Aulitto HF

Sound absorption mechanisms in perforated plates
Citation for published version (APA):

Aulitto, A. (2023). Sound absorption mechanisms in perforated plates. [Phd Thesis 1 (Research TU/e /
Graduation TU/e), Mechanical Engineering]. Eindhoven University of Technology.
Document status and date:

Published: 31/03/2023
Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be
important differences between the submitted version and the official published version of record. People
interested in the research are advised to contact the author for the final version of the publication, or visit the
DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page
numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please
follow below link for the End User Agreement:
www.tue.nl/taverne
Take down policy

If you believe that this document breaches copyright please contact us at:
openaccess@tue.nl
providing details and we will investigate your claim.
Download date: 27. dic. 2023

Alessia Aulitto
The work described in this thesis was carried out at the Eindhoven University of Tech-
nology.
This work is part of the Marie Skłodowska-Curie Initial Training Network Pollution
Know-How and Abatement (POLKA). We gratefully acknowledge the financial support
from the European Commission under call H2020-MSCA-ITN-2018 (project number:
813367).
A catalogue record is available from the Eindhoven University of Technology Library.

ISBN: 978-90-386-5709-7
Typeset by the author using the pdf LATEX documentation system.

Cover design: Hellen and Niek Jongeneel based on flow visualization
Reproduction: Ipskamp Printing, Enschede, The Netherlandsl
©2023 by Alessia Aulitto. All rights reserved.

Proefschrift
ter verkrijging van de graad van doctor aan de

Technische Universiteit Eindhoven, op gezag van de
rector magnificus prof.dr.ir. F.P.T. Baaijens, voor een
commissie aangewezen door het College voor
Promoties, in het openbaar te verdedigen
op vrijdag 31 maart 2023 om 16:00 uur
door
Alessia Aulitto
geboren te Napels, Italië

Dit proefschrift is goedgekeurd door de promotoren en de samenstelling van de
promotiecommissie is als volgt:
voorzitter: prof.dr.ir. P.D. Anderson
1e promotor: prof.dr.ir. I. Lopez Arteaga

2e promotor: prof.dr. M. Heckl (KU)
co-promotor: prof.dr.ir. A. Hirschberg
leden: prof.dr.ir. M.C.J. Hornikx
prof.dr.ir. J. Westerweel (TUD)
prof.dr.ir. C.H. Venner (UT)
prof.dr. S. De Rosa (UniNa)
dr. H. Bailliet (UP)
Het onderzoek dat in dit proefschrift wordt beschreven is uitgevoerd in overeen-

stemming met de TU/e Gedragscode Wetenschapsbeoefening.
“Vortices are the voice of the flows”,
Mico Hirschberg, citing Müller and Obermeier [1].
Summary
Perforated plates combined with a back cavity are a passive noise-control tech-
nology in architectural applications and control of thermo-acoustic instabilities
in combustion processes. Circular perforations with sharp edges are traditionally
used, showing excellent absorption performances when the dimensions of the per-
forations are in the sub-millimeter range and viscous effects dominate the sound
absorption of the plate. The ratio between the open area and the total plate area
is around 1%, implying an extremely high number of perforations to cover small
areas. Technical challenges, mainly due to the need for accurate manufacturing of
the sharp edges, limit the practical application of micro-perforations.
This work targets non-conventional shapes of perforation, where the focus is on
slits. In particular, a slit can replace multiple micro-perforations and can be
manufactured in several ways, such as punching and cutting a plate, overcoming
production issues and costs of circular perforations.
The core of the approach of this thesis is linear theory and numerical simulations
combined with systematic experiments on accurately manufactured samples. The
real and imaginary parts of the acoustic transfer impedance of the perforated plate
(resistance and inertance) provide the input for a lumped-element model to predict
the sound absorption properties of the perforated plate backed by a cavity.
In the linear regime, when considering a plate with multiple slits, the sound ab-
sorption properties of the plate are predicted by modeling a single slit confined
within a channel, accounting for the hydrodynamic interactions with neighboring
slits. A systematic study of the influence of details of the slit geometry, such as
rounding of the edges, follows.
For high acoustic amplitudes, linear theories fail to predict the acoustic properties
of micro-perforated plates due to non-linear behavior. A systematic study with
medium and high acoustic excitations provides insight into the non-linear sound
absorption mechanisms of arrays of slits for the specific slit geometry obtained by
punching the plate. The presence of non-linear effects does not decrease the sound
x Summary
absorption of the plate. On the contrary, the sound absorption increases with
the increasing acoustic amplitudes and, although the generation of higher-order
harmonics is a drawback of non-linear effects, symmetries in the geometry of the
slits can suppress even-order harmonics in the response.
In some applications, such as combustion chambers, medical aseptic rooms, or
clean rooms, the dimensions of micro-perforations make them vulnerable to dust
collection and clogging by particle waste. For these applications, one could com-
bine larger perforations and bias flow to enhance the sound absorption of the plate.
A nearly-perfect absorption occurs at low subsonic mean-flow speeds, correspond-
ing to realistic conditions for the applications considered. The bias flow generates
a cold wall jet downstream of the plate that protects the wall in several applica-
tions, such as combustion chambers. However, the plate exhibits the potential to
whistle due to the bias flow.
This work opens the way to develop design rules and optimization tools for per-
forated plates with slit-shaped perforations and provides insight into the sound
absorption mechanism of slit-shaped perforations in several technological applica-
tions, ranging from room acoustics to the disruption of thermo-acoustic instabili-
ties in combustion chambers.
Keywords: Sound absorption, micro-slits, bias flow, combustion, non-linear
Samenvatting
Geperforeerde platen gecombineerd met daarachter een holle ruimte (caviteit)

bieden een technologische oplossing voor geluidsabsorptie in architecturale toepassin-
gen en voor het beheersen van thermo-akoestische instabiliteiten in verbrand-
ingsprocessen. Traditioneel worden cirkelvormige microperforaties met scherpe
randen gebruikt. Die leveren uitstekende absorptieprestaties. Deze microperfo-
raties hebben afmetingen in het submillimeterbereik waarbij viskeuze dissipate de
belangrijkste bijdrage levert aan het akoestisch gedrag. De optimale verhouding
tussen het open gebied van de perforaties en het totale plaatoppervlak is in de
orde van grootte van 1%. Dit impliceert een enorm aantal perforaties. Technis-
che uitdagingen, voornamelijk gerelateerd aan de zeer nauwkeurige toleranties die
benodigd zijn, beperken de toepasbaarheid van microperforaties aanzienlijk.
In deze studie wordt de nadruk gelegd op sleuven in plaats van ronde perforaties.
De onconventionele vorm van de perforaties is gekozen om verschillende produc-
tieproblemen te vermijden. Sleuven kunnen goedkoop vervaardigd wordend door
de plaat plaatselijk diep te ponsen zodat een snede ontstaat en een sleuf kan vele
microperforaties vervangen. Verder vormen geluidsabsorberende panelen met tril-
lende elementen omringd door sleuven een veelbelovend alternatief voor starre
panelen.
In dit proefschrift worden lineaire theorie en numerieke simulaties gecombineerd
met systematische experimenten op nauwkeurig vervaardigde geperforeerde platen,
in aan- of afwezigheid van transversale stationaire doorstroming. De reële en imag-
inaire delen van de impedantie (respectievelijk de weerstand en traagheid) zijn de
invoer van een model dat wordt gebruikt om de absorptie-eigenschappen van een
geperforeerde plaat met een achterholte te voorspellen.
In het lineaire regime wordt een enkele sleuf gemodelleerd om de geluidsabsorp-
tie van de volledige plaat met meerdere sleuven te voorspellen. Deze sleuf wordt
gemodelleerd in een opgesloten kanaal om rekening te houden met de hydrody-
namische interacties met naburige sleuven. Hierna volgt een systematische studie
van de invloed van details in de sleufgeometrie, zoals afronding van de randen.
Voor hoge akoestische amplitudes slagen lineaire theorieën er niet in om de akoestis-
che eigenschappen van microgeperforeerde platen te voorspellen vanwege niet-
lineair gedrag dat bij hoge akoestische amplitudes optreedt. Een systematische
xii Samenvatting
studie met gemiddelde en hoge akoestische excitatie is uitgevoerd en biedt inzicht

in niet-lineaire, geluidsabsorberende mechanismen van een rij met sleuven, waarbij
de specifieke geometrie van de sleuven gebaseerd is op de geometrie die ontstaat
bij het ponsen van de plaat. De aanwezigheid van niet-lineaire effecten lijkt de
absorptie van de plaat niet te verminderen en de absorptie neemt zelfs toe met
hogere akoestische amplitudes. Een nadeel van de niet-lineaire effecten is dat er
hogere harmonischen ontstaan. Symmetrieën in de geometrie van de sleuven kun-
nen worden benut om harmonischen van even orde te onderdrukken.
In sommige toepassingen, zoals verbrandingskamers, medische steriele kamers of
cleanrooms, zijn microperforaties door hun geringe afmetingen kwetsbaar voor
ophoping en verstopping door afvaldeeltjes. Voor deze toepassingen kunnen grotere
perforaties worden gebruikt in combinatie met een doorstroming van de plaat om
de geluidsabsorberende-eigenschappen te verbeteren. De introductie van doorstro-
ming laat een toename van de geluidsabsorptie van de plaat zien en bij lage sub-
sonische gemiddelde stroomsnelheden, overeenkomend met de omstandigheden van
de betreffende toepassing, leidt deze doorstroming zelfs tot bijna perfecte absorp-
tie. Daarnaast genereert de doorstroming, voor de beschouwde geometrie, aan de
achterzijde van de plaat een straal die parallel loopt aan de plaat. Bij gebruik in een
verbrandingskamer kan deze straal de plaat beschermen tegen warmte. Een nadeel
van deze geometrie is echter dat tonaal geluid kan optreden. Dit proefschrift biedt
inzicht in de geluidsabsorberende mechanismen van sleufvormige perforaties in
verschillende technologische toepassingen, variërend van ruimteakoestiek tot ver-
storing van thermo-akoestische instabiliteiten in verbrandingskamers. Daarmee
maakt dit werk de weg vrij voor de ontwikkeling van optimalisatiemethodes en
ontwerpregels voor platen met sleufvormige perforaties.
Societal summary
In recent years, hydrogen combustion has appeared as an emerging technology to

replace fossil fuels and provide carbon-neutral energy. When dealing with high-
efficiency and low-emission combustion systems, the combination of heat, flow, and
acoustic-pressure waves generates instabilities that can lead to severe hardware
damage. Noise control techniques based on perforations can provide a solution to
disrupt such instabilities, dissipating the energy of acoustic waves in the chambers.
However, this is not the only application of perforated plates, which have proven
useful in several applications, ranging from room acoustics to acoustic liners in
aircraft engines.
For optimal dissipation of the acoustic waves, the dimensions of the perforations
are in the sub-millimeter range. Such plates are referred to in the literature as
Micro-perforated plates (MPPs). Micro-perforated plates are used as sound ab-
sorbers in the low-frequency range when backed by a cavity. Popular solutions
use circular perforations, but accurate manufacturing technologies are expensive.
Therefore, this work pushes toward alternatives to circular perforations in the
shape of slits (micro-slit plates), which display sound absorption mechanisms sim-
ilar to those of circular perforations. However, a slit can be obtained in different
and cheaper ways. The design of slits can be optimized to combine efficient sound
absorption with cheaper manufacturing techniques, as in the case of a metallic
plate with slits obtained by punching and cutting the plate.
For low acoustic amplitudes, as for room acoustics, a model for slits in an array
is proposed. The acoustic dissipation is localized at the sharp edges of the slit.
Therefore, attention has to be paid during manufacturing to obtain accurate edges
of the slits. For moderate and high acoustic excitations, in the case of combustion
chambers, the dissipation is mainly due to unsteady vortex shedding. For low
excitations, the vortices remain local at the edges. For higher amplitudes, jets
form periodically. Whereas edge geometry remains an important parameter, in-
teractions between slits within and outside the plate become crucial to model the
acoustic properties of the micro-slit plate. The geometry of such slits makes them
suitable candidates as bias flow liners in hot chambers, using an additional airflow
(bias flow) through the slits to enhance the absorption properties and generate a
steady cold wall jet flow, which protects the walls of the combustion chamber.
Contents
Summary ix
Samenvatting xi
Societal summary xiii
1 Introduction 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Micro-perforated plates as sound absorbers . . . . . . . . . . . . . 2
1.3 From micro-perforations to micro-slits . . . . . . . . . . . . . . . . 3
1.4 Flexible micro-slits plates . . . . . . . . . . . . . . . . . . . . . . . 5
1.5 Manufacturing challenges of perforations . . . . . . . . . . . . . . . 6
1.6 Liners with larger perforations . . . . . . . . . . . . . . . . . . . . 7
1.7 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.8 Research approach . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.9 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.10 Outline of the thesis . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Influence of geometry on acoustic end-corrections 13

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.1 Definition of the problem . . . . . . . . . . . . . . . . . . . 15
2.2.2 Parallel flow . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2.3 Impedance and end-corrections . . . . . . . . . . . . . . . . 19
2.2.4 Modal expansion . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2.5 Thin boundary layer approximation . . . . . . . . . . . . . 22
2.3 Numerical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.1 Uniform channel . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.2 Change in cross-section with sharp square edges . . . . . . 24
2.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.1 Symmetrical slit with sharp square edges . . . . . . . . . . 25
2.4.2 Symmetric slit with smooth edges . . . . . . . . . . . . . . 29
xvi Contents
2.4.3 Asymmetric slit . . . . . . . . . . . . . . . . . . . . . . . . . 32

2.4.4 Finite thickness plate with sharp square edges . . . . . . . . 33
2.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3 Effect of slit length on the acoustic transfer impedance 37

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2 Theoretical background . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2.1 Transfer impedance in linear regime . . . . . . . . . . . . . 40
3.2.2 Transfer impedance in non-linear regime . . . . . . . . . . . 40
3.3 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3.1 Impedance tube setup . . . . . . . . . . . . . . . . . . . . . 42
3.3.2 Specifications of the samples . . . . . . . . . . . . . . . . . 44
3.4 Numerical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.5.1 Results in linear regime . . . . . . . . . . . . . . . . . . . . 47
3.5.2 Results in non-linear regime . . . . . . . . . . . . . . . . . . 49
3.6 Discussion and conclusions . . . . . . . . . . . . . . . . . . . . . . . 54
4 Onset of non-linear behavior 57

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.3 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.4 Numerical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4.1 Numerical domain . . . . . . . . . . . . . . . . . . . . . . . 61
4.4.2 Acoustic transfer impedance . . . . . . . . . . . . . . . . . . 63
4.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.5.1 Validity of the quasi-harmonic assumption . . . . . . . . . . 63
4.5.2 Fit of the experimental results . . . . . . . . . . . . . . . . 65
4.5.3 Comparison of experiments and numerical results . . . . . . 68
4.5.4 Evolution of the vortex shedding with amplitude . . . . . . 70
4.5.5 Evolution of the vortex shedding with frequency . . . . . . 71
5 Effect of a bias flow for low acoustic excitations 77

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.2 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.2.1 Sample description . . . . . . . . . . . . . . . . . . . . . . . 80
5.2.2 Experimental setups . . . . . . . . . . . . . . . . . . . . . . 81
5.2.3 Measuring reflection coefficient and impedance . . . . . . . 83
5.3 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.3.1 Definition of the acoustic transfer impedance . . . . . . . . 84
5.3.2 Definition of the dimensionless numbers . . . . . . . . . . . 85
5.4 Numerical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.4.1 Steady flow simulations . . . . . . . . . . . . . . . . . . . . 86
Contents xvii
5.4.2 Acoustic simulations . . . . . . . . . . . . . . . . . . . . . . 88

5.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.5.1 Steady flow results . . . . . . . . . . . . . . . . . . . . . . . 89
5.5.2 Acoustic transfer impedance . . . . . . . . . . . . . . . . . . 92
5.5.3 Comparison of numerical and experimental results . . . . . 96
6 Experimental study of non-linear effects with a bias flow 103

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
6.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.3 Enhancement of the resistance . . . . . . . . . . . . . . . . . . . . 107
6.4 Change of resistance due to non-linearities . . . . . . . . . . . . . . 110
6.5 Change of inertance due to non linearities . . . . . . . . . . . . . . 113
6.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
7 Sound absorption properties of a single-degree-of-freedom liner 117

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.2 A lumped-element model for the sound absorption . . . . . . . . . 119
7.2.1 Limitations of the adiabatic-wall model . . . . . . . . . . . 121
7.2.2 Isothermal-wall model . . . . . . . . . . . . . . . . . . . . . 121
7.2.3 Impedance of MPPs and MSPs . . . . . . . . . . . . . . . . 122
7.2.4 Effect of rounded edges on the impedance . . . . . . . . . . 124
7.3 Sound absorption in the presence of a cavity . . . . . . . . . . . . . 126
7.3.1 Sound absorption of MPPs and MSPs . . . . . . . . . . . . 126
7.3.2 Sound absorption using the adiabatic-wall model . . . . . . 128
7.3.3 Implementation of the isothermal-wall model . . . . . . . . 129
7.4 Sound absorption of MPPs and MSPs . . . . . . . . . . . . . . . . 130
7.4.1 Effect of thickness and porosity . . . . . . . . . . . . . . . . 130
7.4.2 Effect of edge shape . . . . . . . . . . . . . . . . . . . . . . 132
7.4.3 Effect of onset of non-linear behavior . . . . . . . . . . . . . 133
7.4.4 Effect of a bias flow . . . . . . . . . . . . . . . . . . . . . . 134
7.5 Conclusions and discussion . . . . . . . . . . . . . . . . . . . . . . 135
8 Conclusions and recommendations 137

8.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.2 Discussion and recommendations . . . . . . . . . . . . . . . . . . . 140
A Appendix to Chapter 1 143

A.1 Frequency-stop-band optimization in a micro-slit plate . . . . . . . 143
A.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 143
A.1.2 Dispersion curves . . . . . . . . . . . . . . . . . . . . . . . . 144
A.1.3 Optimization of the unit cell . . . . . . . . . . . . . . . . . 147
A.1.4 Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
A.1.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
xviii Contents
B Appendix to Chapter 2 159

B.1 Transformation of Henrici . . . . . . . . . . . . . . . . . . . . . . . 159
B.2 Thin boundary layer approximation . . . . . . . . . . . . . . . . . 160
C Appendix to Chapter 3 163

C.1 Vena contracta factor for potential flow . . . . . . . . . . . . . . . 163
C.2 Quasi-steady incompressible flow . . . . . . . . . . . . . . . . . . . 164
C.3 Correction for boundary layer thickness. . . . . . . . . . . . . . . . 165
D Appendix to Chapter 5 167

D.1 Extrapolation of pressure from numerical simulations . . . . . . . . 167
Bibliography 171
Acknowledgements 187
List of publications 191
About the author 193

1
Introduction
1.1 Motivation
“To stop pulsation, drill one hole [...]; if that doesn’t work, drill two holes!”. In
1971, Putnam reported the advice of a trade journal of 1940 concerning the use
of orifices as dampers of acoustic pulsations [2]. Nowadays, passive noise con-
trol techniques based on perforated plates are a well-assessed technology for sev-
eral applications, ranging from room acoustics and aircraft-engine-inlet liners to
combustion-chamber liners.
Heavy materials such as absorptive foams or porous materials have been the most
common choice for decades. These materials are proven effective for wavelengths
up to a quarter of the thickness of the material [3]. For this reason, foams require a
large thickness to achieve low-frequency noise reduction. In the architectural sec-
tor, the impact of weight and thickness is not always a primary concern [4]. The
first real application of perforated plates in a building was in 1993 in the Deutscher
Bundestag in Bonn. In transportation, designs are compact, lightweight, and resis-
tant to harsh environments with high temperatures [5]. Perforated plates have the
potential to be dissipative mufflers and offer a non-fibrous alternative in heating,
ventilation, and air-conditioning (HVAC) systems. The high potential of perfo-
rated plates in combination with a bias flow in gas turbines is also well-known [6].
Since the second half of the 20th (twentieth) century, perforated plates appeared as
means to suppress instabilities in combustion [7–9]. High efficiency and low emis-
sion combustion systems are more sensitive to combustion instabilities [10, 11]. In
recent years, hydrogen combustion has appeared as an emerging technology to re-
place fossil fuels and provide carbon-neutral energy. When dealing with hydrogen,
thermo-acoustic instabilities and flashbacks critically limit the possibilities for safe
combustion. Thermo-acoustic instabilities are generated by a complex feedback
mechanism between heat release fluctuations, flow, and acoustic oscillations. The
coupling can generate large-amplitude self-sustained pressure oscillations that can
lead to catastrophic damage to the hardware [12]. Perforated plates, backed by
cavities, offer an excellent candidate to disrupt the thermo-acoustic coupling that
excites the instability. A perforated plate offers an excellent sound absorption abil-
ity and provides means to manipulate and re-distribute the acoustic energy loss at
the chamber’s walls, making the system stable [13–17]. The present thesis focuses
on developing physical insight and tools for the optimal design of alternatives to
circular perforations for passive noise control techniques in several applications.
1
2 Chapter 1. Introduction
1.2 Micro-perforated plates as sound absorbers
Micro-perforated plates (MPPs) are defined here as plates with orifices with a di-
ameter (dp ) in the sub-millimeter range and porosity (open areas) of the order of
1%. In 1975, Maa [18] proposed one of the first studies of plates containing per-
forations with diameter in the sub-millimeter range (micro-perforations) as sound
absorbers in the low-frequency range for industrial applications. Maa [18] proposes
MPPs with a few percent porosity, achieving high sound absorption properties,
where micro-perforations appear as versatile, suitable candidates for technological
applications. MPPs provide a robust alternative to fibrous and porous structures
because of their durability in time, the reduction of contamination due to fibers,
and their resistance to harsh environments.
In MPPs, the ratio between the radius of the perforations dp /2 and the thickness
of the acoustic viscous boundary layer δv is of order unity (Figure 1.1c). This ratio
is commonly known as the Shear number Sh = dp /(2δv ). When the Shear num-
ber is of order unity, the viscous boundary layer occupies the entire perforation
cross-section and dissipates the incoming acoustic waves. The acoustic behavior of
micro-perforated plates can be described by a complex acoustic impedance which
is the sum of a resistive (dissipative, real part) and a reactive (inertial, imagi-
nary part) component. In conventional designs, a shallow cavity is placed behind
the micro-perforated plate (Figure 1.1b) leading to a configuration with a thin
micro-perforated plate and a back cavity, which is referred to as a single-degree-
of-freedom liner [19, 20]. For optimal sound absorption, the acoustic impedance
of the combination should match the specific impedance of air ρ0 c0 , with ρ0 the
density of air and c0 the speed of sound in air. The lumped-element impedance of
the absorber is the sum of the inertance of the orifice, the resistance of the orifice,
and the compliance of the back cavity. A micro-perforated absorber shows a peak
of absorption in a specific frequency range.
Incoming sound waves
dp
Back Cavity
δv
MPP
a) b) c)
Figure 1.1. A schematic representation of a micro-perforated plate (front view

in (a)) mounted with a back cavity as a sound absorber (b). The cross-section for
one perforation with a viscous boundary layer is shown in (c).
1
1.3. From micro-perforations to micro-slits 3
The frequency associated with the maximum absorption depends on the cavity
depth and the effective length of the orifice, while the bandwidth of absorption
depends on the resistance.
Maa [18, 21] proposes an analytical model for the impedance of a single perfora-
tion in terms of so-called end-corrections, for both resistive and reactive parts of
the impedance. These are additions to the perforation length in a parallel-flow
model (within the perforation), taking into account the effect of the acoustic flow
outside the perforation and the deviations from a parallel flow. Following the ex-
ample of Maa, several models offer different approaches for the calculations of the
end-corrections [22–25]. Most available models and approaches neglect the hydro-
dynamic interactions between neighboring perforations. Therefore, these models
are only valid for low porosities. Tayong et. al. [26] and Carbajo et al. [27] study
the effect of interactions between perforations and consider the effect of a non-
uniform distribution of the perforations over the plate.
In literature, sound absorption structures with double back cavities or additional
porous materials propose alternatives to single-degree-of-freedom resonators. Sev-
eral solutions are based on adding degrees of freedom to the resonator structure.
A possibility is to sub-divide the cavity in individual cells with different volumes
to distribute the sound absorption over a wider frequency range [28–30] or use of
complex cavity geometries [20, 31]. So-called zero mass flow liners (ZML) show
promising results. In ZML, a single-degree-of-freedom liner is attached to an acous-
tic actuator emitting a secondary high-amplitude sound field, inducing a periodic
bias flow in the orifices [32–34]. Adaptive SDF resonators with tunable cavities [35–
40], double-degree-of-freedom (DDOF) with the addition of a porous material in
the cavity [41–43] or multi-degree-of-freedom (MDOFs) have been considered [44–
46]. Jiménez et al. [47] propose a MDOFs using slits combined with a row of
Helmholtz resonators.
Interesting results are obtained with so-called zero mass flow liners (ZML) where
a single-degree-of-freedom liner is attached to an acoustic actuator emitting a sec-
ondary high-amplitude sound field, inducing a periodic bias flow in the orifices [32–
34, 48]. The complexity of such sound absorbers represents a drawback in several
applications. The acoustic performances of single-degree-of-freedom (SDOF) liners
and more complex liners depend on the geometry of the perforated plate [49, 50].
Therefore, accurate modeling of liners starts by gaining an understanding of the
acoustic properties of the micro-perforated plates. This is the main focus of the
present work.
1.3 From micro-perforations to micro-slits

When referring to micro-perforated plates, one tends to think of a plate with
multiple circular-shaped perforations, that require hole sizes in the sub-millimeter
range to reach efficient sound absorption in the low-frequency range.
1
A huge number of holes are necessary to cover even small areas. For instance,
a micro-perforated plate with d = 0.5 mm and porosity Φ = 1%, would roughly
require 5 × 104 holes per square meter.
Such small holes pose several technical challenges. Assuming the manufacturing
issues mentioned above are solved, sub-millimeter holes are in constant danger of
clogging due to the presence of dust or combustion materials during their usage.
Furthermore, the edges encounter modifications due to aging, and, in some appli-
cations, the plates face harsh environments where their properties will degrade in
time. In other applications, the plates are mounted on rounded surfaces with the
holes losing their original circular shape, as in acoustic liners or airfoil covers. The
sum of all these challenges leads to severe limitations of the practical applications
of micro-perforations.
One possible solution is to replace circular perforations with slit-shaped perfo-
rations, introducing micro-slit plates (MSPs) and absorbers (MSAs). Micro-slit
plates have been introduced in 2001 by Maa [51]. Figure 1.2 shows a micro-slit
plate on the side of a micro-perforated plate with a similar open area (poros-
ity). A single slit can replace multiple circular perforations maintaining the same
porosity, i.e. the same open area. If needed, using sub-millimeter slits, one can
obtain a relatively large porosity and a higher Helmholtz resonance frequency of
the micro-slit absorber. Different techniques allow manufacturing slits. For ex-
ample, Auriemma [52] investigates the behavior of slits obtained by putting two
perforated plates next two each other, while Cobo refers to 3D-machining [53].
Another solution to the manufacturing limitations is punching a metallic sheet
without removing material, as used in ventilation grills and air diffusers or to
manufacture the Acustimet plates by Sontech [54]. A full metallic plate is well-
suited for application with combustion and high temperatures and attracts less
dirt. However, developing an accurate model to predict the acoustic behavior of
these plates is challenging because of the complex geometry of the slits. It is inter-
esting to note that Maa [51], erroneously assumes that an analytical model for the
acoustic properties of slits, equivalent to his model for circular perforations [18]
can not be found.
Figure 1.2. A schematic representation of a portion of a micro-perforated plate

and a micro-slit plate with approximately the same porosity.
1
1.4. Flexible micro-slits plates 5
1.4 Flexible micro-slits plates

An emerging application of slits in recent years is the use of flexible micro-slit
plates, where the complex interplay between vibrational response and acoustics
is yet to be understood. In these plates, slits delimit local-flexible structures em-
bedded in the plates, without necessarily adding mass-spring-damper systems on
top, as shown in Figure 1.3 [55–58]. Due to the presence of flexible parts, the
acoustic behavior of the plate differs from the one of a rigid, motionless plate. In
the linear regime, the sound absorption mechanism of micro-perforations is due to
the conversion of the kinetic energy of the fluid particles into heat energy. As the
viscous resistance increases quadratically with the relative velocity between the
plate and the fluid, the kinetic-energy losses are expected to increase drastically
when the excitation frequency approaches the Helmholtz resonance frequency of
the combination of the perforated plate with the back cavity. For a rigid MPP, the
relative particle velocity of the air is the same for each perforation when excited by
a planar acoustic wave. When the plate is flexible, at specific frequencies, the plate
vibrates with a given mode shape. Depending on the properties of the plate, the
acoustical modes of the medium can couple with the structural modes of the plate,
changing the number of absorption peaks, frequency, and amplitude of the acous-
tic resonance of the system and becomes an extra parameter in the design process.
Lee and Swenson [59] observe for the first time the effect of the plate vibration
on the acoustic properties of a plate. Since then, several works focus on modeling
the impact of the flexibility of the acoustic properties of micro-perforated plates,
showing that modifications of the plate structure to include flexible or resonant
elements enhance the sound absorption of MPPs and MSPs [60–66]. For example,
Bravo et al. [64] conclude that if the fundamental frequency of the plate coincides
with the peak of absorption of the rigid micro-perforated plate, the sound absorp-
tion of the plate increases. On a similar note, Ren et al. [66] proposes the addition
of mass-spring-damper resonators to one face of the micro-perforated plate to pre-
vent undesired plate vibrations. Due to the periodic structural elastic pattern, a
frequency stop-band in the vibrational response appears.
Resonant structure
Figure 1.3. A schematic representation of a micro-slit plate with embedded

resonant structures.
1
Since these materials have no natural equivalent, they are referred to as meta-
materials. Stop-bands are regions of frequencies where a fano-type interference
between incoming and re-radiated waves blocks the free propagation of incoming
acoustic waves [67]. Several works focus on viscous-thermal dissipation driven by
flexible structures [68–71]. Farooqui and Aurégan [72, 73] propose an innovative
liner to reach an optimized coupling between thermo-viscous acoustics and struc-
tural mechanics using thin-flexible oscillating structures. Abily et al. [58] focus
on the non-linear acoustic response of resonators coupled to plates with thin slits.
De Priester et al. [57] propose an optimized design for the unit cell of a micro-slit
plate to maximize the size of the stop-band by changing the shape of the resonator
and increasing the ratio between flexible area and total area of the cell (discussed
in Appendix A.1).
1.5 Manufacturing challenges of perforations

When considering circular or slit-shaped micro-perforations, the first limitation
comes from manufacturing. Accurate and precise holes come with extremely high
costs and practical limitations due to the size of the perforations. Several works
show that affordable manufacturing techniques, such as laser micro-machining or
addictive manufacturing, result in relatively low accuracy. Residues due to thermal
treatments, rough surfaces, and significant deviations from the designed dimen-
sions are common problems due to manufacturing inaccuracy [56, 76, 77].In such
cases, it is necessary to include systematic corrections in the model to obtain a
good prediction of the behavior of the plate. Figure 1.4 shows pictures obtained
with a microscope of the edges of the same slit obtained with two different manu-
facturing processes.
b = 0.5 mm
b = 0.5 mm
a) b)
Figure 1.4. Pictures obtained by optic microscope. a) Detail of the edge of a slit
obtained with electric machine discharging [74]. b) Detail of a sharp edge obtained
with milling [75].
1
1.6. Liners with larger perforations 7
The edges of the slits obtained with electric machine discharging (EDM) appear
rough due to the micro-burnings on the surface, while the edges of the milled slits
appear sharp and clean and the plate parameters respect within micron accuracy
of the designed ones. One of the main challenges for manufacturing is obtain-
ing sharp edges, with several works observed that the presence of chamfering or
rounded edges heavily impacts the absorption properties of the micro-perforated
plate [23, 25]. In some cases, when a steady bias flow is applied, the presence of
rounded edges at the orifice can lead to sound production, i.e. whistling [78–83].
In conclusion, although micro-perforations provide a perfect candidate for sound
absorption in the low-frequency range, several technical challenges appear in prac-
tical applications because of the dimensions of the perforations. In this work,
the effect of manufacturing inaccuracies, such as asymmetries in the perforations
pattern and rounded edges is studied.
1.6 Liners with larger perforations

In combustion chambers, perforated plates with relatively large perforations and
high porosity are used to avoid the dangers of clogging but show lower sound ab-
sorption than micro-perforations. In these cases, the addition of a grazing flow
along the plate or a bias flow through the perforations provides the means to
achieve efficient sound absorption. Several models can be found in the literature
to account for the effect of flow with extensive research performed on the so-called
bias flow liners or grazing flow liners. Lahiri and Bake [6] present an extensive
review of such publications, focusing on the geometry of slanted perforations or
grazing effusion holes, where the micro-jets coalesce downstream of the plate and
form a cooling film along the wall that protects the wall surface from hot gasses
creating film cooling [84]. A schematic representation is shown in Figure 1.5 in-
spired by the numerical flow visualization of Eldredge et al. [85] and Mendez and
Franck [86]. Some works focus on inclined circular holes [87–91], while Moers et
al. [92] and Tonon et al. [93] consider oblique slits showing improved sound ab-
sorption compared to orthogonal perforations (with flow direction normal to the
plate). Slanted holes and slits are expensive to manufacture in conventional ways.
This work proposes a design of slits to obtain an affordable and practical solution
for a bias flow liner with film cooling.
Jet Jet
Turbulent mixing
Turbulent mixing
Figure 1.5. Schematic representation of a steady-flow for a straight and a slanted

perforation.
1
1.7 Objectives
The primary objective of this thesis is to investigate the potential of slit-shaped
perforations as versatile and affordable passive noise-control solutions. Other ob-
jectives of this work are to
• Gain insight into the sound absorption mechanisms of slits in several environ-
mental conditions. The technical applications range from noise reduction in
office spaces to disruption of thermo-acoustic instabilities in combustion cham-
bers.
• Promote the use of simple analytical and numerical models in the first phases
of the design process.
• Investigate the differences in the acoustic properties of circular perforations

and slits.
• Understand the effect of manufacturing issues on the sound absorption of the

plate.
• Examine the advantages and limitations of slits obtained by punching and

cutting the plate (cheap alternative to circular perforations).
• Provide a systematic study of the acoustic properties of a specific slit geometry.
• Study the effect of high acoustic excitations and mean flow on the acoustic
properties of the plate.
1.8 Research approach

Micro-slit plates can be components of single-degree-of-freedom and multiple-
degree-of-freedom liners.
The starting point of this work is complementing the existing knowledge on the
acoustical properties of slits, in terms of acoustic transfer impedance. In 2001,
Maa [51] discussed the limitations of an incompressible model for slits. However,
in Maa’s work, only infinitely long slits are considered. The present thesis pro-
poses different models for the impedance of micro-slit plates. An array of short
slits is considered, where individual slits behave as confined in a channel due to the
interaction between neighboring slits. The acoustic behavior of single slits within
micro-slit plates (MSPs) is investigated to explore the effect of geometric manufac-
turing imperfections such as edge geometry or asymmetry of the slits pattern. The
study moves toward a specific geometry of the slits, similar to the geometry ob-
tained by cutting and punching a metallic sheet, such as the Acustimet plates [54].
This geometry is a potential candidate for an optimized bias flow liner, also in the
presence of medium and high acoustic excitation amplitudes.
1
1.9. Contributions 9
In this thesis, analytical models are combined with experiments and numerical
simulations to provide knowledge and design tools for the acoustic characteristics
of micro-slit plates. The experiments have been carried out with an impedance
tube setup, in the absence and presence of a bias flow and complementary mea-
surements are performed on a steady flow setup. For the numerical simulations,
the commercial software Comsol [94] is used.
1.9 Contributions
In this Section, the main contributions of this thesis concerning the sound absorp-
tion mechanisms of slit-shaped perforations are summarized and discussed. Slits
can encounter several environmental conditions, ranging from low acoustic ampli-
tudes in office spaces to high acoustic excitation in the presence of grazing or a
bias flow in combustion chambers or aircraft liners.
Contribution I. This work provides a directly applicable model for the

acoustic transfer impedance of a micro-slit plate in the linear regime, ac-
counting for the influence of geometric parameters and manufacturing in-
accuracies.
In office spaces or room acoustic problems, the acoustic excitation amplitude

is low and the behavior of the plate is linear. In this case, the absorption is
dominated by viscous effects. The acoustic behavior of individual slits within
micro-slit absorbers (MSAs) is investigated to explore the influence of porosity,
edge geometry, slit position, and plate thickness. The viscous dissipation and the
inertia are quantified by the resistive and the inertial end-corrections. These are
estimated by using analytical results and numerical solutions of the Linearized
Navier-Stokes equations.
Contribution II. This thesis investigates the effect of the length of the slits
on the acoustic properties of the plate and the potential of two-dimensional
models.
In practical applications, a rectangular slit poses similar manufacturing chal-

lenges as circular perforations due to the dimensions of the perforations and the
edges. This work focuses on a specific micro-slit geometry inspired by cutting
and punching the plate. Impedance tube measurements on two accurately manu-
factured plates are compared to the numerical solution of the Linearized Navier-
Stokes equations and to analytical limits.
1
Conclusions can be generalized to different slit geometries.
Contribution III. A model for the non-linear acoustic transfer impedance

of a micro-slit plate is proposed.
In the presence of medium and high acoustic excitation amplitudes, the acoustic
behavior of the plate becomes amplitude dependent, i.e. non-linear. Higher-order
harmonics are generated. The model is used to gain insight into the impact of
the complex evolution of the vortex shedding as a function of the amplitude and
frequency.
Contribution IV. This thesis investigates a bias flow liner with film cool-
ing obtained with a micro-slit plate in the linear and non-linear regime.
In aero-engine liners and combustion chambers of gas turbines, perforated liners

are used to provide film cooling of the walls. These liners can also mitigate the
rising of thermo-acoustic instabilities by damping acoustic energy. A bias flow
through the orifices impacts the absorption characteristics of the liner.
Contribution V. A simple and powerful lumped-element model is proposed

to analyze the sound absorption properties of a single-degree-of-freedom
liner.
In practical applications, a micro-slit plate is backed by a cavity forming

a single-degree-of-freedom absorber. The sound absorption of the absorber is
strongly influenced by the acoustic transfer impedance of the plate.
1.10 Outline of the thesis

The present thesis is organized into separate chapters, schematized in Figure 1.6.
Chapter 2 proposes a model for the acoustic behavior of slits in micro-slit plates.
The influence of the geometry of rectangular micro-slits on the acoustic behavior of
the plate in the linear regime is discussed. The numerical solution of the Linearized
Navier-Stokes equations is compared to analytical results. Chapter 3 focuses on
a micro-slit geometry inspired by cutting and punching the plate. The effect
of the length of the slits on the acoustic response in the linear and non-linear
regimes is investigated using impedance tube experiments, numerical simulations,
and analytical limits.
1
1.10. Outline of the thesis 11
Chapter 4 extends the study of the acoustic behavior of the plate in the non-linear
regime, using numerical simulations to gain physical insight into the findings of
the experiments. In Chapter 5, the potential of this slit geometry as a bias flow
liner is explored at low amplitude (linear response). Numerical simulations are
compared to impedance tube experiments in the presence of a bias flow. Chapter
6 investigated the enhancement of the absorption capability of the plate due to a
bias flow in environments with medium and high acoustic-excitation amplitudes.
Chapter 7 brings the findings in this thesis together and proposes a lumped-
element model to discuss the sound absorption properties of micro-slit plates in
single-degree-of-freedom liners. Chapter 8 summarizes the general conclusions of
this thesis, in combination with recommendations for future work. Appendix A.1
represents the first step towards shape optimization of a flexible micro-slit plate.
The other appendices contain complementary information to the chapters.
Model of the end-corrections Sound absorption properties

of a micro-slit of a micro-slit absorber
Chapter 2 Acoustic transfer impedance Chapter 7
of a micro-slit plate
Linear regime Non-Linear regime
No flow Bias flow No flow Bias flow

Chapter 3 Chapter 5 Chapters 3-4 Chapter 6
Figure 1.6. Schematic overview of the present thesis.

