Ahmed 1984

Some Salient Features of the Time -Averaged Ground Vehicle Wake
Author(s): S. R. Ahmed, G. Ramm and G. Faltin

Source: SAE Transactions, Vol. 93, Section 2: 840222––840402 (1984), pp. 473-503
Published by: SAE International
Stable URL: https://www.jstor.org/stable/44434262
Accessed: 15-12-2018 11:01 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
SAE International is collaborating with JSTOR to digitize, preserve and extend access to SAE
Transactions
This content downloaded from 144.122.49.153 on Sat, 15 Dec 2018 11:01:39 UTC
All use subject to https://about.jstor.org/terms
840300
Some Salient Features

of the Time -Averaged Ground
Vehicle Wake
S. R. Ahmed and G. Ramm

Institut für Entwurfs- Aerodynamik, DFVLR
G. Faltin
Techn. Univ. Braunschweig
Abstract
For a basic ground vehicle type of bluff body, the time averaged
wake structure is analysed. At a model length based Reynolds
number of 4,29 million, detailed pressure measurements, wake
survey and force measurements were done in a wind tunnel. Some
flow visualisation results were also obtained. Geometric para-
meter varied was base slant angle. A drag breakdown revealed that
almost 85 % of body drag is pressure drag. Most of this drag is
generated at the rear end.
Wake flow exhibits a triple deck system of horseshoe vortices.

Strength, existence and merging of these vortices depend upon the
base slant angle.
Characteristic features of the wake flow for the low drag and
high drag configurations is described. Relevance of these phe-
nomena to real ground vehicle flow is addressed.
2.473
0096-736X/85/9302-0473$02 . 50
Copyright 1985 Society of Automotive Engineers, Inc.
2.474 S. R. AHMED, ET AL.
1 . Introduction
Ground vehicles can be termed as bluff bodies moving in close

vicinity of the road surface. The shape of such vehicles evolved
over the years under the constraints of aesthetics, operational
safety, service accessibility etc. Effect of these design guidelines
on aerodynamics was not of prime importance in the past.
With the increased concern about future availability of fuel, fuel
economy is an important requirement expected of a modern car or
utility vehicle. Fuel consumption depends, among other factors, on
the aerodynamic drag of the vehicle.
A key feature of the flow field around a vehicle are the regions of
separated flow. Even simple basic vehicle configurations free of
all appendages and having smooth surfaces generate a variety of
quasi two-dimensional and fully three-dimensional regions of sepa-
rated flow. A major contribution to the drag of a vehicle stems
from the pressure drag which is a consequence of flow separation.
In a time averaged sense, the regions of flow separation exhibit

complex kinematic macro structures. Such structures in the wake,
which is the major separated flow region of a vehicle flow field,
determine the drag experienced by the body.
A qualitative understanding of the flow phenomena in bluff body

wakes is available in aeronautical and automotive literature.
Results which could enhance the quantitative insight into the com-
plex interrelation between wake structure, pressure distribution
on body surface, drag and configuration geometry are scarce. Lack
of this information is a major hinderance to the attempts currently
being undertaken to theoretically model vehicle flow fields. This
data can also be used to validate the computational codes developed.
Based on extensive experimental results, this paper attempts to

(1) identify the time-averaged flow structures present in the wake
of a basic vehicle type body, and (2) analyse the effect of body
geometry on wake structure, pressure distribution and drag. Atten-
tion is focussed on rear end geometry and wake structure. This was
the consequence of a drag breakdown analysis, which indicated that
TIME-AVERAGED GROUND VEHICLE 2.475
most of the drag is generated by flow mechanisms at the body rear end.
2 . Experimental investigations
Model
Selection of the configuration used in this study was governed by

the requirement that it should generate the essential features of
a real vehicle flow field, with the exception of that due to rotating
wheels, engine and passenger compartment flow, rough underside
and surface projections. It was conjectured that the model chosen
should generate: a strong three-dimensional displacement flow in
front, relatively uniform flow in the middle, and a large struc-
tured wake at the rear.
The wind tunnel model, Fig. 1 , with an overall length of 1.044 m

