Development of Passive Control Techniques for Flow-Induced Noise

Reduction in Vehicle Cavities

Omar Afifi, Mahmoud Shaaban, Ahmed Omer, Nadim Arafa, Atef Mohany1

AeroAcoustics and Noise Control Laboratory

Faculty of Engineering and Applied Science
University of Ontario Institute of Technology, Oshawa, ON, Canada

Abstract
Noise, Vibration & Harshness (NVH) is becoming more significant in today’s vehicle
manufacturing industry. It is part of the perceived product quality and a key aspect for gaining
market advantage. The AeroAcoustics and Noise Control research group at UOIT has been
actively involved in many projects related to NVH such as noise reduction from automotive
HVAC system, sound characterisation and improvement of cabin pressure release vents, and
control of flow-induced noise in radiators and cavities. At certain speeds, flow over cavities in
vehicles can be a potential source of acute noise when coupled with one of the acoustic
natural frequencies of the cavity, which negatively impacts the perceived quality of the
vehicle. This paper presents a summary of some innovative techniques that were developed at
UOIT to suppress flow-induced noise in cavities. The techniques depend on manipulating the
behavior of the flow shear layer over the cavity by using high frequency vortex generator,
upstream flow spoilers, and surface mounted blocks. The effectiveness of each technique is
evaluated experimentally and a summary of the results is presented in the paper.

Keywords: NVH, Flow-Induced Noise, Passive Control, Cavities, Shear Layer, Sound
Pressure, Vortex Generators, Spoilers.

1 Introduction
Noise, Vibration & Harshness (NVH) is becoming more significant in today’s vehicle
manufacturing industry. It is part of the perceived product quality and a key aspect for gaining
market advantage. For example, the customers’ feeling of their vehicle’s doors in terms of the
effort it takes to close the doors and the sound they produce are important features that could
negatively impact the perceived quality of the vehicle. Likewise, acoustic noise resulting from
automobile components can cause great discomfort and irritation not only to passengers inside
the vehicle but also to pedestrians outside. There are different potential sources of noise in
vehicles such as noise from engine exhaust system, noise from fan and cooling system, noise
from tire’s interaction with road surface, and powertrain noise. The sole purpose of NVH is to
characterize the noise and/or vibration sources to be able to control or eliminate them for the
ultimate comfort of the passengers. In order to obtain market leadership in today’s
competitive automotive market, companies must not only develop high quality products but
also they must develop products with less negative impacts on the users or environment.
Controlling or eliminating the noise generated in vehicles is thus a very crucial matter. The
acceptance criteria for any vehicle in terms of user’s comfort is a proportional to the amount
of noise and vibration the user experiences inside the vehicle, and the amount of annoyance in
any vehicle can be directly related to the levels of sound energy and structural excitation
within that vehicle(1).

1
Email: atef.mohany@uoit.ca
In addition to the above mentioned noise and/or vibration sources, a significant source of
extensive noise generated inside vehicles is of an aeroacoustic nature due to the external air
flow over gaps in the vehicle’s structure at high cruising speeds. For example, external flow
over mirror and/or door gaps of a vehicle can produce whistling noise which at some point
can be substantially loud and irritating to the vehicles’ occupants. This noise is a result of a
coupling between one of the acoustic modes of the gap (i.e. cavity) and one of the
hydrodynamic mode of the shear layer caused by the flow over the gap (i.e. cavity) in a
process known as flow-excited acoustic resonance (2-4). Noise related to air flow over exterior
cavities in vehicle structure have been observed and investigated in the literature (1, 5-6).

When flow passes over a cavity, a boundary layer separation occurs and results in the
formation of a shear layer over the cavity mouth, as shown in Figure 1. The shear layer
formation creates vortices that are carried with the flow to impinge on the downstream edge.
The impingement of these vortices on the downstream edge generates pressure perturbations
at the downstream edge of the cavity that travel upstream and enhance the shear layer
separation at the upstream edge and act as a feedback cycle of oscillations. Moreover, a sound
generation occurs at the downstream edge caused by the vortices impingement. The sound
wave coupling with the flow oscillations creates the flow-excited acoustic resonance. The
geometry of the upstream and downstream edges plays a significant role as the upstream edge
controls the boundary layer separation and the vortices formation in the cavity, while the
downstream edge controls the vortices impingement, the feedback oscillations, and the sound
generation (7-8).

