Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

SAE TECHNICAL
PAPER SERIES 2008-01-0098

Correlation of a CAE Hood Deflection

Prediction Method
Thomas N. Ramsay and Ann Fredelake
Honda R&D Americas, Inc.

Kimberly Stevens
The Ohio State University

Reprinted From: Vehicle Aerodynamics, 2008

(SP-2151)

2008 World Congress

Detroit, Michigan
April 14-17, 2008

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-0790 Web: www.sae.org
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

By mandate of the Engineering Meetings Board, this paper has been approved for SAE publication upon
completion of a peer review process by a minimum of three (3) industry experts under the supervision of
the session organizer.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of SAE.

For permission and licensing requests contact:

SAE Permissions
400 Commonwealth Drive
Warrendale, PA 15096-0001-USA
Email: [email protected]
Tel: 724-772-4028
Fax: 724-776-3036

For multiple print copies contact:

SAE Customer Service

Tel: 877-606-7323 (inside USA and Canada)
Tel: 724-776-4970 (outside USA)
Fax: 724-776-0790
Email: [email protected]

ISSN 0148-7191
Copyright © 2008 SAE International
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
The author is solely responsible for the content of the paper. A process is available by which discussions
will be printed with the paper if it is published in SAE Transactions.

Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

2008-01-0098

Correlation of a CAE Hood Deflection Prediction Method

Thomas N. Ramsay and Ann Fredelake
Honda R&D Americas, Inc.

Kimberly Stevens
The Ohio State University

Copyright © 2008 SAE International

ABSTRACT hood structure only FEA model to determine the wind

induced deflections at a specific point on the hood.
As we continue to create ever-lighter road vehicles, the While the testing was conducted at yaw angles of 0 and -
challenge of balancing weight reduction and structural 10 degrees, due to time constraints and relevance to the
performance also continues. One of the key parts this actual in-house hood lift requirements, only the 0-degree
occurs on is the hood, where lighter materials (e.g. yaw condition was compared in this analysis.
aluminum) have been used. However, the aerodynamic
loads, such as hood lift, are essentially unchanged and Additionally, as an outcome of this correlation study, an
are driven by the front fascia and front grille size and investigation into the measurement apparatus itself was
styling shape. This paper outlines a combination undertaken and is discussed at the end of the paper.
CFD/FEA prediction method for hood deflection
performance at high speeds, by using the surface EXPERIMENTAL TESTING
pressures as boundary conditions for a FEA linear static
deflection analysis. Additionally, custom post-processing In order to correlate the proposed CFD/FEA method, a
methods were developed to enhance flow analysis and full-scale vehicle was tested in a wind tunnel.
understanding. This enabled the modification of existing
test methods to further improve accuracy to real world WINDTUNNEL - Testing was completed at a full-scale
conditions. The application of these analytical methods wind tunnel. This wind tunnel combined many appealing
and their correlation with experimental results are attributes including very low background noise for
discussed in this paper. Finally, this correlation study acoustic testing, climatic capabilities, and an open-jet
initiated a redesign of the hood lift measurement test section (1).
apparatus, which is also discussed.
EXPERIMENTAL SETUP - Lollipop pressure taps were
INTRODUCTION chosen to capture the pressure distribution across the
hood surface due to the relatively flat nature of the hood,
With the increase in gasoline prices over the past few as seen in Figure 1.
years, the interest in creating more fuel efficient vehicles
has risen dramatically, whether by applying advanced
engine technology, reduced vehicle weight, or improved
aerodynamics. In the specific case of reducing the
weight of the hood, which is a non-structural part, by
either redesigning the underlying structure or changing
the material, the design must still maintain in-house
aerodynamic deflection requirements while driving on the
road.

