flap at 25 degrees)
ˆ Mach number = 0.2
ˆ Reynolds number = 4.3E+6 based on mean aerodynamic chord (MAC)
ˆ Geometry provided by the workshop
ˆ MAC = 1.0067 m
ˆ No brackets
ˆ Angles-of-attack: -4, 0, 6, 13, 21, 25, 28, 32, 34 and 37 degrees.
Two different spatial resolutions have been used depending on the lift level, see Table
2.3 for further details.

Walls Wake Far Field Max # of Particles

Resolution 1 0.005 m 0.01 m 1.28 m 25E+6
Resolution 2 0.005 m 0.02 m 1.28 m 10 E+6

Table 2.3: Spatial discretization.

28 Aerodynamics

As it is shown in Figure 2.19, Resolution 1 leads to very large number of elements

at high angles-of-attack, which could not be computed with our resources (an
Intel quad-core workstation with 16 GB of RAM). This is due to a large number
of elements generated for the wake refinement as separation starts to occur and
detached eddies leave the airfoil. For this reason Resolution 2, with a coarser
refinement for the wake, has been used for the angles-of-attack 34 and 37 degrees
(where stall starts to happen) and Resolution 1 for the other angles.
Workshop attendants were asked to run all the simulation as free air (no wind-
tunnel walls or model support systems are to be included) and fully turbulent.
Since there is a symmetry axis on the geometry model, the airfoil has been placed
in the XFlow virtual wind tunnel with the ground wall enabled as a free-slip wall,
which has been used as the plane of symmetry.
The experimental data have been gathered at the 14x22 wind-tunnel at NASA
Langley Research Center [12]. Forces, moments, and Cp distributions have been
provided for free transition and with lower and upper values, which are assumed to
be the range of uncertainty in the wind tunnel measurements.
In Figures 2.20 and 2.21 one can observe the drag and lift coefficients versus the
angle-of-attack α. XFlow shows good agreement within the range 1 to 28 degrees,
predicting accurately both slope and values of the aerodynamic coefficients. The
polar curve in Figure 2.22 is hence matching the experimental results, especially
in the pre-stall region. The pitching moment coefficients have been calculated for
the range 6 to 34 degrees and also lay between the upper and lower limits of the
experimental results, as shown in Figure 2.23.
Eventually, some snapshots of the vorticity flow field are presented in Figures 2.24-
2.25.
2.4 1st AIAA High Lift Prediction Workshop 29

(a)

(b)

Figure 2.19: Example of lattice refinement - Resolution 1: (a) virtual wind tunnel
and far field resolution, (b) resolution near walls and wake refinement.
30 Aerodynamics

Figure 2.20: Drag coefficient versus angle-of-attack.

Figure 2.21: Lift coefficient versus angle-of-attack.

2.4 1st AIAA High Lift Prediction Workshop 31

Figure 2.22: Polar curve.

Figure 2.23: Pitching moment coefficient versus angle-of-attack.

32 Aerodynamics

Figure 2.24: Volumetric rendering of vorticity, AoA = 13, upper side.

Figure 2.25: Volumetric rendering of vorticity, AoA = 13, lower side.

 3
Automotive

Contents
3.1 ASMO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 Ahmed body for Re=4.29E+6 . . . . . . . . . . . . . . . . 39

3.1 ASMO

In automotive aerodynamics it is a common practice to use reference geometries to

validate CFD codes [13]. This section uses the well known ASMO (Aerodynamisches
Studien Modell ) vehicle model, which comprises a square-back rear, smooth surfaces,
boat tailing, underbody diffuser and no pressure induced boundary layer separation.
The geometry does not have a well defined separation line and is characterized by a
low drag shape.
For this validation the 1/5 wind tunnel test model has been adopted, where vehicle’s
length, width and height are 0.81 m, 0.29 m and 0.27 m respectively; and wind tunnel
dimensions are 18 x 6 x 10 m (see Figure 3.1). This corresponds to a blockage ratio
of 0.057%. Experimental data for this test case are available from Daimler Benz and
Volvo model scale wind tunnel.

Figure 3.1: ASMO body and wind tunnel geometries.

34 Automotive

To perform the simulation, the XFlow Virtual wind tunnel is used. The inlet-
velocity is set to 50 m/s (uniform) and the fluid properties are: density ρ = 1 kg
m−3 and dynamic viscosity µ = 1.5 × 10−5 Pa s. The Reynolds number is thus
2.7 × 106 , taking the length of the vehicle as reference.
In this problem it is essential to properly resolve the turbulent wake. Therefore
dynamic wake refinement is applied, so that the specified particle resolution is
automatically adopted in regions with high turbulence, while less turbulent regions
are treated with fewer particles. Particle resolution in the far field is 0.1 m, whereas
scales down to 2.5 mm are resolved in the wake and on the model surface. Turbulence
modelling is approached using a Large Eddy Simulation (LES) approach, together
with a Wall-Adapting Local Eddy-viscosity (WALE) subgrid model for the turbulent
viscosity.

The turbulent wake structure can be observed in Figures 3.2 and 3.3, together with
the instantaneous pressure field and the skin friction distribution in Figure 3.4.

Figure 3.2: Vorticity isosurface.

3.1 ASMO 35

(a)

(b)

(c)

Figure 3.3: Instantaneous velocity field: (a) on the vehicle surface, (b) in the
Y = 0.07 m plane, and (c) in the symmetry plane.
36 Automotive

Figure 3.4: Instantaneous pressure field in the symmetry plane and skin friction
distribution.