had a length : width : height ratio of 3.36 : 1.37 : 1. It consisted
of three parts; a fore body, a mid section and a rear end. Edges
of the fore body were rounded, as indicated in Fig. 1, to achieve
a separation free flow over its surface. Middle section was a box
shaped sharp edged body with a rectangular cross section. A set of
nine interchangeable rear ends enable a base slant variation in
steps of 5°, between the value of 0° and 40°. A tenth rear
variant, with 12.5° base slant was also tested. All rear ends had
same base slant length 1 of 222 mm and had sharp edges. Morel Ml
O
used a similar bluff body to investigate the effect of base slant

on drag behaviour.
One half of the model was instrumented with pressure taps. A total
of 210 taps on the fore body and 83 taps on the mid section were
evenly distributed on the surface. Only three rear ends with slant
angles of 5°, 12.5° and 30° were equipped with pressure taps
5° - rear end had 444, the 12.5°- rear end 430 and the 30° - rea
end 450 pressure holes distributed evenly over one half of its
surface. Scanivalves for acquisition of pressure data were in-
stalled inside the model body.
Wind tunnel and test set-up
The tests were conducted in the DFVLR subsonic wind tunnels at
Braunschweig (pressure measurements, flow visualization) and
Göttingen (wake survey, force measurements) . These facilities,
described in C23 and C3] are open test section, closed return,
wind tunnels with a square 3 m by 3 m nozzle. A test section length
of about 5.8 m is available for the experimental set-up.
The model was fixed on cylindrical stilts 50 mm above a ground

board 3 m wide and 5 m long, Fig . 2 . About 1.35 model length of
ground plane projected in front and 2.43 model length behind the
model. Leading edge of ground board was carefully rounded to
avoid any separations.
All tests, except the flow visualization, were performed at a wind

speed of 60 m/s. This corresponds to a model length based Reynolds
number of 4.29 million. In both tunnels the turbulence intensity
lies below 0.5 per cent.
Wake survey
A ten hole directional probe, Fig. 3, which is a further develop-

ment of the probe used in earlier experiments £4 3, C53, £63, was
employed for the wake survey. On the conical tip of the yawmeter
probe, two pairs of opposing orifices are arranged. One pair is
sensitive primarily to flow yaw and the other to flow incidence.
To determine the flow angularity, yaw rotations are imposed on
the probe till the pressure in the yaw sensing orifice pair is
equalised. In this position, the probe tip axis is pointing
nominally along the direction of local flow yaw. Local flow inci-
dence is computed from the pressure difference of the incidence-
sensitive orifice pair via a calibration curve. In the yaw mode,
the probe shaft is vertical.
Alternately, with probe shaft horizontal, incidence rotations can
be imposed to equalise the pressure in incidence sensing orifice
pair. Yaw angle is determined then, as before, with a calibration
curve. The decision, which orifice pair is used to align the probe
tip depends upon the anticipated incidence and yaw gradients in
the flow.
Mean value of the pressure sensed by the four orifices on the

cylindrical sleeve of probe tip, and the pressure in the tip
orifice are functions of local static and total pressure. Cali-

bration curves are used to evaluate the static and dynamic pressure
from this data. Thus magnitude and direction of local velocity
vector are determined.
The tenth orifice, situated at the probe tip rea

is used to detect flow reversal. If the total pressure measured here
exceeds the value sensed at the probe tip, the probe shaft is ro-
tated 180°. Nomenclature and formulae for evaluating the flow
and velocity components from probe data are summarised in Fig. 3.
The wake survey probe was mounted on a rigid carriage which provides
cartesian translation along the full length, width and height of
the test section. Such movements are remote controlled and digitized
by electronic counters.. A wake scan in a YA-ZA plane is performed
by moving the probe in zA direction with xA and yA positions kep
fixed. The zA -traverse was repeated with a new value of yA which
stepwise increased. During the zA~traverse, the probe halted at
discrete points for a duration of 2s. The analog values recorded
during this period were integrated and were used to compute the
mean pressure values.
Estimated accuracy of flow angle measurement is ±0.4°. Errors
of upto 1 per cent of free stream dynamic pressure are present in
the pressures measured. More details of the accuracy estimates
are given in C 53 .
Pressure and force measurements
Pressure and force measurement results reported here are restricted