In an effort to reduce the amount of experimental testing

- and potential structure changes - required late in a
vehicle's development, a combined CFD/FEA method is
proposed and discussed. The proposed method uses
only the exterior styling shape of a vehicle, which is Figure 1. Side View of Front Clip
available much earlier in the development cycle than the
engine room and underbody, to obtain the aerodynamic In preparation for the accurate and efficient placement of
loads on the hood. These loads are then applied to a the pressure taps during the wind tunnel test, a grid was
mapped on the hood. Due to the curvature of the front
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

region of the hood, this grid was more finely spaced for After the first set of runs, the pressure taps were
the first 250mm of the depth. Spacing of the lollipops removed from the hood, and the tubing and taps were
was set every 50mm in width and depth. After the first moved downstream to the next row of grid points (Figure
250mm, in the flatter portion of the hood, the grid 4). After the first couple iterations, repeatability of the
spacing was relaxed in depth to every 100mm (Figure 2). pressure data was verified by moving the pressure taps
A CMM machine was used to accurately record in space to a previously measured row. Pressure was again
where the lollipops were set. Later this data was used in recorded and compared to the previously recorded data.
the CFD model to match the pressure records in the
CFD model to the identical location on the test vehicle.
All in all there were 16 rows of 25 pressure taps.
However, because of the geometry of hood, the last row
(# 16) had the outer three pressure taps in the same
position as the previous row. This can be seen in the
upper left hand corner of Figure 2.

Figure 4. Pressure Tap Placement during Testing

Lollipops
EXPERIMENTAL RESULTS - The raw pressure data
was converted to pressure coefficients. Figure 5 shows
the pressure coefficient map where the hood edges/gaps
Locations were not taped, which represents on-road conditions.
The pressure coefficients were calculated using the
average freestream velocity and density of the tests and
the reference pressure, which was taken from above the
Figure 2. Pressure Tap (Lollipop) Locations on Top of inlet of the wind tunnel (e.g. in relatively still air).
the Hood

Additionally, the underhood pressure taps were placed

on the outer edges of the hood, where the most
deflection occurred (Figure 3) and only moved once to
determine any front/rear effect on the engine room
pressures.

Figure 5. Pressure Coefficient Map for the Untaped

Condition (testing, blue=low, red=high)

Figure 6 shows the pressure coefficient map where the

Figure 3. Pressure Tap Locations for Underhood
Pressures hood edges/gaps were taped down, which better
represents the modeling conditions.
EXPERIMENTAL TESTING - A prototype vehicle was
placed in the test section with the instrumented hood
applied. The pressure taps were placed on the most
upstream surface of the hood, with the tubing
downstream. Pressure measurements were taken at
160 kph for a 0 degree yaw and -10 degree yaw
condition.

Since the primary purpose of the study was to correlate

the pressures on the hood to the pressures in CFD
model, the hood was tested in two configurations: taped
and untaped. The untaped configuration represented an
on-road vehicle. The taped configuration, where the
surrounding body panels were taped to the hood to Figure 6. Pressure Coefficient Map for the Taped
prevent lift which would not be captured by the proposed Condition (testing, blue=low, red=high)
CFD/FEA method, as well as to replicate the no-gap
nature of a simple exterior-only CFD model.
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

Note that in both Figure 5 and Figure 6, although there

seems to be some calibration offsets applied to taps 12
and 21 (which were noted during testing) the data
appeared as expected: lower pressures - and
corresponding higher lift - near the leading edge of the (CL) 13

-Cp
(UH-L) 26

hood. (UH-R) 27

Whether the taped and untaped conditions closely match

each other was a different matter, entirely. Figure 7
shows the centerline pressure coefficients for the taped 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
and untaped conditions. Note that while the trends of the Row Number

taped and untaped pressures are the same, the largest

difference was on the leading edge. This means that an
Figure 8. Pressure Coefficients for Centerline (CL) and
actual, stand-alone, prediction using the proposed Underhood (UH) Tap Locations (testing)
method would most certainly not be completely accurate,
because the discussed CFD/FEA method does not
MODELING METHOD
capture this dynamic change in hood deflection.
In order to judge the usability of the proposed CFD/FEA
method, a full-scale CFD model was built and solved.

MODEL CONSTRUCTION - The CFD models were

created from styling geometry, then surface and volume
TAPED
meshed. Because this model was being used for direct
-Cp

UNTAPED

comparison to experimental results, the vehicle was

created within a full-scale replica of the wind tunnel
(Figure 9).