Figure 1: Feed-back mechanism for cavity noise

The oscillating modes of the shear layer occur at a dimensionless frequency of St = f l / U ≈

0.5 n. Where St is the Strouhal number, f is the frequency of oscillations, l is the cavity length,
U is the upstream flow velocity, and n is the mode number. The oscillations in the boundary
layer can couple with the acoustic mode of the surroundings and result in an acoustic
resonance of high sound pressure levels (SPL). Several techniques have been developed to
suppress the undesirable effects associated with the flow-excited acoustic resonance and they
can be categorized into active and passive techniques. The active techniques involve control
loops that apply external energy to the system (8-9), while the passive techniques are achieved
by modifying the geometries in a way that changes the flow characteristics (10-12). The main
objective of this paper to summarize some of the techniques that were developed at UOIT for
passive noise control in cavities.

2 Experimental setup
The experimental test section is a wind tunnel consisting of four main components, the bell-
mouth, test section, diffuser and blower, as shown in Figure 2. The blower is driven by a
75HP motor that can achieve a maximum flow velocity of 160 m/s in the test section, which is
sufficiently high to self-excite acoustic resonance in the cavity. The diffuser connects the
blower and the test section together, with a flexible connection to isolate the vibration from
the blower. The test section is made entirely of reconfigurable interlocking blocks of acrylic.
The test section can be reconfigured easily, thus allowing multiple different configurations to
be tested. The dimensions of the test section are 254mm height x 127mm width, and the
stand-alone cavity is 127mm x 127mm x 127 mm (height x width x depth). The cavity can
also be reconfigured to have different upstream edges.

Figure 2: (a) Schematic of the experimental setup (b) CAD of the test section with attachable
cavity.
In order to quantify the level of noise produced by the setup at different configurations and
evaluate the performance of each noise control technique, the sound pressure level (SPL) is
measured using a ¼ inch pressure microphone. The microphone is flush-mounted at the base
of the cavity floor which is the location of pressure antinodes for the cavity acoustic modes.
The flow velocity is measured by a pitot tube. The variable speed motor enables the control of
speed up to 150 m/s for the current setup. A LabVIEW program is used for the data
acquisition and analysis with a sampling rate of 10 kHz and each signal is averaged 100 times
which correspond to 100 seconds in real time.

3 Effectiveness of the Control techniques

The results obtained from the different passive control techniques are presented in the
following sections. Each method is described and evaluated based on the effect of its
attachment and configuration on the sound pressure level of the flow-induced noise generated
in the cavity.

3.1 Upstream edge manipulation

In this case study, the effect of several upstream edge geometries on the acoustic resonance
excitation is reported. Figure 3 shows some of the different configurations that were
considered; including round edges, chamfered edges, vortex generators and spoilers with
different sizes and configurations. Every configuration was examined in the two different
length to depth cavity ratios, L/D = 1 and 1.67, and with slight variation in their own
parameters as spacing, height, radius and angle. The parameters of the configurations were
changed to examine their effect as well on the acoustic pressure. The variations for every
configuration is summarized in Table 1.

As shown in Figures 4, the experiments determined that, compared to the base case (sharp
edge); the round and chamfered edges, with all their variations, were found to delay the
acoustic excitation occurrence to higher velocities with a magnification in the acoustic
pressure. The magnitude of the amplification and the delay in the onset of acoustic resonance
are affected by the radius in the case of the round edges and by the angle in the case of the
chamfered edges. The delay of the onset of acoustic resonance can be useful in many
industrial applications as a method of shifting the resonance out of the operation range.
However, this should be applied carefully since the acoustic pressure produced with the other
round edges and chamfered edges when resonance is materialized is much higher than the
original cavity with sharp edges. As shown in Figure 5, the curved and delta spoilers were
both found to be able to significantly suppress the acoustic pressure to values as low as 250 Pa
and 1250 Pa as compared to approximately 2100Pa recorded for the base case. The ability of
the curved and delta spoilers to suppress the acoustic resonance excitation can be attributed to
these velocity fluctuations that result in preventing the formation of the main vortices
developed by the boundary layer separation, more details cab ne found in Omer et al.(10).
 (I)

( II )

Figure 1 : (I) Tested upstream cavity configurations (A): rectangular sharp edge, (B): round
edge, (C): chamfered edge. (II) Different spoiler configurations (A): delta spoiler,
(B): curved spoiler, (C): straight spoilers with 4mm spacing between teeth, (D):
straight spoilers with 16mm spacing.