Figures 3.5 to 3.8 show the comparison between the XFlow results and the
experimental data for the pressure field along the symmetry plane on the front,
roof, base and underbody surfaces. It can be seen that the comparison with the
measurements is good, although some deviations can be observed especially in the
base pressure, which is slightly underpredicted. However the proper level of the base
pressure is not known exactly, as there is a large difference between both experiments.

Figure 3.5: Front pressure distribution along the symmetry plane.

3.1 ASMO 37

Figure 3.6: Roof pressure distribution along the symmetry plane.

Figure 3.7: Base pressure distribution along the symmetry plane.

Figure 3.8: Underbody pressure distribution along the symmetry plane.

38 Automotive

Typically, drag stabilizes in a characteristic time of the order of the flow travelling the
vehicle length. In 0.1 seconds, the flow has travelled more than six times the whole
body. The time averaged drag between 0.05 and 0.1 seconds of physical simulation
is Cd = 0.166 (see Figure 3.9), in good agreement with the values measured in the
experiments shown in Table 3.1.

Figure 3.9: Overall drag history.

XFlow 0.166
Experiments Volvo 0.158
Experiments Daimler Benz 0.153

Table 3.1: Drag values for ASMO model.

3.2 Ahmed body for Re=4.29E+6 39

3.2 Ahmed body for Re=4.29E+6

The Ahmed body is the most classical benchmark in the automotive industry. It
has been first defined and its characteristics described in the experimental work
of Ahmed [14]. The Ahmed body geometry is described in Figure 3.10. The
experimental measurements at various slant angles (ϕ =[0◦ ,40◦ ]) were conducted
by Ahmed and Ramm in the DFVLR subsonic wind tunnels at Braunschweig and
Göttingen; these having a square nozzle of (3 x 3) m and a length of 5.8 m.

Figure 3.10: Ahmed body geometry, from [15].

The goal of this validation test case is twofold: (i) the validation of the drag curve
against the slant angle; and (ii) the analysis of the mean turbulence structures on
the slant surface of the Ahmed body and in the downstream region. To this end,
the geometry shown in Figure 3.10 has been simulated with XFlow using the 3D
single-phase engine and a virtual wind tunnel of dimensions 8 x 2 x 2 m as a external
domain. The wall ground was activated with zero velocity.
In order to reach a Reynolds number based on the car length equal to 4.29 million,
the simulation parameters have been set according to Table 3.2.
Firstly, the mesh dependency of the drag coefficient is studied to find a trade-off
between accuracy and computing time. In this study the far field resolution is set to
0.08 m, whereas the walls and wake resolution level is varied according to the values
given in Table 3.3.
Figure 3.11 shows the variation of the drag coefficient (Cx ) calculated by XFlow for
the different refinement levels together with the experimental value. It turns out that
the wall resolution level h/22 (0.01 m) gives good results and in an acceptable time.
Figure 3.12 shows the value of the drag coefficient (Cx ) calculated by XFlow for
the different wake refinement levels, keeping the wall-resolution to 0.01m. From
the figure, the best resolution for the wake is also 0.01 m. Henceforth, the spatial
discretization for all cases is set to Adaptive refinement with a resolution of 0.08 m
40 Automotive

Project tree parameter Value

Inlet velocity [Vref ] 60 m s−1
Density [ρ] 1 kg m−3
Dynamic viscosity [µ] 1.46014e-5 Pa s
Car length [Lref ] 1044 m
Reynolds number [Re] 4.29e+6
Slant Angles [ϕ] 0 ; 5 ; 10 ; 12.5 ; 15 ; 20◦ ; 25◦ ; 30◦ ; 32◦ ; 35◦ ; 40◦
◦ ◦ ◦ ◦ ◦

Turbulence Intensity 0.5 %

Simulation time 2.0 s
Courant number 1s

Table 3.2: Simulation settings

ϕ = 35◦ h h/2 h/22 h/24

Resolution [m] 0.04 0.02 0.01 0.005
Initial number of elements 72853 103267 219457 676854

Table 3.3: The initial number of elements for the different resolutions, case ϕ = 35◦

in the far field and 0.01 m around the Ahmed body and wake.

Figure 3.11: Near-wall resolution dependency of the drag coefficient for ϕ= 35◦

The first part of validation is about comparing the Ahmed-body drag-coefficient

value obtained with XFlow with that measured by Ahmed and Ramm [14]. Figure
3.13 shows the drag coefficient obtained with XFlow superposed on the experimental
drag curve by Ahmed and Ramm [14]. XFlow and experimental drag curves agree
with a good accuracy and the minimum drag angle found is for 12.5 degrees slant
angle as expected. An additional point around 29 degrees would be required to
3.2 Ahmed body for Re=4.29E+6 41

confirm the drag peak around 30 degrees.

Figure 3.12: The influence of the wake resolution on the drag coefficient for ϕ = 35◦

Figure 3.13: Drag coefficient against slant angle

The second part of validation is about analysing the main turbulence structures
around the Ahmed Body. According to Franck et al. [16], the experiments show two
critical slant angles: ϕ = 12.5◦ and ϕ= 30◦ , called first and second critical angles
respectively; where the main structure of the time-averaged flow in the near-wake
42 Automotive

changes significantly. Between these two critical angles, the flow is massively 3D
separated, for other slant angles the flow is almost 2D attached to the body. This
behaviour is perfectly reproduced by XFlow as shown in Figure 3.14.
Ahmed and Ramm [14] also provides some pictures of the flow streamlines on the
slanted surface. Figure 3.2 shows both Ahmed and XFlow streamlines for the same
slant angles. It can be seen that the XFlow streamlines present the same main
structures than those captured experimentally [14].
The validation of the Ahmed body case with XFlow confirms its good performance
for aerodynamic applications with high Reynolds numbers. The time required in
XFlow to set up a case is about 5 minutes. The calculation takes about 6-8 hours
on a machine equipped with 16 CPUs. This time is almost constant for all the slant
angles.