0 100 200 300 400 500 600 700 800

Centerline Length

Figure 7. Centerline Pressure Coefficients for Taped and

Untaped Hood Edges/Gaps (testing)

Another important aspect of the testing was to determine

if the pressures in the engine compartment were large
enough to require modeling the underhood flow in
addition to the exterior styling geometry. Figure 8 shows
the pressure coefficients for above hood centerline and Inlet
underhood tap locations. It's reasonably clear from the Velocity
results that, while the engine compartment pressures are
not zero, they can be considered negligible when taking
into account the relatively large hood pressures (near the Figure 9. Virtual Wind Tunnel
leading edge) along with the manpower, computational
cost, and relative inaccuracy early in the development Due to the desired simplicity and ease of application
(because the exterior styling is available much earlier desired for this method, a flat underbody and simplified
than the rest of the vehicle) associated with accurately wheels were applied to the styling surfaces. The vehicle
modeling the engine compartment. This also means that was created with 2.2 million surface elements, with the
the underhood modeling pressure will be uniform and at hood being a mostly quad 5 mm mesh and the rest of
atmospheric pressure. the vehicle (and wind tunnel) being a tria mesh. This
was done because a FEA code is more accurate with a
mostly quad mesh. The entire model contained 17.3
million cells, with a 4 mm prism layer created on the
vehicle itself, another separate prism layer in the wind
tunnel floor and walls, and the rest of the cells being
tetrahedrals.

MODELING SETUP - In order to best match the

experimental results, the numerical wind tunnel was
calibrated just like the actual one: an inlet speed sweep
was conducted and the velocity at a point one meter
above the center of the turntable was recorded. The
inlet velocity and the air properties were set to the
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

calibrated and experimental values, respectively. A

pressure outlet was used to model the resistance
provided by the contents of the engine compartment for
flow through the upper, middle, and lower grille openings.
These pressure values were based on the size of each
opening, the tunnel velocity, and empirical data.

MODELING RESULTS - Since this is primarily intended

as an aerodynamic paper, the CFD portion of this
analysis is discussed and only a brief summary of the
FEA modeling results is included. The CFD model took
7 hours to solve on 24 processors, with convergence
being judged by a drag monitor. Figure 10 shows the
Figure 12. Pressure Coefficient Map (modeling,
static pressure contours on the hood. Note that, as blue=low, red=high)
expected, the leading edge of the hood (bottom edge of
the image) created lower static pressures. The mostly quad hood mesh was then exported and by
use of a custom translator, converted into in-house FEA
units as a hood shell mesh with an accompanying
distributed pressure boundary condition. This was then
imported and connected to a linear-static structural FEA
model, Figure 13.

Figure 10. Hood Static Pressure Contours (modeling,

blue=low, red=high)

A series of points were created on the hood that

matched the pressure tap placement of the testing (see Figure 13. Linear-Static FEA Model
Figure 2) and are shown in Figure 11. The bunching of
the pressure locations near the front of the hood is Note that only the attachment points to the body - the two
clearly evident. It is apparent, again, that the last row hinges and the lock mechanism - were constrained in the
contains fewer points. FEA model. Figure 14 shows the resulting hood
deflection map, where the front outer sections of the
hood exhibited the most deflection.

Figure 11. Numerical Pressure Point Locations

Figure 12 reconfigures the static pressures of Figure 10

using the pressure point locations of Figure 11 and, thus, Figure 14. Hood Deflection Contours (modeling, red =
shows the pressure coefficient map for the modeling high, blue = low)
results. The pressure coefficients were calculated in the
same manner as for the reduced experimental data, with
the reference pressure being obtained from above the
inlet.
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