to the zero yaw onset flow condition. Some experiments conducted
with yawed onset flow in the range of ß = ±10° served mainly a
check of the flow symmetry.
For force measurements, the model was connected to a strain gauge
balance, arranged below the ground plane, by four cylindrical
stilts. Balance assembly was screened from tunnel air flow by a
casing.
Estimated errors in the force and moment measurements are ±0.2 N
and ±0.1 Nm respectively.
2.478
3. Experimental Results s* R* AHMED' ET
Drag and pressure measurements
Variation of drag with base slant angle , for a body of the

type studied here has been already reparted by Morel C1D. Also
Janssen and Hucho £7D, in an earlier paper, described a similar
drag behaviour for a passenger car. Fig. 4 illustrates the drag
breakdown obtained through force and pressure measurements for
the zero yaw onset flow conditions. Even though apparent differ-
ences were present in the model geometry (edge radii, overall
dimensions, stilt positions etc.) and a different ground clear anc«
was used in the test setup, the total drag values obtained are
almost same to those of Morel HID.
Contributions to pressure drag from front part c

J'
*, slant rear
end c *, and
o
vertical rear end
B
base c * were evaluated by inte-
gration of the axial component of the measured pressure over the
surface. Basis for drag breakdown are pressure measurements on
configurations with rear end slant angles of 5°, 12.5° and 30°.
The low drag flow for y = 30° was realised by fixing a splitter
plate vertically on the ground board in the plane of symmetry
behind the model. Between the upstream edge of the splitter plate
and the model base, a gap of about 25 mm was left free.
Numerical values of the drag contributions are given in Table 1

below. Total drag value has been corrected for tare drag of stilts.
Base slant angle cT7 c * c * c*

^ W K o B
(High°Draq) °-3
(Low Drag) ļ0"26
Table 1 . Drag breakdown for three configurations
Results of Fig. 4 and Table 1 clearly demonstrate the small con-

tribution of forebody to the total pressure drag. For the
TIME -AVERAGED GROUND VEHICLE 2.479
configuration studied, its v

£
slant low drag situation), constant at c J'= 0.016. A conclusion,
which can be drawn from this is that the interference between the
rear end and fore body flow is weak; this could be a consequence
of the relatively long mid section. The result is not necessarily
valid for configuration with a short mid section. As the flow
leaving the fore body feeds the flow on the mid and aft sections
of the body, and subsonic flow is considered, interference between
upstream and downstream flow regions is to be anticipated.
Major contribution to the pressure drag comes from the slant and
vertical base surface of rear end. For base slant angle y =
rear end pressure drag is wholly contributed by the flat base;
with increasing values of y , the vertical base area decreases, but
the pressure distribution is changed as well, see Table 2. Thus
there are two overlapping effects, one geometrical, the other fluid
mechanical.
Total Pres- Vertical Base Slant Surface F0^0"

sure Drag
f
cX
5°
12.5°
30°
(High 0.321 0.615 0.29 0.39 0.664 0.050 0.849 0.151
Drag)
_
(Low 0.209 0.615 0.48 0.39 0.426 0.091 0.804 0.196

Drag)
Table 2. Relative Drag Contributions
With y changing from 5° to 30° (see Table 2) , the vertical

decreases from 93 % of front area to 61.5 %; the c * value decreases
from 86 % to 29 % of total pressure drag. For the same variation,
the projected area of slant surface increases from 7 % to 38.5 % of
model front area; the contribution to pressure drag c* increases O
thereby from 5.4 % to 66.4 %. These figures emphasize that the

pressure drag for the basic body considered here is mainly generated
on the slant and vertical base of the rear end.
Also shown in Table 2 are the relative magnitudes of the total
pressure drag cp* a
drag c^. The relativ
to 85 %, with friction drag accounting for the rest of 24 % to
15 %. Considering the "minimum" drag configuration with =
76 % of its total drag is pressure drag and the rest 24 % friction
drag. The high drag configuration with <f = 30° has 85 %
total drag resulting from pressure drag, with the rest 15 % coming
from friction drag. Contribution of friction drag to the overall
drag becomes thus of increasing importance for low drag configu-
rations.
Fig.
surface. Presence of vortices at side edges of the slant surface