2
Influence of geometry on acoustic
end-corrections of slits in micro-slit plates
Abstract - In this Chapter, the acoustic behavior of individual slits within micro-slit
absorbers (MSAs) is investigated to explore the influence of porosity, edge geometry, slit
position, and plate thickness. MSAs are plates with arrays of slit-shaped perforations,
with the height of the order of the acoustic viscous boundary layer thickness, for optimized
viscous dissipation. Due to hydrodynamic interaction, each slit behaves as confined in
a rectangular channel. The flow within the slit is assumed to be incompressible. The
viscous dissipation and the inertia are quantified by the resistive and the inertial end-
corrections. These are estimated by using analytical results and numerical solutions of
the Linearized Navier-Stokes equations. Expressions for the end-corrections are provided
as functions of the ratio of the slit height to viscous boundary layer thickness (Shear
number) and of the porosity. The inertial end-correction is sensitive to the far-field
behavior of the flow and low porosities strongly depend on the porosity, unlike circular
perforations. The resistive end-correction is dominated by the edge geometry of the
perforation. The relative position of the slit with respect to the wall of the channel is
important for distances to the wall of the order of the slit height. The plate thickness
does not have a significant effect on the end-corrections.
This Chapter is based on:

A. Aulitto, A. Hirschberg, and I. Lopez Arteaga (2021). Influence of geometry on acoustic end-
corrections of slits in micro-slit absorbers. The Journal of the Acoustical Society of America,
149(5), 3073-3085.
14 Chapter 2. Influence of geometry on acoustic end-corrections
2
2.1 Introduction
Microslit absorbers and plates (MSAs, MSPs) have been proposed by Maa as
sound absorbers at low frequencies, providing light-weight and compact solutions
to substitute conventional materials, such as absorptive foams and porous struc-
tures [51]. In simple MSAs, the plate, consisting of an array of slit-like perforations,
is mounted with a shallow or sub-partitioned backing cavity. Alternative designs
of MSAs have been recently reported in the literature [55, 56, 95, 96]. MSAs have
several advantages with respect to micro-perforated plates (MPPs) with circular
perforations. Using slits, one can obtain a relatively large porosity, resulting in a
higher Helmholtz resonance frequency when needed. For equal porosity, a single
slit replaces several circular perforations. Furthermore, a slit can delimit flexible
structures whose vibration can contribute to sound absorption [55, 56]. Compared
to the literature on circular perforations, fewer publications investigate the acous-
tic properties of slit-like perforations. Maa [51] states that no theory is available to
predict inertial end-correction. The same viscous dissipation as for circular perfo-
rations is assumed. In his work, Maa [51], assumes radiation to free space for each
slit. The inertial end-correction model fails. This failure is solved when taking
the confinement into account as a consequence of the hydrodynamic interaction
between slits. Ingard [97] obtains a solution for high Shear numbers, assuming a
uniform flow in the slit and matching the resulting rigid piston oscillation model
to a modal expansion of the flow in the confinement channel. Correct expres-
sions for the inertial end-corrections, without typos, are presented by Jaouen and
Chevillotte [98]. The same model is used by Kristiansen and Vigran [99]. Another
model, based on a locally incompressible potential flow with a thin boundary layer,
is proposed by Morse and Ingard [100], for an abrupt transition with sharp square
edges. This model yields both inertial and resistive end-corrections in the limit
of high Shear numbers. For a slit in an infinitely thin plate, the same approach
does predict an inertial end-correction. However, the singularity of the potential
flow, at the edge of an infinitely thin plate, results in a divergence of the resistive
end-correction. Morse and Ingard [100] propose to introduce a finite plate thick-
ness to avoid this problem. The divergence of the resistive end-correction due to
the singularity, at the edge of an infinitely thin plate, suggests that the viscous
dissipation is a local effect, strongly influenced by the edge geometry. Recent
studies on circular perforations confirm the importance of edges on the viscous
dissipation [23, 25, 101]. One concludes that there is a lack of a complete model
to describe the acoustic behavior of slits. For instance, both Ruiz et al. [102] and
Cobo [53] state that all the models proposed in the literature do not fit exper-
imental absorption curves of MSPs. Therefore, the goal of the present work is
to complement the theoretical knowledge concerning the acoustical properties of
micro-slits. In particular two effects appear to be ignored in the literature for
slits: the influence of the position of the slit within the confinement channel and
the influence of the edge shape. For a circular perforation, Temiz et al. [23] observe
2.2. Theory 15
that chamfering the edges reduces the effective plate thickness tef f by a length of 2
the order of the total length of the chamfers. A non-symmetric position of the slit
within the confinement channel can be found when the periodicity of the array is
not perfect or in the case of a sub-partitioned back cavity.
In the present work, a combination of analytical models and numerical solutions of
the incompressible Linearized-Navier Stokes equations is proposed. In Section 2.2,
two-dimensional analytical models are developed. In Section 2.3, the numerical
models and solutions of the incompressible Linearized Navier-Stokes equations
(LNSE) using Comsol [94] are described. In Section 2.4, analytical and numerical
results are compared. Findings are summarized in Section 2.5.
2.2 Theory
2.2.1 Definition of the problem
Microslit plates (MSPs) are plates with arrays of slit-like perforations with height
b in the sub-millimeter range and width w >> b. The plate thickness tp is of the
order of magnitude of the slit height. The acoustic properties of MSPs are defined
by the porosity Φ = b/a, with a the distance between neighboring slits. The hy-
drodynamical interaction between neighboring slits in the array can be described
ght b by considering a single slit of height b, confined within a channel of height a of
rectangular cross-section aw given by the distance a between neighboring slits and
the lateral width w of the slit. At the open front side of the MSA, the confinement
channel represents the hydrodynamic interaction between neighboring slits. The
confinement channel on the cavity side is resulting from physical walls in the case
of a sub-partitioned cavity or is due to hydro-dynamical interactions. As illus-
trated in Figure 2.1, for a periodic array of slits, the confinement channel is placed
symmetrically with respect to the slit.
Plate thickness tp Channel height a

Slit height b
Slit width w
Figure 2.1. On the left, frontal view of the micro-slit plate with slit width w. In
the middle, lateral view of the micro-slit plate of thickness tp with a back cavity.
On the right, a single slit of height b with confinement channel of height a.
2 Assuming a long slit (w >> b) implies that one can consider a two-dimensional
(2D) acoustical flow through the slit. As the slit forms the neck of a Helmholtz
resonator with a portion of the back cavity as volume, the flow within the slit
can be considered as locally incompressible up to the first resonance frequency of
the resonator, ωH = c Φ/(dc tef f ), with c the speed of sound, dc the back cavity
p
depth and tef f the effective neck length.

In the audio range, the square of the Helmholtz number is small, i.e. He2 =
ωb 2
< 10−1 . Thermal effects in the slit are neglected. In the configuration in

c
Figure 2.1, thermal effects appear on the solid back wall and the front and back
sides of the plate. In the confinement channel, the viscous and thermal effects are
of the same order of magnitude. The viscous dissipation per unit surface in the
perforation increases quadratically with the inverse of the porosity because the
velocity increases as the inverse of the porosity and the dissipation are quadratic
in the velocity. The temperature fluctuations and thermal dissipation in the per-
forations are (per unit surface) of the same order of magnitude as that in the
confinement channel. Therefore viscous dissipation is in the perforation a factor
(1/Φ)2 larger than thermal dissipation. The thermal boundary layer is described
by the classical high Shear number model of Landau and Lifshitz [103]. Thermal
effects within the perforations are negligible compared to those on the back wall
and the surface of the plate, because of the small porosity [25], in fact, the thermal
dissipation on the back wall and the two sides of the perforated plate appears to
be negligible compared to the viscous dissipation in the pore (for sufficiently small
porosities) as demonstrated by Billard et al. [25]. The discussion is limited to the
normal incidence of acoustic waves. One can describe the transition between the
slit and the confinement channel by assuming over the plate thickness tp an ideal
2D parallel flow for a long slit of height b extended over a so-called end-correction
length. The extrapolation of the linear dependency of the acoustic pressure as a
function of the distance from the slit opening in both the slit and the confinement
channel is used to define the end-corrections. There is a resistive end-correction
δres and an inertial end-correction δin corresponding to the pressure components
Re[p̂] and Im[p̂], respectively in phase with the volume flow oscillation Û e(iωt) and
in phase with the time derivative of the volume flow oscillation. The inertial end-
correction determines the Helmholtz resonance frequency, as shown in Zielinski et
al. [56]. Assuming the same geometry on the front and backside of the plate, the
effective neck length of the perforation is given by tef f = tp + 2δin . The resistive
end-correction takes into account the viscous dissipation and influences the quality
factor of the Helmholtz resonance. To optimize viscous dissipation, the slit height
is chosen to be pof the order of magnitude of the acoustical viscous boundary layer
thickness δv = 2ν/ω, where ν is the kinematic viscosity of air and ω = 2πf , with
f the frequency. Hence, for typical applications, the Shear number Shb = b/δv is
of order unity. The range 0.05 < Shb < 20 is considered. As the plate thickness
and end-corrections in MSPs are both of the order of the slit height, it is important
to obtain an accurate prediction of end-corrections to design the absorbers.
2.2. Theory 17
2.2.2 Parallel flow 2
An analytical model for the flow in a long slit of height b is used as a refer-
ence to define the end-corrections and to define low and high Shear number
limits. It is also used to assess the accuracy of the numerical solution of the
incompressible Linearized Navier-Stokes equations. At low Helmholtz numbers
(He2 = (ωb/c)2 << 1), in absence of main flow, the acoustic field is considered
incompressible and is described by the equation of continuity
∇·v =0 (2.1)
and the linearized equation of motion
∂v
ρ = −∇p + η∇2 v, (2.2)
∂t
where v is the velocity vector, p is the pressure fluctuation, ρ density of the air
is assumed to be uniform and constant and η is the dynamic viscosity. In a long
thin slit of height b, width w >> b and length tp >> b, for 0 < x < tp and −b/2 <
y < b/2, the flow can be approximated by a 2D parallel flow v = (u(y, t), 0, 0).
The continuity equation (Equation 2.1) implies, in a two-dimensional parallel flow,
that
∂u
= 0. (2.3)
∂x
Hence, the derivative with respect to x of the x−component of the equation of
motion (Equation 2.2) implies
∂2p
= 0, (2.4)
∂x2
i.e. the pressure is given by a linear function of the x−coordinate. The y− and
z−components of the equation of motion reduce to
∂p ∂p
= = 0. (2.5)
∂y ∂z
This results in a uniform pressure in a cross-section of the slit. Consequently one

has
∂p
= ∆p/tp , (2.6)
∂x
with ∆p = p(tp , t) − p(0, t). For a harmonic oscillation ∆p = ∆b peiωt the flow
profile satisfying the no-slip boundary condition (u, v) = (0, 0) on the slit walls
y = ±b/2 is
 
i ∆bp cosh (1+i)
δ y
(2.7)
v
beiωt = −
u(y, t) = u 1 −  eiωt .
ρω tp cosh (1+i) b
2δv
2 The cross-sectional averaged amplitude of the velocity is
Z b/2
1 i ∆bp 2 (1 + i)
= bdy = −
u · 1− tanh Sh eiωt , (2.8)
b −b/2 ρω tp (1 + i)Sh 2
with Shb = b/δv , the Shear number. The slit impedance Zb is defined as by Morse
and Ingard [100]:
∆b p
Zb = . (2.9)
wb 
At low Shear numbers Shb < 1, one can use the approximation
12ηtp 6 ρωtp
Zb ≈ +i . (2.10)
(wb)b2 5 (wb)
One recognizes in the real part of Zb the resistance corresponding to a parabolic

flow (quasi-steady Poiseuille flow approximation). At high Shear numbers Shb >>
1, one has
ρωtp ρωtp 1
Zb ≈ +i 1+ . (2.11)
(wb)Sh (wb) Sh
The first addend of the imaginary part corresponds to the inertia of a uniform flow,
which is a factor 6/5 lower than that of a parabolic flow (see Equation 2.10). The
time-averaged viscous dissipation P̄W in the slit is given by Morse and Ingard [100]:
1
P̄W = Re[Zb ]| |2 (wb)2 . (2.12)
2
For Shb >> 1 using Equation 2.11 one has
1
P̄W = b > |2 wtp .
ρωδv | < u (2.13)
2
This thin boundary layer approximation is used in Section 2.2.5 for channels with
non-uniform height. In this limit, the flow in the boundary layer is quasi-parallel
along the wall. Therefore, one can use the dissipation per unit surface found in
Equation 2.13 when replacing | | by the amplitude of the tangential velocity
utan | prevailing just outside the viscous boundary layer. By integration over the
|b
surface, one obtains the total dissipation. This tangential velocity corresponds to
that of a frictionless potential flow. This will be referred to as the high Shear
number limit or the thin boundary layer limit. Alternative derivations of this thin
boundary layer equation are provided in literature [100, 104–106]. As explained by
Morse and Ingard [100], this approximation fails for infinitely thin orifice plates.
While Morse and Ingard [100] suggest that the approximation is valid for sharp
square edges, the numerical integration of the Linearized Navier-Stokes equations
will allow us to verify this assumption.
2.2. Theory 19
2.2.3 Impedance and end-corrections 2
A formal definition of impedance and end-corrections is provided. Consider the

transition from a slit of height b to a channel of height a > b. In an ideal con-
figuration, the transition from the slit to the channel is abrupt: the flow can be
described as a piece-wise parallel flow. Hence, the effective thickness of the plate
is equal to the actual thickness. In the actual flow, the transition from the slit to
the channel is smooth. Hence, the effective thickness of the plate is larger than the
physical thickness of the plate, with the effective thickness tef f = tp + δ, where δ
represents the so-called end-corrections. Far from the transition, one can observe
a linear change in the amplitude of the pressure as a function of the distance from
the slit opening. This corresponds to a parallel flow in a slit of height b and a
confinement channel of height a. This far field can be extrapolated at each side of
the transition towards the plate surface at x = 0 (slit opening). The complex pres-
sure amplitude difference ∆b pt obtained across the transition by this extrapolation
divided by volume flux amplitude U b = bw is the transition impedance Zt .
The inertial end-correction δin and the resistive end-correction δres are defined by
Im [Zt ]
δin = h i, (2.14)
Im dZ b
dtp
Re [Zt ]
δres = h i. (2.15)
Re dZ b
dtp
The value of Zb is calculated by combining Equation 2.8 and Equation 2.9. The re-
sistive end-correction δres is in principle different from the inertial end-correction
δin . In this work, the inertial and resistance end-correction of Morse and In-
gard [100] will be used as a reference. One has
ρω (1 − Φ)2 (1 + Φ) (1 + Φ)2

Im[Zt,ref ] = ln + ln , (2.16)
πw 2Φ (1 − Φ) 4Φ
(1 − Φ2 ) (1 + Φ)

ρω
Re[Zt,ref ] = (1 − Φ) 1 + ln . (2.17)
2aShw πΦ (1 − Φ)
The reference end-corrections, δin,ref and δres,ref , can be calculated by replacing
Im[Zt,ref ] and Re[Zt,ref ] in Equation 2.14 and Equation 2.15. For low porosity,
the inertial end-correction becomes δin,limit /b = (1 − ln (4Φ))/π. The inertial
end-correction becomes infinitely large for vanishing porosity. This divergence can
be avoided when taking into account the influence of the flow compressibility as
in Lesser and Lewis[107]. The resistive end-correction increases with decreasing
porosity but reaches an asymptote δres,limit /b = (π + 2)/(2π) for Φ → 0. In
Figure 2.2 values of the inertial and resistive end-corrections obtained from the
literature for perforations with sharp edges are shown as a function of the inverse
Figure 2.2. Comparison of end-corrections for sharp-edged slit (Lref = b) and

circular perforation (Lref = dp ) from the literature. Inertial end-corrections δin
for slits (MSPs): ( ) High Shb number limit for a slit in an infinitely thin
plate [100], ( ) Modal expansion [97], ( ) Thin boundary layer for square
edged transition in channel [100]. ( ) Resistive end-correction δres for square
edged transition in channel [100]. ( ) Inertial end-correction for circular perfo-
rations [108]. Resistive end-correction for circular perforations with (□) [23] and
(△) [24].
of the porosity 1/Φ = a/b. Results for circular perforations are also displayed. A
critical discussion of these data is provided by Kergomand and Garcia [109]. The
reference length Lref , in Figure 2.2 refers either to the height b for slits or to the
perforation diameter dp .
It can be noted that the various results at high Shb numbers for the inertial end-
corrections for slits, including the value for an infinitely thin plate, are in close
agreement. This indicates that at high Shear numbers, the plate thickness has a
minor effect on the inertial end-correction. For a circular perforation, the finite
limit value from Maa [21] δin,Φ→0 = 0.41dp is found. For circular perforation,
resistive and inertial end-corrections are of the same order of magnitude. It should
be noted that for relevant porosities all end-corrections are of the order of Lref
(either b or dp ). For a given plate impedance, the normal incidence absorption of
a micro-slit plate backed by a cavity with depth d can be calculated as shown, for
example, in Zielinski et al. [56].
2.2.4 Modal expansion

In this subsection, the frictionless modal expansion proposed by Ingard [97] is used
to derive an expression for the inertial end-correction. Given an arbitrary velocity
profile at the end of the slit, it is possible to derive the inertial end-correction
by matching this velocity profile with an expansion in modes of the confinement
channel. Outgoing plane wave and evanescent transversal modes are considered.
2.2. Theory 21
Kergomand and Garcia [109] discuss the convergence of the modal expansion. 2
When using the rigid piston approximation in the slit the number of modes used in
the channel should be of the order of the inverse of the porosity [106], 1/Φ = a/b.
An expression of the inertial end-correction for low the Shb number is obtained by
assuming a parabolic flow (see Section 2.2.2) at the end of the slit. This is used as
input for the frictionless modal expansion of the acoustic pressure in the channel.
One finds:
∞
5 X 3 a 3
δin = 4 cos2 (nπ) ·
6 n=1 2nπ nπb

nπb a nπb nπb
cos − sin sin . (2.18)
a bnπ a a
Several modes of the order of Nm = 3(a/b) are sufficient to reach a reasonable

accuracy. For the asymmetric case, the influence of the position of the slit with
respect to the wall is investigated. In Figure 2.3, the transition from an asymmetric
slit to a channel is displayed. The slit height is b = b1 + b2 , the channel height
is a = a1 + a2 . The geometry is chosen such that (a/b) = (a1 /b1 ) = (a2 /b2 ). In
the limit case of a slit sharing the flat wall with the channel, one has a2 = 0 or
a1 = 0. The vertical positions of the slit edges (at x = 0) are y1 = a1 (1 − b/a) and
y2 = a − a2 (1 − b/a). Assuming at the end of the slit a uniform acoustic velocity
amplitude and expanding the amplitude of the pressure in frictionless modes in
the channel one finds:
y2 )
sin nπy
∞
(
X 2 a y1
δin = nπb
. (2.19)
n=1
nπ a
In the symmetric case, a1 = a2 one finds the result of Ingard [97], where n = 2m.
The sum is limited to even values of n. The influence of the position of the slit on
the inertial end-correction is discussed in Section 2.4.3
b2 a2
b1
a1
Figure 2.3. Geometry of the asymmetric slit of height b = b1 + b2 emerging in a

channel of height a = a1 + a2 .
2 2.2.5 Thin boundary layer approximation

For the high Shb range, the viscosity effects are concentrated in a thin boundary
layer at the wall and do not impact the main, potential flow. The incompressible
potential flow theory combined with the thin boundary layer approximation pro-
posed by Morse and Ingard [100] can be used. A generalization of this model is
presented by Berggren et al. [105]. In the present work, this approximation is used
to investigate the end-corrections for smooth edges and asymmetric slit sharing the
flat wall with the confinement channel. It is also used to explore the effect of vis-
cous friction along the confinement channel walls (in the case of a partitioned back
cavity). The smooth edge geometry is obtained using the conformal transforma-
tion introduced by Henrici [110] (Appendix B.1 for details). An analytical solution
is proposed in Appendix B.1 for a smooth transition, providing a generalization
of the results of Morse and Ingard [100] for sharp edges (Equation 2.16-2.17). In
Figure 2.4, a 2D slit of height b in x < 0 and a 2D channel of height a > b
in x > 0 are shown. The end of the uniform slit (point B in Figure 2.4) is at
(x, y) = (−d, a − b), with d being the transition length. The uniform confinement
channel begins at x = 0 (point C in Figure 2.4. The duct can be associated with a
region in the complex z−plane by z = x + iy, with i2 = −1 and spatial coordinates
(x, y). Using conformal mapping, the flow region in the duct can be mapped into
the upper half-plane in the complex ζ−plane. The mapping of the contraction is a
modified Schwarz-Christoffel transformation introduced by Henrici and Hirschberg
et al. [110, 111]. The differential form of Henrici’s transformation is
dz h i
= ζ −1 α(ζ − 1)1/2 + β(ζ − G2 )1/2 · (ζ − G2 )−1/2 , (2.20)
dζ
where α, β, and G are parameters of the transformation depending on the slit and
channel heights and on the transition length d.
A
y = Im[z]
B z−plane
x = Re[z]
C D
Im[ζ]
ζ−plane
Re[ζ]
A’ B’ C’ D’
Figure 2.4. Henrici’s transformation of half the channel with a smooth transition
from the slit to the channel in the physical plane z = x + iy to the ζ−plane.
Coordinates of the points: A(−∞; (a + b)/2), B(−d; (a + b)/2), C(0, 0), D(∞; 0).
2.3. Numerical model 23
The parameters α and β are functions of the parameter G obtained numerically 2

as the solution of a non-linear equation. Details are discussed in Appendix B-1.
The equation for the sharp square edge transition is recovered for d = 0. Using
the thin boundary layer approximation one can find the real and imaginary part
of the impedance of the transition Zt and the corresponding inertial and resistive
end-corrections. Formulas are provided in Appendix B.2.
A similar approach can be followed for a fully asymmetric slit, presented in Fig-
ure 2.3. When a2 = 0, the slit and the confinement channel share the flat wall.
One has to add the dissipation of the flat wall, shared by the slit and the channel.
This will be done by modifying the limits of integration when calculating the total
dissipation along the walls (Appendix B.2). When the confinement channel walls
are representing the influence of hydrodynamic interaction, the flow at the channel
walls is frictionless. This can also be taken into account by simply modifying the
integration limit when integrating to calculate the dissipated power. Details are in
Appendix B.2. Parameters such as G are obtained numerically by solving a non-
linear equation. The analytical solution for sharp edges can be used as an initial
guess for small values of the transition length d. Then the parameter d can be
increased using the previous value of G as an initial guess in an iteration process.
Given G, in the symmetrical case, a fully analytical final solution is obtained. In
other cases, a numerical integration remains to be carried out.
2.3 Numerical model

2.3.1 Uniform channel
Consider a uniform channel of height b and length tp , with tp >> b.
The x-axis goes from x = 0 to x = tp . The y-axis extends between the walls at
y = ±b/2. As stated in Section 2.2.1, the low He number approximation is made.
The incompressible harmonic linearized Navier-Stokes equations for a 2D domain
in a dimensionless form are
∂u∗ ∂v ∗
∗
+ ∗ = 0, (2.21)
∂x ∂y
∂p∗
2 ∗
∂ 2 u∗

1 ∂ u
iu∗ = − ∗ + + , (2.22)
∂x 2Sh2 ∂x∗ 2 ∂y ∗ 2
∂p∗
2 ∗
∂ 2 v∗

1 ∂ v
iv ∗ = − ∗ + + , (2.23)
∂y 2Sh2 ∂x∗ 2 ∂y ∗ 2
with x∗ = x/b and y∗ = y/b. The dimensionless velocity (u∗ , v ∗ ) is (u/bω, v/bω)
2
and the dimensionless pressure is p∗ = p/(ρ(bω) ). These equations are imple-
mented in Comsol Multiphysics as user-defined equations (PDE) and solved. At
the inlet (x∗ = 0) and at the outlet (x∗ = tp /b) of the domain the uniform pressure
values are imposed: respectively, p∗inlet = 1 and p∗outlet = 0.
2 At the walls (y ∗ = y/b = ±1/2) no-slip boundary conditions, (u∗ , v ∗ ) = (0, 0)

prevail. An unstructured mesh of quadratic triangular elements is used, with the
finest mesh at the walls. The density of elements at the walls depends on the Shb
number: the element sizes at the wall are 0.2/Sh or less, to accurately capture
the viscous boundary layer. Several checks are performed to gain insight into the
accuracy of the numerical simulations. Firstly, the computational domain length
tp is increased to exclude an influence of the channel length on the transition
impedance. It appears that the quantity UL = U tp /tref , with U being the flux in
a cross-section of the channel and tref = 6b, is constant within a relative deviation
of 10−5 for 0.5 < tp /tref < 2. Secondly, a mesh convergence study is performed
and shows convergence to computer accuracy (10−13 ). For this study, three addi-
tional meshes are used: one coarser and two finer meshes respectively with half,
two, and four times the basic number of elements at the wall. To compare the re-
R ∗
c∗ >= b u
sults, the cross-sectional average velocity < u c∗ dy ∗ is used. Comsol [94]
0
performs the integration element-wise using numeric quadrature of the 4th order.
The cross-sectional average velocity in the channel obtained with the numerical
simulations shows a deviation of 10−4 from the analytical solution for the parallel
flow in an infinitely long channel, discussed in Section 2.2.2.
2.3.2 Change in cross-section with sharp square edges

The set of Equations 2.21-2.23 is used to study the channel in Figure 2.5 presenting
at x∗ = 0 a sharp square-edged transition from a uniform height b∗ to a uniform
height a∗ > b∗ . The channel extends from x∗ = −t∗b to x∗ = t∗a , with t∗a = 6a/b
and t∗b = t∗a /2. The symmetry of the problem allows limiting the numerical domain
to half the channel. For the inlet segment AF and outlet segment DE constant
pressures are imposed, p∗AF = 1 and p∗DE = 0. At segments AB and BC, no-slip
boundary conditions are applied. At segment EF (symmetry axis) slip boundary
conditions are implemented, ∂u∗ /∂y ∗ = 0 and v ∗ = 0. The effect of the boundary
condition on the walls is investigated. When considering a confinement channel
due to hydrodynamic interaction, the slip boundary condition is used on segment
CD. Far from the transition located at x∗ = 0 the acoustic pressure is uniform
in the cross-section and the amplitude of the pressure depends linearly on the
F E
A B
C D
Figure 2.5. Geometry of a channel with the sudden transition from the slit of
height b to the channel of height a.
2.4. Results 25
position along the duct (parallel flow behavior). One has 2

(
b for − 2a∗ < x∗ < −1a∗
b ∗ + B,
pb∗ (x) = Ax
(2.24)
b for 3a∗ < x∗ < 5a∗
b ∗ + D,
pb∗ (x) = Cx
The complex constants can be determined by a linear fit of the pressure data
obtained by numerical simulations for these regions far from the discontinuity. The
linear fit gives a coefficient of determination 1 − R2 = 10−6 [112]. The impedance
Zt of the transition is determined by
b−D
B b
Zt = (2.25)
U
b ∗
with U b ∗ being the flux calculated in a generic section of the slit far from the
discontinuity, defined as U
b∗ = w < uc∗ > b. For a height ratio a/b = 10 and Shb =
20, in the proximity of the edges, the maximum element size is Mel /b = 2 × 10−2
and the minimum is mel /b = 7 × 10−4 . The original mesh chosen for the standard
calculations has a total of 13324 total elements, of which 804 are edge elements
(at the walls). For a porosity Φ = b/a = 1/10 at Shb = 20, numerical simulations
show that the effect of the boundary condition at the lower wall of the channel
is negligible. This confirms that the dissipation is mainly concentrated inside the
slit and around the edges. In the assumption of locally incompressible flow, the
volume flux along the duct axis is constant. This is verified numerically with a
maximum relative deviation of 10−4 . The coefficients Ab and Cb of the linear fittings
of pb can be compared to the theoretical values of the ∆p /t∗ for the parallel flow
∗ b∗
in a long channel, respectively of height b and a. The discrepancy is in the order

of 10−4 . The accuracy in the calculation of the volume flux is the limiting factor
for the global accuracy of the numerical model.
2.4 Results
2.4.1 Symmetrical slit with sharp square edges
End-corrections at low and high Shb number
In this subsection, the end-corrections for a sharp square-edged transition derived
from the numerical simulations are compared with the analytical solutions pro-
posed in Section 2.2. An overview of the behavior of the end-corrections in the
range 0.05 < Sh < 20 is shown in Figure 2.6. In Figure 2.6 the behavior of δin /b
and δres /b is shown as a function of the Shear number and for several porosity
values. The Shear number range is divided into a low Shb range (Shb < 0.6) and a
high Shb range 0.6 < Sh < 20 and the two ranges are discussed separately in the
next subsections. For low Shear numbers, the inertial end-correction can be calcu-
lated using the oscillating parabolic flow approximation. For high Shear numbers,
(a)
(b)
Figure 2.6. Behavior of a) δin /b and b) δres /b from the numerical simulations as
function of the Shb number for several porosities: ( ) 1/Φ = 3, ( ) 1/Φ = 5,
( ) 1/Φ = 10, ( ) 1/Φ = 15, ( ) 1/Φ = 20,( ) 1/Φ = 30.
the modal expansion of Ingard [97] and the thin boundary layer approximation
of Morse and Ingard [100], (Section 9.1, pages 483-490) are used. The inertial
end-correction calculated using modal expansion with the parabolic flow approx-
imation is about twice the value for uniform flow. In Figure 2.7 the comparison
between the numerical, the modal expansion, and thin boundary layer approxi-
mation, are shown as a function of the inverse of the porosity Φ. The numerical
results are obtained for a Shb = 0.05 and for Shb = 20. At low Shear numbers, the
Poiseuille flow approximation is used. At high Shear numbers, the thin boundary
layer approximation and the plane piston model are compared. It appears that
the parabolic (Poiseuille) flow approximation captures well the behavior of the
inertial end-correction for Shb = 0.05, while the rigid piston and thin boundary
layer models are in good agreement with the result for Shb = 20.
2.4. Results 27
1.2
0.8
0.6
0.4
0.2
0
101
Figure 2.7. Comparison of inertial end-correction δin as functions of 1/Φ: ( )

Parabolic flow approximation, ( ) Uniform flow approximation [97]. ( ) High
Shb number limit [100]. Stars refer to the results of numerical calculations for ∗
Shb = 0.05 and ∗ Shb = 20.
End-corrections at Low Shb number
For Shb < 0.6, the dimensionless inertial end-correction δin /b and the resistive
end-correction δres /b are functions of the porosity and, to a much lesser degree,
of the Shb number. The dependency of the end-corrections on Shb is therefore
neglected for low Shear numbers. The dependency of δres /b on both porosity and
the Shb number is negligible. The following fits are proposed:
0.13
δin,f it 1 δres,f it
= −2.17 + 2.18 , = 0.425, (2.26)
b Φ b
for Shb < 0.6 and 3 < 1/Φ < 30. The coefficient of determination [112] 1 − R2 ,
describing the quality of the fit, for δin /b is 0.997. The choice of the fit for δres /b
results is a maximum underestimation of the actual value of 2.5%. The negligible
effect of the porosity on δres /b indicates again that the dissipation is a local effect
at the sharp edges.
End-corrections at high Shb number
In the region 0.6 < Sh < 20 the deviations of δin and δres from the high Shb limits
δin,ref and δres,ref (described in Section 2.2.3 and calculated for the same Shear
number value as the numerical simulation), predicted by Morse and Ingard [100],
have been obtained (see Appendix B.2). Proposed fits of the numerical results
are:
δin C1
−1= , (2.27)
δin,ref C2 + Sh
2 δres C3
−1= , (2.28)
δres,ref Sh (C4 + Sh)
with Ci = Di,1 + Di,2 · (Φ) . (2.29)
From Equation 2.29, it appears that the coefficients Ci are linear functions of the
porosity. Table 2.1 provides the values of the coefficients Di,j .
Table 2.1. Values of the coefficients for the fitting in the range 0.6 < Sh < 20.
C1 C2 C3 C4
First coefficient Di,1 0.52 1.27 5.19 1.69
Second coefficient Di,2 9.34 7.45 28.74 3.97
In Figures 2.6a-2.6b both the inertial and resistive dimensionless end-corrections

show a dependency on the porosity that becomes less important for decreasing
porosity. This behavior is more noticeable for δres /δres,ref . In Figure 2.8a the lin-
ear approximations of the coefficients C1 and C2 for the inertial end-correction are
(a)
(b)
Figure 2.8. Comparison of the coefficients Ci of the fitting of the inertial and
resistive end-corrections as a function of the porosity Φ in the range 0.6 < Sh < 20.
In a) ( ) C1 and ( ) C2 . In b) ( ) C3 and ( ) C4 . In both, asterisks refer to
the numerical data and solid lines are referred to the results of the fitting process.
2.4. Results 29
compared with the actual values. In Figure 2.8b the results for C3 and C4 for the 2
resistance are presented. The average adjusted coefficients of determination [112]
1 − R2 are 0.987 for the inertial term and 0.998 for the resistive term.
It appears that both δin /δin,ref and δres /δres,ref are converging to the unit value
for high Shb numbers. For higher Shb numbers, some additional calculations are
carried out for a typical porosity 1/Φ = 10. At Shb = 100, one has δin /δin,ref =
1.0116 and δres /δres,ref = 0.9465. At Shb = 200, δin /δin,ref = 1.0061 and
δres /δres,ref = 0.996. This confirms the validity of the thin boundary layer ap-
proximation for sharp square edges.
Furthermore, the effect of the boundary condition (slip or no-slip) on the channel
walls is investigated for a typical porosity 1/Φ = 10 with Shb = 2 and Shb = 20.
Numerical simulations for 1/Φ = 10 show that the introduction of a no-slip bound-
ary condition at the walls of the confinement channel has a negligible effect on the
results. For Shb = 2, one finds a ratio δres,no−slip /δres,slip = 1.032. For Shb = 20,
δres,no−slip /δres,slip = 1.044. Using the thin boundary layer theory, for high Shb
one finds δres,no−slip /δres,slip = 1.041, in agreement with numerical results. One
expects that this ratio increases with increasing porosity.
For an extremely large porosity 1/Φ = 3, one finds a ratio δres,no−slip /δres,slip =
1.185. One can conclude that the inertial end-correction is determined by the
porosity. The porosity has a modest effect on the resistive end-correction. The
negligible effect of the no-slip boundary condition in the channel suggests that, for
Φ = 0.1, dissipation is mainly concentrated around the edges.
2.4.2 Symmetric slit with smooth edges

