See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266883948

Flow and Turbulence Structures in the Wake of a Simplified Car Model (Ahmed
Model)

Article in Notes on Numerical Fluid Mechanics · January 2002

DOI: 10.1007/978-3-540-45466-3_39

CITATIONS READS

393 8,813

3 authors, including:

S. Becker
Friedrich-Alexander-University Erlangen-Nürnberg
256 PUBLICATIONS 3,234 CITATIONS

SEE PROFILE

All content following this page was uploaded by S. Becker on 16 January 2015.

The user has requested enhancement of the downloaded file.

 Flow and Turbulence Structures in the Wake of a Simplified Car
Model (Ahmed Model)

H. Lienhart, C. Stoots, S. Becker

Lehrstuhl für Strömungsmechanik (LSTM)

Universität Erlangen-Nürnberg
Cauerstr. 4, 91058 Erlangen, Germany

Summary
The aim of the "Models for Vehicle Aerodynamics" (MOVA) Project is to develop, refine,
and validate the latest generation of turbulence models for selected examples encountered in
vehicle aerodynamics. The validation of turbulence models requires the availability of
detailed experimental data. These quantitative data should cover the most critical flow
regions around a bluff car-shaped body and they should give physical quantities that can
directly be correlated to the results of numerical simulations. Such experimental data were
measured in the LSTM low speed wind tunnel using a 2-component laser-Doppler
anemometer (LDA) mounted on a traversing system and a simplified model of a car (Ahmed
model). Measurements were made for two rear vehicle body slant angles (25° and 35°) at a
bulk air velocity of 40 m/s. This paper serves as a synopsis of the major results of this
experimental investigation.

Introduction
The flow region which presents the major contribution to a car’s drag, and which poses severe
problems to numerical predictions and experimental studies as well, is the wake flow behind
the car. The location at which the flow separates determines the size of the separation zone,
and consequently the drag force. Clearly, a more exact simulation of the wake flow and of the
separation process are essential for the correctness of drag predictions. However, a real-life
automobile is a very complex shape to model or to study experimentally. Therefore the
MOVA consortium partners (TU Delft, University of Manchester, LSTM, Electricite de
France, AVL List, and PSA Peugeot Citroen) agreed to study the vehicle shape employed by
Ahmed [1], known as the Ahmed body. Figure 1 is a schematic of the Ahmed body, with
actual dimensions in mm included. Two different rear body slant angles (25° and 35°) were
considered, which happen to bracket the critical angle of 30° at which separated flow occurs
within the wake of the slant (see Figure 2).
The Ahmed body was mounted in the test section of the LSTM wind tunnel (Figure 3) and
detailed measurements of velocity profiles were made around this body. The experiments
were performed in the LSTM low speed wind tunnel, a closed return facility which can be
configured with an open or a closed test section. The present studies were conducted in a ¾
open test section (i.e., floor, but no sides or ceiling) with a blockage ratio of 4%. The wind
tunnel can generate flow velocities from 3 to 55 m/s with average turbulence intensities of less
than 0.25%. All measurements concerning the Ahmed body were taken at bulk air velocities
of 40 m/s. To ensure constancy of the test section bulk velocity and air temperature, a
computer-based feedback control system was utilised.

Figure 1. Schematic of the bluff body (Ahmed, 1984) shape.

Figure 2. Characteristic drag coefficients for the Ahmed body for various slant angles.
Hot-wire measurements of the velocity profiles 400 mm upstream of the Ahmed body were
obtained to serve as a inlet condition for the numerical simulations. The measurements were
performed by a two-component hot-wire system, which was rotated to obtain the third
component. The results showed that the test section inlet velocity profile was well defined
and controlled, which guaranteed reliable measurements for the model validation database.
Flow visualization using oil streaks was performed for both model shape angles. The
visualization showed complex three-dimensional flow patterns and confirmed earlier findings
that a small change in the slant angle around the critical 30° causes a dramatic change in the
flow pattern. Comparison of Figures 4 and 5 illustrate these changes. The formation of
clinging vortices and flow reversal for a slant angle of 25° is evident in the photograph of oil
streaks in Figure 4. Attached flow is maintained in this case while at a slant angle of 35°
(Figure 5) detached flow is obvious.

Figure 3. Photograph of the Ahmed body mounted in the LSTM low-speed wind tunnel.