is visible for the configurations f= 12.5° and 30° (high
condition) .
Except in the vicinity of the side edges, the flow on the slant
surface of <f = 12.5° configuration appears to be "two dim
This is evident from the parallel isobars running across the
surface. A large portion of the flow coming off the upstream edge
of this surface, experiences a pressure recovery, which results
in the low values of c* and c * (see Table 1).
O
Strong vortices are present in the flow field of<f = 30 configura-
tion experiencing a high drag, Fig. 5. The vortices influence and
shape the flow over the whole slant surface. Low drag flow for
this configuration shows a flow separation at the slant surface
upstream edge.
The usual assumption of a constant base pressure is not confirmed

by the results of Fig. 5. Especially for they = 30° configurati
the pressure on base in influenced by flow on slant and undersur-
face.
Wake structure
Even though the wake flow of a bluff body is basically unsteady,

the time averaged flow exhibits a macrostructure which appears
to govern the pressure drag created at the rear end. Before discus-
sing the quantitative results of the wake survey, the salient
flow features of the wake, deduced from these results, are
illustrated in Fig . 6 . This schematic sketch of the flow phenomena

is the result of present and earlier studies C4l, C5D, C6D.
The shear layer, coming off the slant side edge, rolls up into a
longitudinal vortex, in a manner similar to that observed on side
edge of low aspect ratio wings. At the top and bottom edges of the
flat vertical base, the shear layer rolls up as indicated, into
two recirculatory flow regions A and B, situated one over another.
The flow on the base surface, derived from oil flow pictures, does
not indicate that the flow regions A and B end on the base surface.
Gonsequently the recirculatory flow A and B can be thought of as
being generated through two "horseshoe" vortices situated one
above another in the "separation bubble" indicated by D in Fig. 6.
The "bound" legs of these vortices are approximately parallel to
the base surface; the "trailing" leg of upper vortex A, aligns
itself in direction of onset flow and merges with the vortex C
coming off the slant side edge. Downstream development or dissipa-
tion of the lower vortex B has been difficult to analyse during
the present and previous investigations, so that a conclusive
statement about its behaviour cannot be made. The shear layer
separating at the vertical side edges Of the base seems to split
up, part of it drawn upwards into the "trailing" leg of vortex A
and into vortex C; rest of it probably merges into the "trailing"
leg of vortex B. Streamlines on base surface, schematically shown
in Fig. 6, indicate this process.
As the flow over the slant surface is influenced by the vortex C

coming off the side edge, the strength of vortex A is dependent
upon the strength of vortex C. As long as flow remains attached
over the slant surface, strength of vortex C depends upon the
base slant angle cp ; consequently the strength of vortex A also
depends upon the angle y . Strength of vortex B depends in the first
instance upon the flow conditions in the ground clearance gap. It
is indirectly linked to the base slant angle <j> over the vortices
A and C.
Quantitative data to support the flow module described above is

presented in Figs. 7,8 and 9. For tne base slant angles of <f =
and 25°, the distribut

y. X Z
of symmetry of the wake is shown in Fig. 7 The length of the
pointers equals the magnitude of velocity vector XVz
at the
location considered.
Clearly visible are the recirculatory flows A and B of Fig. 6.

Also the separation bubble boundary D can be identified. Flow
velocities near the model base and in the region where the sepa-
ration bubble "closes" are small and difficult to measure; the
blank regions indicate the area where meaningful results could not
be obtained by the probe.
Fig. 7 illustrates also the change effected in the extent of the
recirculating flows in wake due to a change in the base slant angle.
Where as for y = 5°, both the upper and lower regions are of
comparable order of magnitude, with <j> increased to 25°, the upp
region dominates the flow phenomena in the wake. Also the length
of the separation bubble is almost halved.
A plot of the velocity vector yV in a transverse plane close to

the model base is shown in Fig . 8 for the <f = 5° configuration.
This result is in support of the concept of two horseshoe vortices
in the wake as hypothesized above.
The merging process of the upper horseshoe vortex A (see Fig. 6)