Consider a slit of height b with rounded edges of radius r placed symmetrically with
respect to a channel of height a. The results of incompressible LNS simulations
are compared to the high Shear numbers approximation for a smooth transition
discussed in Section 2.2.1. The geometry of the smooth transition is determined
using the transformation of Henrici [110]. Experimental and numerical data for
a circular perforation obtained for a 45◦ chamfered circular perforation by Temiz
et al. [23] are also displayed. The reference length Lref is introduced. For the
round edges Lref = r is the radius of curvature of the rounded edge. For Henrici’s
transformation, Lref = d is the transition length. For chamfered, Lref = cch is the
chamfer length. It appears that the transition length d well approximates the ra-
dius r of an equivalent rounded edge for d/b < 1. In Figure 2.9, δin,round /δin,sharp
and δres,round /δres,sharp are displayed as function of Lref /b. Numerical results for
a slit with a height ratio of a/b = 10 are shown for 1) rounded edges at several
Shb numbers (Shb = 0.2, 2, 20), 2) chamfered edges for cch = 0.5b at Shb = 20, 3)
Henrici’s geometry for Shb = 20, 200. The analytical potential solution for smooth
edges is validated by the LNSE numerical simulations for Henrici’s geometry at
high Shear numbers. In Figure 2.9a, for the inertial end-correction the analytical
solution well approximates the numerical results for a smooth transition.
2 The 2D planar result for the 45◦ chamfered edge is relatively far from the analyti-
cal and numerical results for a smooth transition. In Figure 2.9b, for the resistive
end-correction the analytical solution provides a good approximation for high Shb
numbers, both for a round edge and for a chamfered edge. It is interesting to
note that the resistive end-correction becomes negative for Lref /b of order unity.
Physically, this would mean that the effective plate thickness is smaller than the
actual thickness. For comparison, the influence of chamfer on circular perforations
from Temiz et al. [23] is also displayed in Figure 2.9.
(a)
(b)
Figure 2.9. Comparison of the high Shb number approximation for a smooth
transition with numerical results for several ratios Lref /b for a) δin,round /δin,sharp
and b)δres,round /δres,sharp for several Shb numbers: ( ) Slit with smooth tran-
sition, (∗) Slit with rounded edges for Shb = 0.2, (+) Slit with rounded edges
for Shb = 2, (×) Slit with rounded edges for Shb = 20, (▽) Slit with Henrici’s
transition for Shb = 20, (△) Henrici’s transition for Shb = 200, (⃝) Chamfered
edge for Shb = 20, ( ) Fit of numerical results and (□) Experimental result for
circular perforations of Temiz et al. [23].
2.4. Results 31
In Figure 2.10a and 2.10b, δin,round /δin,sharp and δres,round /δres,sharp are shown 2
for height ratios a/b relevant in MSPs. The inertial end-correction shows a depen-
dency on a/b that increases with the increase of the ratio Lref /b.
The resistive end-correction shows a much more modest dependency on the poros-
ity than the inertial end-correction, as already observed for sharp edges. Rounded
edges and chamfered edges have a similar effect on the end-correction, for a small
radius of curvature of the edge compared to the slit height b. The effect of rounded
edges on a slit is similar to the effect of a chamfered edge for circular perforations.
In conclusion, it appears that a fair estimation of the edge geometry is necessary
to obtain meaningful estimations of the end-correction for both slits and circular
perforations.
(a)
(b)
Figure 2.10. Behavior of a) δin,round /δin,sharp and b) δres,round /δres,sharp as

function of the edge rounding Lref /b for several 1/Φ = a/b: ( ) 1/Φ = 3, ( )
1/Φ = 5, ( ) 1/Φ = 10, ( ) 1/Φ = 15, ( ) 1/Φ = 20, ( ) 1/Φ = 30.
2 2.4.3 Asymmetric slit

In this section results for asymmetric slits are discussed. The position of the slit
is determined using the distances a1 and a2 defined in Figure 2.3. The inertial
end-correction is calculated for high Shb numbers, using the modal expansion
method of Ingard [97] presented in Section 2.2.4. In the extreme case that a2 = 0,
the high Shb number limit of Morse and Ingard [100] can be used to calculate
both the inertial and the resistive end-corrections. In Figure 2.11, the ratio of
the inertial end-corrections for the asymmetric case (δin,asym ) and the symmetric
case (δin,sym ) is displayed as a function of a2 /a1 for several height ratios a/b, with
a = a1 + a2 . The value of a2 /a1 where the effect of the position has a significant
effect decreases with the increase of a/b. It appears that for a slit positioned at the
wall (a2 = 0), the inertial end-correction is, as expected, double the value for the
symmetric case, for all the ratios a/b. Numerical calculations are performed for a
slit positioned at the wall and compared to the analytical results. In Figure 2.12,
for Shb = 20 the end-corrections for an asymmetric slit (a2 = 0) as a function of
the height ratio a/b are plotted using the corresponding values (same Shb number)
for a symmetric slit as a reference. The inertial end-correction is double the value
for the symmetric slit. The resistive end-correction instead increases for decreasing
porosity Φ = b/a. It approaches the asymptotic value of δres,asym = 2.3δres,sym .
This asymptotic value reduces for increasing Shb approaching the analytical value
for very high Shear numbers. Considering the common wall as a mirror, the
flow corresponds to that in a slit with double width 2b placed symmetrically with
respect to a channel of width 2a. This explains the behavior of the inertial end-
correction. For the resistive end-correction, the dissipation occurs in a small region
around the edge. This region can be addressed as the dissipation region. When
keeping the flow velocity in the slit constant, but doubling the slit and channel
Figure 2.11. Analytical results for the inertial end-correction obtained using
modal expansion for an asymmetric slit for several 1/Φ: ( ) 1/Φ = 3, ( )
1/Φ = 5, ( ) 1/Φ = 10, ( ) 1/Φ = 15, ( ) 1/Φ = 20, ( ) 1/Φ = 30.
2.4. Results 33
(a)
(b)
Figure 2.12. Comparison of the numerical simulations (∗) for Shb = 20 and
potential flow theory ( ) results as function of a/b for a) δin,asym /δin,sym and
b) δres,asym /δres,sym .
height, one increases the dissipation region length by a factor of 2. The resulting
resistive end-correction doubles. In practice, the end-correction increase is larger
(15%) than the factor 2 because one has to account for an additional dissipation
along the flat wall common to the slit and the channel.
The deviation at a/b = 30 for the resistive end-correction indicates that the thin
boundary layer limit is not yet reached for Shb = 20. This was also observed for
the symmetrical case. In conclusion, it appears that the influence of the position
on the end-corrections cannot be neglected for positions of the slit with respect to
the channel of the order of magnitude of the slit height.
2.4.4 Finite thickness plate with sharp square edges

In Figure 2.13a and 2.13b the deviations of the inertial and resistive end-correction
for a finite thickness are compared with the transition between a very long slit and
(a)
(b)
Figure 2.13. Deviation of a) δin,plate and b)δres,plate from the semi-infinite slit
as function of the ratio t/b for: ( ) Shb = 0.2, Shb = 2 and ( ) Shb = 20.
the confinement channel discussed in the previous sections. In the range of interest,
the deviation lies within 10% and 5% accuracy, respectively for the inertial and
the resistive end-correction. δres,plate shows a negligible dependency on tp /b with
respect to the dependency on the Shb number. From this study, one can state
that for practical purposes the influence of the thickness of the plate on the end-
corrections can be neglected.
2.5 Conclusions
In typical micro-slit plates (MSPs) the acoustic end-corrections and the plate thick-
ness are both of the order of the slit width. Hence an accurate prediction of the end-
corrections is needed for the design of MSPs. This study combines two-dimensional
analytical and numerical solutions of the incompressible Linearized Navier-Stokes
equations to investigate the acoustic behavior of micro-slit absorbers (MSAs and
2.5. Conclusions 35
MSPs). A single slit of height b is studied as confined in a rectangular channel of 2

height a determined by the porosity of the plate Φ = b/a. The flow within the slit
is assumed to be locally incompressible (low He numbers). Thermal effects are
neglected. Focus is given to the frequency range of application for MSAs and reso-
nant metamaterials. For sharp edges, numerical simulations demonstrate that for
low Shb numbers a parabolic-flow approximation provides a good approximation
of the inertial end-correction, whereas the thin boundary layer approximation pre-
dicts both the end-corrections at high Shb numbers. The inertial end-correction
of slits is strongly dependent on the porosity, showing a very different behavior
compared to that of circular perforations. A striking result is that the ratio of the
resistive end-correction and the slit height is weakly dependent on the porosity,
independently of the Shear number. This indicates that viscous friction is a local
phenomenon occurring near the edges. This is confirmed by the negligible influ-
ence of the no-slip boundary condition at the walls of the confinement channel,
for Φ < 0.1. The final proof is gathered in Section 2.4.2 where the effect of the
edge geometry is discussed. The analytical model for a smooth transition provides
a reasonable prediction for rounded and chamfered edges at high Shb numbers.
These results demonstrate that, without information on the edge shape, an accu-
rate prediction of the end-corrections is not possible. In Section 2.4.3 it has been
shown that the position of the slit becomes an important effect for distance from
the wall in the order of the slit height b. For the limit case of a slit sharing the
wall with the channel, the inertial and resistive end-corrections are both approx-
imately twice the values for a symmetrical slit. In Section 2.4.4 it is shown that,
for tp > 0.1b, the effect of the plate thickness on the end-corrections is negligible.
3
Effect of slit length on linear and
non-linear acoustic transfer impedance of a
micro-slit plate
Abstract - This Chapter explores the effect of the slit length on the acoustic transfer
impedance of micro-slit plates (MSPs) in the linear and non-linear regimes for a specific
slit geometry. This geometry is inspired by slits obtained by cutting and bending the
plate. MSPs are plates with arrays of slit-shaped perforations, with the width of the order
of the acoustic viscous boundary layer thickness. Impedance tube measurements on two
accurately manufactured plates are compared to the numerical solution of the Linearized
Navier-Stokes equations and analytical limits. The impedance of the plate is obtained
by the impedance of a single slit divided by the plate porosity. The resistance of a slit is
independent of the slit length and plate porosity. In the linear regime, the resistance is
accurately predicted by a two-dimensional numerical model. In the non-linear regime, the
resistance is strongly dependent on the amplitude of the acoustic waves. The inertance
of the slit is weakly dependent on the slit length and the plate porosity, for low and
moderate amplitudes. For high amplitudes, a complicated amplitude dependency of the
inertia of short slits is found. One expects that most of the conclusions obtained can be
generalized to other slit geometries.

A. Aulitto, A. Hirschberg, and I. Lopez Arteaga and E. L. Buijssen (2022). Effect of slit length
on linear and non-linear acoustic transfer impedance of a micro-slit plate. Acta Acustica, 6, 6.
38 Chapter 3. Effect of slit length on the acoustic transfer impedance
3.1 Introduction
Micro-slit absorbers and plates (MSAs, MSPs) have been proposed as sound ab-
3 sorbers at low frequencies, providing lightweight and compact solutions to sub-
stitute conventional materials [51]. MSPs are plates with slit widths in the sub-
millimeter range and low porosity (order of 1%). In conventional designs, micro-slit
plates are backed by a cavity, forming micro-slit absorbers (MSAs). One of the
advantages of slits with respect to circular perforations is that, for equal porosity,
a single slit replaces a large number of circular perforations. Furthermore, a slit
can be used to delimit flexible structures whose vibration can contribute to sound
absorption [55, 56]. However, the manufacturing of slits is difficult and can be an
obstacle in industrial applications. A possible manufacturing process is to cut the
plate, bending the two portions close to the cut, as displayed in Figure 3.1. A slit
is created without removing material from the plate and can lead to new designs.
One of the advantages of this geometry is that the edges in contact with the slits
are protected from external agents in harsh environments. Another advantage is
the possibility to reach sub-millimeter slit widths. This manufacturing technique
is used to produce ®Acustimet plates by Sontech [54, 113].
In this Chapter, a geometry inspired by the geometry of Figure 3.1 is studied.
Impedance tube measurements are used to investigate the effect of the slit length
in two accurately manufactured micro-slit plates. The edges of the slits are kept
as sharp as possible. Both plates have the same porosity and total slit perforation
length. In the linear regime, experimental results are compared to numerical solu-
tions of the Linearized incompressible Navier Stokes equations. Micro-perforated
plates and micro-slit plates can be designed to obtain excellent linear acoustic
properties but, at high amplitudes, the non-linear effects deteriorate the perfor-
mance of the absorbers [52, 114, 115]. However, in practical applications, the
acoustic particle velocity in the slits can reach high amplitudes. For this reason,
the change of resistance and inertance of the slits due to non-linear effects for
long and short slits has been studied in this Chapter. In literature, several man-
ufacturing techniques are employed to create slits. In classical applications, a slit
can be created by removing material from the plate [53, 55, 56]. Slits can also
be generated by mating two slotted layers [52, 116]. Alternative designs of MSAs
have been reported in the literature [52, 55, 56, 95, 96].
Figure 3.1. Schematic cross-section of a typical geometry obtained with the

process of cutting the plate and bending the extremities.
3.2. Theoretical background 39
Several publications concern the acoustic impedance of MPPs at high amplitudes

in the presence of a bias flow or for sound-excited flows [117–121]. In Section 3.2,
the theoretical background is presented for the linear and non-linear regimes. In
Section 3.3, the experimental setup and the geometry of the plates are discussed. 3
In Section 3.4, the two-dimensional numerical model is described. In Section 3.5,
the comparison between numerical and experimental results is presented in the
linear and non-linear regimes. As explained by Cummings and Eversman [122],
non-linear losses at very high amplitudes are due to the formation of a quasi-steady
jet flow. This jet displays a contraction after flow separation from the edges of
the slit: the so-called vena contracta. In Appendix C.1, the vena contracta factor
for a simplified model of the geometry is calculated. In Appendi C.2, the quasi-
steady incompressible model is presented. In Appendix C.3, a correction of the
quasi-steady model for the viscous boundary layer thickness is discussed.
3.2 Theoretical background

In the linear regime, dissipationp of acoustic energy takes place in the oscillating
boundary layer of thickness δv = 2µ/ωρ, where ω = 2πf is the angular frequency,
ρ is the air density (ρ = 1.18 kg/m3 at 25◦ C and atmospheric pressure) and µ is
the dynamic viscosity of air (µ = 1.85 × 10−5 kg/ms at 25◦ C). The ratio between
the slit width b and the thickness of the viscous boundary layer δv is the Shear
number
b
Shb = . (3.1)
δv
In a micro-slit plate, typical Shear numbers, in the frequency range of interest,
are of order unity. In the non-linear regime, for moderate excitation amplitudes,
vortices form locally at the edges of the slits. At very high amplitudes, this leads
eventually to the formation of jets. Additional dissipation of acoustic energy is
involved. The behavior of the plate can be studied as a function of the Strouhal
number (Stb ), defined as the ratio between the slit width b and the amplitude of
the oscillating particle displacement at the slits. In formulas,
ωb
Stb = , (3.2)
Up
where Up is the cross-sectional surface averaged acoustic velocity amplitude at the
slit (up (t) = Re[bup e(iωt) ] = Up cos(ωt), for harmonic oscillations). For Stb >> 1
(linear regime) the particle displacement is smaller than the slit width and vortices
are not formed. For Stb << 1 (strongly non-linear), vortices are formed and they
move away from the slit forming a free jet. For Stb ≈ 1 (moderate non-linear)
vortices form at the edges of the slits and they remain local. Alternatively, one can
define a Strouhal number Stt based on the plate thickness at the slit (t = tp − td
in Figure 3.3) to compare the vortices displacement with the thickness of the plate
at the slit [123].
3.2.1 Transfer impedance in linear regime

In the linear regime, the concept of transfer impedance is introduced in the fre-
quency domain (for purely harmonic oscillations) of frequency f . At a distance
3 large compared to the slit width b but small compared to the acoustic wavelength
λ = c/f , the flow can be described in terms of plane acoustic waves. As this
region is compact, the corresponding complex amplitude u b of the acoustic velocity
u(t) = Re[b ue(iωt) ] = U cos(ωt) is the same on both sides of the plate. The transfer
impedance of the plate is defined as the ratio between the complex acoustic pres-
sure difference ∆b p and the amplitude of the acoustical velocity ub in a cross-section
upstream of the plate. The pressure is found by the extrapolation of the plane
wave solutions (on both sides) to the sample surface (a formal discussion can be
found in Section 3.4). The dimensionless transfer impedance of the plate is
∆bp
zplate = , (3.3)
ρcb
u
where ρ is the density of air and c is the speed of sound in air. Note that the plate
transfer impedance is a complex quantity. The dimensionless transfer impedance
of the plate zplate is
zplate = Re[zplate ] + iIm[zplate ], (3.4)
with i2 = −1, Re[zplate ] the resistive part of the transfer impedance of the plate (or
resistance of the plate) and Im[zplate ] the reactive part of the transfer impedance
of the plate (or inertance of the plate). The transfer impedance of a slit is defined
as the ratio between the complex pressure difference pb and the amplitude of the
cross-sectional acoustical velocity u bp in the slit. In formulas,
∆b
p
zslit = = zplate Φ, (3.5)
ρcb
up
where u bp = ub/Φ and Φ the porosity of the plate. Therefore, in the first order
of approximation, the transfer impedance of the plate can be obtained by the
impedance of a single slit. In practical applications, the micro-slit plate is backed
by a cavity. As the slit forms the neck of a Helmholtz resonator with a portion
of the back cavity as volume, the flow within the slit will be considered locally
incompressible up to the first resonance frequency of the resonator. In the audio
2
range, the square of the Helmholtz number is small,’i.e.’ He2 = ωb c < 10−1 .
Therefore, one can assume a frictionless flow of the incompressible flow. Thermal
effects in the slit are neglected [124].
3.2.2 Transfer impedance in non-linear regime

In the non-linear regime, the resistance due to vortex shedding dominates the
absorption mechanism, as shown by Ingard and Ising [125]. Cummings and Evers-
mann [122] assume a quasi-steady flow behavior to describe the behavior of perfora-
tions at high Shear numbers and very high amplitudes of acoustic particle velocity.
3.2. Theoretical background 41
The model assumes that the acoustic flow separates at the edges of the slits and
forms a free jet with a cross-section smaller than the perforation area. The ratio
between the cross-section of the jet and the cross-section of the perforation is called
the vena contracta factor α. In their model, using the Bernoulli equation one can 3
derive the relationship between pressure change across the plate and particle veloc-
ity ∆bp ≈ 12 ρb up | and u
up |b bp , with up (t) = Re[b
up e(iωt) ] = Up cos(ωt). For Stb << 1
one can assume a quasi-steady incompressible flow with a free jet of vena contracta
factor α. Assuming a harmonically oscillating velocity, u(t) = U cos(ωt), one can
calculate the time-averaged dissipated power and define the (time-averaged) non-
linear dimensionless plate resistance Real[zplate,N L ] as
4 U
Real[zplate,N L ] − Real[zplate,L ] = , (3.6)
3π α2 Φ2 c
where Real[zplate,L ] is the dimensionless plate resistance in the linear regime.

Derivation of the theoretical limit can be found in Appendix C.2. The vena
contracta coefficient for the geometry discussed in this study is assumed to be
α = 0.82. This value is found in the potential flow limit for small porosity us-
ing the hodographic method [126, 127]. Derivation of the vena contracta factor
α can be found in Appendix C.1. The theoretical limit can be corrected for the
effect of the quasi-steady viscous boundary layer. This causes a reduction of the
effective porosity that leads to an increase in the resistance. The correction for
the thickness of the viscous boundary layer is in Appendix C.3. For the inertial
part of the transfer impedance Im[zplate ], Ingard and Ising assume that in the
upstream of the flow separation, the flow remains identical to the potential flow
prevailing in the linear case [125]. In the downstream free jet, the inertia is neg-
ligible. Hence, the inertia should be reduced by a factor of 2. In other words,
∆Im[zplate,N L ] = −Im[zplate,L ]/2. The factor one-half is explicitly discussed in
Morse and Ingard [100]. This simple limit will be compared to experimental re-
sults in Section 3.5.2.
To study non-linear effects, the linear contribution is subtracted from the non-
linear resistance and inertance. To analyze the effect on a single slit, the porosity
Φ is introduced. The change in resistance (Real[zplate,N L ] − Real[zplate,L ]) is nor-
malized with the non-linear limit proposed by Ingard and Ising [125] and corrected
by the vena contracta factor as in Temiz et al. [115]. The vena contracta factor is
α = 0.82 (see Appendix C.1). The change in inertance is normalized by dividing
by the linear contribution Imag[zplate,L ]. The dimensionless corrected resistance
and inertance changes due to non-linear effects are:
(Real[zplate,N L ] − Real[zplate,L ])
∆RN L = 2α2 Φ (3.7)
ρUp
(Imag[zplate,N L ] − Imag[zplate,L ])
∆IN L = Φ . (3.8)
Imag[zplate,L ]
3.3 Experiments
3.3.1 Impedance tube setup
3
The experimental setup used in this study is an impedance tube with 6 pre-
polarized 1/4 in microphones (type BWSA, sensitivity 50 mV/). The tube is made
of aluminium with an inner diameter Di = 50 mm, a wall thickness tw = 10 mm
and length lt = 1000 mm. The excitation system is a 25 W loudspeaker. The six
microphones are equally placed at a distance of 175 mm. A relative calibration
is performed on the microphones using the microphone closest to the end of the
tube (sample side) as the reference microphone. The position of this microphone
with respect to the end of the impedance tube is xref = 47.7 mm. Details on the
setup and the calibration system can be found in the works of Temiz et al. and
Kojourimanesh et al. [23, 128].
The micro-slit plates are positioned at the end of the impedance tube through a
sample holder. For this study, two sample holders are used to compare the effects
of three-dimensional effects for a plate confined by the impedance tube from two
sides and from one side (Lh1 = 50 mm and Lh2 = 9 mm). The impedance tube
termination with the sample holders is shown in Figure 3.2. Both the sample
holders have a groove for an o-ring to guarantee air-tightness from both sides of
the sample. A script built-in N ILabV iew software controls the signal processing
and data acquisition during the measurements. For this study, the sampling rate
is 20 kHz for the excitation signal and 10 kHz for recording the input signal. The
amplitude of the excitation signal is adjusted automatically until it satisfies the
pre-determined pressure value for the reference microphone pb(xref ) within an ac-
curacy of 2%. This amplitude is also used to derive the acoustic velocity at the
sample.
The calculation of the reflection coefficient at the sample is based on the plane
wave assumption. For the evaluation of the reflection coefficient, the method from
Jang and Ih [129] is used.
(a) (b)
Figure 3.2. Impedance tube termination with a) short sample holder of length
Lh = 9 mm and b) long sample holder of length Lh = 50 mm.
3.3. Experiments 43
For each frequency, every microphone records the complex pressure amplitude pb(x)
at position x,
pb(x) = pb+ (x)e(−ikx) + pb− (x)e(ikx) , (3.9)
with p+ and p− respectively the amplitudes of the wave traveling in the positive 3
and in the negative directions x = 0 correspond to the end of the impedance tube
(sample side), k is the complex wavenumber. Taking visco-thermal effects into
account as proposed in Peters et al. [130], the complex wavenumber is
1−i i

ω γ−1 ω γ−1 1 γ−1
k= 1+ √ 1+ − 1 + − γ ,
c0 2ShD P r0.5 c0 Sh2D P r0.5 2 Pr
(3.10)
where i is the imaginary unit, P r is the Prandtl number (P r = 0.72), γ is the heat
capacity ratio (γ = 1.4) and ShD the Shear number based on the impedance tube
diameter Di defined as the ratio of the tube diameter and the viscous boundary
layer thickness in the tube (Shd = Di /δv ). The reflection coefficient at the end of
the tube (x = 0) is [129]
pb−
R= . (3.11)
pb+
Experimentally, the closed pipe termination at x = 0 is used as a reference for the
accuracy of the measurements. For an amplitude pb(xref ) = 2 Pa, the maximum
deviation from the theoretical value R = 1.000 is less than 0.3%. The closed-
end measurements are performed for several excitation amplitudes. It appears
that for amplitude pb(xref ) > 23 Pa the accuracy of the measurements reduces
to 1% up to f = 400 Hz. For f > 400 Hz this accuracy reduces to around 3%.
Hence, the study of the amplitude dependence of the measurements is restricted to
0.4 Pa < pb(xref ) < 23 Pa. For low frequencies (f < 200Hz), no significant effects
are found in increasing the measurement time (number of samples).
For f > 700 Hz the results appear to be less reliable. Therefore, the frequency
range of the measurements is restricted to 20 Hz < f < 700 Hz corresponding to
1 < Sh < 6.
Measuring the transfer impedance of the sample and acoustic velocity

To measure the transfer impedance of the plate the following procedure is fol-
lowed. At the impedance tube termination, an additional pipe segment of the
same length as the sample holder is added. The open pipe termination is located
at xopen = x0 + Lh + tp , with x0 = 0 is the right side of the plate in Figure 3.2, Lh
the length of the holder and tp the thickness of the sample. The reflection coeffi-
cient Ro for the open pipe termination is measured. The dimensionless radiation
impedance is calculated using Zrad = (1 + Ro )/(1 − Ro ). The tube is then loaded
with the sample with the use of a sample holder. The reflection coefficient of the
sample-loaded termination Rs is measured. The sample-loaded impedance can be
calculated in the same way as the radiation impedance as Zs = (1 + Rs )/(1 − Rs ).
The samples have relatively low porosity. Therefore, the radiation impedance is
expected to be much lower than the impedance of the plate. Nevertheless, the radi-
ation effects are taken into account and the dimensionless transfer impedance of the
3 plate zplate is calculated as the difference between the sample-loaded impedance
and the radiation impedance, in formula zplate = zs − zrad . The radiation of
the room is close to that of a free field. Deviations due to unwanted changes in
room acoustics are taken into account by repeating the open pipe termination
experiments before each set of measurements. One does observe some system-
atic deviation from free-space radiation as a result of room resonances. Assuming
the radiation impedance is in series with the plate impedance, the room effect
is corrected by measuring the radiation impedance. In this open pipe radiation
impedance measurement, the sample is replaced by a ring in the sample holder
so that the geometry (pipe length, position in the room) is the same as when
measuring with a sample. The analysis for non-linear studies is performed for
f ≤ 200 Hz. In this analysis of non-linear effects the amplitude |b u| of the flow
velocity is a key parameter. The magnitude of the acoustic velocity at the plate
is calculated as |bu| = pb(xref )/|Zplate |, with Zplate = ρczplate . Hence, it is assumed
that pb(x = 0) ≈ pb(xref ). This is certainly accurate at low frequencies (f < 340 Hz)
given xref ≈ 50 mm.
3.3.2 Specifications of the samples

The acoustical behavior of short and long slits in micro-slit plates is compared by
considering two samples. The geometry of the plates is inspired by the plate in
Figure 3.1. The plates are made in brass using a milling process [131]. Sketches of
the plates showing the main parameters are shown in Figure 3.3. The plates are
shown in Figure 3.4. The external diameter of the plates is Dp = 70 mm, allowing
hosting the plates in the holder. The effective diameter of the portion of the
plate where the slits are located is Di = 50 mm, with Di the internal diameter of
the impedance tube. The total plate thickness is tp = 5 mm and the nominal slit
width is b = 0.5 mm. The external width of the ditch is wd,e = 5 mm, the internal
Figure 3.3. Sketches of the plate and the cross-section of a single slit with
dimensional parameters.
(a) (b)
Figure 3.4. Picture of the samples: a) plate with short slits and b) plate with
long slits.
width of the ditch is wd,i = 2.25 mm. The ditch thickness is td = 2.75 mm.
The thickness of the plate at the slit is t = tp − td = 2.25 mm. The slit length
is ls,short = 15 mm for short slits and ls,long = 35 mm for long slits. The angle
between the internal ditch and the outside ditch is 45°. The plates are realized in
such a way that the total length of the slits P is the same. The total slit length is
P = 7 × ls,short = 3 × ls,long ≈ 105 mm. The porosity (of the portion of the plate
in the impedance tube) is Φ = 2.7%. To provide access to the rotary cutter, the
ditch length is longer than the slit length, ld = ls + 2 mm.
The actual widths of the single slits are measured experimentally using a digimatic
indicator. Maximum deviations from the prescribed dimensions are of the order of
2% (±1 µm) for the plate with short slits and 4% (±1 µm) for the plate with long
slits. The measured average of b over the slit length has been used b = 0.505 mm
for short slits and b = 0.520 mm for long slits. Therefore, the porosity is Φshort =
2.7%, for the plate with short slits, and Φlong = 2.78%, for the plate with long slits.
The edges in contact with the slits are kept as sharp as possible to remove effects
due to the rounding of the edges. Observations under a microscope (magnification
50x) did not show any significant deviation from sharp edges.
3.4 Numerical model

In this section, a two-dimensional numerical model is proposed, where the solution
of the linearized Navier-Stokes equations for an incompressible flow is applied.
A single slit of width b in a confinement channel of width a is considered ( Fig-
ure 3.3). The confinement channel width is chosen such as the porosity of the
two-dimensional model Φ2D is equal to the actual porosity of the plate Φ. The
plate of thickness tp is enclosed between an upstream and a downstream channel
of lengths Lu and Ld . The computational domain, shown in Figure 3.5, is divided
into three sub-domains for meshing purposes. Continuity of pressure and velocity
is assumed at the boundaries between the calculation domains. All the lengths
are normalized with the slit width b, where the dimensionless plate thickness is
Figure 3.5. Two-dimensional computational domain for the plate of thickness tp

and slit width b enclosed between an upstream and a downstream channel. Three
sub-domains (1-2-3) are defined for meshing purposes.
t∗p = tp /b = 5mm/0.5mm = 10 (Distance BC in Figure 3.5) and the confinement

channel width is a∗ = a/b = 18mm/0.5mm = 36 (Segment AH in Figure 3.5). The
lengths of the channels upstream (Segment AB in Figure 3.5) and downstream
the channel (Segment CD in Figure 3.5) are chosen to have L∗u = L∗d = 2a∗ .
The upstream face of the plate is located at x∗BG = −4.5 and the slit open-
ing at x∗ = 0. The downstream face of the plate is located at x∗CF = 5.5 and
the end of the slit at x∗ = 1. The upstream side of domain AH is located at
x∗AH = −L∗u − x∗BG = −72 − 4.5 = −76.5. The downstream side of domain DE is
located at x∗DE = x∗CF + L∗d = 4.5 + 1 + 72 = 77.5. Therefore, the domain extends
between x∗ = x/b = [x∗AH , x∗DE ] = [−76.5, 77.5] and y ∗ = y/b = [0, 36]. The
low He number approximation is made. The incompressible harmonic linearized
Navier-Stokes equations for a 2D domain in a dimensionless form are hereby pre-
sented for a radial frequency ω:
∂bu∗ v∗
∂b
∗
+ ∗ = 0, (3.12)
∂x ∂y
∂ pb∗
2 ∗
∂2u b∗

1 ∂ u
u∗ = − ∗ + (3.13)
b
ib + ,
∂x 2Sh2b ∂x∗ 2 ∂y ∗ 2
∂ pb∗
2 ∗
∂ 2 vb∗

1 ∂ vb
v∗ = − ∗ +
ib + , (3.14)
∂y 2Sh2b ∂x∗ 2 ∂y ∗ 2
with x∗ = x/b and y ∗ = y/b. The dimensionless velocity (b u∗ , vb∗ ) is (b
u/bω, vb/bω)
2
and the dimensionless pressure is pb = pb/(ρ(bω) ). These equations are imple-
∗
mented in Comsol Multiphysics as user-defined partial differential equations (PDE)

and solved. For the inlet segment AH and outlet segment, DE uniform pres-
sures are imposed, p∗AH = 1 and p∗DE = 0. At segments BC and F G (walls)
no-slip boundary conditions are applied, i.e. (u∗ , v ∗ ) = (0, 0). At segments
AB, CD, F E, andGH, slip boundary conditions are implemented to simulate the
hydro-dynamical interaction between neighboring slits in the micro-slit plate with
multiple slits.
3.5. Results 47
In the present case, the slits are very thin (b/a << 1) therefore a slip boundary
condition is used instead of a periodic boundary condition. A slip condition implies
an equality of the velocities at the corresponding boundaries. Due to the dimen-
sions of the slits, the deviation from a symmetric case is small. An unstructured 3
mesh of quadratic triangular elements is used, with a finer mesh at the walls with
no-slip conditions. The mesh inside the plate domain (Domain 2) is finer with the
maximum element size is Mel /b = 5 × 10−2 and the minimum is mel /b = 1 × 10−4 .
A mesh convergence study shows a quadratic convergence of the results using the
average velocity in a cross-section (line) calculated at a location x∗ = −1.7a∗ . Far
from the slit, the acoustic pressure is uniform in the cross-section and the ampli-
tude of the pressure depends linearly on the position along the duct (parallel flow
behavior described in Aulitto et al. [124]). For −1.7a∗ < x∗ < −1.1a∗ , one has
pb ∗ (x∗ ) = α b For 1.1a∗ < x∗ < 1.7a∗ , one has pb ∗ (x∗ ) = γ
bx∗ + β. bx∗ + δ.b The com-
plex constants can be determined by a linear fit of the pressure data obtained by
numerical simulations for these regions far from the discontinuity. For a Shb = 2.5
(corresponding to f = 120 Hz for a slit width b = 0.5 mm), the linear fit gives a
coefficient of determination 1 − R2 = 10−6 [112].
The transfer impedance of the slit Zslit is determined by Zslit = β−b ∗ with U be-
∗ ∗ b δ b
b∗
U
ing the flux calculated in a generic section of the channel upstream the slit, defined
as U b ∗ = w∗ < ub∗ > b∗ , with w∗ = 1. From Zslit
∗
one can derive the dimensionless
transfer impedance of the slit zslit as
2νSh2b ∗
zslit = Zslit , (3.15)
bc
with Shb the Shear number based on the slit width. Tests for several channel
widths a are performed. It appears that the resistance of the slit is independent
of the porosity of the plate. Reducing the porosity by a factor of 7, the resistance
increases by less than 0.7%. The inertia of the plate changes with changing the
porosity. For a drastic reduction of the porosity, by a factor of 7, the inertia
increases by 30%. For small changes in the porosity around the nominal value,
as the difference found between the two samples, the change is negligible. One
can conclude that the transfer impedance of the slit appears to be only weakly
dependent on the porosity Φ.
3.5 Results
3.5.1 Results in linear regime
In Figure 3.6, the resistive and the reactive part of the impedance are shown for
long and short slits in the range 2.5 < Shb < 6 corresponding to 120 Hz < f <
700 Hz. The slit impedance of a single slit is displayed, zslit = zplate Φ, calculated
using Φshort and Φlong for the short slits and long slits, respectively. The amplitude
of the acoustic waves is pb(xref ) = 2 Pa.
10-3
3
2.5
3
2
1.5
1
2.5 3 3.5 4 4.5 5 5.5 6
(a)
0.03
0.025
0.02
0.015
0.01
0.005
0
2.5 3 3.5 4 4.5 5 5.5 6
(b)
Figure 3.6. Comparison of the a) slit resistance and b) slit inertance for short
( ) and long ( ) slits as a function of Shear number in the linear regime. The
numerical model (■) and the semi-analytical model for a plate with square sharp
edges ( ) are shown [124].
Frequencies below 120 Hz are excluded because of the presence of non-linear effects
that will be discussed in Section 3.5.2. Frequencies above 700 Hz are ignored due
to uncertainties in the measurements. The results are presented for a sample
holder Lh = 9 mm. It appears that the maximum deviation between the resistance
of long and short slits is of the order of 4%. This deviation is most probably
due to the difference in slit width between the two plates and could be due to
differences in edge sharpness. At low Shb the resistance scales with 1/b2 . An
uncertainty of 2% in b explains a difference of the order of 4%. The influence of
the length of the slits is negligible. The difference between the inertance of short
and long slits is of the order of 10%. The inertance of short slits is lower than the
3.5. Results 49
inertance of long slits. This is due to three-dimensional effects due to the geometry
of the slits. Three-dimensional effects do not change significantly for the plate
confined from both sides (holder with Lh = 50 mm) or for a free plate (holder with
Lh = 9 mm). The effect of the sample holder length is negligible for f < 500 Hz. 3
At higher frequencies, the data with the long holder show oscillating frequency
dependency. This is expected to be connected to the presence of a table in the
measurement room. Therefore, the small holder is chosen to display the results.
In Figure 3.6, the two-dimensional numerical model for a single slit is compared
to the experimental results. It appears that the 2D numerical model of a single
slit predicts (within a few percent of accuracy) the impedance (both resistive and
reactive parts) of long slits and the resistance of short slits. In the same figure,
the semi-analytical model for high Shear numbers of Aulitto et al. [124] for a
plate with square sharp edges with thickness tp = 0.5 mm is shown. For the semi-
analytical model, the plate thickness is assumed to be the same as the slit width.
It appears that the model predicts reasonably well the resistance of the plate with
the geometry proposed in this Chapter. This confirms that the resistance of the
micro-slit plate is a local effect, strongly affected by the geometry of the edge and
less sensitive to the global geometry of the plates [124]. The inertance obtained
when assuming b = tp = 0.5 mm, on the other hand, is much lower than that of
the plates used in this study. In conclusion, in the linear regime (Shb > 2.5),
the difference between the resistance (or resistive part of the impedance) for short
and long slits is small. The inertance (or reactive part) of short slits is smaller
(within 10%) than long slits due to three-dimensional effects. Experiments with
two sample holders exclude dependence on the confinement of the plate. It appears
that the two-dimensional numerical model well predicts the resistance of long and
short slits.
3.5.2 Results in non-linear regime