Figure 3.14: Time-averaged flow for the different slant angles (critical angles are
outlined in red).
3.2 Ahmed body for Re=4.29E+6 43

Figure 3.15: Streamlines on the rear surface for ϕ = 12.5◦ , 25◦ and 30◦ , from [14].
44 Automotive
 4
Free surface

Contents
4.1 NACA-0024 hydrofoil . . . . . . . . . . . . . . . . . . . . . 45
4.2 Dam-Break 3D . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3 SYSSER50 hydrodynamics . . . . . . . . . . . . . . . . . . 53

4.1 NACA-0024 hydrofoil

In order to validate the XFlow free-surface engine, it is used to simulate the test-case
of a surface-piercing NACA-0024 hydrofoil. Numerical results are compared with
the experimental data reported by Metcalf et al. in [17].
The experiments were conducted in a towing tank for three different flow regimes,
namely: F r = 0.19, 0.37, 0.55; F r being the Froude number based on the NACA
chord length (L):
Vref
Fr = √ (4.1)
gL
Vref represents the NACA velocity and g the gravity. To facilitate the flow field
measurements, the NACA-0024 foil chord (L = 1.2m) and draft (T = 1.5m) were
designed to achieve a large separation region and no-wave effects and 2D flow at
large depths [17]. Some photographs of the experiments are shown in Figure 4.1.
The XFlow 3-D Free-Surface external solver has been used to reproduce numerically
the experimental test case corresponding with a Froude number of 0.37. To reach
such a Froude number, the simulation parameters have been set according to Table
4.1; temporal and spatial discretisation details are also given in Table 4.1.
46 Free surface

(a) Bow region (b) Front-quarter view

(c) Side view of wave profile (d) Wake region

Figure 4.1: Photographs of the NACA 0024 wave field for F r = 0.37, from [17].

Project tree parameter Value

Gravity [g] 9.81 m s−2
Inlet velocity [Vref ] (1.27, 0, 0) m s−1
Density [ρ] 998.3 kg m−3
Dynamic viscosity [µ] 0.001 Pa s
NACA-0024 chord length [Lref ] 1.2 m
NACA-0024 behaviour Fixed
Simulation time 20.0 s
Courant number 1s
Resolved scale 0.2 m
Refinement algorithm Adaptive refinement
Wake refinement threshold Automatic
Wake resolution 0.0125 m
Shape refinement 0.0125 m

Table 4.1: Relevant simulation settings

4.1 NACA-0024 hydrofoil 47

(a) Experimental data

(b) Experimental and numerical averaged contour overlapped

Figure 4.2: Wave profiles and near-field elevations close to the foil surface for
F r = 0.37.

Figure 4.2(a) shows the contours of the z-coordinate on the free surface, illustrating
the varying height of the free surface along the hydrofoil. In Figure 4.2(b) the
computed free surface (averaged over the period [0s; 19.7s]) is compared with
experimental data, finding a good match.
48 Free surface

(a) Experimental data

(b) Numerical averaged

Figure 4.3: Free-surface elevation - mean for F r = 0.37.

Figures 4.3(a) and Figure 4.3(b) show the experimental and numerical mean free-
surface elevation contours on the x-y plane. Qualitative agreement is found.
4.2 Dam-Break 3D 49

4.2 Dam-Break 3D

The dambreak is a popular validation case for free surface flows, since no special
in-/out-flow boundary conditions are needed. In this section, XFlow free-surface-
internal engine is validated by comparing the XFlow results with the experimental
data from the Maritime Research Institute Netherlands (MARIN) [18].

Figure 4.4: Measurement positions for water heights and pressures in the dambreak
experiment

The experiment is performed using a large tank (3.22 x 1 x 1 m) with an open roof.
The right part of the tank is first closed by a door. Behind the door 0.55 m of water
is waiting to flow into the tank when the door is opened. This is done by releasing a
weight, which almost instantaneously pulls the door up. In the tank a box has been
placed that represents a scale model of a container on the deck of a ship. During
the experiment measurements have been performed of water heights, pressures and
forces. In Figure 4.4 the positions of the measured quantities are shown. Four
vertical height probes have been used; one in the reservoir and the other three in
the tank. The box was covered by eight pressure sensors, four on the front of the
box and four on the top [18]. Two images of the experiment are shown in Figure
4.5.

Figure 4.5: Pictures taken during the experiment, at time 0.4 (left) and 0.56 s
(right)[18].

Table 4.2 shows the relevant parameters used to run the XFlow simulation, whereas
50 Free surface

Project tree parameter Value

Engine 3D Free Surface Internal (Isothermal)
Gravitational potential Off
External acceleration laws 9.81 m s−2
Initial liquid function if (x < 0, 0, y < 0.55)
Initial velocity field (0 0 0) m s−1
Initial gauge pressure 0 Pa
Density [ρ] 1000 kg m−3
Dynamic viscosity [µ] 0.001 Pa s
Tank behaviour Fixed (Wall boundaries)
Simulation time 7.2 s
Courant number 1
Resolved scale 0.01 m
Refinement algorithm Disabled

Table 4.2: Relevant simulation settings

Figure 4.6: XFlow free-surface at time 0.4 s (left) and 0.56 s (right).