COMPARISON AND VALIDATION measured for in-house requirements, which is

determined from the field of view of the driver, and is
In order to determine accuracy of the proposed method, located on the side edge and near the front of the hood.
two methods of comparison were conducted: Figure 17 shows a comparison of the deflection
aerodynamic and structural. The aerodynamic measurements, both from the testing and the modeling.
comparison simply will determine if the CFD pressure The accuracy is 2% and 9% for the left and right sides,
loads on the exterior surface of the hood were correctly respectively. Both the experimental an computational
captured. The structural comparison will determine the results were for a single point on the side edge and small
resulting FEA model captured the resulting deflection distance from the front edge of the hood. This point
trends found in the actual hood deflection testing. coincides with in-house standard testing procedures and
is, not coincidentally, near the farthest point from the
AERODYNAMIC COMPARISON - While the testing and supports (hinges and lock).
modeling images from Figure 6 and Figure 12 show very
good similarity, a direct comparison between the testing Test Model

and numerical hood pressures is compared in Figure 15,

which plots the centerline pressures coefficients for the
testing (taped) and the modeling results. Figure 16 also

Hood Deflection
shows the pressure coefficient curves for the left edge
points. Note that the both sets of curves almost lay on
top of one another. However, it should also be noted
that only near the front portion of the hood are the
pressures coefficients dramatically bigger than the
(measured) underhood pressure coefficients (from
Figure 8). While numerical correlation - based on curve- Left Side Right Side

to-curve comparison for Figure 15 or Figure 16 - would Figure 17. Deflection Measurements for Testing and
be meaningless, the point-to-point errors are in the range Modeling
of 5-8% near (but not on) the leading edge of the hood.
MEASUREMENT APPARATUS DESIGN

While doing this correlation investigation, it became

apparent that the sometimes-used method of measuring
the deflections externally (whether in the wind tunnel or
TEST
on the test track) could be improved upon. Basically, the
-Cp

MODEL

laser measurement apparatus had exhibited some flow-

induced vibrations, which led to noisy measurement
data. While it is outside the original scope of the
investigation, it was decided by the authors to include in
0 100 200 300 400 500 600 700 800 this paper to show how CAE methods can sometimes
Centerline Length
lead to improved experimental methods. It was not
always possible to measure the hood deflection from
Figure 15. Pressure Coefficients on Hood Centerline for within the engine compartment due to obstructions, and
Testing and Modeling Results external measurements were sometimes required.
However, this put the measurement device in the flow
TEST
field which could deflect the apparatus and potentially
MODEL
change the local flow field, which, in turn, would alter the
pressure on the hood. Figure 18 shows the hood lift
measurement apparatus, position, and parts, where the
laser measurement device is in green. The apparatus
-Cp

was judged to be at least 15 years old and note that it

has a wedge-type deflector.

0 100 200 300 400 500 600 700 800

Edge Length

Figure 16. Pressure Coefficients on Left Hood Edge for

Testing and Modeling Results

STRUCTURAL COMPARISON - While the aerodynamic

comparison turned out quite well, in order to complete its
use, a structural comparison should also be conducted.
The point of deflection comparison was the same as Figure 18. Hood Lift Measurement Apparatus
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

Figure 19 shows the negative static pressures (thus the Figure 22 shows the local negative pressures on the
leading edge shows "high" values) on the hood for a hood and clearly the apparatus was affecting the local
CFD solution with and without the measurement flow field. Basically, the target was to get the apparatus
apparatus. While it shows that the apparatus did have to cause the least amount of change to the local
an effect on the hood pressure distribution, the true pressure field on the hood itself.
effect of the wedge-apparatus was probably pretty
localized; however, there is always room for
improvement and the design of a more aerodynamic - No
and, thus, less intrusive - measurement apparatus was Apparatus
undertaken.

With
Apparatus

Figure 22. Local Negative Pressures on the Hood

To that end, the primary goal was to raise the apparatus

up and to streamline the shape. Figure 23 shows the
current wedge and two conceptual designs: one is an
Figure 19. Hood Negative Pressures
airfoil and the other adds a splitter plate. Both are
angled at 10-degrees to better align with the flow field.
While, ultimately, a local pressure distribution was used
to compare the various designs, some near-apparatus
path lines and apparatus forces were also evaluated for
design direction. Figure 20 shows these two methods.
Note that the path lines are actually deflected about 10-
degrees as they passed the area of interest.
Additionally, the hood was remeshed to create a local
rectangular region that could be analyzed separately.
Figure 23. Current and Proposed Apparatus Designs

Figure 24 shows the effect of a simple change in shape,

to an airfoil, on the path lines near the apparatus. This
airfoil shape clearly produced a flow field more like no
apparatus (which replicates the on-road condition).