with the vortex C coming off the side edge is depicted in the
results of Fig.
half of the symmetric transverse planes at xÄ/l =-0.077, -0.19

and -0.479 for they» = 25 configuration*. The cross section
boundary of the bubble is shown shaded in Figs. 9 a and b. The
contour of the bubble edge was evaluated by assuming it to be
situated at points where the total pressure coefficient cy equals 0.1 .
Region of reversed flow, is cross hatched.
Formation of the side vortex is clearly visible in Fig. 9 a; it is

also seen that its core is fed by the separation bubble. Velocity
vectors in the cross hatched region indicate the existence of an
upper and lower region of reversed flow; the axis of the upper
*
W
obtain as the high drag creating flow could not be maintained over
a long period of time in wind tunnel tests.
*#-
Note the difference in scale of the velocity vectors plotted.
TIME- AVERAGED GROUND VEHICLE 2.483
region is curved upwards in direction of the core of side edge
vortex. Further downstream, atxA/l = -0.19, the separation

narrows down, and the side edge vortex core, isolated, lies above
the separation bubble "boundary".
Merging of side edge vortex and upper separation bubble vortex
takes place close to the model base; after it, the edge vortex
and separation bubble appear as separate entities; this seems to
be the case especially where a strong side edge vortex is generated,
as in the configuration with <f> = 25°. Still further downstre
xA/l = -0.479, the separation bubble closes, and the overall f
is dominated by the downwash inducing vortex, Fig. 9c.
Fig. 10 illustrates schematically the features analysed for the

high drag flow situation on the = 30° rear end. For this base
slant configuration, the flow in the middle part of the slant
surface separates at the upstream edge; the presence łof strong side
edge vortices prevents a lateral widening of this separated flow.
A closed, half elliptic region of circulatory flow "E" , flanked
by 2 triangular attached flow regions "F", is present on the slant
surface. This may be assumed to be generated by a fourth vortex,
whose axis is aligned with the core of the circulatory flow E in
Fig. 10. The "trailing" leg of this vortex merges with the vortex
coming off the side edge at the leading edge/side edge junction.
Thus the core of the side edge vortex is fed with low energy
material from the separated flow region on the slant base surface.
The presence of this separation region lowers the level of pressure

prevalent on the complete surface of base slant; this contributes
to the dramatic rise of the pressure drag evidenced by the results
of Fig. 4, Table 1, and the isobar plots of Fig. 5.
Actually, the flow separation, in the region denoted by E in Fig. 10,

is already initiated at still lower values of cj? . This is shown in
the oil flow picture series of Fig. 1 1 . Formation of side edge
vortex with a primary and secondary vortex formation can also be
observed. Configurations with base slant angles slightly less than
30°, appear to generate the three individual vortices A, B
of Figs. 6 and 10. With <f> slightly above 30°, the separatio
on the slant surface joins the separation bubble of the base, so
2-484 S. R. AHMED, ET AL.
that the vortices A and E can no longer be considered as separate.
This merging of the separation regions, probably triggered by
seemingly insignificant disturbances in the oncoming flow, results
in the switch over to the low drag type of flow in case of the
= 30° configuration. This low drag flow is characterised
absence of the strong side edge vortices. Proof of this is shown
in the total pressure isobar contours of Fig. 12. The pressure
measurement was done at xA =0, i.e. at the base, just above the
slant surface downstream edge.
In Fig. 12a, the side edge vortex and the region of separated flow
in the middle are clearly noticeable. The flow observed is that
for a = 30° configuration under high drag condition. Fig
illustrates the isobar contour for the low drag situation. Only
a weak trend of the flow to turn around the side edge can be
detected. Otherwise the flow appears to be separated over the
complete slant surface.
Inspite of the absence of the side edge vortices, which as dis-

cussed above shape the flow mechanisms in the wake, a similar
cross flow field was observed in the far field for both flow
situations, as shown in Fig. 13.