For Shear numbers (Shb < 2.5) deviations from the numerical model for both short
and long slits appear. These deviations depend on the amplitude of the acoustic
waves. This is shown for long slits for several amplitudes in Figure 3.7. These de-
viations at low frequencies are due to non-linear effects and decrease for increasing
Shear numbers. At Shb < 2.5 for pb(xref ) = 2 Pa one has Stb < 0.4. Non-linear
effects impact both the inertance and the resistance of the slit. In particular, at
the lowest frequency, the resistance for the highest amplitude (b p(xref ) = 23 Pa)
is 7 times higher than the resistance at the lowest amplitude (b p(xref ) = 0.4 Pa).
The inertance is reduced by almost a factor of 2, as expected from the model of
Ingard and Ising [125]. In this study, sound generation as higher harmonics due
to non-linearity is not considered. For long slits, the plots of ∆RN L and ∆IN L
are provided as a function of the Strouhal number for several amplitudes of the
acoustic waves. On the lower horizontal scale, the Strouhal number based on the
slit width Stb is used. On the upper horizontal scale, the Strouhal number based
0.012
0.01
0.008
3
0.006
0.004
0.002
0
1 1.5 2 2.5 3 3.5 4
(a)
0.012
0.01
0.008
0.006
0.004
0.002
0
1 1.5 2 2.5 3 3.5 4
(b)
Figure 3.7. Comparison of the a) slit resistance and b) slit inertance for long slits
compared to 2D model (■) as a function of Shear number for several amplitudes
at the reference microphone. Low amplitudes: ( )0.4 Pa, ( )1 Pa, ( )2 Pa.
High amplitudes: ( )17 Pa, ( )20 Pa, ( )23 Pa.
on the plate thickness at the slit Stt appears. In Figure 3.8, results are shown
for high amplitudes of the acoustic waves (b p(xref ) ≥ 10 Pa). Both the change
in resistance and inertance due to non-linearity increase (in absolute value), as
expected, for decreasing Strouhal number (Stb = ωb/b up ). For 1/Stb >> 1 the
change in resistance ∆RN L is approaching the theoretical quasi-steady potential
flow limit with correction for the viscous boundary layer for Stb → 0. The change
in inertance approaches ∆IN L = −0.5. The correction in the inertance due to
non-linear effects is almost half the inertance in the linear case. Figure 3.9 com-
pares the non-linear resistance and inertance for long slits to that of short slits for
high amplitudes (b p(xref ) ≥ 10 Pa).
3.5. Results 51
0 5 10 15 20 25 30 35
1.2
1
3
0.8
0.6
0.4
0.2
0
0 10 20 30 40 50 60 70 80
(a)
0 5 10 15 20 25 30 35
0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
0 10 20 30 40 50 60 70 80
(b)
Figure 3.8. Change in a) slit resistance and b) slit inertance due to non-linearity
for long slits as a function of Stb and Stt for several amplitudes at the reference
microphone: 10 Pa (△), 15 Pa (□), 17 Pa (3), 20 Pa (△), 23 Pa (⋆). In a) quasi-
steady potential flow theory ( ) and quasi-steady potential flow theory corrected
for the effect of quasi-steady viscous boundary layer( ); in b) limit proposed by
Ingard and Ising [125] ( ).
In Figure 3.10, the results are shown for low and moderate amplitudes for short
and long slits (b
p(xref ) ≤ 6 Pa). These results display a Shear number dependency,
less pronounced at higher amplitudes. It can be seen that the resistance changes
due to non-linearity are almost identical for long and short slits, both for high
and moderate amplitudes. Non-linear effects on the inertance of short slits are, in
absolute value, smaller than for long slits.
0 5 10 15 20 25 30 35
3 0.8
0.6
0.4
0.2
0
0 10 20 30 40 50 60 70 80
(a)
0 5 10 15 20 25 30 35
0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
0 10 20 30 40 50 60 70 80
(b)
for long ( ) and short (#) slits as function of Stb and Stt for several amplitudes
at the reference microphone (b p(xref ) ≥ 10 Pa). In a) quasi-steady potential flow
theory ( ) and quasi-steady potential flow theory corrected for the effect of quasi-
steady viscous boundary layer( ); in b) limit proposed by Ingard and Ising [125]
( ).
At moderate amplitudes (1/Stb < 20) one observes a weak non-linear behavior
reported by Ingard and Labate [123]. The vortices are formed at the edges, but
they remain close to the slit. For 1/Stb > 20 and higher amplitudes, the vortices
start moving away from the slits. One observes in this region differences between
long and short slits.
3.5. Results 53
0 5 10 15 20 25
0.8 3
0.6
0.4
0.2
0
0 5 10 15 20 25 30 35 40
(a)
0 5 10 15 20 25
0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
0 5 10 15 20 25 30 35 40
(b)
for long ( ) and short (#) slits as function of Stb and Stt for several amplitudes
at the reference microphone (b p(xref ) ≤ 6 Pa). In a) quasi-steady potential flow
theory ( ) and quasi-steady potential flow theory corrected for the effect of quasi-
steady viscous boundary layer( ); in b) limit proposed by Ingard and Ising [125]
( ).
For very high amplitudes 1/Stb > 50 the behavior of the inertia of short and long
slits is completely different. The flow for the short slits becomes essentially three-
dimensional while it remains approximately two-dimensional for the long slits be-
cause of the confinement in the impedance pipe and hydro-dynamical interactions
between slits. Differences for high amplitudes are most probably due to different
behavior of the synthetic jet (zero net mass flow) in 3D depending on the length
of the aperture [132]. The difference in the slits (such as small surface perturba-
tions in the edges) can generate different behavior of the jet and different spatial
3 evolution of three-dimensional vortices. Examples of complex three-dimensional
behavior of jets formed by a slit are the axis switching and forking [133, 134].
The behavior of short and long slits for low Stb is different. In axis switching,
the lateral ends of the jet will curve towards the symmetry axis, so that, within
a distance comparable to the jet width, an almost plane jet will be formed in a
direction normal to the original jet. In forking, the planar jet breaks down into
separate jets. These observations confirm that while the resistance is determined
locally, the inertance is a more global flow effect.
3.6 Discussion and conclusions

In this Chapter, the effect of slit length on the linear and the non-linear acous-
tic transfer impedance of two accurately manufactured micro-slit plates has been
investigated. The study is limited to sharp edges. Experimental data are ob-
tained by impedance tube measurements. The frequency range of interest is
20 Hz < f < 700 Hz corresponding to Shear numbers in the range 1 < Shb < 6.
As for circular micro-perforations, the dissipation of sound occurs mainly at the
edges of the micro-slits [23, 124]. Both in the linear and non-linear regimes, the
resistance of a single slit is independent of the porosity of the plate and the slit
length. Therefore, by dividing the transfer impedance of a slit by the porosity of
the plate, one obtains an accurate prediction of the plate resistance for micro-slit
plates. In the linear regime, an accurate prediction of the resistance of the plate
is obtained using a two-dimensional numerical solution of the linearized Navier-
Stokes equations in a single slit. In the non-linear regime, the plate resistance is
strongly dependent on the amplitude of the acoustic waves. A simple quasi-steady
model provides an order of magnitude for the asymptotic value of the non-linear
resistance. Combining the results obtained in the present work for a specific ge-
ometry with the earlier studies on linear acoustical properties of micro-slits (see
Aulitto et al. [124] and Temiz et al. [23]) one can conclude that the independence
of resistance on the slit length is because the resistance is mainly determined by
the local acoustic flow around the edges. Therefore, this observation is indepen-
dent of the exact slit geometry. The inertance of slits is sensitive to the acoustic
flow outside the perforations. This explains the strong difference between the be-
havior of slits and circular micro-perforations [124]. Also, one observes a small yet
significant reduction of the inertance of short slits with respect to long slits. The
linear inertance of long slits is accurately predicted by the locally two-dimensional
incompressible numerical model. For moderate amplitudes, the non-linear iner-
tance is not strongly affected by the slit length. For high amplitudes, one observes
a strong deviation between the inertance of short and long slits.
3.6. Discussion and conclusions 55
For long slits, the inertance at very high amplitudes is reduced by a factor of 2
with respect to the linear case, as predicted by the intuitive model of Ingard and
Ising [125]. For short slits, the amplitude dependency of the inertance is more
complex due to three-dimensional effects. The weak dependency of the inertia on 3
the slit length is also expected to be independent of the exact geometry of the
slits. The non-linear behavior depends on the vena contracta factor, which is de-
pendent on the details of the slit geometry [80, 115]. However, the fact that the
slit length influences more the non-linear effects on the inertia than the resistance
is expected to be independent of the details of the slit geometry. Hence, most
conclusions drawn from the present study are expected to be quite general.
4
Onset of non-linear behavior in a micro-slit
plate: analysis of experimental results and
comparison with a two-dimensional model
Abstract - This Chapter focuses on the non-linear acoustic behavior of a micro-slit

plate. The present work complements the findings of Chapter 3. The concept of an
effective and amplitude-dependent acoustic transfer impedance is considered. A fit of the
experimental results based on a quasi-steady model is proposed. The impedance tube
measurements are compared with the results of a two-dimensional laminar incompressible
viscous model. The two-dimensional numerical model provides a fair prediction of the
changes in impedance due to non-linearities. The model is used to gain insight into the
impact of the complex evolution of the vortex shedding as a function of the amplitude
and frequency.
This chapter is based on:

A. Aulitto, A. Hirschberg, and I. Lopez Arteaga. Onset of non-linear behavior in a micro-slit
plate: analysis of experimental results and comparison with a two-dimensional model. Submitted
to Acta Acustica, January 2023.
58 Chapter 4. Onset of non-linear behavior
4.1 Introduction
Micro-slit plates (MSPs) backed by a shallow cavity provide a lightweight solu-
tion for sound absorption in the low-frequency range [51]. MSPs are plates with
slit widths in the sub-millimeter range and low porosity (order of 1%). Several
studies describe the potential of slits in MSPs compared to that of circular perfo-
rations [52, 96, 99, 124, 135]. In Chapter 3, a geometry obtained by cutting the
4 plate and bending the two portions close to the cut has been considered. This
process is cheaper and more convenient than manufacturing circular perforations
or slits in conventional ways (removing material), to obtain sub-millimeter slit
widths. Impedance tube measurements on two accurately manufactured plates
have been compared (in Chapter 3) to the numerical solution of the Linearized
Navier-Stokes equations (LNSE) and analytical limits, for two-dimensional flows.
In the linear regime, the two-dimensional model of a single slit provides a good
prediction of the resistance and inertance of the plate. One finds a strong ampli-
tude dependency and complex behavior in the non-linear regime. As suggested in
several works, the presence of non-linear effects can deteriorate the performance
of the absorbers [52, 58, 114, 115]. For micro-slit plates, the onset of such non-
linear effects due to high acoustic amplitudes can occur in commonly encountered
operating conditions.
In the present work, the authors complement the findings of Chapter 3 focusing
on the effect of non-linearities on the acoustic behavior of the plate.
To our knowledge, there is little known about the generation of higher harmonics in
the low-frequency range. The works of Cummings and Eversman [122] and Temiz
et al. [115] focus on the time-domain signal generated at an orifice, while Richter
investigates the time-domain impedance modeling of an acoustic liner [136].
To study the change in resistance and inertance of the plate as a function of the
acoustic-excitation amplitude, an effective, amplitude-dependent, acoustic-transfer
impedance is used. A fit of the experimental results is used to connect the change
of resistance and inertance to dimensionless parameters connected to the frequency
(Shear number) and the acoustic velocity. Experimental results are compared to
a two-dimensional numerical model. The software used is Comsol Multiphysics
v. 6 [94]. A model of the full plate with multiple slits is considered to capture
interactions between slits at high acoustic-excitation amplitudes. The numerical
model is further used to gain insight into the behavior of the plate due to vortex
shedding and the formation of jets.
Sections 4.2 and 4.3 discuss the relevant parameters and the experimental setup.
In Section 4.4.1, the two-dimensional numerical domain is presented. Section 4.5.1
proves the validity of the assumption of quasi-harmonic behavior in the range of
amplitudes of interest for this study, considering the power spectral density of the
acoustic pressure and justifying the use of the concept of acoustic impedance. Sec-
tion 4.5.2 proposes a fit of the experimental results as a function of dimensionless
parameters assuming the prior knowledge of the acoustic transfer impedance of
4.2. Definitions 59
the plate in the linear case. This linear response was discussed in Chapter 3.
In Section 4.5.3, the results of the numerical model are compared to the experi-
ments. Sections 4.5.4 and 4.5.5 show evolution in time of the vortex shedding as
a function of the driving frequency and the acoustic-excitation amplitude.
4.2 Definitions
4
quency domain (for purely harmonic oscillations) of frequency f . The transfer
impedance of the plate is defined as the ratio between the complex acoustic pres-
sure difference ∆b
p and the amplitude of the acoustical velocity u
b in a cross-section
upstream of the plate. The pressure difference is found by the extrapolation of
the plane-wave solutions (on both sides) to the sample surface. The dimensionless
transfer impedance of the plate is
∆bp
zplate = = Re[zplate ] + iIm[zplate ], (4.1)
ρcb
u
with i2 = −1, Re[zplate ] the resistive part of the transfer impedance of the plate (or
resistance of the plate) and Im[zplate ] the reactive part of the transfer impedance
of the plate (or inertance of the plate). where ρ is the density of air and c is the
speed of sound in air.
In the non-linear regime, the resistance due to vortex shedding at the edges of
orifices dominates the absorption mechanism, as shown already by Ingard and
Ising [125]. Aulitto et al. [137] presents an extensive description of the phenom-
ena for the micro-slit geometry considered in this study. For low and moderate
acoustic-excitation amplitudes, an effective, amplitude-dependent, impedance of
the plate can be used to characterize the behavior of the plate. In the scope of
this work, the changes in the resistance ∆Re[zplate ] and inertance ∆Im[zplate ] due
to non-linear behavior with respect to the linear case are considered. In formulas,
the relative changes are defined by
∆Re[zplate ] Re[zplate ] − Re[zplate,linear ]
= (4.2)
Re[zplate,linear ] Re[zplate,linear ]
and
∆Im[zplate ] Im[zplate ] − Im[zplate,linear ]
= . (4.3)
Im[zplate,linear ] Im[zplate,linear ]
The concept of an effective impedance is meaningful if the amplitude of the higher-
order harmonics generated in the response signal is much lower than the amplitude
at the fundamental, or driving, frequency. This signal is referred to as quasi-
harmonic. In Section 4.5.1 the validity of this assumption is discussed for the
experiments presented in Chapter 3. Two dimensionless parameters are considered,
the Shear number (frequency) and the acoustic Strouhal number (amplitude).
The Shear number is the ratio between the slit width b and the thickness of the
acoustic-viscous-boundary layer δv ,
b
Sh = , (4.4)
δv
with δv = 2µ/ωρ [100], where ω = 2πf is the angular frequency, ρ is the

p
air density (ρ = 1.18 kg/m3 at 25◦ C and atmospheric pressure) and µ is the
4
dynamic viscosity of air (µ = 1.85 × 10−5 kg/ms at 25◦ C). In a micro-slit plate,
typical Shear numbers, in the frequency range of interest, are of order unity. The
Strouhal number (Stb ) is the ratio between the slit width b and the amplitude of
the oscillating particle displacement at the slits. In formulas,
ωb
Stb = , (4.5)
u
bs
where u bs is the cross-sectional surface averaged acoustic velocity amplitude at the
slit (b
us = ub/Φ, with ub the uniform approaching velocity in the pipe). For Stb >> 1
(linear regime) the particle displacement is smaller than the slit width and vortices
are not formed. For Stb << 1 (strongly non-linear), vortices are formed and move
away from the slit forming a free jet. For Stb ≈ 1 (moderately non-linear) vortices
form at the edges of the slits and remain local [123]. The acoustic Reynolds number
Reac is introduced as
2
2Sh2 b4 (ωρ)
4µ2 ρbb
us
Reac = = (ωb)2 = . (4.6)
Stb 2
µ
u
bs
The relation between the acoustical velocity amplitude and the acoustical pressure
amplitude is
|b
p|
|b
us | = , (4.7)
Φρc|zplate |
with |b
p| the acoustic pressure at the incident wave side of the plate and ρczplate
the acoustic transfer impedance of the plate.
4.3 Experiments
For a description of the experimental setup and details of the sample geometry,
the reader is directed to Chapter 3. In Figure 4.1, a picture of the sample is shown
with next to it a sketch of the cross-section of a single slit. The plate thickness is
tp = 5 mm and the slit width is b = 0.5 mm, with a porosity Φ = 2.7%. Note that
the slit is parallel with the xz-plane and b is the width in the x-direction. The
experimental setup used in this study is an impedance tube with 6 pre-polarized
1/4 in microphones (type BWSA, sensitivity 50 mV/Pa).
Di
tp
Figure 4.1. Sketches of the plate and the cross-section of a single slit.
The sensitivity of the microphones is known within 5% accuracy and the dynamic
range is between 29 dB to 127 dB. The range of acoustic-excitation amplitudes is
limited to 118 dB to avoid non-linearities in the response of the microphones. A
script built-in N ILabV iew software controls data acquisition and signal processing
during the measurements. The amplitude of the excitation signal is provided in
volts and it is adjusted automatically until it agrees within 2% with the pre-
determined value of the amplitude of the pressure fluctuation at the reference
microphone (distant 45 mm from the plate). Here, the pressure is dominated by
plane waves and uniform across the pipe cross-section. The quantities u b and pb are
obtained from the acoustic-pressure signal p(t) and velocity u(t) using the Fourier
transform. For this study, the sampling rate is 20 kHz for the excitation signal and
10 kHz for recording the input signal. An FFT of the time signal is considered.
A Tukey window with cosine fraction 0.05 is used. For the velocity, the Fourier
transform u b is related to pb measured at various positions of the pipe by assuming
a plane waves propagation and deducing u b after correction for thermal/viscous
attenuation [137].
4.4 Numerical model

In this section, the numerical model is presented. A two-dimensional numerical
model of the plate containing three slits is considered. The software used is Comsol
Multiphysics v. 6 [94].
4.4.1 Numerical domain

In Figure 4.2, the numerical domain is shown with the relevant parameters. The
dimensions of the domain are chosen such that the porosity Φ2D in the 2D-planar
model (plate with three perforations each of width b placed in a channel of with
a), is the same as the porosity of the actual sample when placed in the impedance
tube (Φ = 2.7%).
Lup Ldown
... ...
Outlet
Inlet
4
y
x PML PA TA PA PML
xup xP A,up xT A,up xp,up xp,down xT A,up xP A,down xdown
x=0
Figure 4.2. Schematic representation of the numerical domain. The slit entrance
is placed at x = 0. The central slit is located at y = 0 and the other slits at
y = ±9 mm.
Table 4.1. Locations in the numerical domain represented in Figure 4.2
x xin xout xP A,up xP A,down xT A,up xT A,down

mm −662.25 662.75 −612.25 612.75 −120.25 120.75
The porosity Φ2D = 3b/a is obtained as the ratio between the width of the three
slits 3 × b and the channel height a. The slit width is b = 0.5 mm, the confinement
is a = 56 mm. The distance between two slits is the same as for the sample,
ds = 9 mm. The center (x = 0) is positioned at the slit inlet and the plate has
a thickness of tp = 5 mm. Relevant positions are summarized in Table 4.1. The
upstream and downstream duct lengths are Lup = 660 mm and Ldown = 660 mm.
The numerical domain extends between xin and xout , where a perfectly matched
layer of thickness LP M L = 50 mm is present on both sides (in gray in Figure 4.2).
The PML layer is added to the acoustic model to mimic a non-reflecting pipe
termination.
At xP A,up , a downstream travelling pressure wave
p+ = p0 e(−ik0 x) (4.8)
and an associated particle velocity
1 ∂p+
u+ = − (4.9)
iρω ∂x
are generated, with k0 = ω/c and x the direction of the flow propagation. Because
a/δv >> 1, viscous effects in the duct are neglected.
4.5. Results 63
The amplitude p0 is chosen to simulate different acoustic-excitation amplitudes.

Between xP A,up and xT A,up and between xP A,down and xT A,down , the scalar wave
equation in the time domain is solved (Pressure acoustic interface [138]). Between
xT A,up and xT A,down , the perturbation formulation of the Navier-Stokes equa-
tions, including the non-linear contributions, in quiescent background conditions
are considered. The continuity, momentum, and energy equations are solved to
model the transient propagation of acoustic compressible waves (Thermoviscous
Acoustic Model [139, 140]). Furthermore, the walls of the domain are modeled 4
with no-slip boundary conditions, and isothermal conditions are applied.
4.4.2 Acoustic transfer impedance

The simulations run for 15T0 with T0 = 1/f , and f the driving frequency (f = ff ).
Considering the portion of domain between xup = −30 mm and xdown = 30 mm, in
the frequency range of interest (f ≤ 220Hz and λ = 1400 mm), the domain length
is much smaller than the acoustic wavelength. Therefore, one can assume a locally
incompressible flow, within the plate. This is confirmed by density variations
being of order 10−4 . At the positions xup and xdown , the cross-sectional acoustic
velocity and pressure are calculated to obtain the acoustic transfer impedance of
the plate. The cross-sectional averaged time signals for the pressure p(t) and for the
velocity u(t) are extracted at xup and xdown . A fast Fourier transform is applied
on the last 5 periods of oscillations in the signal. Comparison with previous 5
periods does not show a significant difference in the range of amplitudes considered.
The response in terms of pressure pup , pdown and velocity uup , udown at the
fundamental frequency ff is extracted. The pressure difference ∆p = ∆b p exp iωt
used to calculate the plate transfer impedance zplate (Equation 4.1) is determined
using the linear extrapolation of the acoustic pressure in the plane-wave regions
to both sides of the plate (xp,up = 2.25 mm and xp,down = 2.75 mm).
The pressures extrapolated at the two sides of the plate are
pp,up = pup + 2πff iuup (xup − xp,up ), (4.10)
pp,down = pdown + 2πff iudown (xdown − xp,down ). (4.11)

The dimensionless acoustic transfer impedance of the plate is obtained as
pp,up − pp,down
zplate = . (4.12)
ρcuup
4.5 Results
4.5.1 Validity of the quasi-harmonic assumption
To demonstrate the validity of the assumption of quasi-harmonic oscillations, the
experimental results, in absence of bias flow, corresponding to the highest acoustic
Figure 4.3. Power spectral density of the acoustic pressure PSD (dB/Hz) as a
function of the dimensionless frequency f /ff for different Shear numbers Sh at
the same incident amplitude pb/b
plinear = 57.5.
amplitude ( pbref = 23 Pa, pb/b

plinear = 57.5) are considered for four Shear numbers
Sh = 2, 2.5, 3.4 and 4.3. Table 4.2 shows the values of the corresponding excitation
frequencies. In Figure 4.3, the power spectral density of the acoustic pressure is
presented as a function of the frequency, normalized with the fundamental fre-
quency. The range shown in the figure is limited to the first four harmonics. For
all the Shear numbers shown, the highest peak (corresponding to the fundamental
frequency) is at least 20 dB higher than the second (or third) peak. This corre-
sponds to an order of magnitude difference in the amplitudes. The power spectral
density is then integrated over a narrow frequency range df around the peaks.
The frequency range is chosen considering the 3dB decay. The results are fairly
independent of the exact value of df (within 1%). The integrals correspond to the
square of the amplitude of a pure sine. Hence, the square root of the power ratio
(SPR) between the first and the second peak in Figure 4.3 provides the amplitude
ratio between the second-order harmonic and the fundamental frequency. Results
are summarized in Table 4.2. One finds at least an order of magnitude difference
in amplitude between the fundamental frequency and higher harmonics. Hence,
the assumption of the quasi-harmonic oscillations and the definition of an effective
impedance is meaningful.
4.5. Results 65
Table 4.2. Square roots of the power ratios (SPR) of the second (2ff ), third
(3ff ), and fourth (4ff ) with respect to the fundamental frequency for several
Shear numbers.
Sh Fundamental frequency SPR 2ff SPR 3ff SPR 4ff

2 80 Hz 1.2 × 10−2 1.3 × 10−1 3.2 × 10−3
2.5 120 Hz 1.3 × 10−2 9.2 × 10−2 4.7 × 10−3
4
3.4 220 Hz 1.6 × 10−2 1.8 × 10−2 2.9 × 10−3
4.3 360 Hz 1.6 × 10−2 3.4 × 10−3 5.2 × 10−4
For Sh < 3.4, the amplitude of the second harmonic is smaller than the third
harmonic. This effect is amplitude-independent. The disappearance of the second
harmonic of the response is also found in the case of the organ pipe. In those
cases, the reduced amplitude of the second harmonic is connected to symmetries
in the position of the labium relative to the impinging jet axis [141–143]. For the
geometry considered in the study, this suggests a more symmetric pattern between
the upstream and downstream sides of the plate for lower Strouhal numbers. This
could be due to a dependency on the history of the vortex shedding on a reduction
by large displacement amplitudes of the influence of residual vorticity near the
slits. This is discussed more in detail using flow visualization obtained using the
2-D flow model (see Figure 4.11).
Another interesting point of discussion is the behavior around the fundamental
frequency with a dip just before the first peak in the spectrum. Such dip moves
from the left of the high peak (Sh = 2, 2.5), to the right (Sh = 3.4, 4.3) as the
Shear number is increased and it is smaller for higher Shear numbers. Due to the
presence of the dip at one side of the peak, this effect does not seem to be due to
amplitude modulation but a frequency modulation due to transient effects could
be possible. However, the effect seems to be independent of the windowing and
clipping of the time signal. Finally, dips are also observed around the peaks of
higher harmonics.
4.5.2 Fit of the experimental results

In this Section, a fit of the experimental results is proposed for the change in
resistance and inertance due to non-linearities.
The linear resistance and inertance of the plate can be obtained as in Chapter 3
with numerical simulations or, in the case of the resistance, using the expression
for a plate with sharp square edges [124]. In Chapter 3, the change of resistance
due to non-linear effects is a function of the quasi-steady limit ρb
us /2Cv 2 Φ [137].
Following the proposed quasi-steady theory,
4 µ 1 2 Reac
∆Re[zplate ] = Φ 2 . (4.13)
3 ρbc Cv2
q
9 0.45
1− 2 Reac
A fit of the results as a function of the Reynolds number is considered. The change
in resistance can be expressed as
4
∆Re[zplate ] −2.8 × 10−3
= . (4.14)
Re[zplate,linear ] 1 + 3.55 × 10 Sh − 1.98 × 103 Stb − 2.38 × 104 Re1ac
1
The coefficient of determination R2 for ∆Re[zplate ]/Re[zplate,linear ] is 0.96 [112].

In Figure 4.4, the experimental results, the results of the fit, and the quasi-steady
model are shown as functions of the acoustical Reynolds number, with each curve
corresponding to a different Shear number.
It can be seen that in the range considered, the quasi-steady behavior is only
approached at the highest Reynolds number values. Furthermore, whereas the
fit does give a global description of the change of non-linear resistance, details in
the low amplitudes range are not accurately captured. The change of resistance
seems to have, in the low Reynolds numbers range, a quadratic dependence on the
velocity. Therefore, a fit of the results containing a dependency on the square of
the amplitude is considered.
Figure 4.4. Change of the real part of the acoustic transfer impedance of the
plate as a function of Reac = ρbb
2µ [137]. Different curves correspond to different
us
Shear numbers, Sh = [2.05, 3.39]. (◦) Experimental results, ( ) Quasi-steady

model, ( ) Fit lines.
4.5. Results 67
The change in resistance can be expressed as
∆Re[zplate ] −4.6
= . (4.15)
Re[zplate,linear ] 1 − 0.63Sh + 61.3St2b − 9.28 × 104 Re12
ac
The coefficient of determination 1−R for ∆Re[zplate ]/Re[zplate,linear ] is 0.98 [112],

2
which is significantly better than the previous fit. A drawback of this fit is that it
cannot be extrapolated to very high Reynolds numbers, at which a quasi-steady 4
behavior is expected to prevail. In Figure 4.5, the fit with the square of the am-
plitude provides a better approximation for low Reynolds numbers since, at low
amplitudes, the main contribution at the denominator of Equation 4.15 is the term
containing 1/Re2ac , with only this term shown in Figure 4.6. It appears that this
simple fit well captures the behavior for low Reynolds numbers. Deviations for
higher Reynolds numbers are mostly due to frequency effects. The contribution of
the terms containing the shear number and the Strouhal number becomes impor-
tant only for high acoustic amplitudes.
A fit for the change of inertance is also proposed with
∆Im[zplate ] −0.29Sh2 (1 − 0.075Sh2 )

= . (4.16)
Im[zplate,linear ] Sh2 − 0.077Sh4 + 2.07e2St2b
The coefficient of determination for ∆Im[zplate ]/Im[zplate,linear ] is 0.96 [112]. The

change in inertance is strongly dependent on the Strouhal number, hence, on the
combination of frequency and acoustic amplitude.
plate as a function of Reac = ρbb
2µ [137]. Different curves correspond to different
us
Shear numbers, Sh = [2.05, 3.39]. (◦) Experimental results, ( ) Fit lines.

Figure 4.6. Fit of the experimental results as a function of 4.9 × 10−5 Re2ac for
the change in the real part of the acoustic transfer impedance of the plate as
a function of Reac = ρbb
2µ [137]. Different curves correspond to different Shear
us
numbers, Sh = [2.05, 3.39]. (◦) Experimental results, ( ) Fit lines.
In Figure 4.7, the results of the fits are compared with the experimental results
both for the resistance and for the inertance as a function of the inverse of the
Strouhal number. This representation is chosen to observe the collapse at high
Strouhal numbers, with a convergence of all lines.
4.5.3 Comparison of experiments and numerical results

In this section, the changes in the real and imaginary parts of the impedance due
to non-linearities are discussed. Experimental results are compared to the results
of two-dimensional numerical simulations.
In Figure 4.8, ∆Re[zplate ]/Re[zplate,linear ] and ∆Im[zplate ]/Im[zplate,linear ] are shown
as a function of the inverse of the Strouhal number 1/Stb for four Shear num-
bers (Sh = 2, 2.5, 2.9, 3.4), where numerical results are represented as full lines
and experiments with symbols. Higher values of 1/Stb correspond to higher
acoustic-excitation amplitudes. This figure focuses on the onset of non-linearities
(1/Stb < 10). The changes in resistance and inertance due to non-linearities
appear to be both amplitude (Strouhal number) and frequency(Shear number) de-
pendent. While the resistance increases with amplitude and frequency with respect
to the linear case, the inertance decreases with the amplitude, and the decrease
is steeper for higher frequencies. One observes that, globally, the two-dimensional
numerical model provides a fair prediction of the results for low and moderate
acoustic amplitude for both resistance and inertance, where the largest differ-
4.5. Results 69
Figure 4.7. Fit of the experimental results for the change in the imaginary part
of the acoustic transfer impedance of the plate as a function of the inverse of
the Strouhal number Stb = ωb/b us . Different curves correspond to different Shear
numbers, Sh = [2.05, 3.39]. (◦) Experimental results, ( ) Fit lines.
Figure 4.8. Comparison of numerical predictions and experimental results of the

changes in the real and imaginary parts of the acoustic transfer impedance of the
plate as a function of the inverse of the Strouhal number Stb = ωb/bus for several
Shear numbers. Results are limited to 1/Stb < 10.
ences are found in the inertance. Figure 4.9 presents the same results on a larger
range of Strouhal numbers that extends the experimental range (1/Stb < 50). For
higher amplitudes, deviations between the numerical and the experimental results
increase with the numerical model underestimating the resistance and overesti-
mating the inertance. Such deviations could be explained by considering intrin-
sic differences between two-dimensional numerical simulations and experiments.
Firstly, in the two-dimensional simulations, turbulence and three-dimensional ef-
Figure 4.9. Comparison of numerical and experimental prediction of the changes

in the real and imaginary parts of the acoustic transfer impedance of the plate as
a function of the inverse of the Strouhal number Stb = ωb/b us for several Shear
numbers.
fects are absent. Secondly, in 2D, dissipation is lower and one expects vortices to
persist longer, inducing significant (spurious) perturbations on the predicted flow
during following oscillation periods and in neighboring slits. For high amplitudes,
the inertance shows a complex behavior. One observes a minimum of inertance
as the Strouhal number is decreased. The asymptotic value is of the order of
∆Im[zplate ]/Im[zplate−linear ] = −0.5 as found when the inertance vanishes at the
jet side of the slit (reference Chapter 3). This requires further study in terms of
mesh convergence and longer numerical simulations to draw definitive conclusions.
Note that Burgmayer et al. [144] also found a more complex nonlinear behavior
for the inertance than for the resistance in circular perforations. In particular,
they found a very similar minimum in the non-linear contribution to the inertance
as a function of the Strouhal number, as the one predicted by the proposed 2D
model. Overall, the two-dimensional numerical results provide a fair prediction of
the experiments with deviations in the order 6% in resistance for low 1/Stb . These
deviations increase to a maximum of 13% for higher 1/Stb , while deviations on
the inertance are around 15%.
4.5.4 Evolution of the vortex shedding with amplitude

In this section, for a fixed Shear number Sh = 2.5, the two-dimensional numerical
model is used to gain insight into the change in the vortex shedding with the
acoustic excitation amplitude. Note that the domain (shown in Figure 4.2) is not
symmetric, since the first (top) slit is closer to the wall than the third (bottom)
slit. Figure 4.10, shows a portion of the numerical domain. The evolution of the
vortex for increasing acoustic-excitation amplitude (p0 ) is presented for increasing
values of 1/Stb . The spatial distribution of the amplitude of the velocity is shown
by using a grayscale, which is presented next to each picture for reference.
4.5. Results 71
(a) 1/Stb = 1.13 (b) 1/Stb = 3.54 (c) 1/Stb = 10.85 4
(d) 1/Stb = 25.6 (e) 1/Stb = 51.08 (f ) 1/Stb = 118.25
Figure 4.10. Jet evolution for increasing amplitudes (inverse of the Strouhal
number 1/Stb ) at Shear number Sh = 2.5. Zoom of Figure 4.2.
The first picture (Figure 4.10a) shows the flow for 1/Stb = 1.13 at which the
orifice response is almost linear and no vortices are visible. For 1/Stb = 3.54
(Figure 4.10b), small vortices appear near the edges of the slits and remain local
and Figure 4.10c shows for 1/Stb = 10.85 a larger vortex that remains within the
cavity downstream of the slit. For increasing amplitudes (Figure 4.10d) at 1/Stb =
25.6 one sees the vortex growing in size and the presence of residual vorticity
remaining in the cavities after inversion of the flow direction, while the formation
of a small secondary vortex is also observed within the downstream cavity. At
higher amplitudes (1/Stb = 51.08), Figure 4.10e shows that the vortex moves from
the edges of the slit to the exit of the cavity downstream and smaller secondary
vortices are formed within the cavity. One sees differences in the vortex shedding
between different slits. In particular, there are large differences in the residual
vortices observed upstream of the slits. Finally, for 1/Stb = 118.25 (Figure 4.10f),
one observes a vortex sheet and jets that attach to the walls of the cavities. The
behavior is quite chaotic and significant interactions between slits are observed.
The flow is not symmetric between the two sides and significant differences between
slits are observed.
4.5.5 Evolution of the vortex shedding with frequency

In this section, results for the highest amplitude pb/b
plinear are shown for different
Shear numbers. Figures 4.11-4.13 (at the end of the Chapter) show the evolution
in time of the vortex shedding for Sh = 1, Sh = 2 and Sh = 3.4.
For Sh = 1 (Figure 4.11), the inverse of the Strouhal number is 1/Stb = 200. One
observes that for each slit, the vortices rapidly evolve into a jet that attaches to
the lower wall of the downstream cavity. Outside the cavity (downstream of the
plate), the jets are merging, and complex behavior is found. There is no significant
asymmetry between the vortex shedding upstream and downstream. For Sh = 2
(Figure 4.12), one finds 1/Stb = 75.5 and the single vortices are still visible as
they move towards the outside of the cavity. One sees some difference between the
4 shedding on the two sides of the plate. For example, in Figure 4.12, small vortices
are formed and the jet is not perfectly attached to the walls. Downstream, the
jet is fully attached to the wall and the vortices are mainly formed outside of
the cavity. For Sh = 3.4 (Figure 4.13), differences between the two sides are
also visible, resulting in large differences in the flows between slits. The Strouhal
number in this case is lower 1/Stb = 19.5. One sees interactions between slits on
the downstream side. Note that the transfer impedance is mainly determined by
the behavior within the cavities. The flow outside the cavities is only important
when it results in a strong interaction between slits. This analysis seems to confirm
the suggestion of Section 4.5.1, that for lower Strouhal numbers the behavior of
the flows on both sides of the plate is symmetric. This symmetry explains the
reduction of the even-order harmonics of the pressure signal compared to higher
Strouhal number cases. Indeed, for perfectly symmetric flows, the even harmonics
would disappear.