Figure 4. Oil / soot streak flow visualization Figure 5. Oil / soot streak flow visualization
of the Ahmed body rear for a 25° slant angle. of the Ahmed body rear for a 35° slant angle.

A two-component laser-Doppler anemometer (LDA) was installed on an existing three-

dimensional computer controlled traversing system (rotation of the sending/receiving optics
offer an additional axis of traversing). The LDA was composed of DANTEC fiberoptic-based
optics and electronics. Two different laser wavelengths were used: 514.5 nm (green) and 488
nm (blue). The system was of backscatter orientation, the only optical configuration possible
for these measurements due to the proximity of the measurement grid to the body surface.
The laser was a water-cooled, 5 watt argon-ion Spectra Physics Model 2060. Beam splitting
and frequency shifting were provided by DANTEC FiberFlow optics. Signal analysis and
signal processing were accomplished via a DANTEC Model 57N20 Burst Spectrum
Analyzers (BSA) and DANTEC BSA Flow software. A PC computer provided measurement
control and data acquisition / storage. LDA measurements were made for all three
components of velocity in the symmetry plane from upstream of the Ahmed model to some
distance downstream behind the closure of the wake. LDA measurements were also made in
several transverse planes in the wake.
 LDA Measurement Results
Figure 6 attempts to show the different measurement positions employed in this effort. In
summary, there were 7,500 discrete measurement positions located in 13 unique planes.

400
200

y [mm]
0
-200
-400
-1500 -1000 -500 0 500
x [mm]

Figure 6. Measurement planes utilized to characterize the velocity distribution near the body
surface.
Because higher order statistical moments (i.e., Reynold's stresses, u 'u 'u ' , u 'u 'v ' , etc.) were of
interest, each measurement location had to be sampled twice -- first for the U and V
components of velocity, then for the U and W components. Preliminary statistical analysis
indicated that approximately 40,000 measurement realizations were necessary at each location
for statistically significant results. This translates to a maximum measurement time of
approximately 5 minutes at each location. Following the procedure outlined in [2], for a 95%

TKE: 5 10 15 20 25 30 35 40 45 50

U,V,W / U ∞ = 1

600

-1500
z [mm]

400
-1000
200
-500
x [m 0
m] 0 200
0 ]
500 m
-200 [m
y
Figure 7. Overview of velocity distributions and TKE around the 25° slant Ahmed body.
confidence interval the statistical uncertainty in the outer flow (far from the Ahmed body)
mean velocity was less than 0.005% of the local mean velocity. Within the wake of the
Ahmed body, where turbulence intensities are very high and mean velocities can approach
zero, an estimate of mean velocity measurement uncertainty is rather arbitrary. An order of
magnitude estimate for the calculated 95% confidence interval was 1% for mean and 1.5% for
rms quantities. The LDA system could measure turbulence intensities as small as 0.8%.
Figure 7 serves as a good overview of typical velocity distributions around the 25° slant
Ahmed body. The mean velocity vectors along the line of symmetry of the rear slant indicate
that no separation of the flow is occurring. The two counter-rotating trailing vortices are
shown in the three transverse planes of turbulent kinetic energy (TKE) contour plots. Peaks in
TKE occur in the centers of the vortices. These vortices are responsible for maintaining
attached flow at the slant up to a slant angle of approximately 30°. Vortical structures extend
more than 500 mm beyond the end of the Ahmed body.
800 reference vector 800 reference vector
slant angle 25 degree U;W / U ∞ = 1 slant angle 35 degree U;W / U ∞ = 1

600 600
z [mm]
z [mm]

400 400

200 200

0 0
-400 -200 0 200 400 600 -400 -200 0 200 400 600
x [mm] x [mm]

Figure 8. Comparison of the recirculating wake region and attached / detached flow for the
two different rear slant angles.