In the velocity distribution shown for the high drag flow (at the
station x /i =-0.47 9) , a strong downwash creating vortex with a
A
narrow core is to be seen, Fig. 13 a. In the low drag flow
(Fig. 13 b), the roll up process of the shear layer coming
off the rear end periphery is not yet complete. The reason
that also in this case, a weaker but still downwash creating
circulatory flow is generated, seems to be the following: the
flow separating at the upstream edge of the slant surface induces
a downward tilt to the oncoming flow off the upper surface. Flow
coming off the side and lower edges of the rear end separates
farther downstream, so that a downwards and inwards (from slant
edge) tendency is imparted to the flow coming off the rear end.
Relevance of the flow phenomena described for the wake śtructu.re

and drag behaviour of the idealized vehicle type body studied here
for a real vehicle can be judged by comparing the present results
with some results obtained earlier C61 on a quarter scale vehicle
TIME -AVERAGED GROUND VEHICLE 2.485
model, having same overall dimensions. With the help of interchange-

able upper rear ends, the base slant angle was varied systematically
over the same range as in the present tests. Also the length of
the base slant in plane of symmetry was same and equal to 222 mm.
The plan form of the slant surface, however, changed with the base
slant angel (f , as the model used in C6D had a curved side and roof
surface. Another important difference to the present model was that
the rear part of model undersurface was slightly upswept, creating
a diffusor type of flow in the gap between model undersurface and
ground board.
Fig. 14 shows photographic evidence, taken from C63, of the two

recirculatory flow regions discussed above and denoted with A and B
in Fig. 6. A thin smoke tube was projected vertically through
the ground plane in the wake region, just below and above the
separation bubble edge.
Drag results obtained for the model in C6H, shown in Fig . 1 5 e

confirm the trend indicated by the present results. Also the high
drag value for both models is obtained at the same base angle of
<f = 30°.
Effect of undersurface upsweep has a similar influence on the wake
development, as a base slant. A base slant imparts a downwards
and inwards trend to the separating shear layer at top. An under-
surface upsweep, imparts in a similar fashion, an upwards and
inwards trend to the flow coming from beneath the vehicle. If the
upsweep effect dominates, the longitudinal vortices created down-
stream have a rotation sense which generates an upwash. Fig. 14a
and d depict the situation where either the undersurface upsweep
or the base slant dictates the final sense of rotation of the
vortices in the wake. The cross over point, where a change in the
sense of rotation takes place, lies between the base slant values
of = 10° and 15°. For this "optimum" value, the clearly de
vortex motion seen to be present at <f -values of 5° and 25°
down. The cross flow present at the "optimum" base slant angle is
consequently anticipated to be weak. This phenomena is apparently
associated with the aerodynamic drag; result of Fig. 1 4 e show that
the drag value is lowest at a base slant angle of 12.5°.
Coming back to the wake flow module hypothesized in Fig. 6, the

observations made above, lead to the following explanation. For
the optimum low drag configuration, the strength of both horse-
shoe vortices A and B becomes equal. This can lead to a merging
process, resulting in a ring of vortex, housed in the separation
bubble emanating at the base. The weaker side edge vortex, appears
to develop uninhibited by the flow phenomena inside the separation
bubble. As the mutual strengthening of vortices A and C through
merging is absent, the resulting cross flow in the downstream
also weakens.
A result to substantiate the ring vortex formation concept,

described above, is given in Fig . 16. In Fig. 16a, the cross flow
velocity vectors V^z, and in Fig. 16b the total pre

at a downstream station x^/1 = -0 . 1 1 are shown. Reg
flow is represented by a cross hatch. Referred to the vertical
base area, the upper and lower regions of reverse flow are approx-
imately equal in magnitude.
4 . Cone lusions
1. For the basic bluff vehicle type of body considered, upto