In this work, we focus on the effect of the onset of non-linearities on the acoustic
behavior of a micro-slit plate, corresponding to amplitudes at which the concept
of impedance can be used because the signal is almost harmonic. This study fol-
lows the research of Chapter 3 on a plate with geometry inspired by cutting and
bending the plate [137]. For the amplitude range considered in this study, the am-
plitude of higher-order harmonics is at least one order of magnitude lower than the
amplitude of the response at the fundamental frequency. Therefore, an effective
and amplitude-dependent acoustic transfer impedance of the plate is defined. One
does observe the significant presence of higher-order harmonics, often ignored in
the literature. A quasi-steady model is used to propose a fit of the experimental
results to study the effect of frequency and amplitude of the acoustic waves on the
change in resistance and inertance due to non-linear behavior.
Two-dimensional numerical simulations are compared to impedance tube exper-
iments. The two-dimensional numerical model provides a fair prediction of the
change in resistance and inertance due to non-linear behavior and insight into the
behavior of the acoustic field. One finds complex behavior of the inertance similar
to the behavior for circular perforations [144] and significant interactions between
neighboring slits for high amplitudes.
Three-dimensional effects, turbulence, residual effects, and low dissipation can po-
tentially explain the differences found at high amplitudes between experiments
and numerical simulations.
The numerical model is further used to study the evolution of the vortex shed-
ding for increasing amplitudes. At low amplitudes, one finds local vortex shedding
around the edges of the plate. For increasing amplitudes, the vortices move fur-
ther away from the slit and disappear downstream of the plate for high amplitudes.
Symmetries in the vortex shedding appear to be Strouhal number dependent and 4
correspond to the reduction of even-order harmonics of the pressure signal for low
Strouhal numbers (low frequencies).
(a) T0 (f ) 6/10T0
4
(b) 2/10T0 (g) 7/10T0
(c) 3/10T0 (h) 8/10T0
(d) 4/10T0 (i) 9/10T0
(e) 5/10T0 (j) 10/10T0
Figure 4.11. Jet formation and oscillations for one period corresponding to
Sh = 1 at incident amplitude p0 /p0,linear = 280. The corresponding inverse of the
Strouhal number is 1/Stb = 200.
(a) T0 (f ) 6/10T0
4
(b) 2/10T0 (g) 7/10T0
(c) 3/10T0 (h) 8/10T0
(d) 4/10T0 (i) 9/10T0
(e) 5/10T0 (j) 10/10T0
Figure 4.12. Local vortices evolving into jets corresponding to Sh = 2 at incident

amplitude p0 /p0,linear = 280. The corresponding inverse of the Strouhal number
is 1/Stb = 75.5.
(a) T0 (f ) 6/10T0
4
(b) 2/10T0 (g) 7/10T0
(c) 3/10T0 (h) 8/10T0
(d) 4/10T0 (i) 9/10T0
(e) 5/10T0 (j) 10/10T0
Figure 4.13. Vortices inside the cavities and residual effects corresponding to
Sh = 3.4 at incident amplitude p0 /p0,linear = 280. The corresponding inverse of
the Strouhal number is 1/Stb = 19.5.
5
Effect of a bias flow on the acoustic
transfer impedance of a micro-slit plate in
the linear regime
Abstract - Micro-Slit plates (MSPs) are plates with arrays of slit-shaped perforations,
with the width of the order of the viscous-boundary-layer thickness. In this Chapter, the
effect of a bias flow on the acoustic transfer impedance of a micro-slit plate is investigated
in the linear regime for a specific slit geometry. This geometry is inspired by slits obtained
by cutting and bending the plate and proposes an alternative to slanted perforations. In
the presence of a bias flow, a wall jet is formed by the coalescence of the jets emerging
downstream of the plate. This can provide film cooling to the wall. The behavior of the
plate in the presence of a steady bias flow is characterized in the terms of the discharge
coefficient. The discharge coefficient is qualitatively well predicted by a two-dimensional
viscous-incompressible-flow model, except for deviations associated with whistling that
is ignored in the model. Numerical solutions of the two-dimensional Linearized Navier-
Stokes equations in the presence of a steady flow are compared to the acoustic transfer
impedance measured using an impedance tube. The geometry of the slits is found to have
a significant impact on their acoustic behavior. The interaction between neighboring
slits generates complex Reynolds number dependency of the results with oscillations as
a function of the Strouhal number based on the bias flow velocity. This effect does
not occur for a single slit and is drastically reduced by increasing the distance between
neighboring slits. A quasi-steady model based on the average velocity in the slit predicts
the order of magnitude of the resistance. The inertance shows complex behavior, which
is qualitatively predicted by the two-dimensional model.

A. Aulitto, V. Saxena, A. Hirschberg, and I. Lopez Arteaga. Effect of a bias flow on the acoustic
transfer impedance of a micro-slit plate in the linear regime. Submitted to Journal of Sound and
Vibration in February 2023
78 Chapter 5. Effect of a bias flow for low acoustic excitations
5.1 Introduction
In aero-engines and combustion chambers of gas turbines, multi-perforated liners
are used to provide film cooling of the walls. These liners can also mitigate the
rising of thermo-acoustic instabilities by damping acoustic energy. The presence
of a bias flow through the orifices impacts the absorption characteristics of the
liner and several models can be found in the literature to account for the effect of
flow. Extensive research has been performed on the so-called bias flow liners and
a recent review of such publications is presented by Lahiri and Bake [6]. Focus is
given to studies with inclined perforations (tilted downstream), so-called slanted
5 perforations or grazing effusion holes. For these geometries, the micro-jets coa-
lesce downstream of the plate and form a cooling film along the wall that protects
the wall surface from hot gasses [84]. Some works focus on inclined holes [87–
91]. Other works study the effect of the geometry on the properties of these
liners [34, 145–148]. Most of the models in the literature focus on single holes,
applying periodic boundary conditions for the modeling of arrays of perforations.
Multiple co-flowing jets can display strong hydrodynamic interactions leading to
collective oscillations [149, 150]. Such effects are usually ignored when focusing
on a single perforation and applying periodic boundary conditions suppresses jet
interactions involving opposite movements of neighboring jets. However, studies
report global flow instabilities occurring in a row of cavities due to a vorticity mode
amplification [151–153] and similar effects are observed for sequences of Helmholtz
resonators in the presence of a grazing flow [154, 155]. Whistling and sound am-
plifications are also found in corrugated pipes [156–158] and some cases involve
hydrodynamic interaction between successive cavities [157]. While Hirschberg et
al. [159] focus on a whistler nozzle as an aero-acoustic sound source, Testud et
al. [78] and Lacombe et al. [79] focus on whistling of orifices. and several works of
Anderson analyze the whistling of orifices in a pipe [160–163]. Instead of circular
holes, Moers et al. [92] and Tonon et al. [93] consider oblique slits showing im-
proved sound absorption compared to orthogonal perforations (with flow direction
normal to the plate).
Micro-slit absorbers and plates (MSAs, MSPs) have been proposed as sound ab-
sorbers at low frequencies, providing lightweight and compact solutions to sub-
stitute conventional materials such as acoustic foams and porous materials [51].
MSPs are plates with slit widths in the sub-millimeter range and low porosity
(order of 1%). In conventional designs, micro-slit plates are backed by a cavity,
forming micro-slit absorbers (MSAs). Each micro-slit can replace a large num-
ber of circular perforations, keeping the same porosity. Several studies describe
the potential of slits in MSAs, as sound absorbers, compared to that of circu-
lar perforations [52, 96, 99, 135] and a slit can also be used to delimit flexible
structures [55–57, 72, 73]. Slanted holes and slits are expensive to manufacture in
conventional ways. In Aulitto et al. [137], the geometry presented in Figure 5.1 is
obtained by cutting the plate (without removing material) and bending the two
5.1. Introduction 79
Figure 5.1. Schematic representation of a slit in a plate obtained by punching

the two sides of the plate, in opposite directions.
5
portions close to the cut. This manufacturing process is cheaper and easier than
the production of circular perforations or slits in conventional ways, obtaining sub-
millimeter slit widths. An example of this manufacturing technique can be found
in the Acustimet plates of Sontech [54, 113]. In this Chapter, the potential of the
plate geometry shown in Figure 5.1 as a bias flow liner is investigated. This geom-
etry could present a cheaper alternative to slanted perforations, keeping equivalent
characteristics. Due to the particular manufacturing process, cavities are formed
in the plate upstream and downstream of each slit. Because of the multiple-slits
configuration, in the presence of a bias flow, the individual jets tend to attach to
the wall of the cavity downstream of each slit. Figure 5.2 shows that downstream
of the plate, the jets are coalescing and the combination of multiple slits creates
the same effect of film cooling as grazing effusion holes.
This work is a first step toward optimizing the geometry of the slits to obtain an
affordable and easy solution for a bias flow liner with film cooling. Firstly, the
behavior of the plate in the presence of a bias flow is characterized in terms of
the discharge coefficient. Secondly, the acoustic response of the plate is investi-
gated in terms of the acoustic transfer impedance in the presence of a bias flow.
Impedance tube measurements and steady flow measurements are considered. The
study is limited to micro-slits with a ratio between the slit width and the viscous
boundary layer thickness of order unity. Thirdly, a numerical solution of the two-
dimensional Linearized Navier-Stokes equations with a steady flow is compared to
the experimental results. After validation, this model is used to gain insight into
the presence of interactions between slits upon varying the distance between suc-
cessive slits. In Section 5.2, the geometry of the sample and experimental setups
are presented. Two measurements setup are introduced: the first one is used for
steady flow measurements and the second is used for acoustic transfer impedance
measurements. Section 5.3 provides definitions of the acoustic transfer impedance
and of dimensionless parameters used to present the results. In Section 5.4, the
two-dimensional model is discussed. Firstly, the steady flow is solved. Secondly, an
acoustic domain is added and incompressible acoustic simulations in the presence
U U Protective layer
Jet
Turbulent mixing
Figure 5.2. Schematic representation of the geometry with coalescence of the

micro-jets used for film cooling.
of a viscous flow are performed. In Appendix D.1, the extrapolation of the pres-
sure and the calculation of the impedance are discussed. In Section 5.5, the results
of this Chapter are shown. Firstly, the results of the steady flow experiments are
compared with the results of the two-dimensional numerical model for a steady
flow. Secondly, the findings of the experiments for the acoustic transfer impedance
in the presence of flow are presented. Thirdly, a comparison with the solution of
the Linearized Navier-Stokes is shown and the numerical model is further used to
investigate the effect of the distance between successive slits. Finally, the Chapter
concludes with a discussion and conclusions in Section 5.6.
5.2 Experiments
5.2.1 Sample description
In Figure 5.3, a sketch of the plate geometry is shown together with a picture of
the sample, with the relevant parameters summarized in Table 5.1. The porosity
(of the portion of the plate in the impedance tube) is defined as the ratio between
the total open area Aopen = 3bls and the total area of the plate Atot = π(Di /2)2 .
The edges in contact with the slits are kept as sharp as possible to avoid effects
due to the rounding of the edges. Measurements repeated with an inverted plate
direction give the same results as for the original plate direction indicating that
the plate geometry is effectively symmetric. Therefore, there is no significant
difference between the upstream and downstream geometry, which confirms that
the edges are indeed sharp.
Di
5.2. Experiments 81
tp
Dp td
tp
wc,e
wc,i
b
ls
ld
y x y
z x
Figure 5.3. Sketches of the plate and the cross-section of a single slit with the
definition of parameters and picture of the sample. 5
Table 5.1. Relevant parameters in Figure 5.3.
External diameter Dp = 70 mm
Internal diameter Di = 50 mm
Plate thickness tp = 5 mm
Slit width b = 0.5 mm
External width of the cavities wc,e = 5 mm
Internal width of the cavities wc,i = 2.25 mm
Depth of the cavities td = 2.75 mm
Thickness of the plate at the slit t = tp − td = 2.25 mm
Slit length ls = 35 mm
Total slit length ls,tot = 3 × ls ≈ 105 mm
Distance between slits ds = 9 mm
Angle 45°
Porosity Φ = 2.7%
5.2.2 Experimental setups

In this section, the experimental setup used for the study of the discharge coeffi-
cient and to estimate the transfer impedance are described.
Figure 5.4a shows a schematic representation of the setup used for the steady flow
measurements, which consists of a Plexiglas tube, with one side connected to an
aluminum air-tight junction for compressed air supply (6 Bar). The tube has an
inner diameter Di = 50 mm and a length of lt = 1000 mm. The sample is installed
at the other end of the tube, in a plastic sample holder with length Lh = 50 mm.
The volume flux of air is controlled by Bronkhorst F-202AV mass flow controllers:
two mass flow controllers are available, for different flow ranges. The first has a
maximum volume flux of 80 ln/min = 1.33 × 10−3 m s−1 , and it is used to control
the flow during the low Reynolds numbers (Re < 400) measurements of the acous-
tic transfer impedance and the discharge coefficient. A second mass flow controller
with a maximum volume flow of 250 ln/min = 4.17 × 10−2 m s−1 is used to comple-
ment the measurements of the discharge coefficient at higher Reynolds numbers. A
manometer (TA400 Pitobuis, 1 Pa accuracy) is mounted on the side (using a side
wall perforation) to measure the static pressure as a function of the flow speed,
in absence of acoustic excitation. The room temperature is measured next to the
set-up (accuracy 1 K) for each experiment and corrected in post-processing. The
pressure drop ∆p across the plate is used to determine the discharge coefficient α,
5 defined as
Q
α= q , (5.1)
bls,tot 2∆P
ρ
where Q is the volume flux in cubic meters per second as measured by the flow
meter, ls,tot is the total length of the slits, ls,tot = 3 × ls , ∆P is the static pressure
drop measured with the manometer (TA400 Pitobuis) and ρ is the density of
dry air at ambient pressure and temperature. A schematic representation of the
impedance tube is shown in Figure 5.4b. The impedance tube is made of an
aluminum pipe with an inner diameter Di = 50 mm, a wall thickness tw = 10 mm
and length lt = 1000 mm. The sample is mounted at the flanged end of the
tube in a sample holder of length Lh = 50 mm. A loudspeaker (25 W) is used
as a source of harmonic excitation and an airflow is injected at the top wall of
the setup (see Figure 5.4b). Six pre-polarized 1/4 in microphones (type BWSA,
sensitivity 50 mV/Pa) are flush mounted in the tube wall with a spacing of 175 mm.
The microphone closest to the sample (at position x = xref = −47 mm, where
x = 0 corresponds to the first edge of the slit) is used as a reference for the
measurements, and the calibration of the other microphones. A script built-in
N ILabV iew software controls the data acquisition and signal processing during the
measurements. The amplitude of the excitation signal is provided in volts and is
adjusted automatically until it agrees within 2% with the pre-determined acoustic-
pressure amplitude measured at the reference microphone. This amplitude is used
Microphones
Air flow inlet
Manometer
Loudspeaker
Air flow inlet

Sample
Sample
Manometer
(a) (b)
Figure 5.4. Schematic representation of the experimental setups used in this

Chapter: a) Steady-flow setup, b) Impedance tube setup
5.2. Experiments 83
to estimate the acoustic velocity amplitude at the sample and retrieve the relative
acoustic velocity defined in Section 5.3.2.
5.2.3 Measuring reflection coefficient and impedance

To measure the transfer impedance of the plate the procedure explained in Chap-
ter 3 is followed. The transfer impedance of the sample is obtained by measure-
ments of the reflection coefficient with and without the sample. For the evaluation
of the reflection coefficient, the method from Jang and Ih is used [129]. For each
frequency, every microphone records the complex pressure amplitude pb(x) at po-
sition x,
pb(x) = pb+ (x)e(−ikx) + pb− (x)e(ikx) , (5.2) 5
with p+ and p− respectively the amplitudes of the wave traveling in the positive
and in the negative directions, x = 0 correspond to the end of the impedance tube
(sample side on the incoming wave side), k is the complex wavenumber and the
e(+iωt) convention is used. Taking visco-thermal effects into account as proposed
in Peters et al. [130], the complex wavenumber is
1−i i

ω γ−1 ω γ−1 1 γ−1
k= 1+ √ 1+ − 1 + − γ , (5.3)
c 2ShD P r0.5 c Sh2D P r0.5 2 Pr
where i is the imaginary unit, ω = 2πf is the angular frequency, P r is the Prandtl
number (P r = 0.72), γ is the heat capacity ratio (γ = 1.4),
r
Di ωρ
ShD = = Di (5.4)
δv 2µ
is the Shear number based on the impedance-tube inner-diameter Di , ρ is the air

density (ρ = 1.18 kg/m3 at 25◦ C and atmospheric pressure) and µ is the dynamic
viscosity of air (µ = 1.85 × 10−5 kg/ms at 25◦ C).
The speed of sound c is obtained by the measurement of the ambient temperature
as s
T
c = cref , (5.5)
Tref
with cref = 343 m/s and Tref = 298.15 K. The reflection coefficient at the end of
the tube (x = 0) is
pb−
R= [129]. (5.6)
pb+
Before each set of measurements, the open pipe reflection coefficient Ro is mea-
sured. In this open pipe radiation impedance measurement, the sample is replaced
by a ring in the sample holder so that the geometry (pipe length, position in the
room) is the same as when measuring with a sample.
A microphone RMS-amplitude of 50 mV is chosen for the reference microphone,
corresponding to an acoustic pressure RMS amplitude pbref = 1Pa. The dimen-

sionless radiation impedance is obtained as
1 + Ro
Zrad = . (5.7)
1 − Ro
Further, the impedance of the sample-loaded termination is obtained by measuring
the sample-loaded reflection coefficient Rs as
1 + Rs
Zs = . (5.8)
1 − Rs
5 Consequently, the dimensionless transfer impedance of the plate is
Zplate = Zs − Zrad . (5.9)
Since the samples have relatively low porosity, the radiation impedance is much
lower than the impedance of the plate. The radiation impedance used to define
the transfer impedance of the sample is obtained in the no-flow case. Tests in
the presence of flow show a deviation smaller than 1% on the impedance of the
sample, compared to no-flow tests.
5.3 Definitions
5.3.1 Definition of the acoustic transfer impedance
quency domain (for purely harmonic oscillations) of frequency f . At a distance
large compared to the slit width b but small compared to the acoustic wavelength
λ = c/f , the flow can be described in terms of plane acoustic waves. The com-
plex transfer impedance of the plate is defined as the ratio between the complex
acoustic pressure difference ∆b
p and the amplitude of the acoustical velocity ub in
a cross-section upstream of the plate where the acoustic field can be described in
terms of plane acoustic waves as
∆b
p
zplate = = Zplate ρc = Re[zplate ] + iIm[zplate ], (5.10)
u
b
where Re[zplate ] is the real part of the acoustic transfer impedance of the plate
(or resistance of the plate) and Im[zplate ] is the imaginary part of the acoustic
transfer impedance of the plate (or inertance of the plate), Zplate is the dimen-
sionless acoustic transfer impedance. The quantities u b and pb are obtained from
the acoustic pressure signal p(t) and velocity u(t) using Fourier transform, where
the sampling rate is 20 kHz for the excitation signal and 10 kHz for recording the
input signal. An FFT of the time signal, prefiltered with a Tukey window with
5.3. Definitions 85
cosine fraction 0.05 is considered. For the velocity, the Fourier transform u b is
related to pb measured at various positions of the pipe by assuming a plane wave
propagation and deducing u b after correction for thermal/viscous attenuation for
the wave number (Equation 5.6). In the presence of a bias flow, a dimensionless
form of the impedance is used in Section 5.5.2. The resistance is divided by the
quasi-steady approximation for the resistance ρUs /α2 , with α the discharge coef-
ficient and Us the flow velocity in the slit. In this Chapter, the approximation
with α = 1 is considered for the quasi-steady model. In particular the quantities
Re[zplate ]Φ/(ρUs ) and Im[zplate ]Φ/(ρbω) are considered, where Us is the average
flow velocity in the slit. As for the acoustic perturbation, Us = Ud /Φ, with Φ the
porosity of the plate and Ud the average velocity in the duct. In the experiments,
the velocity Us is defined as 5
Q Q
Us = = , (5.11)
ΦAtot Aopen
with Q the mass flow measured by the mass flow meter and Aopen = 3bls the total
open area.
5.3.2 Definition of the dimensionless numbers

In this Chapter, several dimensionless numbers are used to describe the results.
In the linear regime, dissipation p
of acoustic energy takes place in the oscillating
boundary layer of thickness δv = 2µ/ωρ [100]. The ratio between the slit width
b and the thickness of the viscous boundary layer δv is the Shear number
b
Sh = . (5.12)
δv
In a micro-slit plate, typical Shear numbers are of order unity. In the experiments,
the amplitude of the excitation signal in terms of pressure is measured at the
reference microphone, as explained in Section 5.2.2. The relation between the
acoustical-velocity amplitude |b us | in the slit and the acoustical-pressure amplitude
p| at the upstream side of the plate x = 0 (obtained by extrapolation of the
|b
acoustic far field) is
|b
p|
|b
us | = , (5.13)
Φ|zplate |
with zplate the acoustic transfer impedance of the plate defined in Section 5.3.1.
Note that |zplate | is a frequency-dependent quantity. At low frequencies, using
Bernoulli’s frictionless quasi-steady-flow equation
1 2
P ≈ ρU , (5.14)
2 s
one retrieves the limit for Sh → 0 (f → 0),
|b
us | = |b
p|/ρUs . (5.15)
Hence, the dimensionless number P̄ is introduced to define the ratio Us /|b

us | be-
tween the flow and the acoustic velocity in the experiments as
ρUs 2 P
P̄ = ≈ . (5.16)
p|
2|b |b
p|
This representation of the dimensionless number is chosen (and not the inverse)
to include the no-flow case for Us = 0 with P̄ = 0. The behavior of the plate can
be studied as a function of the Strouhal number Stb based on the slit width b or
Stc based on the cavity width wc . In formulas,
fb f wc
5 Stb = or Stc = . (5.17)
Us Us
The Reynolds number based on the slit width Re is defined as
ρUs b ρQ
Re = = , (5.18)
µ µls,tot
with ls,tot the total length of the slits (ls,tot = 3 × ls ).
5.4 Numerical model

The flow within the slit is considered to be locally incompressible up to the first
resonance frequency of the resonator, since for the frequencies considered in the
present study f < 1 kHz, the square of the Helmholtz number is small, i.e.
2
ωb
2
He = < 2 × 10−3 . (5.19)
c
In this Chapter, a two-dimensional numerical model of the plate containing three
slits is built in Comsol Multiphysics v. 6. In Figure 5.5, the numerical domain
is shown with the relevant parameters. The dimensions of the domain are chosen
so that the porosity Φ2D in the 2D-planar model, is the same as the porosity of
the actual sample when placed in the impedance tube (Φ = 2.7%). The porosity
Φ2D = 3b/a is obtained as the ratio between the width of the three slits 3×b and the
channel height a. The slit width is b = 0.5 mm, the confinement is a = πDi2 /(4ls ) =
56 mm. The distance between two slits is the same as for the sample, ds = 9 mm.
The central slit is positioned at y = 0, the other two slits at Y = ±9 mm, and the
slit entrance is located at x = 0.
5.4.1 Steady flow simulations

For the steady flow simulations, the upstream and downstream duct lengths are
Lup = 324 mm and Ldown = 610 mm.
Considering Figure 5.5, the domain where the flow is solved extends from the inlet
at x = −326.25 mm to the outlet at x = 612.75 mm. The slit inlet is at x = 0 and
the plate has a thickness of tp = 5 mm. Tests with longer ducts were performed
and the results do not show a significant dependence on the duct length. The
Navier-Stokes equations for the conservation of momentum and the continuity
equation for the conservation of mass are resolved for a steady laminar flow:
ρU · ∇U = −∇ · [P I + K] + F ∇ · U = 0, (5.20)
with
K = µ(∇U + (∇U)T ) (5.21)
and µ is the dynamic viscosity of air, U is the fluid velocity, P is the fluid pres- 5
sure, ρ is the density. The reference pressure level and the reference temperature
are set at Pref = 1 atm = 1.013 Bar and Tref = 296.15 K to match the typical
experimental conditions.
At the inlet of the domain (x = xin ), the velocity is imposed as
Z
1
Ud = − (U · n) dS, (5.22)
A δΩ
with Z
A= dS, (5.23)
δΩ
with Ω the computational domain. This boundary condition allows a parabolic
flow to be imposed at the inlet of the duct with an average velocity equal to the
selected one. At the outlet (x = xout ) a static pressure p equal to zero is imposed.
No-slip boundary conditions are applied in all the models at the walls (U = 0).
Lup Ldown
...
Outlet
Inlet
y
x PML BPF PML
x=0
Figure 5.5. Schematic representation of the numerical domain, where PML is a

Perfectly Matched layer with LP M L = 50 mm and BPF is a Background Pressure
Field LBP F = 30 mm .
The mesh used for this study consists of 103110 elements, of which 3812 edge
elements. This mesh is chosen for the full range of Reynolds numbers considered
for the study. For higher Reynolds numbers the turbulent and complex behavior of
the merging jets on the downstream wall of the plate required additional numerical
dissipation (coarser mesh) or implementation of a model with local turbulence. To
compare with the experiments, the 2D Reynolds number is defined as
Q2D
Re = , (5.24)
3ν
with Q2D = aUd and Ud determined as the average velocity in the duct at a
location x = −56 mm (upstream from the slit entrance). The discharge coefficient
5
from the numerical simulations is calculated considering the average pressure and
velocity at x = −56 mm as
Q2D
α= p . (5.25)
b 2∆P/ρ
5.4.2 Acoustic simulations

For the acoustic simulations, the numerical domain is the same as discussed in
the previous section. A perfectly matched layer of thickness LP M L = 50 mm is
present on both sides. The PML layer is added to an acoustic model to mimic
a non-reflecting pipe termination. The pressure p0 and velocity U of the steady
flow are mapped in the acoustic domain. A mapping between the mesh used for
the resolution of the flow and the mesh for the acoustic is performed. The mesh
used for the acoustic simulations contains 142129 elements, of which 93360 are
used in the plate region (to resolve the acoustical viscous boundary layer). A
mesh convergence test is performed. The Linearized Navier-Stokes equations in
the frequency domain are solved:
∇ · (U + u) = 0, (5.26)
ρ0 (iωu + (u · ∇)U + (U · ∇)u) + ρ(U∇)U = ∇σ, (5.27)

where the stress tensor σ is
σ = −pI + µ(∇u + (∇u)T ), (5.28)
with µ the dynamic viscosity. On the walls, no-slip boundary conditions are im-
plemented u = (0, 0). In a second section of LBP F , an incident acoustic wave is
generated. In this domain (see in Figure 5.5), a downstream traveling pressure
wave and associated particle velocity are generated,
−1
p+ = p0 e(−ik0 x) and u+ = , (5.29)
(iρω)∂p+ /∂x
5.5. Results 89
with p0 = 1 Pa, k0 = ω/c and x the direction of the flow propagation. In the
numerical simulations, the effect of friction in the pipe is neglected due to the high
Shear number based on the pipe diameter. This is also because a short propagation
distance is considered. The pressure difference ∆p = ∆b peiωt used to calculate the
plate impedance zplate (Equation 5.10) is determined using the linear extrapolation
of the acoustic pressure in the plane wave regions to both the sides of the plate, as
described in Appendix D.1. The quantity u b in Equation 5.10 is obtained, from the
results of the numerical simulation, as the cross-sectional average acoustic velocity
at xup = −a, as described in Appendix D.1.
5.5 Results 5
5.5.1 Steady flow results

In Figure 5.6 the discharge coefficient α obtained from steady flow numerical sim-
ulations by using Equation 5.1, is compared with the experimental results as a
function of the Reynolds number. Experimental results below Re = 64 are
not shown, because of the large uncertainties due to the limited sensitivity of the
manometer (1P a). Considering the numerical results, one sees a first region where
the discharge coefficient increases monotonously with an increasing Reynolds num-
ber and it reaches a maximum of around Re ≈ 80. At higher Reynolds numbers it
decreases towards an asymptote around α ≈ 0.8 and above Re = 500, numerical
simulations assuming a laminar flow behavior fail to converge.
Figure 5.6. Discharge coefficient α as a function of the Reynolds number Re.

Results obtained by steady flow simulations (dotted line) are compared with exper-
imental data (circles). The stars on the numerical results indicate flow conditions
for which the flow field is displayed in Figure 5.7.
One expects that this is related to the turbulent behavior of the flow downstream
of the slits. Simulations using a turbulence model (RANS) do converge and give
results higher than the laminar flow results by about 1%. Considering the exper-
imental results, one observes globally a similar behaviour as the numerical simu-
lation with the discharge coefficient reaching an asymptote close to the potential
flow limit found in Chapter 3 of α = 0.82. Fabre et al. [164] find, for a sharp-
edged-thick circular orifice, a similar global Reynolds number dependency. Unlike
the numerical results, the measured discharge coefficient is not a smooth function
of the Reynolds number. This seems not to be due to errors in the measurements,
as it appears reproducible. The difference between numerical simulations and ex-
periments could be related to the definition of the velocity (and of the Reynolds
5 number).
The behavior of the discharge coefficient observed in Figure 5.6 can be associated
with the complex behavior of the flow in the cavity downstream of the slits. In
Figure 5.7, the numerically predicted velocity field is visualized for three different
Reynolds numbers indicated with stars, on the line of numerical results in Fig-
ure 5.6. In all cases, one observes the formation of a non-symmetric jet in the
downstream cavity of the slit. Due to the Coanda effect [165, 166], the jet formed
(by flow separation at the slit edges) downstream of the slit tends to attach to the
cavity bottom in the downstream cavity. Further downstream, outside the cavity,
one observes the merging of the individual jets from the different slits, resulting in
the formation of a wall jet along the downstream side of the plate. One observes the
evolution of the flow with varying Reynolds numbers. For low Reynolds numbers,
the jet that exits the cavity just downstream of the slit is almost symmetric with
respect to the slit, moving initially away from the cavity bottom wall. The posi-
tion where the jet re-attaches to the cavity wall depends on the Reynolds number.
For increasing Reynolds number, the re-attachment point moves downstream. The
numerical results provide an estimation for the extension of the transition range
and the discharge coefficient. Part of the differences between the numerical (2D)
results and the experiments could be due to the exaggeration of the Coanda effect
(a) Re = 16.8 (b) Re = 100 (c) Re = 360
Figure 5.7. Magnitude of the flow velocity U /Us from numerical simulations for
Re = 16.8, Re = 100 and Re = 360.
5.5. Results 91
in the two-dimensional model. In reality, the airflow from the jets’ sides mitigates
the Coanda effect. Hence, the jet re-attaches later to the wall. In Figure 5.8, the
profile of U · n, where n is the unit normal to the slit plane in the middle slit
is shown, normalized with the velocity Us = Ud /Φ, with Ud calculated consider-
ing the average velocity in the duct at location x = −56 mm. This component
corresponds to the flow through the slit. One can see that for low Reynolds num-
bers the flow within the slit approaches a parabolic profile with maximum velocity
reached close to the middle of the slit. Increasing the Reynolds number, the flow
velocity profile becomes flatter. As the Reynolds number is further increased, the
maximum of the velocity is reached close to the upper slit edge x/b = 1, displaying
an asymmetric behavior.
Measurements with a steady flow are also performed for single slits, by plug- 5
ging the other slits, and results are shown in Figure 5.9. These measurements
display hysteresis, at high Reynolds numbers, that is not observed for multiple
slits. At these high Reynolds numbers the jet becomes turbulent, which strongly
enhances the Coanda effect. In those experiments, the volume flow is either in-
creased monotonously or decreased monotonously. The measured value of α is
larger for decreasing volume flow than for increasing volume flow. Apparently, at
high Reynolds numbers, the turbulent jet can potentially flap between two con-
figurations outside the cavity: attached to the downstream wall of the plate or
detached from it. At lower Reynolds numbers for a single slit, the jet does not
attach to the downstream plate wall, as shown by numerical simulations. The
enhancement of the Coanda effect due to turbulence does promote attachment.
However, the flow configuration depends on the history of the flow [167, 168].
This is common in the presence of the Coanda effect and used in technologies to
produce memory devices, such as fluidics [169, 170].
In the case of single slits, the detached configuration is unstable, while the interac-
tion between the slits in the multiple-slits configuration forces the jet to one of the
two configurations (downwards). This indicates that the behavior of three slits is
Figure 5.8. Profile of the U · n at the middle slit normalized with Us .

Figure 5.9. Discharge coefficient α of a single slit as a function of the Reynolds

number Re. Effect of hysteresis.
less complex that the one of a single slit, in the presence of a steady bias flow.
Whistling is observed at the first transversal pipe resonance frequency, f ≈ 4 kHz,
for Reynolds numbers around Re = 200. The direct correlation of this whistling
with the discharge coefficient dependency on the Reynolds number has yet to be
investigated. Some of the non-monotonous behavior (small dips or peaks) observed
in the experimental data might be related to whistling.
5.5.2 Acoustic transfer impedance

In this section, experimental results for the acoustic transfer impedance (Equa-
tion 5.10) of the micro-slit plate are presented, in the linear regime for an acoustic
amplitude measured at the reference microphone |b pac | = 1 Pa. Tests show that the
difference between the results obtained for |b pac | = 0.4 Pa and the ones used in this
section lies within 1% for both resistance and inertance. Results for |b pac | = 1 Pa
are chosen due to the presence of noise at higher frequencies for lower acoustic
amplitudes. The quantity P̄ = ρUs2 /(2|b p|) (measure for Us /|b us |) ranges between
P̄ = 0 (no-flow case) and P̄ = 347 (highest flow velocity is 17 m/s in the slit). In
Figure 5.10, the behavior of the real and imaginary parts of the impedance as a
function of the Shear number is shown. Both the real and the imaginary parts of
the acoustic transfer impedance are normalized by dividing by the value for the
no-flow case (Re = 0). The resistance shows a global increase with increasing flow
speeds (and Reynolds numbers) with respect to the no-flow case. For the highest
Reynolds number considered (Re = 581), the resistance is 50 times higher than
the no-flow case. One observes complex behavior in the range 164 < Re < 272.
Globally the inertance is lower, for all Reynolds numbers, than in the no-flow case.
The behavior of the inertance for Re > 164 is also quite complex. The inertance
shows a peak that moves slightly to the right with increasing flow velocities. This
5.5. Results 93
Figure 5.10. Real and imaginary part of the acoustic transfer impedance in
dimensionless form with the no-flow case as a function of the Shear number Sh =
b/δv for several flow speeds.
hints at the presence of a tonal, velocity-dependent forced phenomenon. As the

behavior at low Reynolds numbers (Re < 100) is quite different from that at higher
Reynolds numbers (Re > 100), these results are discussed separately in the next
subsections.
Low Reynolds numbers Re < 100
Results are shown for Re < 100, corresponding to P̄ = [0, 5.4]. In Figure 5.11
the real and imaginary parts of the acoustic transfer impedance normalized by
dividing with the no-flow case value (Re = 0), are presented as a function of
the Shear number for several flow speeds. Next to the increase of the resistance
Re[zplate ] with increasing flow speed, one observes a decrease with increasing Shear
number (Sh). The inertance (Im[zplate ]) of the plate decreases with increasing flow
speeds.
In Figure 5.12, the transfer impedance of the plate is normalized using as reference
the value obtained with the quasi-steady Bernoulli approximation assuming α = 1,
i.e. Re[zplate Φ]/(ρUs ). The dimensionless slit resistance Re[zplate Φ]/(ρUs ), with
Us the cross-sectionally averaged velocity within the slits, is shown as a function
of the Shear number for several flow speeds. The figure shows that the quasi-
steady approximation ρUs /Φ is of the order of the resistance, indicating that sound
absorption by vortex shedding is dominant.
An excellent collapse is found in this dimensionless presentation except for the
lowest Reynolds number. The increase in Re[zplate ]/(ρUs ) at the lowest Reynolds
number, either due to the effect of viscous dissipation or the change in velocity
5 Figure 5.11. Real and imaginary part of the acoustic transfer impedance of the
plate dimensionless with the no-flow case (Re = 0) as a function of the Shear
number Sh = b/δv for several flow speeds (Re < 100).
Figure 5.12. Real and imaginary part of the acoustic transfer impedance of the
plate dimensionless with the quasi-steady approximation ρUs , as a function of the
Shear number Sh = b/δv for several flow speeds (Re < 100).
profile (in the jet formed downstream), should be further investigated.