Figure 4.6 shows the numerical free surface corresponding to the same instants of
time at which the pictures in Figure 4.5 were taken. Comparing these two figures, it
can be observed that there is a good qualitative agreement between numerical and
experimental free-surface shapes. For a quantitative validation, the experimental
pressure on the box surface measured by eight sensors (four on the front and four
on the top) is next compared with the XFlow results.
Figures 4.7 and 4.8 show the quantitative comparison of the experimental and
numerical pressure at the sensors positions. Figure 4.7 shows a very good agreement
between numerical and experimental data agree at the box front measurement
points. As shown in Figure 4.8, this agreement is also found at the sensor positions
on the top of the box, though it is less accurate.
4.2 Dam-Break 3D 51

Figure 4.7: Static Pressure at four points of the box front (P1, P2, P3, P4)
52 Free surface

Figure 4.8: Static Pressure at four points of the box top (P5, P6, P7, P8)
4.3 SYSSER50 hydrodynamics 53

4.3 SYSSER50 hydrodynamics

The Delft Systematic Yacht Hull Series (DSYHS) has been started in 1973 by
professor J. Gerritsma, in cooperation with N. Newman and J. Kerwin of the
Massachusetts Institute of Technology (MIT). The aim is to perform measurements
on a series of systematically varied yacht hulls, in order to assess the influence of the
varied parameter on the hydrodynamic performance (resistance, seakeeping etc.) of
a sailing yacht hull. The model SYSSER50 is part of this series and its experimental
data is available on DSYHS Database website.
The boat length is 2.375 m, the width is 0.737 m, and the thickness is 0.336 m.
The boat is located into the XFlow water channel of dimensions (15, 2, 4) m. The
illustration of the problem setup is depicted Figure 4.9. The aim of the study is
to recover the hull resistance force and compare it with the experimental data from
DSYHS. Besides, it has been took the opportunity to check the wave pattern angle
that is left behind by a ship at sea. This angle was first determined by Lord Kelvin
[19], where the value is θ = 19.28◦ .

Figure 4.9: Water flow around the SYSSER50 geometry in the XFlow water channel.

The flow model used is free surface external, and the fluid gravity is set to -9.81
m/s in Y direction. The inlet is set as constant velocity and height, providing flat
sea conditions. The case is run for several velocities, as indicated Table 4.3. The
hull geometry behavior in XFlow is set to Rigid body dynamics with one degree of
freedom on the heave motion. The pitch angle is fixed and set to the value found
in the DSYHS data. Initial heave position is set to 0 m in Y direction, and will
stabilize after the transient period according to the gravity applied on the geometry
weight and the inlet water velocity.
The simulation time is set to 7 seconds, with automatic time step. The refinement
algorithm is set to Adaptive refinement in order to refine dynamically the wake, the
54 Free surface

Free-stream velocities vref 1.109 m s−1

1.700 m s−1
1.990 m s−1
2.439 m s−1
2.880 m s−1
3.099 m s−1
Density ρ 998 kg m−3
Dynamic viscosity µ 0.001 Pa s
Reference area Aref 0.917 m2
Geometry behavior Rigid body dynamics
Geometry density 156 kg m−3
Geometry mass 47.5 kg

Table 4.3: Simulation conditions.

moving hull, and the free surface. The resolutions are detailed in Table 4.4, and the
resulting lattice structure is shown Figure 4.10.

Refinement Algorithm Walls Wake/Free Surface Far Field

Adaptive refinement 0.015625 m 0.015625 m 0.5 m

Table 4.4: Spatial discretization.

Figure 4.10: Initial lattice structure.

The simulations run on 12 cores in about 30h for every velocity point. The force
histories are converging after 3 seconds of simulation time for low water speeds,
and after 6 seconds for medium and high water speeds. The convergence can be
appreciated Figure 4.11 and is quite stable. This transient period is the time required
for the boat to stabilize the heave, and obviously takes longer when the water speed
is higher.
4.3 SYSSER50 hydrodynamics 55

Figure 4.11: Hull resistance force histories for 1.109, 1.99, 2.439, and 3.099 m/s.

Since XFlow is unsteady, the hull resistance force must be time-averaged. The forces
are averaged for every velocity point, starting from the end of the transient period
until the end of the simulation. The comparison of the force in Newton and the
force coefficient Cx against the velocity is shown on Figure 4.12, and compared with
experimental results. The correlation with experimental data is good, the resistance
force curve trend is similar since the linear slope is reproduced by XFlow. The low
velocity point also shows a lower curve slope as expected, and the higher velocity
points also decrease despite of being slightly overestimated.
Finally, the wave pattern angle is compared against the theoretical value determined
by Kelvin in the Figure 4.11. For the inlet velocity of 2.439m/s XFlow provides an
accurate wave pattern angle, θ = 19.50◦ , where the relative error is 1.14%. The rest
of velocities have a similar wave pattern angle.
56 Free surface

Figure 4.12: Comparison of forces and coefficients against experimental data.

Figure 4.13: Comparison of wave pattern Kevin angle with theoretical value for
2.439 m/s.
 5
Acoustics

Contents
5.1 Simple Expansion Chamber . . . . . . . . . . . . . . . . . 57
5.2 Helmholtz resonator: Flute . . . . . . . . . . . . . . . . . 61

5.1 Simple Expansion Chamber

Reactive mufflers operate by the destructive interference of the acoustic waves

propagating within them. The simplest type of reactive muffler is the simple
expansion chamber, shown in Figure 5.1.
The aim of the present test-case is to validate the XFlow acoustics analysis by
predicting the transmission loss in a simple expansion chamber. The accuracy
of XFlow results will be determined by comparing the numerical results with the
experimental data of Selamet and Radavich [20] for the geometry described in Figure
5.1.