85 % of the total drag is pressure drag. Rest is friction drag.
2. With attached flow prevalent over its surface, the forebody

contributes a maximum of 9 % to the pressure drag. Rest of the
pressure drag is generated at the rear end.
3. The time-averaged structure of the wake exhibits a pair of

horseshoe vortices, situated one above another in the sepa-
ration bubble at the vehicle base. Vortices, coming off the
slant side edges are also present.
4. Strength of side edge vortices and the horseshoe vortices in

the separation bubble is mainly determined by base slant angle.
5. A low drag rear end configuration induces a weak transverse

flow in its wake.
6. High drag generating flow is characterized by strong side edge

vortices, a separation bubble on the base slant surface, and
TIME- AVERAGED GROUND VEHICLE 2.487
the separation bubble emanating from the vertical rear end
base.
7. High drag flow, described under point 6 above, is unstable;

the switch over to the stable low drag flow is accompanied by
disapperance of strong vertical motion in the wake.
5 . Acknowledgement
The authors wish to thank Deutsche Forschungs- und Versuchsan-

stalt für Luft- und Raumfahrt (DFVLR) for the permission to
publish this paper. They are indebted to the staff of the DFVLR
wind tunnel department at Braunschweig and Göttingen for the
cooperation and help offered.
6. References
CID T. Morel: The Effect of Base Slant on the Flow Pattern and
Drag of three-dimensional Bodies with Blunt Ends.

Proceedings of Symposium on Aerodynamic Drag Mechanisms of
Bluff Bodies and Road Vehicles, (Editors G. Sovran et al.),
Plenum Press, New York, 1978, pp. 191 - 226.
C 2D H. Trienes: Der Normalwindkanal der Deutschen Forschungs-

anstalt für Luft- und Raumfahrt (DFL) in Braunschweig.
Zeitschrift für Flugwissenschaften, 12 (1964), 4, pp. 135-142
C3D F.W. Riegels, W. Wuest: Der 3-m Windkanal der Aerodynamischen

Versuchsanstalt Göttingen.
Zeitschrift für Flugwissenschaften, 9 (1961), pp. 222 - 228
C4d S.R. Ahmed, W. Baumert: The Structure of Wake Flow Behind

Road Vehicles.
Symposium on Aerodynamics of Transportation, (Editors

T. Morel et al.), ASME, New York, 1979, pp. 93 - 103
C5D S.R. Ahmed: Wake Structure of Typical Automobile Shapes.

ASME Journal of Fluids Engineering, Vol. 103, 1981, pp. 162-169
C6D S.R. Ahmed: Influence of Base Slant on Wake Structure and

Drag of Road Vehicles.
ASME Journal of Fluids Engineering, Vol. 105, 1983, pp. 429-434
2.488
S. R. AHMED. ET AL.
C7D W.-H. Hucho: The Aerodynamic Drag of Cars. Current Under-

standing, Unresolyed Problems and Future Prospects.
Symposium Proceedings, ref. MU, pp. 7 - 40
7 . Nomenc latur e
b model width ( = 389 mm)
cß* vertical base pressure drag coefficient, based on

F and qTO (Fig. 4)
c =(P -P ) /q total pressure coefficient

g T 00 00
cR* forebody pressure drag coefficient, based on F and

<im (Fig. 4)
cp=(P-Pj qœ static pressure coefficient

c * friction
K
drag coefficient
oo
based on Fand q
Cg* slant surface pressure drag coefficient, based

on F and q (Fig. 4)
oo
cw=W/ (qœF) drag coefficient

F projected frontal area of model
h, h* model height ( = 288 mm) and h

1 model length ( = 1044 mm)
lg base slant length ( = 222 mm) , Fig. 1

P, Pa local and free stream static pressure
Pip local total pressure

2
qoo= 2 voo free stream dynamic pressure
VXA ' vya' VZA velocitY components in XA, and ZA directions,

(Fig. 3)
VXZ ' VYZ resultant of V^, VZA or VyA, VZA velocity

components
V free stream velocity

00
W drag force
X- » Y„ . Z„ cartesian coordinates defined in Fig. 6

A » A' . A
a.,ß flow angles, defined in Fig. 3

p density
base slant angle

f
Fig. 1 Wind tunnel model
Fig. 2 Experimental set-up in wind tunnel
Fig. 3 Probe head details and nomenclature
Fig. 4 Variation of Drag with base slant angle
2,492 S. R. AHMED, ET AL.
Fig. 5 Static pressure isobars on rear end surface
Fig. 6 Horseshoe vortex system in wake (schematic)
g* 7 Velocity distribution in wake central

plane ( <f = 5U and 25°)
Fig. 8 Cross flow veloc
Fig. 9 Cross flow velocity distribution at three downstream

stations in the wake (y=25°)
Fig. 10 Schematic representation of high

drag flow (<^ = 30 )
Fig. 11 Flow pattern on rear end slant surface
TIME-AVERAGED GROUND VEHICLE
Fig. 12 Total
High drag an
2.500
Fig. 13 Cross flow velocity distribution in far wake.