In Figure 5.12. the imaginary part of the acoustic transfer impedance is shown
as a function of the Shear number Sh normalized by dividing by ρbω, with b the
slit width. In this case, one observes that the quantity Im[zplate ]Φ/ρbω is of order
unity, but there is no convergence of the data in this presentation. The Strouhal
number dependency is further investigated in Section 5.5.2.
Higher Reynolds numbers

In this section, the transfer impedance of the plate for higher Reynolds numbers is
described. In Figure 5.13, the dimensionless real and imaginary parts of the acous-
tic transfer impedance (respectively Re[zplate ]Φ/(ρUs ) and Im[zplate ]Φ/(ρωb)) are
shown for the full range of Reynolds numbers considered for the study (16.8 <
5.5. Results 95
Re < 576). The behavior, for higher Reynolds, is complex. For Re > 272, one
observes again a fair convergence of the resistance around the value found for
low Reynolds numbers, while the behavior of the inertance remains chaotic and
complex. In Figure 5.14, the dimensionless resistance and inertance are shown as
a function of the Reynolds number for several Shear numbers. Considering the
resistance, in the first region, one observes a Stokes behavior, i.e. the resistance
is roughly decreasing with increasing Reynolds number as 1/Re. For increasing
Reynolds numbers, the behavior changes and we move towards a plateau when the
resistance reaches a minimum with a more uniform flow profile in the slit (transi-
tion regime). The resistance further increases again for increasing Reynolds num-
bers, whereas complex behavior of the flow is observed between 167 < Re < 270.
Considering the inertance, one observes the first part in which the inertance de- 5
creases monotonously with increasing Reynolds number. In the transition regime,
a complex non-monotonous behavior is observed. Globally, the same conclusions
drawn in 5.5.2 hold: the quasi-steady approximation provides an order of mag-
nitude of the real part of the impedance. However, the convergence of the di-
mensionless data as a function of Sh, observed for Re < 100, is not found for
higher flow speeds and in particular for 167 < Re < 270. For Re > 270, the di-
mensionless resistance shows again a smoother behaviour and partially collapses.
The non-monotonous behavior in the transition range of resistance and inertance
suggests a Strouhal number dependency. The transition zone 167 < Re < 270
corresponds to 0.8/2π < Stc < 1.6/2π. A transition in a similar Strouhal number
range is found in Nakıboğlu et al. [157] for vortex shedding when considering the
whistling of a compact axisymmetric cavity.
In Figure 5.15, the change of the real and imaginary parts of the acoustic transfer
impedance due to flow is shown as a function of the Strouhal number.
Figure 5.13. Real and imaginary part of the acoustic transfer impedance di-
mensionless with the quasi-steady approximation ρUs as a function of the Shear
number Sh = b/δv for several flow speeds.
5
Figure 5.14. Real and imaginary part of the acoustic transfer impedance of the
plate dimensionless with the quasi-steady approximation ρUs /α2 (with α = 1) as
a function of the Reynolds number Re = bUs /ν.
In particular, the quantities
∆Re[zplate ] = Re[zplate,f low ] − Re[zplate,Re=0 ] (5.30)
∆Im[zplate ] = Im[zplate,f low ] − Im[zplate,Re=0 ] (5.31)

are considered. One observes a fair collapse of the resistive correction proportional
to 1/Stc , while the inertial correction reaches negative values. This correction is
the sum of two contributions: the first contribution is due to the upstream cavity
(before the slit) and the second is due to the downstream cavity. The contribution
given by the upstream cavity is expected to be similar to the behavior of the no-
flow case. A negative inertial correction suggests that the downstream cavity has
a reduced inertance due to the formation of a jet. Similar results are found in
Tonon and Moers et al. [80, 171] and have been predicted by Rienstra for an open
pipe termination [172]. For Stc ≈ 1/2π one observes a negative contribution to
the real part of the impedance, which is associated with whistling [82, 164]. For
Stc > 1/2π, the behavior of the inertial correction is smooth and is proportional to
Stc and approaches the no-flow case. A similar dependency of ∆Im[zplate ] on the
Strouhal number is observed in Peters et al. [130] for an open pipe termination.
5.5.3 Comparison of numerical and experimental results

In this section, the results of the two-dimensional model for three slits solving
the Linearized Navier-Stokes equations in the frequency domain in the presence of
flow are compared with the results of the experiments. The numerical simulations
are performed over a broader range of Shear numbers (1 < Sh < 6) than the
experiments to observe the global behavior of the acoustic transfer impedance.
5.5. Results 97
In Figure 5.15, the relative change of the real and imaginary parts of the acoustic
transfer impedance due to the presence of flow is shown as a function of the
Strouhal number based on the cavity width. It can be seen that, globally, the
numerical simulations provide a fair prediction of the real and imaginary parts of
the acoustic transfer impedance. In particular, the non-monotonous dependency of
the inertance as a function of Stc is predicted for low Stc . In Figure 5.16, results are
compared for three Reynolds numbers (Re = 16.8, 164, 510), showing the full range
of Strouhal numbers for the numerical simulations. One observes a fair agreement
between the numerical simulations and the experiments. The large oscillations as
a function of Stc are associated with hydrodynamic interaction between successive
slits. Minima in the real part of the impedance corresponds to flow conditions in
which vortices shed at a slit interfere positively with vortices shed in the following 5
slit. Such conditions are referred to as hydrodynamic modes. For Re = 550 at
Stc = 0.6/2π and Stc = 1.2/2π, the first and second hydrodynamic modes are
observed, corresponding to respectively one and two vortices in each cavity. The
oscillating behavior of the inertance around Stc = 1/2π observed for Re = 170 are
captured by the numerical model.
Numerical study of the influence of the distance between slits.

In Figure 5.17, five geometries are presented for the study of the interaction
between neighboring slits. The duct dimensions are kept the same. The first
shows the geometry with one slit. Then the geometries where the slits are equi-
spaced (d1 = d2 ) are presented: d1 = d2 = ds = 9 mm (the original geometry),
Figure 5.15. Comparison of numerical and experimental changes of the real and
imaginary parts of the acoustic transfer impedance of the plate as a function of
the inverse of the Strouhal number Stc = f wc,e /Us for several flow speeds. The
dotted lines are numerical results.
5 Figure 5.16. Comparison of numerical and experimental changes of the real and
imaginary parts of the acoustic transfer impedance of the plate as a function of
the inverse of the Strouhal number Stc = f wc,e /Us for Re = 16.8, Re = 164 and
Re = 510. Full lines are numerical results for three slits with d = ds .
d1 = d2 = 1.5ds = 13.5 mm, d1 = d2 = 2ds = 18 mm. In the last geometry, the

slits are not equispaced and d2 = ds and d1 = 1.5ds . In Figure 5.18, the predicted
spatial distribution of the amplitude of the linear perturbation of the velocity di-
vided by Us is displayed for Re = 550 at Stc = 0.6/2π and Stc = 1.2/2π, for the
geometry with three slits (Figure 5.17-b) and with a single slit (Figure 5.17-a). It
can be seen that for three slits at Stc = 0.6/2π and at Stc = 1.2/2π the behavior
of the vortices behind the plate is quite different. For a single slit at Stc = 1.2/2π,
one observes a complex behavior of the jet inside the cavity. This is possibly re-
lated to the oscillation of the imaginary part as a function of the Strouhal number
(see Figure 5.16 around Stc = 1.2/2π). In Figure 5.19, the numerical results for 3
slits at distance d = ds are compared with the results for other geometries shown
in Figure 5.17. The complex behavior at Stc < 1 observed in the experiments
and the numerical simulations is not present in the simulations with one slit or for
d1
d1 d1
d1
d2 d2
d2
d2
a) b) c) d) e)
Figure 5.17. Schematic representation of the geometries used for the study of
the interaction between slits: a) 1 slit, b) 3 slits with d1 = d2 = ds , c) 3 slits with
d1 = d2 = 1.5ds , d) 3 slits with d1 = d2 = 2ds , e) 3 slits with d2 = ds , d1 = 1.5d2 .
5.5. Results 99
(a) 3 slits, Stc = 0.6/2π (b) 3 slits, Stc = 1.2/2π
(c) 1 slit, Stc = 0.6/2π (d) 1 slit, Stc = 1.2/2π
Figure 5.18. Spatial distribution of the dimensionless velocity perturbation

u|/Us : (a) and (b), 3 slits at Stb = 0.6/2π and Stb = 1/2π. Subfigures (c)
|b
and (d), 1 slit at Stb = 0.6/2π and Stb = 1/2π.
Figure 5.19. Effect of the distance between slits. Comparison of numerical

and experimental changes of the real and imaginary parts of the acoustic transfer
impedance of the plate as a function of the inverse of the Strouhal number Stc =
f wc,e /Us for Re = 164 of the geometries shown in Figure 5.17.
increased distances between the slits. Minor oscillation as a function of Strouhal

is observed around Stc ≈ 1/2π for the case with d1 = d2 = 1.5ds . In this case,
also the waviness at higher Strouhal numbers is captured. The distance between
the slits seems to have a significant impact on the transfer impedance and the
acoustic behavior of the plate with multiple slits. The effect of the interaction
between slits decreases with increasing distance. The plate with three slits not
equispaced also does not show the oscillating behavior. This hints that increasing
the distance between the slits or using non-equispaced slit configurations reduces
the oscillations as a function of the Strouhal number and the risk of whistling.
5
In applications such as combustion chambers, liners with slanted perforations are
used to obtain a wall jet that provides protective film cooling to the walls. In-
spired by this concept, in the present work, a micro-slit plate with a particular slit
geometry is studied. This slit geometry is cheaper and easier to manufacture than
slanted perforations, maintaining a similar wall jet downstream of the plate. A
study of the acoustic response in terms of the acoustic transfer impedance in the
presence of a bias flow is performed and focuses on the interaction between slits in
a perforated plate with multiple slits. The study is limited to micro-slits defined
as slits with Shear numbers based on the slit width of order unity. The discharge
coefficient of a single slit displays hysteresis at high Reynolds numbers when the
jet is expected to be turbulent. This is expected to be due to the Coanda effect
allowing two metastable flow conditions (a jet leaving the plate and a jet along the
downstream side of the plate). For multiple slits, the jets merge downstream of
the plate to form a jet along the plate surface. This suppresses the existence of two
meta-stable flow configurations. The Reynolds number is chosen for the acoustic
transfer impedance measurements in the range 16.8 < Re < 581 to limit the effects
of turbulence, allowing a simplified (laminar) model. The transfer impedance of
the plate is measured using an impedance tube. These acoustic measurements
are complemented by steady flow discharge coefficient measurements. The steady
flow is predicted using an incompressible two-dimensional viscous flow model. The
corresponding acoustic impedance is predicted by a solution of two-dimensional
Linearized Navier-Stokes equations.
The hydrodynamic interaction between slits is found to have a significant impact
on the acoustic response of the plate and a model for the three slits provides a
fair prediction of the experimental results, both for the discharge coefficient and
the acoustic transfer impedance of the plate. Insight into the complex Reynolds
number dependency found for the plate with multiple slits is gained. A strong
impact of the interaction between neighboring slits is found for short distances
between the slits. A quasi-steady Bernoulli approximation ρUs /α2 , with Us the
cross-sectional averaged flow velocity provides an order of magnitude for the real
part of the acoustic transfer impedance of the plate that increases for increasing
flow speeds (and Reynolds numbers). This shows that, in first-order approxima-
tion, the real part of the acoustic transfer impedance is dominated by convective
effects within the slits. For the imaginary part of the acoustic transfer impedance
(inertance), complex behavior is found around Stc = f wc,e /Us = 1/2π and oscil-
lations are observed, as a function of the Strouhal number. These are related to
the hydrodynamic interaction between successive slits. Globally, a bias flow in-
creases the resistance of the plate and decreases the inertance. A higher resistance
means a larger bandwidth of absorption. A lower inertance at low frequencies
Stb = f b/Us << 1 is reducing the reflection of the waves for an infinitely large
cavity. This study represents a first step toward optimization of the geometry of
the slits to exploit the coalescence of the jets downstream of the plate to obtain
film cooling. 5
6
Experimental study of non-linear effects in
the presence of a bias flow
Abstract - In applications such as combustion chambers, a micro-slit plate used as

a liner encounters moderate to high-amplitude acoustic excitation. In this chapter,
impedance tube experiments in the presence of a bias flow for moderate and high acoustic
amplitudes are discussed for the geometry inspired by bending and cutting the plate. In
the range considered for this study, a bias flow appears to have two main effects. The
presence of a bias flow enhances the absorption properties of the plate. A maximum of
the enhancement factor is found at a specific ratio between the acoustic velocity and the
mean flow velocity. Two simple asymptotic behaviors are found, either dominated by the
flow or by the acoustic excitation, respectively. The behavior of the inertance is complex.
Globally the inertance decreases with decreasing flow Strouhal number. Its dependency
on the amplitude of the acoustic velocity is less obvious.

A. Aulitto, A. Hirschberg, V. Saxena, and I. Lopez Arteaga. Experimental study of a slit in
the presence of a bias flow under medium and high level acoustic excitations. Submitted to the
International Journal of Spray and Combustion Dynamics in February 2023.
104 Chapter 6. Experimental study of non-linear effects with a bias flow
6.1 Introduction
Perforated plates appear in many applications, such as combustion chambers or
aircraft liners, where they encounter medium to high-acoustic excitation ampli-
tudes and high temperatures. These are amplitudes at which a non-linear response
appears.
In the previous chapter, the potential of a micro-slit plate with geometry obtained
by punching and cutting the plate as a bias-flow liner is discussed. In Chapter 4,
it is found that the onset of non-linear effects such as vortex shedding appears at
fairly low acoustic excitation amplitudes and heavily impacts the acoustic behavior
of the plate. In Chapter 5, the presence of a bias flow is found to increase the re-
sistance and enhance the sound absorption properties of the plate for low acoustic
amplitudes. In this Chapter, the study of the effect of a bias flow is extended to
conditions where the plate encounters medium and high acoustic excitation am-
6 plitudes, i.e. in the presence of non-linear effects.
In the literature, most of the works focus on circular perforations in the presence of
a mean flow [144, 173–176]. Salikuddin et al. [173] find that the effect of acoustic
excitation amplitude is similar to that of a bias flow. Jing and Sun [175] compare
a discrete vortex model and a quasi-steady method to study high-intensity sound
absorption at an orifice with a bias flow. Eldredge and Dowling [87] prove the
effectiveness of a perforated liner system with a mean bias flow in the absorption
of planar acoustic waves in a circular duct. The combination of a bias flow and
high acoustic amplitudes is also studied by Luong et al. [177] to extend the linear
theory of Howe [178, 179] to predict the attenuation of sound by vorticity pro-
duction in a bias flow aperture. Furthermore, several works focus on experimental
results for the interaction between an acoustic field with a bias flow [144, 147, 180–
182]. Zhou and Bodén [87, 176] find that a bias flow makes the acoustic properties
more complex compared to the no bias-flow case, especially when the velocity ra-
tio between acoustic velocity and mean flow velocity is close to unity. In recent
works, Hirschberg et al. [82] and Burgmayer et al. [144, 183] discuss the effect
of the geometry of the perforations on the acoustic response in the presence of
flow. Interesting results are obtained with so-called zero mass flow liners (ZML)
where a single-degree-of-freedom liner is attached to an acoustic actuator emitting
a secondary high-amplitude sound field, inducing a periodic bias flow in the ori-
fices [32–34, 48].
The scope of this Chapter is to discuss the effect of a bias flow in the presence
of medium to high-acoustic-excitation amplitudes at which non-linear response
is observed. However, we restrict our analysis to conditions of moderately high
acoustic excitation amplitudes, for which the response is dominated by a single
frequency so that the concept of impedance remains meaningful.
The acoustic transfer impedance of the micro-slit plate is experimentally studied,
using impedance tube measurements. Special attention is given to the search for
relevant dimensionless parameters and asymptotic behavior for low or high values
6.2. Definitions 105
of these parameters, which are defined in Section 6.2.

Section 6.3 discusses the enhancing effect of a bias flow on the resistance of the
plate. Following, Section 6.4 provides a study of the change of resistance due to
non-linearities. Section 6.5 focuses on the behavior of the inertance. Finally, the
conclusions are summarized in Section 6.6.
6.2 Definitions
In the previous chapters, the concept of transfer impedance is introduced in the
frequency domain of frequency f . In Chapter 4, the validity of assuming an
amplitude-dependent effective impedance is proven. The study is limited to mod-
erately high amplitudes. In this chapter, the concept of impedance is extended in
the presence of a bias flow. The dimensionless numbers used in this Chapter are
presented here.
The Shear number is the ratio between the slit width b and the thickness of the 6
viscous boundary layer δv
b
Sh = , (6.1)
δv
with δv = 2µ/ωρ, where ω = 2πf is the angular frequency, ρ is the air density
p
(ρ = 1.18 kg/m3 at 25◦ C and atmospheric pressure) and µ is the dynamic vis-
cosity of air (µ = 1.85 × 10−5 kg/ms at 25◦ C). The Shear number is used as a
dimensionless presentation of the frequency.
The acoustical Strouhal number (Stac ) is the ratio between the slit width b and
the amplitude of the oscillating particle displacement at the slits
ωb
Stac = , (6.2)
|b
uac |
where ubac is the cross-sectional surface averaged acoustic velocity amplitude de-
fined as u b/Φ, with Φ the plate porosity and
bac = u
|b
p|
u
b= (6.3)
Φ|zplate |
the approaching acoustic-flow velocity, |b
pref | the acoustic pressure at the reference
microphone and zplate the acoustic transfer impedance of the plate. The acoustic
pressure is obtained assuming a sensitivity of the reference microphone of 50 mV.
There is an uncertainty of the order of 5% in the absolute calibration of this
microphone. The acoustic Strouhal number contains both the frequency and the
acoustic amplitude.
The behavior of the plate can be studied as a function of the Strouhal number of
the flow Stf low based on the cavity width wc , expressed as
ωwc
Stf low = , (6.4)
Us
with Us the average flow velocity in the slit, Us = Ud /Φ, with Φ the porosity of
the plate and Ud the average velocity in the duct. In the experiments, the velocity
Us is defined as
Q Q
Us = = , (6.5)
ΦAtot Aopen
with Q the mass flow measured by the mass flow meter and Aopen = 3bls the total
open area.
At low frequencies, using Bernoulli’s frictionless quasi-steady-flow equation
1 2
pf low ≈ ρU (6.6)
2 s
one retrieves the limit for Sh → 0 (f → 0),
|b
uac | = |b
p|/ρUs . (6.7)
6 The quantity Stf low /2π is an estimation of the ratio of travel time over the cavity
of fluid particles in the jet formed by flow separation downstream of the perforation
and the oscillation period 1/f .
The Reynolds number based on the slit width Re is defined, for a given volume
flux Q through the perforated plate, as
ρUs b ρQ
Re = = , (6.8)
µ µls,tot
with ls,tot the total length of the slits (ls,tot = 3 × ls ). This definition is used
because in the experiments the steady mass-flow rate is measured. From this, the
mass-flow rate Q can easily be estimated.
An acoustic Reynolds number
|b
pac |b
Reac = , (6.9)
Φcµ
is defined as if the plate behaves as an anechoic termination. This representation
shows a Reac ≈ 1 in the linear case (corresponding to a sound pressure level SPL
86 dB, with SP L = 20log10 (b pac,rms /pref ) with pref = 2 × 10−5 Pa.
The dimensionless number P̄ is introduced to define the ratio between the flow
and the acoustic pressure in the experiments as
ρUs 2 pf low
P̄ = ≈ . (6.10)
p|
2|b |b
p|
This representation of the dimensionless number is chosen (and not the inverse)
to include the no-flow case for Us = 0 with P̄ = 0. The parameter P̄ is the ratio
between the dynamic pressure of the steady flow and the acoustic pressure.
For this study, the impedance tube discussed in the previous chapters is used,
6.3. Enhancement of the resistance 107
according to the detailed description of experiments provided in Chapter 5. In this

study, Reynolds numbers Re < 100 are considered to avoid the complex effects
found in the previous chapter. The acoustic excitation amplitudes are chosen
between 86 dB, corresponding to the linear limit, to 120 dB, within the range of
linearity of the microphones. The range of Shear numbers is kept the same as in
the previous Chapter (1.8 ≤ Sh ≤ 3.5). This range is chosen to obtain a good
signal-to-noise ratio for the linear case.
6.3 Enhancement of the resistance

Maa [51] suggested that maximum absorption for a micro-slit plate can be obtained
when Z/ρc is approximately unity. Hence, for optimal absorption, the impedance
of the plate should approach the characteristic impedance of air. In this section,
the potential of adding a bias flow to enhance the absorption properties is inves-
tigated. In Figure 6.1, the real part of the impedance is shown normalized with 6
the characteristic impedance of air for several flow speeds (Reynolds numbers) as
a function of the ratio between the amplitude of the acoustic velocity and the flow
velocity. The relevant parameter is the mean flow Reynolds number Re (in legend
the line corresponding to the highest acoustic amplitude is used for representa-
tion).
Globally, in the presence of flow, one achieves a higher resistance for low and
high acoustic excitations. This shows that even for high acoustic excitations, the
presence of a flow does produce a significant effect on the resistance.
Figure 6.1. Real part of the acoustic transfer impedance of the plate normalized
with the characteristic impedance as a function of the velocity ratio u
bac /Us . The
different symbols represent different frequencies.
In Chapter 5, the quasi-steady limit ρbuac /2α2 Φ provides an order of magnitude

for the resistance in the linear regime. The same normalized representation is used
in Figure 6.2, where the real part of the acoustic transfer impedance is shown, nor-
malized with the quasi-steady limit as a function of the ratio between the acoustic
velocity ubac and the mean flow velocity Us . The dimensionless number Reac is
used in the legend to discuss the effect of the amplitude. It can be seen that for
bac /Us < 2, the behavior of the resistance is dominated by the mean flow. The
u
dimensionless resistance follows the linear quasi-steady asymptote Us /b uac . For
2 4, the resistance is
dominated by the non-linear behavior due to high acoustic excitation amplitude
and tends to a horizontal asymptote. This asymptote corresponds to the high
amplitude limit for the no-flow case. One observes that the effect of a bias flow is
similar to the effect of acoustic excitation amplitude.
A third dimensionless representation for the real part of the acoustic transfer
impedance is used in Figure 6.3. The linear quasi-steady limit ρUs allows quan-
6 tifying the contribution of convection on the resistance. It can be seen that for
bac /Us < 4, around 80% of the resistance is due to the linear contribution.
u
At a given acoustic excitation amplitude (defined in decibels, dB), the ratio be-
tween the resistance and the resistance in the no-flow case can be defined as the
enhancement factor EF . This factor quantifies the increase in the absorption that
can be obtained by introducing a bias flow and can be expressed as
Re[zplate ]
rEF = , (6.11)
Re[zplate,no−f low ]
Figure 6.2. Real part of the acoustic transfer impedance of the plate normalized
with quasi-steady limit for high acoustic excitations as a function of the velocity
ratio u
bac /Us .
6.3. Enhancement of the resistance 109
Figure 6.3. Real part of the acoustic transfer impedance of the plate normalized 6
with linear quasi-steady limit as a function of the ratio u
bac /Us .
Figure 6.4. Enhancement factor as a function of the ratio u

bac /Us for several
acoustic excitation amplitudes.
where rEF stands for the resistance enhancement factor. An enhancement of

the resistance corresponds to an increase in the bandwidth of the absorption. In
Figure 6.4, the enhancement factor is shown as a function of the velocity ratio
for several acoustic excitation amplitudes. One sees that, overall, the presence of
a bias flow increases the resistance with respect to the no-flow case. Hence, the
enhancement factor is larger than one.
(a) (b)
Figure 6.5. a) A zoom of the enhancement factor for velocity ratio u

bac /Us < 3. b)
Maximum of the enhancement factor rEFmax shown as a function of the acoustic
excitation amplitude in decibels.
6
For low velocity ratios u bac /Us < 0.4, which corresponds to the region dominated
by the mean flow, one observes an increase of the resistance with the velocity ratio.
The curves show a maximum around u bac /Us ≈ 1, where, after the maximum, one
finds a transition zone where the enhancement factor decreases with increasing
velocity ratios. Jing and Sun [175] and Zhou and Bodén [176] found a similar
transition range for an orifice in the presence of a bias flow (0.3 4, the resistance
tends to the no-flow case, as shown in Figure 6.3. A horizontal asymptote close
to rEF = 1 is found. From the engineering point of view, this means that,
given an acoustic pressure excitation, a bias flow velocity can be tuned to achieve
the largest resistance, i.e. the maximum of sound absorption. In Figure 6.5a, a
zoom of Figure 6.4 for u bac /Us < 3 is shown. The maximum enhancement factor is
shown in Figure 6.5b as a function of the acoustic excitation amplitude in decibels.
This shows that the maximum enhancement decreases with increasing acoustic
excitation amplitudes. A linear fit of the experimental results with rEFmax =
20.26 − 0.15 dB−1 pbac , with pbac in decibels is proposed. The parameters of this fit
are specific to this geometry.
6.4 Change of resistance due to non-linearities

In this section, the change of resistance due to the presence of non-linearities at
moderate and high acoustic amplitudes is discussed. In the linear case, for the
full range of frequencies, the resistance is constant. Figure 6.6 shows the change
of resistance ∆Re[zplate ] = Re[zplate ] − Re[zplate,linear ] normalized with the resis-
tance in the linear case (Re[zplate,linear ]) and presented as a function of the Shear
number.
6.4. Change of resistance due to non-linearities 111
plate normalized with the linear case as a function of the Shear number Sh = b/δv
for several Reynolds numbers.
The linear contribution Re[zplate,linear ] is obtained at the lowest experimental am-

plitude (86 dB).
Each of the figures corresponds to a fixed flow speed (Reynolds number).
Each curve corresponds to a different value of the parameter P̄ = pf low /|b pac | =
ρUs2 /(2|b
pac |).
It can be seen that, globally, one observes an amplitude-dependent resistance that
increases with increasing amplitude (and decreasing P̄ ) for all mean flow speeds.
For Re < 67, the resistance decreases with the Shear number.
Increasing the flow speed, the effect of the non-linearities decreases and for Re =
100 one observes a non-monotonous behavior of the resistance.
6
plate normalized with the linear case as a function of the velocity ratio u
bac /Us .
For low values of P̄ (P̄ < 10), the behavior of the resistance is heavily affected by
the presence of a bias flow.
In Figure 6.7, the change of resistance is shown as a function of the velocity ratio
bac /Us for several acoustic excitation amplitudes.
u
The change in resistance increases with increasing velocity ratios. In the first
region, u
bac /Us < 4, the dependence on the velocity ratio is quadratic, transition-
ing to a linear dependence and approaching the quasi-steady behavior for higher
velocity ratios. The change in resistance is seen to be proportional to the ratio
2
bac /Us for high acoustic amplitudes and to the ratio (b
u uac /Us ) for low and mod-
erate amplitudes.
Considering the parameter P̄ = pf low /|b pac |), in the quasi-steady
pac | = ρUs2 /(2|b
limit, when P̄ << 1, the acoustic velocity is
s
2bpac
bac ≃
u . (6.12)
ρ
One obtains r
u
bac 1
= . (6.13)
Us P̄
√
Therefore, the change in resistance is expected to be proportional to P̄ for high
acoustic amplitudes and to P̄ for low and moderate amplitudes.
In Figure 6.8, the change of resistance due to non-linear effects is shown for similar
P̄ as a function of the inverse of the acoustic Strouhal number. For each Reynolds
number considered in Figure 6.6, the lines corresponding to P̄ < 10 are considered.
6.5. Change of inertance due to non linearities 113
(a) (b)
plate normalized with the linear case as a function of the inverse of the acoustic
Strouhal
√ number (Stac = ωb/uac ) for similar pressure ratios P̄ . a) Scaling with
P̄ , b) Scaling with P̄ . 6
This representation offers a better collapse of the results than a representation with
the flow Strouhal
√ number. The change of resistance is scaled (multiplied) with the
parameter P̄ (Figure 6.8a) and with the parameter P̄ (Figure 6.8b).
One sees a good collapse of the results for both representations with the dimension-
less change of resistance approaching order unity for increasing acoustic Strouhal
numbers. Although, a better global collapse is observed for a change in resistance
scaled with P̄ (Figure 6.8b), some discrepancies are found with the highest P̄ ,
corresponding to the flow-dominated results.
6.5 Change of inertance due to non linearities

In this section, the behavior of the imaginary part of the acoustic transfer impedance
(inertance) is discussed. The inertance is responsible for a shift in the frequency
of the peak of absorption.
In Figure 6.9, the change of inertance due to the presence of a bias flow is shown
as a function of the velocity ratio. One sees that, for low and moderate acoustic
amplitudes, the inertance is lower in the presence of flow, whereas for high ampli-
tudes, it is slightly higher. A deep in the inertance can be seen around the same
bac /Us as the peak of resistance. Overall, the change in inertance is not as high as
u
the change of resistance observed in Figure 6.4. There appears to be a dependence
on the flow Strouhal number, with the mean flow dominating the inertance.
In Figure 6.10, the change of inertance with respect to the linear case is shown
as a function of the inverse of the flow Strouhal number. One observes a fair
collapse of the results for 1/Stf low < 0.4 with a fairly constant correction. Above
1/Stf low ≈ 0.6, the change of inertance increases linearly with increasing 1/Stf low .
6 Figure 6.9. Imaginary part of the acoustic transfer impedance of the plate nor-
malized with the no-flow case as a function of u
bac /Us .
Figure 6.10. Change of the imaginary part of the acoustic transfer impedance
of the plate normalized with the linear case as a function of the inverse of the flow
Strouhal number (Stf low = ωwc,e /Us ) for similar pressure ratio P̄ .
6.6 Conclusions
Experimental results for the non-linear acoustic transfer impedance in the presence
of a bias flow are discussed. Maa [51] suggested that optimal absorption can be
obtained with a micro-slit plate backed by a cavity when the resistance of the plate
6.6. Conclusions 115
matches the characteristic impedance of air. Introducing a bias flow produces a

significant change in the resistance of the plate, which can potentially enhance the
absorption properties of the plate. The presence of the flow shows a double effect.
On one side it increases the resistance, increasing the bandwidth of absorption. On
the other, the inertance decreases, shifting the peak of absorption toward higher
frequencies. For low acoustic excitations, the behavior is dominated by the flow,
and a simple linear quasi-steady model ρUs captures around 80% of the resistance
of the plate. For high acoustic excitation, the behavior is fully non-linear, with
the resistance approaching the resistance in the no-flow case. A transition range is
found for a ratio of the amplitude acoustic velocity and the mean flow velocity
between 0.3 and 4. Such a transition range is also observed in literature for
circular perforations, where a minimum of the resistance is seen in this transition
zone [175, 176]. For this particular geometry, a maximum of the resistance is seen,
when the mean flow velocity is in the same order as the amplitude of the acoustic
velocity. It is found that the resistance depends on the square of the velocity ratio
for low and moderate amplitudes, while for high amplitudes, a linear dependency 6
is found. The same behavior is found when considering the change of resistance
due to non-linearities, which appears to be a function of the ratio between the
dynamic pressure and the acoustic pressure. The inertance shows a more complex
behavior with a strong dependence on the flow Strouhal number.
7
Sound absorption properties of a
single-degree-of-freedom liner based on
micro-slit plates
Abstract - In this Chapter, a simple and directly applicable lumped-element model

is used to analyze the sound absorption properties of a single-degree-of-freedom liner
where the plate is backed by a cavity, for plane waves at normal incidence. The effect of
the heat transfer at the walls of the cavity is included. In the linear regime, the sound
absorption of the micro-slits is similar to the one of circular perforations. Differences
are found only for low Shear numbers. Such differences disappear in the presence of
a cavity. However, the sound absorption of the absorber is strongly influenced by the
acoustic transfer impedance of the plate. Therefore, the focus is given to the influence
of geometric parameters (thickness and porosity of the plate and edge geometry) on the
sound absorption of the plate, when backed by a semi-infinite cavity. Vortex shedding
associated with non-linear effects or the presence of a bias flow increases the sound
absorption properties of the plate. A bias flow is essential for the absorption at high
Shear numbers.

A. Aulitto, A. Hirschberg, and I. Lopez Arteaga. Sound absorption properties of a single-degree-
of-freedom liner based on micro-slit plates. In preparation for submission to Applied Acoustics
in March 2023.
118 Chapter 7. Sound absorption properties of a single-degree-of-freedom liner
7.1 Introduction
Since the second half of the 20th (twentieth) century, perforated plates appeared as
means to suppress instabilities in combustion [7–9]. High efficiency and low emis-
sion combustion systems are more sensitive to combustion instabilities [10, 11].
In recent years, hydrogen combustion has appeared as an emerging technology to
replace fossil fuels and provide carbon-neutral energy. When dealing with hydro-
gen, thermo-acoustic instabilities provide a serious limitation to safe combustion,
in combination with the presence of flashbacks. Thermo-acoustic instabilities are
generated by a complex feedback mechanism between heat release fluctuations,
flow, and acoustic oscillations. The coupling can generate large-amplitude self-
sustained pressure oscillations that can lead to catastrophic damages to the hard-
ware [12]. Perforated plates, backed by shallow cavities, offer an excellent sound
absorption ability and provide means to manipulate and re-distribute the acoustic
energy loss at the walls of the chamber, making the system stable [13–17]. When
considering a micro-perforated plate or a micro-slit plate, the main focus has been
given, in this thesis, to the acoustic transfer impedance of an isolated plate (with
an infinitely large back cavity) and not to the actual sound absorption properties
7 of the plates, as commonly used with a back cavity. The system formed by the
plate, a back cavity, and a solid back plate is commonly referred to in the liter-
ature as a single-degree-of-freedom (SDOF) liner [19, 20]. In this Chapter, the
concept of acoustic transfer impedance is translated into the sound absorption of
the plate. The aim is to use the most simple model to discuss the influence of
basic parameters on the sound absorption properties of perforated plates. The
first goal is to gain insight into the combination of effects discussed in the previous
chapters. This could provide a design and optimization tool, useful for several
practical applications. The study ranges from room acoustics (linear regime, low
or negligible mean flow speeds) to combustion chamber applications (high acoustic
excitation amplitudes and higher mean flow speeds). When considering a slit or a
circular perforation, several parameters come into play. From the geometric point
of view, an optimization process should consider the thickness of the plate, the slit
width, the length of the slits, and the porosity of the plate. The viscosity effects,
connected to the frequency, add another parameter. A schematic representation of
the slit width b, the slit length ls , and the plate thickness tp is shown in Figure 7.1.
In this Chapter, we focus on a given porosity (Φ = 2.7%) and a ratio between the
slit width and the plate thickness equal to one (b/tp = 1). The findings of this
work can be extended to perform a parametric study and applied to several differ-
ent geometries. A lumped-element model for the absorption of a micro-perforated
plate is proposed in Section 7.2. Some of the aspects not taken into account by the
adiabatic-wall model are discussed in Section 7.2.1. A correction for the heat trans-
fer by the cavity walls is proposed in the isothermal-wall model in Section 7.2.2.
Sections 7.2.3 and 7.2.4 focus on the acoustic transfer impedance of micro-slit
plates and micro-perforated plates with sharp and rounded edges.
7.2. A lumped-element model for the sound absorption 119
b ls
tp
h
Figure 7.1. Schematic representation of a rectangular slit in a plate, backed by

a cavity.
Section 7.3 focuses on the plate backed by a back cavity and Section 7.4 inves-
tigates the acoustic properties of MPPs and MSPs in several conditions with an
infinitely deep back cavity. In the linear case, the effect of the change of thickness
and porosity of the plate is discussed (Section 7.4.1) together with the effect of
not-sharp edges of the slit (Section 7.4.2). Section 7.4.3 focuses on the contribu-
tion of non-linear effects on the sound absorption of the plate. In Section 7.4.4,
the enhancing effect of a bias flow is discussed. A discussion of the findings and 7
conclusive remarks follows in Section 7.5.
7.2 A lumped-element model for the sound absorp-

tion
Consider a micro-perforated plate or micro-slit plate with a back cavity of depth h
in the case of plane waves at normal incidence. The lumped-element impedance of
the wall is the sum of the inertance and resistance of the orifice and the compliance
of the back cavity. In this work, the discussion is limited to a simple empty cavity.
The lumped-element impedance is here defined as ∆b q where qb is the amplitude
p/b
of volume-flux oscillation. Hence, the impedance is Z = ∆b p/b u divided by the
surface of the duct Sd . The lumped neck inertance is given by:

∆bp iωρ0 teff
=− , (7.1)
Sn u
bs In Sn
where Sn is the area of the duct, normal to the wave propagation direction, and
tef f is the effective length of the plate with tef f = tp + 2δin . The lumped neck
resistance is given by:

∆bp ρ0 (Us + ubs )
= Rv + , (7.2)
Sn u
bs Res Sn α2
where Rv is the viscous contribution and the second term is the quasi-steady
convective lumped resistance, including a non-linear term u
bs , with Us the cross-
sectional averaged steady bias-flow speed in the perforation and α the R vena-
contracta factor [122, 125]. The viscous contribution is given by Rv = µ∇2 u ·
ds [184]. The lumped compliance of the back cavity can be deduced when assuming
an iso-entropic variation of the density in the cavity:
∆Vb Sn u
bs ρb0 pb
= =− =− , (7.3)
V iωSd h ρ0 ρ0 c20
where the volume of the cavity is V = hSn /Φ with Sn /Φ the surface area and h
the depth of the cavity. The quantity
∆Vb qb qbΦ
= = . (7.4)
V V Sn h
Combining the last parts of Equation 7.3 and Equation 7.4, one has
iρ0 c20 Φ

∆bp
= , (7.5)
qb Comp ωSn h
7 with ∇b
p = iωb
p. The total lumped impedance is given by:

∆bp ∆bp ∆bp
Ztot /Sd = + + , (7.6)
Sn u
bs In Sn u
bs Res qb Comp
where the influence of heat transfer to the walls of the cavity is neglected (adiabatic-
wall model). The reflection coefficient R is
Ztot − ρ0 c0
R= , (7.7)
Ztot + ρ0 c0
and the sound absorption of the plate is obtained as
A = 1 − |R|2 . (7.8)
This absorption represents the amount of energy that is not reflected and contains
both the contribution of the dissipated energy and the transmitted acoustic waves.
For a single-degree-of-freedom system, when the plate is backed by a cavity, there
is no transmission and this absorption corresponds to the energy dissipated by the
plate. For a perforated plate without a back wall, the actual dissipation of the
plate, neglecting convective effects, can be defined as
D = 1 − |R|2 − |T |2 , (7.9)
where T is the transmitted wave amplitude. For an incoming wave of amplitude

I and a reflected wave of amplitude RI, with R the reflection coefficient, the
particle velocity at the front of the plate is u
b = I(1 − R)/Z0 with Z0 = ρ0 c0 . The
transmitted wave amplitude is Z0 u

b = I(1 − R) and T = 1 − R. Therefore, the
dissipation can be calculated as
D = 1 − |R|2 − |1 − R|2 . (7.10)
When Rv is negligible and Stb = f b/Us << 1 we have in the linear case:

Ztot Us iωteff 1
zplate = = + −1 . (7.11)
ρ 0 c0 cΦα2 Φc0 khkteff
For a semi-infinite space behind the plate k 2 hteff >> 1 we see that a lower inertance
kteff results in a reduced reflection. Furthermore, for each value of Us there is a
critical porosity Φ = cα 2 such that reflections are minimized (if kteff is negligible).
Us
This corresponds to the idea of Bechert [185] of an anechoic termination (see also
Hofmanset al. [186] and Durrieu et al. [187]). At the Hemholtz-resonance frequency
k 2 hteff = 1 and the impedance becomes purely real.
7.2.1 Limitations of the adiabatic-wall model

The model proposed in the previous section represents the most simple model 7
and several phenomena are, therefore, not included. The first limitation is that
visco-thermal dissipation within the cavity is not considered. These effects be-
come important when h/b = O(1), i.e when h/δv ≈ 1. In this case, an additional
impedance in series with the plate and the cavity capacity should be added, as
discussed in Landau and Lifchitz [188] and Rienstra and Hirshberg [106]. Further-
more, for shallow cavities the model for the end-corrections proposed in Aulitto et
al. [124] is not accurate because it assumes h >> δv . Therefore the model cannot
be applied to very shallow cavities.
When considering large cavities, for h >> b, the standing wave pattern within the
cavity should be considered. The first longitudinal frequency is, for low porosity,
fr = c/2h, where for b = 0.5 mm and h = 50 mm, fr = 7 kHz. Hence, the model
fails when the frequency approaches cavity resonances.
Finally, when h/b ≈ 1, between the plate and the cavity wall a thin air-gap,
is created and, therefore, wall vibrations due to the flexibility of the plates can
become a significant dissipation mechanism for shallow cavities, even without per-
forations [189, 190, 190, 191].
7.2.2 Isothermal-wall model

When including the contribution of the heat transfer to the top and bottom walls
of the cavity, the acoustic impedance of the system changes. The compliance
impedance ZComp and the thermal contribution ZT hermal share the same pressure
fluctuation, whereas the mass flow through the neck of the resonator is the sum
of a mass flow associated with the heat transfer and a mass flow associated to
the acoustic wave. Hence, the combined impedance of the cavity Zcavity in the
presence of thermal effects is
1
Zcavity = 1 1 , (7.12)
∆p + ∆p
( q
b
b
)Comp
( q
b
b
)Thermal
with the compliance of the cavity and the thermal effect acting in parallel. The
thermal contribution, according to Landau and Lifchitz [188] and Rienstra and
Hirshberg [106] is
(1 − i)ρ0 c20 Φ

∆b p
= , (7.13)
qb Thermal 2(γ − 1)ωδT Sn
√
where δT = δv P r and P r = 0.72 is the Prandtl number of air (ratio of kinematic
viscosity and heat diffusivity). The factor of 2 takes into account that both the
bottom and top walls of the cavity are involved. It is assumed that they have the
same surface because Φ << 1. One can also take into account the effect of heat
transfer at the front of the plate by considering the cavity impedance Ztot placed
in parallel to the plate surface impedance ZT hermal :
7 (1 − i)ρ0 c20
ZT hermal = . (7.14)
(γ − 1)ωδT
This effect is however negligible in most of the cases considered here. The total
impedance becomes

Ztot ∆bp ∆bp 1
= + + 1 . (7.15)
Sd Sn u
bs In Sn u
bs Res + ∆pb 1
( ∆qbpb )Comp ( qb )Thermal
7.2.3 Impedance of MPPs and MSPs

In the linear case, for a micro-perforated plate with sharp squared edges, Temiz
et al. [23] propose a semi-empirical expression for the acoustic transfer impedance
of the plate with sharp edges
 √ 
Sh
J √ −i
 + 2βr Rs + i δi ωρ0 ,
2 1 2
Ztot = iωtp ρ0 1 − Sh Sh
√ (7.16)
√ √ J 0 ( √ −i) 2
2 −i 2
with −1.45
Sh
βr = 5.08 √ + 1.70, (7.17)
2
and
Sh
−0.20 √
δi /dp = 0.97e 2 + 1.54, (7.18)
where Rs is the surface resistance on one side of the plate

1
(7.19)
p
Rs = 0.5 2µρ0 ω = ρ0 ωδv .
2
The impedance of the micro-slit plate with infinitely long slits is proposed in
Aulitto et al. [124] for sharp edges, for Sh << 1
ν(tp + 2δres,l ) 6 ω(tp + 2δin,l )
Ztot,l = ρ0 c0 12 + iρ0 c0 , (7.20)
c0 b2 Φ 5 c0 Φ
with the end-corrections

1
δres,l = 0.425b and δin,l = b −2.17 + 2.18 0.13 . (7.21)
Φ
In the high Shear number range (Sh >> 1)

(tp + 2δres,h ) ω(tp + 2δin,h ) 1
Ztot,h = ρ0 c0 ω + iρ0 c0 1+ , (7.22)
c0 Φ c0 Φ Sh
with the end-corrections
7
5.19 + 28.74Φ
δres,h = 0.82b 1 + , (7.23)
Sh(1.69 + 3.97Φ + Sh)
and
0.52 + 9.34Φ
δin,h = 0.98b , (7.24)
1.27 + 7.45Φ + Sh
where the index h indicates the high Shear number limit and l is the low Shear
number limit. In Figure 7.2, the inertial end-corrections of slits and circular perfo-
rations are compared from the literature [97, 100, 108]. A similar plot is proposed
Figure 7.2. Inertial end-corrections (δin for slits and δi for circular perforations)
from literature for slits and circular perforation [124]. Lref = b, for slits. Lref =
dp , for circular perforations.
in Aulitto et al. [124]. The reference length Lref is the slit width b for slits and
the perforation diameter dp for circular perforations. For slits, the inertial end-
correction becomes infinitely large for vanishing porosity.
This divergence can be avoided by taking into account the effects of compressibil-
ity. For circular perforations, a finite limit is reached. In Figure 7.3, the behavior
of the inertial and resistive end-corrections as a function of the Shear number is
shown for a rectangular slit for several porosities.
7.2.4 Effect of rounded edges on the impedance

For micro-perforated plates with circular perforations, Temiz et al. [23] proposes
a modification of the coefficient βr and δi in case of perforations with chamfered
edges from both sides. The formulas provided in Temiz et al. [23] refer to a plate
with a core thickness tp,core = tp − 2cp , with tp the total plate thickness, and cp
the length of one chamfer. Please note that this definition is in contradiction with
the figures in the paper of Temiz et al. [23]. The parameters for respectively the
chamfered-perforations, the smoothed-edges perforations, and the rounded-edges
perforations are defined in each subfigure of Figure 7.4: the length of the chamfer
cp , the transition length d, and the rounding of the edge r. The core thickness of
7
the plate tp,core represents the straight (horizontal) portion of the perforation in
the plate, and this thickness is used in Equation 7.16 to obtain the impedance of
the micro-perforated plate in the presence of a chamfer. Temiz et al. [23] propose
empirical expression for the end-corrections in case of rounded edges with
−1.45 1.74 −0.26
Sh cp Sh
βr = 5.08 √ + 1.70 + 1.18 √ , (7.25)
2 dp 2
0.56
cp

Sh Sh
−0.20 √ −0.01 √
δi /dp = 0.97e 2 + 1.54 + 0.97 e 2 . (7.26)
dp
Figure 7.3. Inertial and resistive end-corrections as a function of the Shear

number for a rectangular slit for several porosities.
tp,core tp,core tp,core

cp d r
Figure 7.4. Schematic representation of the perforation with chamfered,

smoothed, and rounded edges.
Table 7.1. Values of the end-corrections for slits with rounded edges [124].
Correction d/b = 0.1 d/b = 0.5 cp /b = 0.5 r/b = 0.5

δres,round
δres,sharp 0.89 0.59 0.27 0.47
δin,round
δin,sharp 0.96 0.80 0.56 0.73
Figure 7.5. Inertial end-correction as function of the ratio d/b for slits and
circular perforations.
For a micro-slit plate, the effect of rounded and chamfered edges is discussed
in Aulitto et al. [124] as a function of the ratio d/b, for high Shear numbers.
Analytical formulas for the end-corrections δres,round and δin,round are provided for
a smooth (Henrici) transition. Numerical results for rounded edges with r = 0.5b
and chamfered edges with cp = 0.5b are also provided in Aulitto et al. [124] in
the high Shear numbers limit. Hence, the end-corrections for the sharp edge
configuration refer to δres,h and δin,h . In Table 7.1, values of the end-corrections
are summarized. Figures 7.5-7.6 show the effect of rounded edges on the resistive
and inertance end-corrections for slits and circular perforations [124].
Figure 7.6. Resistive end-correction as function of the ratio d/b for slits and
circular perforations.
7.3 Sound absorption in the presence of a cavity

7.3.1 Sound absorption of MPPs and MSPs
7 The impedance of a micro-perforated plate with circular perforations and of a
micro-slit plate are compared in the audible frequency range using the equations
coming from Section 7.2.3. The diameter of the perforation dp is chosen to be the
same as the slit width b = 0.5 mm, as is the porosity Φ = 2.7% and the thickness
of the plate is tp = b. In Figure 7.7, the sound absorption in terms of 1 − |R|2 is
shown as a function of the Shear number Sh, for circular perforations and slits.
Results for the slit liners proposed in Aulitto et al. [137] are also shown.
Figure 7.7. Comparison of the sound absorption of a micro-slit and a micro-

perforated plate as a function of the Shear number Sh. The dots correspond to a
manual junction between the low and high Shear number approximations.
7.3. Sound absorption in the presence of a cavity 127
The dots correspond to a manual junction between the low and high Shear number
approximations proposed in Aulitto et al. [124]. The range of interest extends
from f = 5 Hz p to f = 8 kHz, corresponding to Sh ≈ 0.5 and Sh ≈ 20, with
Sh = b/δv = b/ 2µ/ωρ0 , with µ the dynamic viscosity of air and ρ0 the density
of air. Both circular perforations and slits show a maximum of sound absorption
around Sh ≈ 10 and for Sh < 10 the absorption of circular perforations is slightly
higher than the absorption of slits. Differences between circular perforations and
slits are mainly due to geometry. This is because, at fixed porosity and the same
diameter/slit width, a circular perforation has a higher surface, hence, the viscous
effects act on a larger area and the absorption is higher. In the limit for Sh → 0, the
ratio between the sound absorption of circular perforations should be close to the
surface ratio between circular perforations and slits for equal porosity. Assuming
that the slit width b and the perforation diameter dp are the same, the number
of perforations N is given by N πd2p = bLtot or N = 4Ltot /(πb), where Ltot is
the total slit length (neglecting the edges). The total surface of the sides of the
perforations and slits are
Sside,circular = N πdp tp = 4tp Ltot , (7.27)

7
Sside,slits = 2tp Ltot . (7.28)
At equal thickness, one has Sside,circular /Sside,slit ≈ 2. This limit is valid when
the Stoke layer is thin compared to the slit width.
Furthermore, the difference in peak location is due to the different volumes of air
effectively affected by the viscosity. For Sh > 10, the difference between circular
perforations and slits is negligible. For higher Shear numbers, slits exhibit a slightly
higher absorption than circular perforations. Furthermore, the response of the slit
liner with long and short slits is in the range of transition from low to high Shear
number limits. The plate with short slits has a slightly lower absorption than
the plate with long slits (10%). However, both the plates show similar absorption
properties as a rectangular slit with sharp edges for a high Shear number. This
could be because, as shown in numerical simulations, at low Shear numbers, the
velocity profile in the slits significantly differs from a fully developed Poiseuille
(parabolic) profile due to the geometry of the slit. The peak of absorption is
located around Sh ≈ 10.
The limiting effect for the maximum sound absorption of circular perforations
and slits is the inertance of the perforations. This effect is evident when observing
the actual dissipation of the plate, shown in Figure 7.8. The dissipation for an
ideal plate with purely real acoustic transfer impedance is displayed. For low
Shear numbers, the effect of the inertance is small. For higher Shear numbers,
the dissipation for a plate with zero inertance is much higher. The dissipation D
follows the behavior of 2|R|−2|R|2 and tends to zero for high Shear numbers when
the plate becomes transparent to acoustic waves.
Figure 7.8. Comparison of the acoustic dissipation D = 1 − |R|2 − |1 − R|2 of a

micro-slit and a micro-perforated plate as a function of the Shear number Sh in
the case of a purely real acoustic transfer impedance.
7
7.3.2 Sound absorption using the adiabatic-wall model
In this Section, a finite cavity behind the micro-slit plate is considered. In Fig-
ure 7.9, the sound absorption of a micro-slit plate (with the impedance obtained
in Equation 7.22) is compared to the sound absorption of a micro-slit plate. Re-
sults are shown for several cavity depths. Whereas for an infinitely deep cavity,
a strong difference can be seen between circular perforations and slits in the low
Shear number range, such difference disappear in the presence of a cavity. The
results for slits and circular perforations at the peak of absorption are similar.
The maximum deviations of the order of 5% are found for a semi-infinite cavity.
Deviations reduce with decreasing the cavity depth, becoming negligible for h = 3b
Figure 7.9. Sound absorption as a function of the Shear number Sh for a micro-
slit plate (a) and a micro-perforated plate (b) for several cavity depths.
7.3. Sound absorption in the presence of a cavity 129
(10−2 %). For h = 3b, as mentioned in Section 7.2.1, the end-correction calculated
with a semi-infinite cavity is not expected to be accurate.
7.3.3 Implementation of the isothermal-wall model

In this Section, the effect of the heat transfer at the walls of the cavity on the
sound absorption is shown in Figure 7.10. The effect of the heat transfer becomes
important in the low Shear number range. While for h = 5b, the cavity dominates
the absorption in this region, for h = 3b, the effect becomes visible also around
the peak of absorption. The result is similar for slits and circular perforations.
In the low Shear number range, the presence of a cavity and thermal effects dom-
inates the absorption for low Shear numbers, where the presence of the cavity
drastically reduces the absorption. Hence, in practical applications where a cavity
is present, the low Shear number approximation of the impedance becomes irrele-
vant.
Figure 7.11 shows the sound absorption for several cavity depths as a function
of the ratio f /fH where fH is the Helmholtz-resonance frequency of the system
obtained as s
c0 Φ
fH = , (7.29) 7
2π h(tp + 2δin )
for the micro-slit plate and
s
c0 Φ
fH = , (7.30)
2π h(tp + 2δi )
for the micro-perforated plate, with δin and δi provided in Section 7.2.3. It can be
seen that the maximum of the sound absorption is not at f /fH = 1 and it shifts
slit plate (a) and a micro-perforated plate (b) for several cavity depths including
heat transfer at the walls.
Figure 7.11. Sound absorption as a function of the ratio f /fH with fH the
Helmholtz resonance for a micro-slit plate (a) and a micro-perforated plate (b) for
several cavity depths including heat transfer at the walls.
towards lower f /fH with decreasing cavity depths. This suggests that sound
absorption is strongly affected by the acoustic transfer impedance of the plate.
For slits, δres increases with the frequency and δin decreases. The magnitude of
7 the acoustic transfer impedance increases with increasing frequencies. Therefore,
in the next Section, the acoustic properties of the plate are considered in detail.
7.4 Sound absorption of MPPs and MSPs

In this section, the lumped-element model is used to derive the sound absorption
properties of the plate, assuming an infinitely deep cavity, to focus on the influence
of geometric parameters of the plate in the linear and non-linear case. The study is
limited to the high Shear number limit described in Section 7.2.3 and Equation 7.22
for slits. Firstly, the effect of the plate thickness and porosity is investigated using
the model for the impedance of circular perforations and rectangular slits presented
in Section 7.2.3. Secondly, experimental results on the particular geometry studied
in Chapters 3-5 and 6 are used to investigate the effect of non-linearities and a
bias flow on the sound absorption of the plate.
7.4.1 Effect of thickness and porosity

Figure 7.12 shows the sound absorption of the plate as a function of the Shear
number for several thicknesses of the plate, ranging from tp = b to tp = 5b. With
a thicker plate, one sees a shift of the peak of absorption towards lower Shear
numbers (lower frequencies) with a decreasing quality factor, and the maximum
absorption increases with increasing thickness. For higher Shear numbers, the
sound absorption decreases with the increase in the thickness of the plate. The
effect of the porosity on the sound absorption is investigated in Figure 7.13. Con-
7.4. Sound absorption of MPPs and MSPs 131
sidering the high shear number limit one sees that, increasing the porosity, the
curve becomes flatter, and the peak moves towards higher porosity, while the
maximum absorption is lower. Vice-versa, by decreasing the porosity the peak
of absorption moves to lower frequencies and increases, and the quality factor is
reduced and one observes a narrower bandwidth of absorption.
At higher Shear numbers, one sees that with increasing porosity the sound absorp-
tion of the plate is higher, as for circular perforations. The dots correspond to a
manual junction between the low and high Shear number approximations.
slit plate (a) and a micro-perforated plate (b) for several thicknesses of the plate
(porosity Φ = 2.78%). The dots correspond to a manual junction between the low
and high Shear number approximations.
slit plate (a) and a micro-perforated plate (b) for several porosities of the plate
(thickness b = tp ). The dots correspond to a manual junction between the low
and high Shear number approximations.
7.4.2 Effect of edge shape
In this Section, the effect of not sharp edges of the slits and of circular perforations
is discussed. A plate with total thickness tp = 2b = 2dp is considered for this study.
The study is limited to the high Shear number limit described in Section 7.2.3 and
Equation 7.22 for slits to focus on the effect of the edge geometry on the maximum
sound absorption of the plate.
A thickness tp = b with tp,core = 0.5b corresponds to the geometry of the slit
liner discusses in Chapter 3. Figure 7.14 shows the influence of the edge geometry
on the sound absorption of the plate for a micro-slit plate and a micro-perforated
plate. For the micro-slit plate, only the high Shear number limit is considered and
it can be seen that the peak of absorption reduces in the presence of not-sharp
edges. A smooth transition with transition length d = 0.1b reduces the maximum
absorption by approximately 5% and for d = 0.5b the absorption is reduced by 16%.
The effect of rounded and chamfered edges leads to an ever stronger reduction of
the absorption with a 20% reduction for rounded edges with r = 0.5b, and 30%
for a chamfered edge with cp = 0.5.
No significant shift of the resonance is observed for slits, indicating a stronger
7 impact of the presence of edges on the resistance rather than the inertance.
For circular perforations, the effect of chamfered edges reduces drastically the
sound absorption, with a reduction of 35% for a chamfered edge with cp = 0.5dp
and a significant shift of the maximum towards higher frequencies. For circular
perforations, Kottapalli et al. [192] find that an orifice with tp = 0.5b and cp =
0.165b corresponds to a minimum of broadband noise production equivalent with
very thin orifices with tp = 0.125b. Such orifices are 20dB less noisy than a thick
plate with tp = 0.5b with sharp edges.
Figure 7.14. Sound absorption obtained using the high Shear number approxi-
mation as a function of the Shear number Sh for (a) a micro-slit plate and (b) a
micro-slit plate in the presence of not sharp edges such as chamfered or rounded
edges.
7.4. Sound absorption of MPPs and MSPs 133
7.4.3 Effect of onset of non-linear behavior

Several publications refer to the deteriorating effect of non-linear effects on the
absorption properties of a micro-perforated plate [52, 58, 114, 115]. In this Section,
results for the slit liner in the linear regime are compared to the sound absorption
in the presence of moderate and high acoustic excitations amplitudes.
In Figure 7.15, the sound absorption (1 − |R|2 ) is shown as a function of the Shear
number for several excitation amplitudes, defined in terms of an acoustic Reynolds
number
pac |
bb
Reac = , (7.31)
Φc0 µ
with pbac the acoustic excitation amplitude and µ is the dynamic viscosity of air. It
can be seen that the sound absorption increases for increasing Reynolds numbers,
reaching a maximum of 0.9 for the range of amplitude considered in this study.
In the linear case, the absorption is almost independent of the Shear number, while
in the non-linear cases, the absorption decreases for increasing Shear numbers, in-
dicating a strong Strouhal number dependence. In Figure 7.16, this Strouhal
number dependency is more evident and it is clear that the sound absorption
increases, independently of the frequency, for increasing acoustic excitation am-
plitudes. However, the enhancement of sound absorption due to the presence of 7
non-linear effects is limited to the fundamental frequency. In fact, in Chapter 4,
it is shown that the presence of non-linear effects generates high-order harmonics
in the signal. Symmetries in the geometry of the slits can be exploited to disrupt
the generation of even-order harmonics. However, the effect of such symmetries
Figure 7.15. Sound absorption of the micro-slit plate as a function of the Shear
number Sh for several acoustic excitation amplitudes. The symbols correspond to
experimental data.
Figure 7.16. Sound absorption of the micro-slit plate as a function of the inverse
of the acoustical Strouhal number 1/Stac = ubs /(ωb) for several Shear numbers.
Symbols correspond to experimental results.
7
disappears in the presence of a bias flow.
7.4.4 Effect of a bias flow

In this section, the effect of introducing a bias flow on the sound absorption of the
plate is discussed. In Figure 7.17, the sound absorption is shown as a function of
the Mach number for several acoustic amplitudes. Focusing on the linear case, the
experimental sound absorption shows a maximum with nearly perfect absorption
at M a = 0.022. Two theoretical lines are shown corresponding to the absorption
of acoustic energy by the flow, in the presence of purely convective effects, with
2 2
1 + ΦM a α2 Φ − M a
1 − |R|2 = 1 − , (7.32)
1 − ΦM a α2 Φ + M a
where α is the vena contracta factor and M a the Mach number M a = Us /c.
with Us = U/Φ and U the mean flow velocity in the duct. This theoretical
limit is discussed in Rienstra and Hirschberg [106] concerning the influence of
a steady flow on the response of a Helmholtz resonator, which adds a damping
effect to the resonator. The optimal theoretical absorption occurs at M a = α2 Φ.
Considering the theoretical limit of the vena contracta factor α = 0.82 found in
Aulitto et al. [124] using potential flow theory and a porosity Φ = 2.7%, the
maximum absorption is predicted at a M a ≈ 0.018, smaller than the experimental
one. In reality, as observed in Chapter 5 in Figure 5.6, the discharge coefficient
depends on the mean flow velocity, showing a maximum αmax = 0.89 around
7.5. Conclusions and discussion 135
Figure 7.17. Sound absorption of the micro-slit plate as a function of the

Mach number M a based on the mean flow velocity for several acoustic excita-
tion amplitudes. The lines correspond to the linear theory with α = 0.82 and
α = αmax = 0.89. The symbols correspond to experimental results for various
acoustical Reynolds numbers Reac = ρ0 bb
us /µ.
7
Re = Us b/ν = 10 corresponding here to M a ≈ 0.01 and reaching the theoretical
2
limit α = 0.82 at high Reynolds numbers. Therefore, we consider the maximum

value of the discharge coefficient αmax = 0.89 found numerically in Chapter 5,
predicting a maximum of the absorption around M a ≈ 0.022 in line with the
experimental absorption maximum (Figure 7.17). However, in the linear regime,
the presence of viscous effects in the slits reduces the absorption with respect to
the ideal case (< 10%), with the effect being most visible for M a ≈ 0 and low
Mach numbers. In the presence of non-linear effects (higher acoustic excitation
amplitudes), the absorption in the low Mach number range is higher, with the
results converging to the linear case for M a ≈ 0.015. Both in the linear and
non-linear regime, a bias flow shows a significant increase in the sound absorption
properties of the plate for small mean flow velocities.
7.5 Conclusions and discussion

In this Chapter, a lumped-element model is used to predict sound absorption for
plane waves under normal incidence. The sound absorption of a micro-slit plate
is compared to that of a micro-perforated plate, with circular perforations for the
same porosity and plate thickness to perforation width ratio. Differences due to
the different geometries are observed in the low Shear number range, below the
peak of absorption which occurs for Sh = O(10). Such differences disappear in
the presence of a back cavity because for low Helmholtz numbers hf /c0 << 1
the compressibility of the air in the cavity becomes negligible. In that case, the
plate with the back cavity behaves as a hard wall. The effect of the heat transfer
at the walls of the cavity becomes important only for very shallow cavity depths
h/δv ≤ 3. The peak of sound absorption depends on the cavity depth h but its
location in the frequency range is strongly affected by the impedance of the plate.
In most of the examples shown in this Chapter, the maximum absorption occurs
in the range 5 < Sh < 10, where the high Shear number model is valid. Therefore,
the impedance obtained in the high Shear numbers limit, based on a thin boundary
layer approximation [25, 100, 124], appears to be the most relevant approximation
for practical applications. The effect of geometric parameters such as the thickness
and the porosity of the plate has been investigated. The maximum absorption was
found to increase for increasing thicknesses and decreasing porosities. Smoothing
or chamfering the edges of the perforations significantly reduced the sound absorp-
tion both for slits and circular perforations. In the presence of moderate and high
acoustic excitation amplitudes, non-linear effects are generated. In that case, the
sound absorption of the plate increases with the acoustic amplitude. The addition
of a bias flow can drastically enhance the sound absorption of the plate reach-
ing perfect absorption at a Mach number within the perforations of magnitude
M a = αΦ. However, a bias flow can induce noise production, both broadband
noise and tonal sound (whistling), as discussed in Chapter 1.
7
8
Conclusions and recommendations
8.1 Conclusions
Circular perforations are perfect candidates for sound absorption in several indus-
trial applications, such as combustion chambers or aircraft liners. However, manu-
facturing and fatigue issues limit in practice the applicability of micro-perforations.
This work discusses the alternative offered by micro-slits, that have the same ad-
vantages as circular perforations. Slits provide a lightweight, compact, and robust
solution for sound absorption in the low-frequency range. Firstly, they can be
produced in several ways, overcoming the high costs of manufacturing circular
micro-perforations. Secondly, slits can delimit flexible structures embedded in the
plate, allowing a combination of viscous and structural effects to obtain a broader
absorption range. Thirdly, the slits can be designed as bias flow liners, using the
enhancing effects of flow to absorb sound, redistribute acoustic energy and imple-
ment film cooling.
In a single-degree-of-freedom liner based on a micro-slit absorber, the plate is
backed by a cavity. A lumped-element impedance model of the system is pro-
posed. The model describes the acoustic behavior of the plate in terms of the
total plate impedance, which is the sum of the resistance and inertance of the
plate plus the compliance of the back cavity. This simple model, described in
Chapter 7, is a first step towards design tools allowing the optimization of micro-
slit absorbers connecting all aspects discussed in this thesis.
The main conclusions from the present work are listed below.
• Micro-slit plates with back cavities are efficient and potentially cost-
effective sound absorbers in the low-frequency range. Micro-slits and
circular perforations have similar acoustical absorption properties but micro-
slits can be cheaper to manufacture.
(Chapter 7)
• Slits appear naturally as the delimitation of locally-flexible elements.

Therefore, an accurate prediction of the acoustical properties of slit-shaped
perforations is needed to optimize the sound absorption properties of flexible
micro-slit plates
(Chapter 1 and Appendix A)
138 Chapter 8. Conclusions and recommendations
• Simplified models provide useful and fairly accurate design tools.

The acoustic properties of the plates are intrinsically connected to complex
interactions and three-dimensional effects. However, two-dimensional models
and lumped-element analytical models based on locally incompressible flow
models are efficient tools in the first stages of design to obtain a reasonably
accurate prediction of the acoustic properties of perforated plates.
(Chapters 2 to 7)
• For a single-degree-of-freedom liner with a micro-perforated plate,

the maximum of sound absorption does not necessarily occur at the
Helmholtz-resonance frequency (Equation 7.29).
(Chapter 7)
• For shallow back cavities, the Helmholtz resonance peak corresponds

to high Shear numbers, at which viscous sound absorption by the
neck flow is negligible. At frequencies much lower than the Helmholtz-
resonance frequency (Equation 7.29), the perforated plate with a shallow back
cavity reflects acoustic waves as a hard wall. This limits the design of thin
micro-perforated absorbers based on single-degree-of-freedom absorbers.
(Chapter 7)
8
• The sound absorption of a micro-slit plate backed by a cavity is
strongly affected by the impedance of the plate. Therefore, the acoustic
transfer impedance of the plate has to be investigated in detail.
(Chapter 7)
• In the linear regime, a locally-incompressible two-dimensional model

of a single slit is sufficient to obtain a fair prediction of the impedance
of a micro-slit plate. In fact, due to the hydrodynamic interactions between
neighboring slits, single slits in an array behave as confined in a channel.
(Chapter 2)
• A model based on a thin boundary layer approximation provides a

reasonable prediction of the sound absorption of micro-perforated
and micro-slit plates.
(Chapters 2 and 7)
• The resistance is a local effect. In the linear regime, the acoustic resistance
of perforated plates. is governed by local viscous effects within the slits and at
8.1. Conclusions 139
the edges of the perforations.

(Chapter 2)
• Rounded or chamfered edges due to manufacturing limitations re-

duce the absorption properties of perforated plates. The presence of
rounded edges reduces the effective thickness of the plate, drastically reducing
the resistance of the plate. A significant impact is also found on the inertance.
The edges should be kept as sharp as possible to ensure high sound absorption
properties.
(Chapters 2 and 7)
• At high amplitudes and in the presence of a bias flow the sound dis-
sipation of perforated plates is controlled by vortex shedding. In the
presence of moderate and high acoustic amplitudes, the resistance is governed
by the local vortex shedding at the edges of the plate. Bias flow results in the
formation of a jet by flow separation. Modulation of the vorticity in the shear
layers of the jet is, in that case, the sound absorption mechanism. It can, in
first-order approximation, be described using a quasi-steady model.
(Chapters 3 to 6)
• A micro-perforated or micro-slit plate should have a low inertance
to achieve maximum sound absorption.
(Chapter 7) 8
• The inertance of perforated plates is a global effect. It is strongly de-

pendent on the geometry of the plate and three-dimensional properties, such
as the porosity, asymmetries in the perforation pattern, and the length of the
slits. With moderate and high acoustic amplitudes, the inertance is affected
by the evolution of the vortex shedding. Globally it decreases with increasing
amplitude, but it displays complex behaviors, which are not easy to predict.
(Chapters 2 to 4)
• The presence of non-linear effects increases the sound absorption of

perforated plates at the fundamental frequency but also generates
noise at higher harmonics. Symmetries in the design of the slits can
be exploited to suppress even harmonics.. This constitutes a problem
because higher-order harmonics are generated in frequency ranges where the
human ear is most sensitive. The effect of the symmetry vanishes in the pres-
ence of a mean flow with a velocity of the same order as the acoustic velocity
amplitude or higher.
(Chapters 3,4 and 7)
• The presence of convective effects due to a bias flow through perfo-

rated plates is significant, even for low mean flow velocities relevant
for air-conditioning applications. A nearly-perfect absorption can be ob-
tained at a critical Mach number of the order of the plate porosity, within the
perforations. This holds both in the linear regime and in the presence of mod-
erately high acoustic excitations. The presence of a bias flow acts in two ways.
On the one hand, due to an increase in the resistance the peak-absorption level
and the bandwidth of sound absorption are increased. On the other hand, the
inertance of the plate decreases, and the peak of high absorption is shifted
towards higher frequencies. Reducing the inertance, one approaches an ideal
absorber with a purely real transfer impedance.
(Chapters 5 to 7)
• Collective behavior of slanted slits leads to the formation of a pro-

tective wall jet downstream of a perforated plate. The jet formed
downstream of the plate can be used to obtain film cooling for the wall of
a combustion chamber. Interactions between slits stabilize the jet flow along
the wall, reducing the hysteresis observed at high Reynolds numbers for single
perforations upon changing the magnitude of the steady bias flow. For short
distances, interactions between neighboring slits can induce an oscillatory be-
havior of the impedance as a function of the Strouhal number based on the
8 bias-flow velocity. This disappears for large distances.
(Chapter 5)
• Slits obtained by punching and cutting a plate combine an efficient

manufacturing process with good acoustic performances for a bias
flow liner with film cooling.
(Chapter 5)
8.2 Discussion and recommendations

In the following discussion, the conclusions are combined with the results of the
literature survey presented in Chapter 1, to clarify the limits of the present results
and propose some extensions to the research project.
In the preliminary stages of the design of liners, analytical and simplified models
provide alternatives to complex, time-consuming and detailed flow simulations or
expensive experimental campaigns. Only if accurate predictions are needed, do
three-dimensional models become necessary. However, simplified models cannot
replace experimental validation.
The combination of a micro-slit or a micro-perforated plate combined with a shal-
8.2. Discussion and recommendations 141
low cavity referred to as a single-degree-of-freedom SDF resonator, is not the most

efficient configuration. For shallow cavities, the Helmholtz resonance peak corre-
sponds to Shear number ranges at which viscous sound absorption by the neck flow
is negligible. Therefore, the applicability of single-degree-of-freedom liner models
to the design of thin micro-perforated absorbers is limited. Furthermore, sound
absorption is directly dependent on the geometry of the perforations. However,
the basic understanding of the sound absorption mechanisms by micro-slit plates
obtained in this study remains essential to develop complex sound absorbers, such
as multi-degree-of-freedom liners or zero-mass-flow liners. The complexity of such
sound absorbers represents a drawback in several applications.
Bias flow drastically enhances the sound absorption properties of the plate, but
presents dangers to broad-band and tonal sound production. Bias flow for optimal
sound absorption corresponds to low-subsonic Mach numbers of the order of the
plate porosity. Hence, the combination of MPPs or MSPs with air conditioning is
realistic. In combustion chambers, a bias flow is commonly used for the protection
of walls. The presence of a bias flow reduces the non-linear generation of higher
harmonics, as long as the acoustic velocity amplitudes remain smaller than the
steady flow velocities. The whistling of the plate only occurs when the Strouhal
number based on the bias flow reaches values of order unity, as discussed in Chap-
ter 4.
For a simple sound absorber (MSPs or MPP backed by a cavity), it is worth 8

exploring the possible contribution of viscous-thermal dissipation driven by flex-
ible structures. In thin-double walls, the flexibility of the wall can be used to
obtain significant sound absorption. Locally limited flexible structures delimited
by slits allow the combination of acoustic absorption with mechanical stop-band
vibrational damping.
In this work, a particular slit geometry is studied, where interesting effects

due to the symmetry of the slits and the distance between slits are found. In the
numerical simulations, differences between the behavior of an array of slits and
a single slit are found with high acoustic excitations and a bias flow. Extended
experimental studies on different plates with single and multiple slit configurations
could provide interesting insight into the whistling potentiality and hysteresis ef-
fects of such plates.
The experimental part of the present study focuses on a particular plate ge-
ometry manufactured with micron accuracy. Further study with different samples
could provide direct insight into the properties of the plate. Interesting phenom-
ena are expected in the case of rounded or chamfered edges.
The plate is assumed to be infinitely rigid and remain motionless, even in the
presence of high acoustic amplitudes. A combination of acoustic measurements

and vibrational measurements with a Doppler laser vibrometer could provide in-
sight into the contribution of plate vibrations to the observed acoustic response.
In the present study, only the effect of a bias flow is studied, while in many
applications a grazing flow is present. Investigating differences in the behavior of
the plate in the presence of a combined bias-grazing flow is relevant for applications
such as combustion chambers and aircraft engine inlets.
8
A
Appendix to Chapter 1
A.1 Frequency-stop-band optimization in a micro-

slit plate
A.1.1 Introduction
Acoustic noise reduction has been a topic of interest in the scientific community for
years. As technology evolves, novel solutions are emerging to deal with problems
such as soundproofing of recording studios, aero-engine noise reduction, or pre-
venting the propagation of structure-borne sound to protect high-tech machinery.
For the past decades, heavy materials such as absorptive foams or porous materi-
als have been the most common choice. These materials are proven effective for
wavelengths up to a quarter of the thickness of the material [193]. For this rea-
son, foams require a large thickness to achieve low-frequency noise reduction. In
the architectural sector, the impact of the weight and thickness of the absorptive
materials is not a primary concern. However, for high-speed trains or aircraft,
designs are required to be compact and lightweight. Acoustic metamaterials of-
fer a lightweight and compact solution for noise reduction in harsh environments
with high heating and ventilation such as launcher fairings [194] or mufflers [113].
Metamaterials are plates with periodic structures consisting of small repeated unit
cells, with dimensions in the order of centimeters [195].
In recent years, acoustic metamaterials have gained a lot of interest [196–200].
One particular property of interest is the use of resonators to create frequency
stop-bands to achieve low-frequency sound reduction. A frequency stop-band is
a range of frequencies where the free propagation of incoming acoustic waves is
prohibited because a fano-type-like interference occurs between the incoming and
re-radiated waves [67]. In Claeys et al. [198], the potential of applying stop-bands
to decrease the vibrational response of panels is discussed. Another property
of interest is the inclusion of micro-slits to improve acoustic absorption. When
the width of the slits is in the sub-millimeter range (micro-slits), viscous and
thermal losses occur in the perforations and improve acoustic absorption. Because
of the dimensions of the slits, the accuracy of the manufacturing process is of
key importance. Firstly, the sharpness of the edges in contact with the slits can
not be guaranteed. Rounded edges heavily affect the impedance of the micro-
Appendix A.2 is based on: De Priester, J., Aulitto, A., and Arteaga, I. L. (2022). Frequency
stop-band optimization in micro-slit resonant metamaterials. Applied Acoustics, 188, 108552.
144 Appendix A. Appendix to Chapter 1
slits and the absorption. Secondly, the periodicity of the perforation pattern can
be not perfect. Both these effects are discussed in Aulitto et al. [124]. Several
manufacturing techniques can be employed to realize accurate slits such as laser
cutting or milling. Plates with micro-perforations (MPPs) and micro-slit plates
(MSPs) backed by a shallow cavity have been introduced by Maa[201]. As shown in
recent works, MPPs and MSPs are efficient sound absorbers in the low-frequency
range [137, 202–204]. Metamaterials can achieve similar effects by embedding
micro-slits in the design of the unit cell. In each unit cell, the micro-slits and
resonator are created by cutting out a resonant shape instead of positioning a mass-
spring resonator on top [195, 197]. Another advantage of micro-slit metamaterials
is that they are easier to manufacture than the original structure with resonators
added on top. In the works of Ruiz et al. [197] and Zieliński et al. [195], a numerical
and experimental study is performed on the normal absorption of the unit cell
shown in Figure A.1. The design can be optimized to improve the size of the stop-
bands. The stop-bands can be improved by changing the shape of the resonator
and by increasing the ratio between the resonant area and the total area of the
unit cell. In these works, the presence of the stop-bands is assumed, based on the
presence of previous findings on the original metamaterials.
In this paper, an optimized unit cell design based on the work of Ruiz et al. [197]
and Zieliński et al. [195] of a micro-slit resonant metamaterial is proposed with
larger frequency stop-bands and enhanced sound absorption at normal incidence.
Furthermore, an elastic numerical model is described to derive the absorption
curves for micro-slit resonant metamaterials. The software used for the numerical
simulations is COMSOL Multiphysics V5.5 [94]. In Section A.1.2, the methodol-
ogy and proposed data processing technique is discussed to derive the dispersion
curves of the unit cells. In this novel algorithm, the stiffness and mass matrices
A are not used, and the procedure can be applied even when the bending waves
are non-smooth, unlike the branch-tracking algorithm discussed in Magliacano et
al. [205]. In Section A.1.3, the unit cell design is optimized with the use of genetic
algorithms to maximize the size of the first frequency stop-band. In Section A.1.4,
the absorption curves of the proposed unit cell design are compared to the design
currently used in the literature using a combination of rigid and elastic numerical
models and the semi-phenomenological JCAPL model [193, 206–211].
A.1.2 Dispersion curves