50 m/s

TKE: 0 20 40 60 80 100 120 140 160 180 200

600
z [mm]

400

200

0
-1000 -500 0 500
x [mm]

Figure 9. Contours of TKE along the symmetry plane of the 25° slant Ahmed body
Figure 8 is a direct comparison of the region of attachment / detachment and recirculation at
the rear of the Ahmed body for the two slant angles of interest. These measurements were
made along the plane of symmetry of the Ahmed body. The intention of this figure is to
indicate the major differences in the near-wake flow fields for the two slant angles studied.
Z

Y
X

500

250

500 0
250 500
y

0
[m

0
m

-500 ]
-250 x [mm
]

-1000

50 m/s x = 00 mm 50 m/s x = 80 mm
U [m/s]: -10 -5 0 5 10 15 20 25 30 35 40 U [m/s]: -10 -5 0 5 10 15 20 25 30 35 40

400 400
z [mm]

z [mm]

200 200

0 0
-200 0 200 -200 0 200
y [mm] y [mm]
50 [m/s] x = 200 mm 50 [m/s] x = 500 mm
U [m/s]: -10 -5 0 5 10 15 20 25 30 35 40 U [m/s]: -10 -5 0 5 10 15 20 25 30 35 40

400 400
z [mm]

z [mm]

200 200

0 0
-200 0 200 -200 0 200
y [mm] y [mm]

Figure 10. Development of the counter-rotating trailing vortex system within the wake of the
25° slant Ahmed body.
More detailed near-wake measurements were made but omitted from this figure to avoid
obscuring the wake structures. The only additional information provided by the more detailed
measurements concerns a very small recirculation region in the upper part of the 25° slant
surface. Beyond this region, the flow then reattaches to the slanted surface, then develops a
second, larger recirculation region along the lowermost part of the rear. Observations for the
35° case are quite different, however. The flow detaches along the slanted surface and
develops a single, significantly larger recirculation region within the wake.
50 m/s 50 m/s
x = 200 mm x = 200 mm
uuv: -500 -400 -300 -200 -100 0 100 200 300 400 500 uuw: -1000 -850 -700 -550 -400 -250 -100 50 200 350 500

400 400
z [mm]

z [mm]
200 200

0 0
-200 0 200 -200 0 200
y [mm] y [mm]
50 m/s 50 m/s
x = 200 mm x = 200 mm
uvv: -500 -450 -400 -350 -300 -250 -200 -150 -100 -50 0 uww: -1000 -850 -700 -550 -400 -250 -100 50 200 350 500

400 400
z [mm]

z [mm]

200 200

0 0
-200 0 200 -200 0 200
y [mm] y [mm]

Figure 11. Profiles of higher-order statistical moments within the wake of the 25° Ahmed
body.
Mean velocity vectors and profiles of turbulent kinetic energy (TKE) for the symmetry plane
of the Ahmed 25° degree slant angle body are displayed in Figure 9. Peak TKE values are
within the small recirculation zone at the rear of the body. Note that the Ahmed body only
disturbs the outer flow for a relatively small distance from its surface.
Finally, Figure 10 shows the downstream development of the counter-rotating trailing vortex
system produced by the 25° slant Ahmed body. The contours represent regions of constant
magnitude downstream velocity. At 80 mm downstream from the trailing edge of the Ahmed
 body, there is a large and strong region of recirculation back towards the Ahmed surface.
Although the recirculation has disappeared by x = 200 mm, there is still a large streamwise
velocity deficit. At x = 500 mm the location of the cores of the trailing vortices can still be
distinguished by deficits in streamwise velocity.
Figure 11 is an example of profiles of second - and third - order moments for the wake region
of the Ahmed body. The contours of various moments display good symmetry

Conclusions
Flow visualization, hot wire, and LDA measurements have been performed around an Ahmed
body for two different rear slant angles: 25° and 35°. These two slant angles bracket a critical
instability when the flow detaches from the slanted surface. LDA measurements include all
three components of mean and RMS velocity as well as most second (Reynolds stresses) and
third order moments. The measurements clearly show the differences in flow attachment and
recirculation for the two different slant angles considered. The quantitative information
provided by these measurements should prove invaluable for developing, testing, and
validating computer models of the aerodynamics of vehicular wake regions

Acknowledgements
The support of the presented work by the European Community under the Industrial &
Materials Technologies Program (Brite-EuRam III), contract number BRPR-CT98-0624, is
gratefully acknowledged. Special thanks to Dr. Ahmed and the DLR for providing us with the
wind tunnel model.

References
[1] Ahmed, S.R., Ramm G.: Some Salient Features of the Time-Averaged Ground Vehicle
Wake. SAE Technical Paper 840300, 1984.
[2] Bendat, J.S., Piersol, A.G.: Random Data Analysis And Measurement Procedures.
John Wiley & Sons, New York, 1986.

View publication stats