Figure 4: Acoustic pressure of round and chamfered upstream edges compared to the base
case.
 Figure 5: Acoustic pressure of curved and Delta spoilers compared to the base case.

Table 1 : Characteristic dimension of edges and spoilers

Geometry (edge/spoiler) Characteristic dimension Variation

Round Edge radius (r) 12.7 and 25.4 mm

Chamfered edge angle (Ѳ) 107, 120, 135 and 150

degrees

Delta spoiler height (h) of the teeth 5.9, 8.2, 12 and 16 mm

Curved spoiler height (h) of the teeth 5.9, 8.2, 12 and 16 mm

Straight spoiler (4mm spacing) height (h) of the teeth 5.9 and 8.2 mm

Straight spoiler (16mm spacing) height (h) of the teeth 5.9 and 8.2 mm

3.2 High frequency vortex generator

In this study case, the effect of placing a control cylinder, or high frequency vortex generator
on the acoustic resonance is presented. The exact location and diameter of the control cylinder
is also investigated. The experiments were performed with the same experimental setup
mentioned earlier in section 3.1. Two cavities were used one with length to depth ratio of 1.0
and the other with length to depth ratio of 1.67. The control cylinder was placed in 16
different locations upstream of the flow. Figure 6 shows a side view of the test section and the
different locations that were investigated. Three different cylinder diameters (3.81, 4.57 and
6.35mm) were also investigated to test the diameter effect on the performance of the cylinder.
Figure 6 : Side view of the test section and the different locations of the control cylinder.

Figure 7: Effect of diameter on acoustic pressure.

As shown in figure 7, the experiments indicate that the passive technique of using a control
cylinder, can be effective in suppressing acoustic resonance. Placing the cylinder near the
upstream edge of a cavity can be very effective. However, it should be tuned carefully. The
best location and diameter for the control cylinder was found to be at the location (-25.4, 12.7)
with a diameter of 6.35 mm. This location and diameter configuration is able to keep the
acoustic pressure below 140 Pa, as opposed to 2000 Pa in the base case (without using the
control cylinder) and around 1600 Pa and 1000 Pa for the control cylinders with diameters 3.8
mm and 4.57 mm respectively and at the same location. The introduction of the high
frequency vortex generator upstream of the cavity disturbs the shear layer and decreases its
correlation length, which in turns, weakens the acoustic resonance excitation. More details
about this technique can be found in Omer and Mohany(11), and Omer(12).

3.3 Attachment of blocks upstream.

Surface mounted blocks is another passive method for controlling acoustic resonance, placing
a surface mounted block in a flow causes complex patterns of vorticity downstream. The
effect of placing square surface mounted blocks, with various dimensions (h and w) and at
different distances upstream (d), on the intensity of the acoustic resonance in shallow
rectangular cavities is investigated. The experiments were performed in the setup shown in
Figure 8, with a cavity of length to width ratio of 1 (square cavity). The response of the
system was reported using the dimensionless pressure parameter (P*) and the flow velocity
was indicated in a reduced form (Ur). Figure 9 below shows the response of the cavity with no
blocks (base case), n = 1, 2, and 3 corresponds to the first, second and third shear layer mode
respectively.

Figure 8: Schematic of test section and blocks parameters.

The blocks used in the experiment had a height of h = 0.0191 m and six different width to
height ratios (w/h = 1, 2, 3, 4, 5, and 6.66). Three different locations for the blocks were also
investigated at d/h = 0, 3, and 6.The locations were selected specifically to investigate the
effect of the reattachment position of the lateral flow around the block sides. The acoustic
pressure was recorded using the flush-mounted microphone at the center-base of the cavity
and velocity of up to 160m/s was reached. The cavity response without a block attachment
was used as a base case for all other cases.

The experiment showed that blocks of width to height ratios between 2 and 4, causes
significant acoustic pressure attenuation of up to 30 dB. It was also reported that the optimum
location for the blocks is at 3h (or three times the height of the block) upstream the cavity.
Wide blocks that nearly filled the entire cross section of the wind-tunnel were reported to
affect the sound pressure levels negatively, resulting in an increasing noise levels. More
details about this technique can be found in Shaaban and Mohany(13, 14)

Figure 9: (a) Response of cavity with no blocks and (b) Noise reduction achieved at different
locations.