Parameter Value
d1 4.859 cm
d2 15.318 cm
l 54 cm
l/d2 3.525

Figure 5.1: Schematic of the muffler geometry

The transmission loss in such an expansion chamber is given by:

pinc
T L = 20log10 (5.1)
ptra
pinc and ptra being the incident and transmitted pressure respectively [21].
58 Acoustics

In this work, XFlow is used to find the ratio pinc /ptra and hence the transmission
loss. The relevant parameters of the setup used to run the case are summarised in
Table 5.1. Please note that the inlet boundary condition is a conditional function
that sets a sinusoidal velocity of frequency 3200 Hz and an amplitude of 0.05 ms−1
and generates a pressure wave; the dissipation of which in the muffler is the object
of this study.

Project tree parameter Value

Engine 3D Single phase internal (Isothermal)
Acoustics analysis On
Initial velocity field (0 0 0) m s−1
Initial gauge pressure 0 Pa
Reference area Front
Reference velocity Custom (0.05 ms−1 )
Density [ρ] 1.225 kg m−3
Dynamic viscosity [µ] 3.34e-5 Pa s
Enable bulk viscosity On (3.34e-5 Pa s)
Muffler inlet - Velocity if(t < 0.0003125,0.05 ∗ sin(2 ∗ pi ∗ 3200 ∗ t),0)
Muffler outlet - Gauge pressure outlet 0 P a (Backflow: On)
Muffler shell - Wall Free slip
Simulation time 0.035 s
Courant number 1
Resolved scale 0.0025 m
Refinement algorithm Disabled

Table 5.1: Relevant simulation settings

Figure 5.2 shows the muffler numerical domain. In the figure, the position of two
sensors is indicated - one at the muffler inlet and the other at the outlet. These
sensors allow us to monitor the numerical value of pinc and ptra .

Figure 5.2: Muffler geometry showing the sensors position

Replacing the values of pinc and ptra into Equation (5.1), we obtain the transmission
5.1 Simple Expansion Chamber 59

loss predicted by XFlow. This is shown in Figure 5.3 together with the experimental
data of Selamet and Radavich [20]. From the figure, it can be concluded that
XFlow is able to accurately predict the acoustics behaviour of such a muffler.

Figure 5.3: Transmission loss comparison between XFlow results and experimental
data [20]
60 Acoustics
5.2 Helmholtz resonator: Flute 61

5.2 Helmholtz resonator: Flute

The sound generation in a small stopped flue pipe with recorder-like proportions
was simulated using XFlow. The flute under study is similar to the one studied by
Kühnelt in [22, 23]; its geometry is shown in Figure 5.4 and its relevant dimensions
are reported in Table 5.2.

Figure 5.4: Flute geometry - Inlet in green and outlet in blue

Flute components Nomenclature Value

Inlet flue L3 × L2 × L1 15 × 4.5× 1.2 (mm)
Mouth opening length L4 3.75 (mm)
Resonant cavity L 5 × L6 × L7 60 × 7 × 7 (mm)

Table 5.2: Geometry description

According to Helmholtz theory, the resonance frequency (fH ) of a cavity is given

by: r
cs A
fH = (5.2)
2π Vo L
where cs is the speed of sound, A represents the cross-sectional area of the neck, L
stands for the neck length and Vo is the cavity volume. In particular, from Table
5.2 we have: A = 0.0012 ∗ 0.0045 m2 , L = 0.00375 m, Vo = 0.007 ∗ 0.007 ∗ 0.06 m3
and cs = 340 m/s; replacing the corresponding values into Eq. (5.2) it is found that
the theoretical resonance-frequency of this flute is 1197Hz [23].
This flute has been simulated with XFlow to validate its acoustic analysis (Acoustic
analysis is available in Expert mode, Preferences > Application > Application
mode: Expert mode). The case setup is summarised in Table 5.3 and the velocity
contours inside the flute at t = 0.01 s and t = 0.0106 s are shown in Figure 5.5.
As indicated in Table 5.3, a Probe has been placed at the end of the resonant cavity
to measure the static-pressure variation in time. This is shown in Figure 5.6. The
compression-expansion cycles at the probe position are clearly observed in the figure.
This information is further shown in Figure 5.7 in terms of Sound Pressure Level
(SPL) versus frequency.
62 Acoustics

Project tree parameter Value

Engine 3D Single phase internal (Isothermal)
Acoustics analysis On
Initial velocity field (0 0 0) m s−1
Initial gauge pressure 0 Pa
Density [ρ] 1.205 kg m−3
Dynamic viscosity [µ] 1.7894e-5 Pa s
Flute inlet - Velocity 7.76 ms−1
Flute outlet - Gauge pressure outlet 0 Pa
Simulation time 0.02 s
Courant number 1
Resolved scale 0.00022 m
Refinement algorithm Disabled
Probe position (0.074, -0.003, 0) m

Table 5.3: Relevant simulation settings

Figure 5.5: Velocity contours inside the flute at t = 0.01 s and t = 0.0106 s
5.2 Helmholtz resonator: Flute 63

Figure 5.6: Helmholtz resonator - Static pressure measured at the probe position

Figure 5.7: Helmholtz resonator - Sound pressure level vs frequency at the probe
position

Figure 5.7 shows that the fundamental frequency of the cavity appears at about
1100 Hz which is in close agreement with 1197 Hz - the theoretical value predicted
by Helmholtz formula.
64 Acoustics
 6
Heat transfer

Contents
6.1 Natural convection in a cavity . . . . . . . . . . . . . . . . 65

6.1 Natural convection in a cavity

Natural convection in a cavity is a classical test case for the modelling of buoyancy
driven flows. A schematic of the problem is shown in Figure 6.1. It consists of a
fluid enclosed in a square cavity with different wall temperatures. This leads to a
thermal gradient that gives rise to the natural convection of the fluid. Horizontal
walls are considered adiabatic.