High drag and low drag flow ( <f> = 30 )
Fig. 14 Flow visualisation through smoke injection

in wake central plane (y>= 5 )
Fig. 15 (a,b,c,d) cross flow velocity distribution for
different base slant angles y, and (e) drag varia-

tion with y
Fig. 16 Cross flow velocity distribution and isobars in
wake of low drag configuration (y= 12.5°)

Ahmed 1984

Uploaded by

Copyright:

Available Formats

Ahmed 1984

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ahmed 1984

Uploaded by

Copyright:

Available Formats

Some Salient Features of the Time -Averaged Ground Vehicle Wake

Author(s): S. R. Ahmed, G. Ramm and G. Faltin

Some Salient Features

S. R. Ahmed and G. Ramm

Wake flow exhibits a triple deck system of horseshoe vortices.

Ground vehicles can be termed as bluff bodies moving in close

In a time averaged sense, the regions of flow separation exhibit

A qualitative understanding of the flow phenomena in bluff body

Based on extensive experimental results, this paper attempts to

Selection of the configuration used in this study was governed by

The wind tunnel model, Fig. 1 , with an overall length of 1.044 m

used a similar bluff body to investigate the effect of base slant

Wind tunnel and test set-up

The tests were conducted in the DFVLR subsonic wind tunnels at

Braunschweig (pressure measurements, flow visualization) and

The model was fixed on cylindrical stilts 50 mm above a ground

All tests, except the flow visualization, were performed at a wind

A ten hole directional probe, Fig. 3, which is a further develop-

Mean value of the pressure sensed by the four orifices on the

orifice are functions of local static and total pressure. Cali-

The tenth orifice, situated at the probe tip rea

Pressure and force measurements

Pressure and force measurement results reported here are restricted

3. Experimental Results s* R* AHMED' ET

Drag and pressure measurements

Variation of drag with base slant angle , for a body of the

Contributions to pressure drag from front part c

Numerical values of the drag contributions are given in Table 1

Base slant angle cT7 c * c * c*

Results of Fig. 4 and Table 1 clearly demonstrate the small con-

configuration studied, its v

Total Pres- Vertical Base Slant Surface F0^0"

(Low 0.209 0.615 0.48 0.39 0.426 0.091 0.804 0.196

With y changing from 5° to 30° (see Table 2) , the vertical

thereby from 5.4 % to 66.4 %. These figures emphasize that the

Also shown in Table 2 are the relative magnitudes of the total

surface. Presence of vortices at side edges of the slant surface

The usual assumption of a constant base pressure is not confirmed

Even though the wake flow of a bluff body is basically unsteady,

illustrated in Fig . 6 . This schematic sketch of the flow phenomena

As the flow over the slant surface is influenced by the vortex C

Quantitative data to support the flow module described above is

and 25°, the distribut

Clearly visible are the recirculatory flows A and B of Fig. 6.

A plot of the velocity vector yV in a transverse plane close to

The merging process of the upper horseshoe vortex A (see Fig. 6)

half of the symmetric transverse planes at xÄ/l =-0.077, -0.19

Formation of the side vortex is clearly visible in Fig. 9 a; it is

Note the difference in scale of the velocity vectors plotted.

region is curved upwards in direction of the core of side edge

vortex. Further downstream, atxA/l = -0.19, the separation

Fig. 10 illustrates schematically the features analysed for the

The presence of this separation region lowers the level of pressure

Actually, the flow separation, in the region denoted by E in Fig. 10,

Inspite of the absence of the side edge vortices, which as dis-

situations, as shown in Fig. 13.

Relevance of the flow phenomena described for the wake śtructu.re

model, having same overall dimensions. With the help of interchange-

Fig. 14 shows photographic evidence, taken from C63, of the two

Drag results obtained for the model in C6H, shown in Fig . 1 5 e