Figure 20. Localized Pressure and Path Lines for

Apparatus Evaluation

Figure 21 shows these near-apparatus path lines

(rotated 10-degrees) and it was clear that the current
wedge deflector was creating a relatively large blockage,
but also set at a non-optimal angle to the local flow field.

Figure 24. Path Lines Comparison for Apparatus

Concepts

Figure 25 shows the local negative pressures for the

different apparatus designs. Note that while all the
Figure 21. Near Apparatus Path Lines with and without designs clearly affected the local pressures, the airfoil +
the Wedge Apparatus splitter design minimized size of the effected region (e.g.
the leading and trailing sections of the local region best
match the non apparatus target). The correlation
 Downloaded from SAE International by American Univ of Beirut, Monday, July 30, 2018

coefficients, as compared to the no apparatus target, for pressures - for this single vehicle - were correlated within
the designs in Figure 25 are: wedge = 0.811, airfoil = 0.5% for the taped condition and judged OK to use as an
0.880, and airfoil + splitter = 0.889. input to a FEA model. A full underhood pressure model
is judged not necessary due to the relatively low
underhood pressures measured (again, for this vehicle)
None and the relatively high suction pressures near the area of
interest (the front corners of the hood). The method is
validated within 2% and 9% for left/right hood deflections.
The overall modeling method needs more refinement,
Wedge but can be used for styling direction as it impacts hood
(Current) lift. More investigation into the FEA portion of the
method and the static/dynamic (taped/untaped)
conditions is warranted and planned. Nevertheless it
Airfoil bears repeating that this method would be best for A-to-
Only B comparisons and not absolute values for deflection.

Regarding the hood lift measurement apparatus design,

Airfoil + it was clear from the pressure map, path lines, and
Splitter
generated forces, that the conceptual yawed airfoil +
splitter configurations would be the best choice for a new
apparatus design. Future efforts will be to convert this
Figure 25. Local Negative Pressures for Different concept into a working design (e.g. determine its
Apparatus Designs manufacturability), evaluate it and put it into use.
The blockage and angle of attack of the apparatus would It, again, should be noted that this desired method is to
also produce forces on the apparatus itself. In other be used very early in development when full vehicle
words, the apparatus would be deflected backwards and geometry might not be available. With regard to future
outboard, thereby increasing the length from the laser to efforts in the area of hood lift, the authors acknowledge a
the hood due to the apparatus geometry and position. single correlation point does not a method make, and the
investigation discussed herein has initiated a more
Figure 26 shows the computational drag, yaw, and lift comprehensive accounting of the hood lift phenomena,
forces generated by the various apparatus including, but not limited to, different vehicle types, as
configurations. Note that it was both a change in shape well as both experimental and computational methods
and orientation (10 degree yaw) that created the most that allow for better correlation.
dramatic change in the forces on the apparatus. While
not correlated with testing, these forces yielded another
metric to help determine the proper direction for the
apparatus design.
ACKNOWLEDGMENTS

The authors would like to thank Mike Shelly at Honda

R&D Americas, Inc., for setting up and solving the FEA
models.
Wedge @ 0 deg
Force

Airfoil @ 0 deg
Airfoil @ 10 deg REFERENCES
Airfoil+Splitter @ 10 deg

1. Walter, J., Duell, E., Martindale, B., Arnette, S.,

Nagle, P., Gulker, W, Wallis, S, and Williams, J.,
"The Driveability Test Facility Wind Tunnel No. 8",
Drag Yaw Lift SAE 2001-01-025 (2001).

Figure 26. Computational Forces on the Various CONTACT

Apparatus Configurations
Tom Ramsay, Honda R&D Americas, Inc., 21001 State
CONCLUSION Route 739, Raymond, OH 43067, USA

The authors acknowledge that the proposed method to

model the hood deflection is not completely accurate and
it doesn't 100% represent the on-road condition because
it is an external-only CFD model. However, the hood