Figure 6.1: Geometry and boundary conditions.

XFlow models the buoyancy forces according to the Boussinesq approximation, that
assumes the linear variation of the density as a function of the temperature θ:

ρ = ρ0 [1 − α(θ − θ0 )]
66 Heat transfer

and that the thermodynamic properties of incompressible fluids are constant except
when considering the body force ρg in the momentum equation. Hence, the
governing equations of the fluid motion are as follows:

Mass conservation: ∇·v =0 (6.1)

dv
Momentum conservation: ρ0 = ∇ · σ + ρ0 [1 − α(θ − θ0 )]g (6.2)
dt
dθ
Enthalpy conservation: ρ0 cp = k∇ · (∇θ) + Φ (6.3)
dt
where α is the thermal expansion coefficient, cp the specific heat at constant pressure,
k the thermal diffusion coefficient, θ0 a reference temperature, ρ0 = ρ(θ0 ), and Φ is
the viscous heat dissipation.
Benchmark solutions for this natural convection problem were published by de Vahl
Davis [24]. These solutions are compared with the numerical results obtained with
XFlow for several spatial resolutions (20x20, 40x40, 80x80, and 160x160 lattice
points) and for two Rayleigh numbers (Ra = 103 , 106 ):

ρ g α ∆θ L3
Ra = (6.4)
µκ

∆θ = |θ1 − θ2 | being the temperature difference between the two vertical walls
and L = 1 m being the cavity side length. The fluid properties for the first case
(Ra = 103 ) are given in Table 6.1.

Case Ra = 103 Ra = 106

ρ0 1 kg m−3 1 kg m−3
g -10 m s−2 -10 m s−2
α 0.1 K−1 0.1 K−1
∆θ 1K 1K
µ 0.0266458 Pa s 0.0008426 Pa s
k 0.0375293 kg m s−3 K−1 0.0011868 kg m s−3 K−1
cp 1 m2 s−2 K−1 1 m2 s−2 K−1

Table 6.1: Fluid properties

µ cp
The values of µ, k and cp are chosen to match the air Prandtl number: P r = =
k
0.71. In both cases, the viscous heat dissipation has been neglected as it is done in
[24].
The comparison is made in terms of the the following dimensionless variables at
steady state:
Z 1
x y vx L vy L ∂θ
X= , Y = , Vx = , Vy = , and N u0 = dy
L L κ κ 0 ∂x X=0
6.1 Natural convection in a cavity 67

k
where κ = is the thermal diffusivity of the fluid.
ρ cp
Tables 6.2 and 6.3 show the comparison of XFlow results with the solution in [24]
for Ra = 103 and Ra = 106 , where:
ˆ N u0 , average heat flux in the hot vertical wall
ˆ max Vx , maximum value of the non-dimensional horizontal velocity on the
vertical centreline and its location Ymax
ˆ max Vy , maximum value of the non-dimensional vertical velocity on the
horizontal centreline and its location Xmax
There is a good agreement in both cases although the accuracy decreases for the
highest Rayleigh number. This is consistent with the findings of most contributors
reported in [24]. In this test, the maximum and minimum locations are cell centered,
none of the interpolation techniques suggested by [25] have been employed.
Velocity, vorticity and temperature contours resulting from the XFlow simulations
are shown in Figures 6.2 and 6.3 for Ra = 103 and Ra = 106 respectively.
For this case it is essential to: (i) deactivate the viscous heat dissipation
Φ ( Environment > Engine > Advanced Options > Enable viscous term in energy
equation:Off); (ii) do not use turbulence model (Environment > Engine >
Turbulence settings/ Turbulence generation: Off); and (iii) set a unity Courant
number(Co ≈ 1), as large Courant values would introduce an excessive
compressibility in the flow and the results will not correspond to the Boussinesq
incompressible reference solution.
68 Heat transfer

20x20 40x40 80x80 160x160 Ref.[24]

max Vx 3.5428 3.6325 3.6422 3.6468 3.649
at Ymax 0.825 0.8125 0.8187 0.8156 0.813
max Vy 3.6359 3.6941 3.7023 3.7012 3.697
at Xmax 0.175 0.1875 0.1812 0.1781 0.178
N u0 1.0916 1.1072 1.1129 1.1155 1.117

Table 6.2: Case Ra = 103 . Comparison of XFlow results with the solution in [25]
for different spatial resolutions.

(a) (b)

(c) (d)

Figure 6.2: Result fields for Ra = 103 : (a) temperature, (b) horizontal velocity, (c)
vertical velocity, and (d) vorticity.
6.1 Natural convection in a cavity 69

20x20 40x40 80x80 160x160 Ref.[24]

max Vx 54.448 57.750 63.805 64.902 64.630
at Ymax 0.875 0.863 0.856 0.847 0.850
max Vy 127.259 193.479 215.178 219.817 219.360
at Xmax 0.075 0.038 0.044 0.041 0.038
N u0 5.924 7.975 8.688 8.823 8.817

Table 6.3: Case Ra = 106 . Comparison of XFlow results with the solution in [25]
for different spatial resolutions.