4 Conclusion
Flow–excited acoustic resonance is a noise source of high sound pressure levels that may
occur in some automobile components. Edge manipulation, vortex generators, and surface
mounted blocks are passive noise control methods that can be effectively used in automotive
industry to reduce noise generated in vehicles’ cavities and enhance passenger experience. In
this paper, the effectiveness of each of these methods in suppressing the flow-excited acoustic
resonance is presented.

For the edge geometry manipulation, it was found that the round and chamfered edges delay
the occurrence of acoustic resonance to higher velocities with a magnification in the acoustic
pressure. The curved and delta spoilers were both found to be able to significantly suppress
the acoustic resonance excitation. As For the vortex generator or the control cylinder it was
found that this technique can be effective in suppressing acoustic resonance and the maximum
noise reduction is obtained when the control cylinder was placed at the location (-25.4, 12.7)
with a diameter of 6.35 mm, which was able to keep the acoustic pressure below 140 Pa.
Lastly, in the third case study, it was reported that blocks of width to height ratio (w/h)
between 2 and 4, cause significant acoustic pressure attenuation of up to 30dB, and that the
optimum location for the blocks is at 3h (or three times the height of the block) upstream of
the cavity.

Passive control techniques such as the ones discussed in the paper can offer significant
attenuation for applications with relatively high frequencies noise. For low frequencies, active
noise control techniques become more effective and easier to adapt. Active noise control
techniques such as flow blowing or sucking are very promising for automotive noise
applications with low frequencies. The effectiveness of this technique in addition to other
active noise control techniques are currently investigated at the AeroAcoustics and Noise
Control Laboratory at UOIT.
References
(1) Z. Yang, Z. Gu, J. Tu, G. Dong, Y. Wang, Numerical analysis and passive control of a car side
buffeting noise based on Scale-Adaptive Simulation, In Journal of Applied Acoustics, 79(1), pp 23-
34, 2014.

(2) A. Mohany and S. Ziada, Flow-excited acoustic resonance of two tandem cylinders in cross-flow, In
Journal of Fluids and Structures, 21(1), pp 103 – 119, 2005.

(3) A. Mohany and S. Ziada, Measurements of the dynamic lift force acting on a circular cylinder in
cross-flow and exposed to acoustic resonance, In Journal of Fluids and Structures, 27 (8), pp 1149 –
1164, 2011.

(4) A. Mohany, D. Arthurs, M. Bolduc, M. Hassan, S. Ziada, Numerical and experimental investigation
of flow-acoustics resonance of side-by-side cylinders in a duct, In Journal of Fluids and Structures,
48, pp 316 – 331, 2014.

(5) George, Albert R. Automobile aerodynamic noise. No. 900315. SAE Technical Paper, 1990.

(6) H.Kook, L.Mongeaus, S.I Zorea and D. V. Brown, Analysis of the interior pressure oscillations
induced by flow over vehicle openings, In Noise control Engineering Journal, 45(1), pp. 222-234,
1997

(7) Rockwell, D., and E. Naudascher. "Self-sustained oscillations of impinging free shear
layers." Annual Review of Fluid Mechanics 11.1 (1979): 67-94.

(8) L. Cattafesta and D. Williams, Review of Active Control of Flow-Induced Cavity Resonance, In 33rd
AIAA Fluid Dynamics Conference, pp 1–20, 2003.

(9) S. Ziada, H. Ng, C.E. Blake, Flow excited resonance of a confined shallow cavity in low Mach
number flow and its control. In Journal of Fluids and Structures, 18(1), pp 79–92, 2003.

(10) A. Omer, N. Arafa, A. Mohany, M. Hassn, The Effect of Upstream Edge Geometry on the Acoustic
Resonance Excitation in Shallow Rectangular Cavities, ASME 2014 Pressure Vessels and Piping
Conference. Anaheim, California, July 2014.

(11) A. Omer A. and A. Mohany, The effect of high frequency vortex generator on the acoustic resonance
excitation in shallow rectangular cavities, In Journal of Canadian Acoustics, 42(3), 2014.

(12) A. Omer. Passive Methods for suppressing acoustic resonance excitation in shallow rectangular
cavities, MSc. Thesis, University of Ontario Institute of Technology, Canada, 2014.

(13) M.Shaaban and A.Mohany, Control of acoustic resonance in shallow rectangular cavities using
surface mounted blocks, In Journal of Canadian Acoustics, 42(3), 2014.

(14) M. Shaaban and A. Mohany, Passive Control of Flow-Excited Acoustic Resonance in Rectangular
Cavities Using Upstream Mounted Blocks, In Journal of Experiments in Fluids, in press, 2015.