(a) (b)

(c) (d)

Figure 6.3: Result fields for Ra = 106 : (a) temperature, (b) horizontal velocity, (c)
vertical velocity, and (d) vorticity.
70 Heat transfer
 7
Multi-phase flows

Contents
7.1 Rayleigh-Taylor . . . . . . . . . . . . . . . . . . . . . . . . 71

7.1 Rayleigh-Taylor

When heavy fluid lies above lighter, the equilibrium in unstable and a small
perturbation of the interface from the horizontal will grow with time, producing
the phenomenon known as Rayleigh-Taylor instability [26]. This instability is a
prototype problem for computational studies of multi-phase flows.
The problem consists of two layers of fluid initially at rest in the rectangular domain
Ω = (−d/2, d/2)×(−2d, 2d), see Figure 7.1. The flow is characterized by the density
difference between the two fluids and their effective viscosity.

Figure 7.1: Initial configuration and physical properties of the fluids.

The density difference is represented by the Atwood number At = (ρA − ρB )/(ρA +

ρB ). The Reynolds number is defined as Re = ρA d3/2 g 1/2 /µ, where d is the reference
length, g the gravity acceleration and µ the dynamic viscosity of the fluids (assumed
uniform).
72 Multi-phase flows

Figure 7.2: Vertical position of spike and bubble vs. time. Comparison of
XFlow solution with reference ones.

The growth and evolution of Rayleigh–Taylor instability has been investigated

among others by Tryggvason [27] for inviscid incompressible flows, and by Guermond
& Quartapelle [28] and Ding et al. [29] for viscous flows. None of these studies has
taken into account surface tension.
We compare the XFlow results with those of [28] and [29] at At = 0.5 and Re =
3000. The initial position of the perturbed interface is y(x) = −0.1 d cos(2πx/d).
Computations are carried out on a 200 × 800 grid and the time step is automatically
set to 0.000144 s. Free-slip condition is enforced at all walls. The tracking of the
interface is done using the marker-and-cell method.
Results on the vertical position of the tip of the falling and rising fluid (spike and
bubble, respectively) are shown in Figure 7.2. XFlow solution is in good agreement
with the reference results [28, 29].
The evolution of the instability is shown in Figure 7.3 at dimensionless times t̃ =0, 1,
√
1.25, 1.5, 1.75, 2, 2.25, 2.5, where t̃ = t g At. Around t̃ = 1.5 the heavy fluid begins
to roll up into two counter-rotating vortices (see also Figure 7.4). Later, around
t̃ = 2, these two vortices become unstable and a pair of secondary vortices appear
at the tails of the roll-ups. The roll-ups and vortices in the heavy fluid spike are due
to the Kelvin–Helmholtz instability. The shapes of the fluid interface obtained with
XFlow compare well with those of the reference results [28, 29].
7.1 Rayleigh-Taylor 73

t̃ = 0 t̃ = 1 t̃ = 1.25 t̃ = 1.5

t̃ = 1.75 t̃ = 2 t̃ = 2.25 t̃ = 2.5

Figure 7.3: Rayleigh-Taylor instability evolution.

74 Multi-phase flows

Figure 7.4: Velocity field at t̃ = 1.5 s.

 References

[1] C.-H. Bruneau and M. Saad. The 2d lid-driven cavity problem revisited.
Computers & Fluids, 35:326–348, 2006.

[2] U. Ghia, K.N. Ghia, and C.T. Shin. High-Re solutions for incompressible
flow using the Navier-Stokes equations and a multigrid method. Journal of
Computational Physics, 48:387–411, 1982.

[3] Wim M Van Rees, Anthony Leonard, DI Pullin, and Petros Koumoutsakos.
A comparison of vortex and pseudo-spectral methods for the simulation of
periodic vortical flows at high reynolds numbers. Journal of Computational
Physics, 230(8):2794–2805, 2011.

[4] C. Rumsey, R. Biedron, and J. Thomas. CFL3D: Its history and some recent
applications. Technical report, NASA TM-112861, 1997.

[5] D.P. Lockard, L.-S. Luo, S.D. Milder, and B.A. Singer. Evaluation of
PowerFLOW for Aerodynamic Applications. Journal of Statistical Physics,
107(1/2):423–478, 2002.

[6] D. Somers. Design and experimental results for the s825 airfoil. Technical
report, National Renewable Energy Laboratory, 2005.

[7] R. J. McGhee, W. D. Beasley, and J. M. Foster. Recent modifications and

calibration of the langley low-turbulence pressure tunnel. Technical report,
NASA TP-2328, 1984.

[8] P. A. Baranov, S. V. Guvernyuk, M. A. Zubin, and S. A. Isaev. Numerical and

physical modeling of the circulation in a vortex cell in the wall of a rectilinear
channel. Fluid Dynamics, 35:663–673, 2000.
76 Multi-phase flows

[9] S.A. Isaev, S.V. Guvernyuk, M.A. Zubin, and Y.S. Prigorodov. Numerical and
physical modeling of a low-velocity air flow in a channel with a circular vortex
cell. Journal of Engineering Physics and Thermophysics, 73:337–344, 2000.

[10] R. Donelli, P. Iannelli, S. Chernyshenko, A. Iollo, and L. Zannetti. Flow models

for a vortex cell. AIAA Journal, 47:451–467, 2009.

[11] C. Rumsey. The 1st AIAA CFD High Lift Prediction Workshop (HiLiftPW-
1), NASA Langley Research Center, http://hiliftpw.larc.nasa.gov/index-
workshop1.html.

[12] C.B. Mc Ginley, L.N. Jinkins, R.D. Watson, and A. Bertelrud. 3-D High-Lift
Flow-Physics Experiment - Transition Measurements. AIAA Paper, 2005-5148,
2005.

[13] G. Le Good and K. Garry. On the use of reference models in automotive

aerodynamics. SAE paper, 2004-01-1308.

[14] S.R. Ahmed and G. Ramm. Some Salient Features of the Time-Averaged
Ground Vehicle Wake. SAE-Paper 840300, 1984.

[15] C. Hinterberger, M. Garcı́a-Villalba, and W. Rodi. Large eddy simulation of

flow around the Ahmed body, pages 77–87. Springer Berlin Heidelberg, Berlin,
Heidelberg, 2004.

[16] G. Franck, N. Nigro, M. A. Storti, and J. D’Elia. Numerical simulation of

the flow around the Ahmed vehicle model. Latin American applied research,
39:295–295, 2009.

[17] B. Metcalf, J. Longo, S. Ghosh, and F. Stern. Unsteady free-surface wave-

induced boundary-layer separation for a surface-piercing NACA 0024 foil:
Towing tank experiments. Journal of Fluids and Structures, 22:77–98, 2006.

[18] K.M.T. Kleefsman, G. Fekken, A.E.P. Veldman, B. Iwanowski, and B. Buchner.

A Volume-of-Fluid based simulation method for wave impact problems. Journal
of Computational Physics, 206:363–393, 2005.

[19] J. J. Stoker. Water Waves; The Mathematical Theory with Applications. 1992.

[20] A. Selamet and P.M. Radavich. The effect of length on the acoustic attenuation
performance of concentric expansion chambers: An analytical, computational
and experimental investigation. Journal of Sound and Vibration, 201(4):407–
426, 1997.

[21] J. M. Middelberg, T. J. Barber, S. S. Leong, K.P. Byrne, and E. Leonardi.

Computational fluid dynamics analysis of the acoustic performance of various
7.1 Rayleigh-Taylor 77

simple expansion chamber mufflers. Proceedings of ACOUSTICS,, pages 123–

128, 3-5 November 2004, Gold Coast, Australia.

[22] Helmut Kı̈¿½hnelt. Simulating the mechanism of sound generation in flutes using
the lattice Boltzmann method. Proceedings of the Stockholm Music Acoustics
Conference, pages SMAC–1 – SMAC–4, August 6-9 2003, Stockholm Sweden.

[23] W. Kausel and H. Kı̈¿½hnelt. Modelling wave and fluid propagation in wind
instruments. Methods and Applications.

[24] G. DeVahl Davis and I.P. Jones. Natural convection in a square cavity: a
comparison exercise. International Journal for Numerical Methods in Fluids,
3:227–248, 1983.

[25] G. de Vahl Davis. Natural convection of air in a square cavity: a benchmark

numerical solution. International Journal for Numerical Methods in Fluids,
3:249–264, 1983.

[26] D.H. Sharp. An overview of Rayleigh-Taylor instability. Physica D, 12:3–18,

1984.

[27] G. Tryggvason. Numerical simulations of the Rayleigh-Taylor instability.

Journal of Computational Physics, 75:253–282, 1988.

[28] J.-L. Guermond and L. Quartapelle. A projection FEM for variable density
incompressible flows. Journal of Computational Physics, 165:167ı̈¿½188, 2000.

[29] H. Ding, P. Spelt, and C. Shu. Diffuse interface model for incompressible
two-phase flows with large density ratios. Journal of Computational Physics,
226:2078ı̈¿½2095, 2007.
 or registered trademarks of Dassault Systèmes, a French “société européenne” (Versailles Commercial Register # B 322 306 440), or its subsidiaries in the United States and/or other countries. All other trademarks are owned by their respective owners. Use of any Dassault
©2020 Dassault Systèmes. All rights reserved. 3DEXPERIENCE, the Compass icon, the 3DS logo, CATIA, BIOVIA, GEOVIA, SOLIDWORKS, 3DVIA, ENOVIA, EXALEAD, NETVIBES, MEDIDATA, CENTRIC PLM, 3DEXCITE, SIMULIA, DELMIA, and IFWE are commercial trademarks

Systèmes or its subsidiaries trademarks is subject to their express written approval.

Our 3DEXPERIENCE® platform powers our brand applications, serving 11 industries, and provides
a rich portfolio of industry solution experiences.
Dassault Systèmes, the 3DEXPERIENCE Company, is a catalyst for human progress. We provide business and people with collaborative virtual environments
to imagine sustainable innovations. By creating ‘virtual experience twins’ of the real world with our 3DEXPERIENCE platform and applications, our
customers push the boundaries of innovation, learning and production.

Dassault Systèmes’ 20,000 employees are bringing value to more than 270,000 customers of all sizes, in all industries, in more than 140 countries. For
more information, visit www.3ds.com.

Americas Europe/Middle East/Africa Asia-Pacific

Dassault Systèmes Dassault Systèmes Dassault Systèmes K.K.
175 Wyman Street 10, rue Marcel Dassault ThinkPark Tower
Waltham, Massachusetts CS 40501 2-1-1 Osaki, Shinagawa-ku,
02451-1223 78946 Vélizy-Villacoublay Cedex Tokyo 141-6020
USA France Japan