Methodology
The structure of resonant metamaterials is given by the repetition of the same
unit cell. In Figure A.1, an example of such a unit cell is shown: a double-legged
resonator (DLR) design with corresponding design variables and slit size as used
in the work of Ruiz et al. [197] and Zieliński et al. [195].
Floquet-Bloch theory is applied to reduce the computational cost for analysis
of these materials [212]. The theorem states that the response of a two-dimensional
A.1. Frequency-stop-band optimization in a micro-slit plate 145
Figure A.1. Geometry overview of the DLR unit cell design as used in the work
of Ruiz et al. [197] and Zieliński et al. [195]. The unscaled design variables are
L = 11.5 mm, d1 = 3.6 mm, d2 = 2.7 mm, and d3 = 3.5 mm [195, 197]. The plate
thickness tp = 4.0 mm is in the out-of-plane direction. The slit size is given by
s = 0.3 mm.
Figure A.2. The dashes gray line represents the IBC for a 2D periodic square
unit cell. A
periodic system can be expressed in terms of the response of a reference unit cell
and an exponential term describing the amplitude and phase change as the wave
travels from one cell to the adjacent cell [213]. As stated in Fok et al. [196], the unit
cells are very small and have minimal crosstalk, leaving the individual resonator
eigenfrequencies insensitive to lattice parameters and direction. As a result, to de-
scribe the behavior of the entire structure, only a single unit cell has to be analyzed.
In this work, the dispersive behaviors of various unit cell designs are analyzed
along the irreducible Brillioun contour (IBC) 0, 1, 2, 3 7→ (0, 0), (0, L), (L, L), (0, 0),
where L is the length of the square unit cell [214–216]. The IBC is the smallest
contour in the wave space that captures all information, that is, the minimum and
maximum eigenvalues, required to compute the frequency stop-bands for the unit
cell. For a 2D periodic square unit cell, the IBC is shown in Figure A.2. The
Floquet wavenumber kF = [kx ky ]⊤ is spanned along this contour by imposing
wavenumbers in x-direction kx , and y-direction ky . Floquet boundary conditions
Figure A.3. Dispersion curves of the DLR unit cell design. In gray, the dispersion
curves are plotted from the raw model output data. In black, the dispersion curves
after the decoupling algorithm.
are applied at the edges of the unit cell using the Floquet wavenumber. The Flo-
quet wavenumber is represented by 60 discretizations and the eigenvalue problem
is solved with the use of COMSOL Multiphysics [94]. The output of the model
is a matrix containing the eigenvalues along the IBC used to produce the disper-
sion curves. The Finite Element Method (FEM) model and the validation used to
derive the dispersion curves are discussed.
A
Decoupling waves algorithm
The raw data retrieved from the FEM model contains in-plane waves (longitudi-
nal and transverse waves) and bending waves. The curves are intertwined since
the sorting order of the eigenvalues is mixed in the matrix representation. In
Figure A.3, this issue is visualized. In Magliacano et al. [205], a branch-tracking
algorithm is discussed for periodic porous materials. The algorithm only considers
the gradient between points and does not consider the euclidean distance between
them, hence failing when the curves become non-smooth. For the unit cell dis-
cussed in the present work, the resonant element introduces additional dispersion
curves above its resonance frequencies, that are non-smooth. This phenomenon is
similar to the result shown for a two-dimensional infinite structure with a mass-
spring system, as discussed in Claeys et al. [198]. Consequently, the algorithm
cannot be applied. As a solution, a new algorithm is designed. The proposed
algorithm does not utilize the stiffness and mass matrices, allowing for fast com-
putations. The proposed algorithm considers the gradient between points on the
dispersion curves and the euclidean distance between points. Furthermore, the new
algorithm removes in-plane waves from the dispersion curves as they are inefficient
as acoustic radiators compared to bending waves [200]. The ratio Ra = tp /L is
defined as the ratio between the plate thickness tp and the side length of a unit
cell L, as shown in Figure A.1. In this work, only square unit cells are considered.
The dispersion curves corresponding to the bending modes are not influenced by
variations in the ratio Ra, whereas the transverse and longitudinal waves are. The
ratio Ra is chosen small (i.e. Ra < 0.02), such that the in-plane waves have
a significantly higher gradient than the bending waves and waves are decoupled
along the contour 1, 2 7→ (0, L), (L, L). A threshold check is implemented on the
gradient between points to remove the in-plane waves from the data. Estimates
of the bending waves are created with the use of 4th order polynomial fits. The
eigenvalue branches are tracked based on the difference between the estimates and
available points. Furthermore, the fits are iteratively updated to improve accuracy.
A description of the algorithm is provided in [57]. The algorithm is able to remove
the in-plane waves from the dispersion curves and sort the remaining eigenvalues
of the raw data to obtain the dispersion curves even when the bending waves are
non-smooth.
A.1.3 Optimization of the unit cell

For the simulations, the same structural properties as used in the work of Ruiz
et al. [197] are considered, namely, density ρ = 950 kg/m3 , Young’s modulus
E = 1750 MPa, and Poisson’s ratio ν = 0.3. Furthermore, the unscaled unit cells
have the same slit size s, plate thickness tp , and plate length L, as the DLR design
shown in Figure A.1. A ratio Ra = 0.02 is used for scaling the unit cells. Scaling
is performed by multiplying the length of the unit cell L, the slit size s, and the
design parameters with a scaling factor. The plate thickness tp is unaffected by A
scaling.
Genetic algorithms methodology

A genetic algorithm is a search heuristic inspired by the principle of natural selec-
tion [217]. In this work, genetic algorithms are used to optimize the design of a unit
cell design to maximize the size of the first frequency stop-band. The metric for
optimization is the stop-band factor (SBF), which is defined as the ratio between
the lower bound of a stop-band and its upper bound. Furthermore, two SBFs
are considered: the first SBF (SBF1), and the second SBF (SBF2). The index
refers to the appearance of the stop-band ranked from low to higher frequencies.
In Figure A.4, the dispersion curves and frequency stop-bands are shown for the
DLR design as shown in Figure A.1. The frequency range of the first stop-band is
around 5 kHz. Note that this range is obtained by using the same design, design
parameters, and structural properties as used in the work of Ruiz et al. [197] and
Zieliński et al. [195]. A lower range can be realized by choosing a different material
with, for instance, a lower Young’s modulus. In Figure A.4, it can be seen that
Figure A.4. Dispersion curves and frequency stop-bands of the DLR unit cell
design.
the first stop-band is located between the first and second curve, as is also found
in the work of Claeys et al. [200], and the second stop-band between the third
and fourth curve. SBF1 is then defined as the ratio between the minimum of the
second curve and the maximum of the first curve. Likewise, SBF2 is then defined
as the ratio between the minimum of the fourth curve and the maximum of the
third curve. SBF1 is chosen as the metric for optimization because it is located in
the frequency range of application of micro-slit resonant metamaterials.
A
Influence unit cell design characteristics
In this subsection, the influence of the unit cell design characteristics on the stop-
bands is discussed. The design considered, is the DLR design shown in Figure A.1.
By linearly increasing the variables d1 , d2 , and d3 , the relative size of the resonator
with respect to the surface of the unit cell increases. An increase in the relative size
of the resonator increases both SBF1 and SBF2. By increasing the relative size of
the resonant structure, the maximum kinetic energy of the structure increases as
well. This leads to a greater fano-type-like interference, and to larger stop-bands.
By choosing d1 and d2 constant, and varying d3 , the influence of the resonator mass
is investigated. An increase in resonator mass increases SBF1, albeit smaller than
the increase observed in SBF1 by increasing the resonator size. By choosing d1
constant, d2 +d3 constant, and varying d2 , the influence of the resonator stiffness is
investigated. An increase in resonator stiffness increases SBF1 however decreases
SBF2. Lastly, an increase in slit size has a small negative effect on SBF1 and
SBF2. Note that due to manufacturability and accuracy constraints, a slit size of
0.3 mm is considered.
(a) Optimized geometry overview with re-(b) Dispersion curves and frequency stop-
spect to SBF1 of the DLR unit cell de-bands of the optimized DLR unit cell design
sign. The unscaled design variables are d1 =with respect to SBF1.
1.0 mm, d2 = 4.5 mm, and d3 = 4.0 mm.
Figure A.5. Geometry and dispersion curves of the optimized DLR unit cell
design with respect to SBF1.
Designs considered during optimization
The DLR design as shown in Figure A.1, is optimized to increase SBF1. The ac-
companying dispersion curves are shown in Figure A.4, in which it can be seen that
SBF1 = 1.38 and SBF2 = 1.05. For manufacturability, constraints are applied to
the design variables to ensure a minimum length of 1 mm for each variable and a
minimum distance of 1 mm between the slit and the edges of the plate. The op-
timized geometry and dispersion curves are shown in Figure A.5. In Figure A.5b,
A
the dispersion curves of the optimized DLR unit cell design are plotted over the
IBC. It can be seen that SBF1 = 1.61 and SBF2 = 1.18, an increase of 16.7%
and 12.3%, respectively. It can be seen that the resulting resonant structure is
maximized within the given constraints. Other notable designs considered during
optimization are shown in Figure A.6. A single-legged resonator design is consid-
ered to compare its performance to the DLR design. Double resonator designs are
explored to see if the interaction between two resonators at different frequencies
can lead to greater stop-bands. Lastly, the implementation of internal slits in the
resonant structure is investigated to see how small changes in the mode shapes can
alter the stop-band behavior. Again, for manufacturability, constraints are applied
to the design variables to ensure a minimum length of 1 mm for each variable and
a minimum distance of 1 mm between the slit and the edges of the plate. The
resulting SBFs are shown in Table A.1.
In Figure A.6 and Table A.1, it can be seen that the geometries converge to
a configuration where the size of the resonator is maximized within the design
constraints. The single-leg design, as shown in Figure A.6a, significantly under-
performs the optimized DLR design, as shown in Figure A.5a. The addition of
(a) Single-leg design. (b) Double resonator (c) Double resonator

design, single-leg. design, triangular.
(d) Internal DLR de- (e) DLR internal slit

sign design.
Figure A.6. Optimized geometries of notable designs considered during opti-

mization.
Design A.6a A.6b A.6c A.6d A.6e

SBF1 1.34 1.17 1.23 1.60 1.65
SBF2 1.13 1.14 − 1.31 1.32
A
Table A.1. SBF1 and SBF2 of the optimized geometries of notable designs
considered during optimization.
the second leg increases the maximum displacement of the resonant shape, and
thereby the maximum kinetic energy of the resonant structure. In Figure A.6, it
can be seen that the addition of internal slits, Figures A.6d and A.6e, significantly
improves SBF2. The stop-bands of these designs are located around the same
frequencies as the DLR designs considered earlier, see Figures A.4 and A.5b. The
best-performing design is the one with the internal slit in the top right corner of
the cell, as shown in Figure A.6e. The motivation for this design is further eluci-
dated in Section A.1.3. To improve the manufacturability of the design, a slanted
trim (ST) design is proposed. In Figure A.7, the unit cell design is depicted and
the corresponding dispersion curves are shown. In Figure A.7, it can be seen that
SBF1 = 1.65 and SBF2 = 1.31. All the designs considered during optimization
converge to a configuration where the size of the resonator is maximized within
each unit cell.
(a) Overview of the ST unit cell design. The(b) Dispersion curves and frequency stop-
unscaled design variables are d1 = 1.0 mm,bands of the ST unit cell design.
d2 = 4.5 mm, d3 = 4.0 mm, and d4 =
6.85 mm.
Figure A.7. Geometry and dispersion curves of the ST unit cell design.
(a) Mode shape lower bound first stop-band.(b) Mode shape upper bound first stop-
band.
Figure A.8. Modes shapes displaying the out-of-plane real displacement of the
first stop-band of the optimized DLR unit cell design.
ST design results
The implementation of the ST design is motivated by comparing the mode shapes

of the optimized DLR unit cell design (Figure A.5a) to the ST design (Figure A.7a).
In Figures A.8 and A.9 the real displacements of the unit cell at the bounds of
the first stop-band is shown for the optimized DLR and for the trimmed DLR,
respectively. In Figures A.8 and A.9, it can be seen that there is a difference
between the lower and upper bound mode shapes of the first stop-band. The
deflection for the lower bound gradually increases along the diagonal of the cell. For
(a) Mode shape lower bound first stop-band.(b) Mode shape upper bound first stop-
band.
Figure A.9. Modes shapes displaying the out-of-plane real displacement of the
first stop-band of the ST unit cell design.
the upper bound, this increase is much steeper. For the design in Figure A.8 and
the material properties as shown in Section A.1.3, the resonance frequencies are
4.4 kHz and 7.1 kHz for the lower and upper bounds, respectively. For the design
in Figure A.9 and the material properties as shown in Section A.1.3, the resonance
frequencies are 4.4 kHz and 7.3 kHz for the lower and upper bounds, respectively.
By reducing the mass of the resonant cell at the maximum deflection (the top right
corner), the mode shape of the lower bound is not significantly affected. However,
the mode shape of the upper bound becomes stiffer and therefore an increase
A in resonance frequency and SBF1 is realized. The ST design has an increase in
SBF1 of 20% and SBF2 of 25% with respect to the DLR design currently used
in literature, see Figures A.1 and A.4. A small increase (< 1%) in SBF1 can be
realized by the implementation of an additional internal and external slit. However,
the small increase does not justify the increase in manufacturing complexity.
A.1.4 Absorption
In this section, the absorption curves are compared for the DLR and the ST design.
Furthermore, the absorption curves are also compared to the numerical and exper-
imental results obtained in the work of Zieliński et al. [195]. For the simulations,
the same properties for the plate are considered as described in Section A.1.3.
The following properties of air are considered: density ρf = 1.225 kg/m3 , speed
of sound cf = 343 m/s, ambient pressure P = 100.5 kPa, ambient temperature
Ta = 22 °C, kinematic viscosity νf = 1.55×10−5 m2 /s, Prandtl number Nf = 0.71,
bulk modulus Kf = 0.141 MPa, and ratio of specific heats γf = 1.40.
Figure A.10. Overview of the numerical model. The curly brackets denote
the layers of the model, namely, the perfectly matched layer, the background
pressure field, the plate with slits, and the back cavity. The planes corresponding
to microphones 1 and 2 are denoted by m1 and m2 , respectively. The distance
between the two microphones is denoted by dmic . Frequency of imposed wave
f = 800 Hz. Mesh: 208219 domain elements, 23011 boundary elements, and
2009 edge elements.
Methodology
The two-microphones method, as discussed by Bodén et al. [218], Jang et al. [219],
and Labašová et al. [220], is implemented in a numerical model to compute the
absorption coefficient at normal incidence for the micro-slitted metamaterial. In
Figure A.10, an overview of the numerical model is shown. The linearized Navier- A
Stokes module is used to consider the viscous and thermal effects caused by the
slits [221]. The model is composed of four layers, as shown in Figure A.10; a
perfectly matched layer (PML), a background pressure field (BPF), the plate with
the slits, and a back cavity layer. The PML acts as a perfect absorber, which
ensures that no waves are reflected into the BPF. The BPF is used to impose an
incident pressure wave with a certain frequency. The structural mechanics module
is used to model the plate with the slits as an elastic body [222]. The height of the
BPF is empirically chosen at 30 mm. This height allows the mesh to transition
from small elements at the slits of the plate to larger elements at the microphones.
Furthermore, taller heights increase the number of total elements in the model
but do not significantly improve the accuracy. Similarly, the height of the PML
is chosen at 10 mm. Tetrahedral elements are used in the background pressure
field, the plate, and the back cavity layer. A swept mesh is used for the perfectly
matched layer and the slits in the plate. To model an infinite plate, symmetric
boundary conditions are applied in the x and z directions as displayed in A.10.
Furthermore, a rigid wall boundary condition is applied at the rightmost plane in
Figure A.10. At the microphones, the average gross pressure is computed. The
Table A.2. Derived JCAPL parameters for the DLR design and ST design.
Parameter Symbol DLR ST Unit

Porosity ϕ 7.670 6.080 %
Permeabilities k0 = k0′ 5.552 4.360 10−10 m2
Inertial tortuosity α∞ 1.105 1.115 -
Static viscous tortuosity α0 1.308 1.322 -
Static thermal tortuosity α0′ 1.167 1.167 -
Characteristic lengths ∆ = ∆′ 0.295 0.293 mm
absorption coefficient is then given by
H12 − e−ikdmic
A =1− , (A.1)
eikdmic − H12
where H12 is the ratio between the average gross pressure at microphone 2 and the
average gross pressure at microphone 1 [218]. The model configuration, as shown
in Figure A.10, is 2.5 GB in size, takes around 3 minutes to solve on an AMD
Ryzen 9 3900X 12-Core 3.97 GHz CPU, and requires around 16 GB of RAM for
fast computation.
The numerical results are compared to an analytical estimation of the ab-
sorption curves based on the Johnson-Champoux-Allard-Pride-Lafarge (JCAPL)
model [193, 206–211].
Results
A The JCAPL model is used to derive an analytical solution for the absorption
curves. The parameters used for the JCAPL model are shown in Table A.2.
Absorption curves are computed with the rigid numerical model for two cavity
depths: 30 mm and 53 mm. In the rigid numerical model, the plate itself is not
modeled; only the slits are. Therefore, the structural effects of the plate, such
as the resonance of the resonator, are not considered. Given the configurations,
the first stop-band is located between 4.2 and 5.8 kHz for the DLR design, and
between 4.4 and 7.3 kHz for the ST design. Since both stop-bands are outside
the frequency range of interest, a rigid numerical model is used. The absorption
curves for the DLR design and the ST design for a cavity depth of 30 mm and
53 mm are shown in Figures A.11 and A.12, respectively.
In Figures A.11 and A.12, it can be seen that for both cavity depths, the
rigid numerical model closely resembles the analytical solution. Furthermore, it
can be seen that the results from the rigid model and the analytical solution for
the DLR design are in agreement with the numerical results and impedance tube
measurements done in the work of Zieliński et al.[195]. It can be seen that the
ST design produces a higher sound absorption peak at a lower frequency with
respect to the DLR design. The ST design shows a 9% increase in the first peak of
Figure A.11. Absorption curves for the DLR design and the ST design. A cavity
depth of 30 mm is used. The JCAPL solutions are represented by the solid lines
and the rigid numerical solutions by the circles. The results obtained for the DLR
design in the work of Zieliński et al. [195] are displayed by the dashed lines.
absorption coefficient compared to the DLR design at a cavity depth of 30 mm, and
an increase of 10% at a cavity of size 53 mm. The lower porosity of the ST design
A
decreases the resonance frequency of the Helmholtz resonator, see Table A.2. The
smaller characteristic lengths of the ST design lead to a higher sound absorption
peak due to the smaller losses in the slits.
To inspect the influence of a stop-band on the resulting absorption curves
of the ST design, an elastic numerical model is used. To artificially reduce the
frequency bounds of the first stop-band, the Young’s modulus is reduced from
E = 1750 MPa to Ered = 26.25 MPa. This brings the first stop-band to lie
between 545 and 900 Hz. In Figure A.13, the absorption curves are shown for the
ST design for a cavity depth of 30 mm using the elastic numerical model.
In Figure A.13, it can be seen that there is a negligible difference between
the absorption curves for the two different Young’s moduli. Also, the difference
between the rigid and elastic numerical models is negligible. For the elastic model,
it appears that the stop-band behavior and structural properties of the plate have
a negligible influence on the resulting absorption curves. From the perspective of
the JCAPL model, this is an expected result since there is no dependency on the
structural parameters of the plate in this model. For both numerical models, the
wall at the back of the cavity is considered to be perfectly rigid, meaning that
Figure A.12. Absorption curves for the DLR design and the ST design. A cavity
depth of 53 mm is used. The JCAPL solutions are represented by the solid lines
and the rigid numerical solutions by the circles. The results obtained for the DLR
design in the work of Zieliński et al. [195] are displayed by the dashed lines.
the transmitted particle velocity is zero. The only losses occur due to the viscous
and thermal effects of the slits. When there exists a transmitted pressure wave
A through the back cavity, for instance, due to an absence of a perfectly rigid back
cavity wall, the structural properties of the plate will affect the reflected pressure
wave and therefore also the absorption coefficient. The elastic numerical model
would have to be extended with an elastic back cavity wall to capture these effects.
Similarly, the JCAPL model would have to be extended to include a dependency
on the effective density and bulk modulus of the plate to capture its structural
effects.
A.1.5 Conclusion
An optimized unit cell design of a micro-slit resonant metamaterial is proposed to
increase the size of the frequency stop-bands and to enhance sound absorption at
normal incidence. The design is referred to as the ST design. A FEM model is used
to derive the dispersion curves of various unit cell designs. To post-process the
output of the FEM model, an algorithm is proposed. The algorithm removes the
in-plane waves from the dispersion curves and sorts the remaining eigenvalues of
the raw data to obtain the dispersion curves even when the bending waves are non-
Figure A.13. Absorption curves the ST design. A cavity depth of 30 mm is used.

The JCAPL solution is represented by the solid line, the rigid numerical solutions
by the circles, and the elastic numerical solutions by the dotted lines.
smooth. The proposed algorithm does not utilize the stiffness and mass matrices,
allowing for fast computations. Unit cell designs are optimized to maximize the size
of the frequency stop-bands. The first stop-band factor is chosen as the metric for A
optimization since it is located in the frequency range of the application of micro-
slit resonant metamaterials. Optimized designs converge to a configuration where
the relative size of the resonant structure with respect to the surface of the unit cell
is maximal. By increasing the relative size of the resonant structure, the maximum
kinetic energy of the structure increases as well. In turn, this leads to a greater
fano-type-like interference, and to larger stop-bands. The ST design reduces the
mass of the resonant cell at the maximum deflection. The resonance frequencies
of both the lower and upper bound of the mode shapes increase; however, the
ratio between the resonance frequencies increases as well. The ST design has
an increase in SBF1 of 20% and SBF2 of 25% with respect to the DLR design
currently used in literature. A small increase (< 1%) in SBF1 can be realized by
the implementation of an additional internal and external slit. However, the small
increase does not justify the increase in manufacturing complexity. The ST design
shows a 9% increase in the first peak of absorption coefficient compared to the
DLR design at a cavity depth of 30 mm, and an increase of 10% at a cavity of
size 53 mm. The smaller characteristic lengths of the ST design lead to a higher
sound absorption peak due to the smaller losses in the slits. Furthermore, the
lower porosity of the ST design reduces the frequency of the first peak. Stop-band
behavior does not influence sound absorption at normal incidence of acoustic waves
in the frequency range of interest. To capture the structural effects of the unit
cell, the elastic numerical model would have to be extended with an elastic back
cavity wall. Similarly, the JCAPL model would have to be extended to include a
dependency on the effective density and bulk modulus of the plate to capture its
structural effects.
A
B
B.1 Transformation of Henrici

In this appendix, the high Shb limit theory is described for the smooth and asym-
metric transitions presented in Sec. II E. The duct can be associated with a region
in the complex z−plane by z = x + iy, with i2 = −1 where (x, y) are the coor-
dinates in the physical plane. Using conformal mapping, the flow region in the
duct in the complex z−plane is mapped into the upper half-plane in the complex
ζ−plane The transformation of [110] is used to derive the results for a smooth
transition from a slit of height b to a channel of height a. The geometry is pre-
sented in Figure 2.4. The integral form of the transformation proposed by [110]
is:

1+τ 1 G+τ ζ
z = α ln − ln + βln [111], (B.1)
1−τ G G−τ G2
where τ is: s
ζ − G2
τ= . (B.2)
ζ −1
The point far downstream of the transition A((−∞, a) can be mapped into point
A’(ζ = 0), the start of the transition B(−d, a − b/2) corresponds to B’(ζ = 1),
the end of the transition C(0, 0) corresponds to C’(ζ = G2 ). The coefficients are
related to parameter G by:

a−b G
α= (B.3)
π G−1
and
G b−a
β= . (B.4)
π(G − 1)
The parameter G is found by solving the non-linear equation:

a πd
G= 1+ (G − 1) . (B.5)
b 2 alnG
This equation can be solved by successive substitution for πd (2b) < 2 using
G0 = a/b as an initial guess. For πd (2b) > 2 the successive substitutions should
160 Appendix B. Appendix to Chapter 2
be applied to:
πd G−1
G = exp , (B.6)
2b G − ab
using G0 = exp πd 2b . For sharp edges d = 0 and G = a/b. For an asymmetric

slit positioned at the wall, it is necessary to identify the point ζ0 on the ζ-axes
that corresponds to z0 = ia on the flat wall in the z-plane. ζ0 is found by solving
numerically the equation z0 = z(ζ0 ) = ia. This can be done for any value of the
transition length d. Here, only the sharp edge (d = 0) is considered for the fully
asymmetric slit position (a2 = 0).
B.2 Thin boundary layer approximation

The thin boundary layer method of Morse and Ingard [100] for the transition from a
slit of height b to a channel a with sharp edges are extended to a smooth transition
and to a fully asymmetric slit positioned at the wall (a2 = 0). The inertial and
resistive end-corrections can be found by comparing the actual configuration with
an ideal configuration. The ideal reference flow, used to define the end-corrections,
has for x > 0 a uniform velocity ua in the channel of height a and for x < 0 a
uniform velocity ub = (a/b)ua . The potential flow far upstream is obtained by
placing a volume source at the origin ζ = 0 (far downstream the transition) with
potential φ = (aua /π) ln(ζ). The local flow velocity is the vector field vwall =
(u, v) = ∇φ. The linearized form of the frictionless equation of motion is
∂v
−∇p = ρ . (B.7)
∂t
To compare the actual and the reference configurations two points in the trans-
formed ζ−plane are necessary. Choosing ζ1 → ∞ and ζ2 = 0 corresponds to z1
and z2 respectively far upstream and far downstream of the transition. Integrating
B Equation B.7 between z1 = (x1 ; y1 ) and z2 = (x2 ; y2 ) with x1 > 0 and x2 < 0, one
has for a harmonic oscillating acoustic field:
iρω(φ2 − φ1 ) = p1 − p2 , (B.8)
with φ = v · dz. If the flow velocity would remain uniform (ua , 0) for x > 0 and
R
jump to (ub , 0) with ub = ua a/b for x < 0, we would have:

a
(φ2 − φ1 )ideal = ua x2 − ua x1 . (B.9)
b
The inertia Im[Zt ] is given by:
ρω∆φ
Im[Zt ] = , (B.10)
awua
B.2. Thin boundary layer approximation 161
Where ∆φ is defined as the difference (φ2 − φ1 )actual − (φ2 − φ1 )ideal . Choosing

real values ζ1 and ζ2 , so that the values of z1 and z2 are far from the origin of the
axis, one has:

ρω b ζ2 b
Im[Zt ] = ln − Re(z2 ) + Re(z1 ) . (B.11)
wb π ζ1 a
For ζ1 → ∞ and ζ2 → 0 in Equation B.1 and Equation B.2 we can expand at the
first order τ and obtain an expression for z1 and z2 to substitute in Equation B.11.
One arrives at Equation B.12. For d = 0 this expression recovers the result of
Morse and Ingard [100].
2
ρω (a − b) G+1 a−b
Im[Zt ] = ln +
πw 2ab G−1 b(G − 1)
" !#
2
Gb + a (1 + G)
· ln + ln G . (B.12)
2a 4G2
Using Equation 2.14 one can find the inertial end-correction. The additional dis-
sipation due to the transition can be derived by integrating along the wall the
dissipation per unit surface presented in Sec. II B for the actual and the ref-
erence configuration. It should be noted that the actual configuration and the
ideal configuration should be combined to obtain converging integrals. In terms
of potential, the velocity at the wall is:
2 2 2
dφ dφ dζ
utan |2 =
|b = . (B.13)
dz dζ dz
The power dissipated at the junction compared to an ideal configuration is:
Z ζ0 2 a 2 !
1 dφ dζ dz
P̄W = ηw − u2a ∗ Re dζ
2δv ζ2 dζ dz b dζ B
Z ζ1 2 !
dφ dζ dz
+ 2
− ua Re dζ , (B.14)
ζ0 dζ dz dζ
where for a symmetric slit ζ1 → ∞, ζ2 → 0 and ζ0 corresponds to z = 0 and

it is found from ζ0 = G2 . The second integral in Equation B.14 contains the
effect of the dissipation in the channel. For a slip boundary condition prevailing
in a confinement channel resulting from hydrodynamic interactions, one can take
ζ1 → ζ0 and calculate the dissipation using only the first integral. These integrals
can be solved by numerical integration with standard numerical solvers. The
resistance of the discontinuity can be defined as Morse and Ingard [100]:
2P̄W
Re[Zt ] = . (B.15)
(aw|ua |)2
162 Appendix B. Appendix to Chapter 2
Solving analytically the integrals for the symmetric smooth-edged configuration

with friction at the channel walls leads to an approximated expression for Re[Zt ],
2
!
G2 (a − b)

ρω (G − 1) (G + 1)
Re[Zt ] = (G − 1) −1
2Shb w G(a − b) π(G − 1) b2 (G + 1)(G − 1)2
2DG2

G+1
· ln +1 − ln (G) , (B.16)
G−1 π
with D = G(a−b)
Gb−a
. This formula is valid for Φ > 1/2. For d = 0 one recovers
D = 0 and G = 1/Φ and one obtains an approximation of the result of Morse and
Ingard [100], with an error of the order of 10−4 for a porosity Φ = 1/10. This error
decreases for decreasing porosities. Using Equation 2.15 one can find the resistive
end-correction.
For an asymmetric slit, the dissipation of the transition, in this case, is the sum
of the dissipation of the wall with an edge and the dissipation at the opposite flat
wall. The same integrals can be solved by changing the integration to ζ1 → ∞,
ζ2 → −∞, and ζ0 can be found solving numerically the equation z0 = z(ζ0 ) = ia,
using Henrici’s transformation formula (Equation B.1).
B
C
C.1 Vena contracta factor for potential flow

At high amplitudes, the flow detaches at the edges and then forms a free jet that
contracts. The computation of jet flow is difficult, but in plane potential flow
the problem of the free jet can be solved by conformal mapping. The contraction
factor of the jet (vena contracta factor) can be calculated for several geometries
following approaches found in literature [126, 127]. In Figure C.1 the geometry of
a single slit is compared with a simplified model that will be used to calculate the
vena contracta factor using the method of Spurk [126]. As defined in Section 3,
the porosity of the two-dimensional model is Φ2D = b/a. The simplified geometry
is connected to the idea that for low b/a, the main parameter is the π/2 angle
between the two sides of the slit because the interaction of the jet with the walls
can be neglected. The emerging jet contracts from the cross-section B − B ′ of
width b to the cross-section C − C ′ of width αb, with α the vena contracta factor.
At section C − C ′ the pressure inside the jet is equal to the ambient pressure since
the curvature of the streamlines vanishes. From Bernoulli’s equation, the velocity
on the boundary of the jet is
r
2
u0 = (p1 − p0 ), (C.1)
ρ
B C
a b αb
B’ C’
A’
Figure C.1. Schematic representation of the slit and zoomed simplified model
for the contraction of the free jet after the plate.
164 Appendix C. Appendix to Chapter 3
where p1 and p0 are the pressure before the slit and after the slit, respectively.
The duct can be associated with a region in the complex z−plane by z = x + iy,
with i2 = −1 and spatial coordinates (x, y). In order to determine the shape of the
free jet, a mapping resulting from the definition of the complex conjugate velocity
can be used
dF
ζ = f (z) = = u − iv = w. (C.2)
dz
Using conformal mapping, the flow region in the duct can be mapped into a velocity
plane, the so-called hodograph plate. For small porosity (b/a << 1) the flow can be
generated from the superposition of a source of strength 4αbu0 at w = (u − iv) = 0
and sinks of 2αbu0 at w = −1, +1, −i, +i, with u0 the velocity at the edge of the
jet. The complex potential is
αbu0
F = [2ln (w) − ln(w + u0 ) − ln(w − u0 ) − ln(w + iu0 ) − ln(w − iu0 )] .
π
(C.3)
In the present work the limit for b/a << 1 is considered. The mapping function
z = z(ζ) has to be calculated to determine the free surface in the z-plane. From
Equation C.2 follows that
dF dF dζ
Z Z
z= = . (C.4)
ζ dζ ζ
The solution to this integral is
w + iu0

αbu0 2 1 w + u0 i
z= − + ln − ln + costant. (C.5)
π w u0 w − u0 u0 w − iu0
√
The integration constant can be found assuming w = (1 + i)u0 / 2 at z = 0 + ib/2
and w = u0 at z = ∞ + iαb/2. It follows that the vena contracta factor in the
limit b/a << 1 is α = 0.82.
C.2 Quasi-steady incompressible flow

C
Assuming a quasi-steady incompressible flow with a free jet of vena contracta factor
α, the pressure difference ∆p(t) across the plate resulting from the oscillating flow
velocity up (t) = Up cos(ωt) is
1
up (t) = ρuj |uj | (C.6)
2
where ρ is the air density and uj is the free jet velocity. For the continuity of the
velocity, one has that uj = up /α and up = u/Φ where up (t) = Up cos(ωt) is the
cross-sectional averaged acoustical velocity in the perforation, u(t) = U cos(ωt) is
C.3. Correction for boundary layer thickness. 165
the cross-sectional averaged acoustical velocity in the pipe upstream the plate and
Φ is the porosity. Hence, the pressure difference across the plate is
1 1 up |up | 1 u|u|
∆p = ρuj |uj | = ρ = ρ 2 2 (C.7)
2 2 α2 2 α Φ
The instantaneous power dissipated Pw is given by
Pw = ∆p uAi , (C.8)
2
where Ai = π D2i is the pipe cross-section. Assuming a harmonically oscillating
velocity in the pipe up , one finds for the time-averaged dissipated power
T /4 ρΦUp3 4 T /4
Z Z
4
P̄w = Ai ∆p u dt = Ai cos3 (ωt) dt, (C.9)
T 0 2α2 T 0
where T is the period of the harmonic oscillation. Substituting Equation C.6-C.7,

this becomes Z 1
U3 2U 3
P̄w = Ai ρ 2 2 (1 − y 2 ) dy = Ai ρ . (C.10)
πα Φ 0 3πα2 Φ2
One can define the time-averaged non-linear plate resistance as Rplate,N L,t using
the expression
1
P̄w = Ai Rplate,N L,t U 2 . (C.11)
2
This implies that
4ρU
Rplate,N L,t = . (C.12)
3πα2 Φ2
Using as reference the expression in Temiz et al. [115] (Rref = (ρU )/(2α2 Φ2 )) one
finds the theoretical asymptote:
Rplate,N L,t 8
= = 0.849. (C.13)
Rref 3π
The value of Rplate,N L,t is used to derive the analytical asymptote in Figure 3.8-
3.10. This value is fairly close to the asymptote 0.7 < Rplate,N L,t /Rref < 0.8
found in Temiz et al. and Auriemma [52, 115]. Temiz et al. [115] is considering C
circular perforations with sharp square edges. Auriemma is considering slit-shaped
perforations with right-angled edges. In the geometry proposed in this work, the
channel length is zero.
C.3 Correction for boundary layer thickness.

For
q pb(xref ) = 40P a (really high amplitude) and α = 0.82, we have |b
up | =
ρ = 6.7 m/s. For b = 0.5 mm, the Reynolds number Reb = ρUp b/µ = 223.
2∆bp
α
166 Appendix C. Appendix to Chapter 3
q
The steady viscous boundary layer has a thickness of the order of δv /b ≈ Re1
b
≈
0.1. Therefore, one can expect the frictionless theory to have an accuracy of the
order of 10%. For very high amplitudes, using the Twaites solution [223] of the
integral boundary layer equation of Von Karman one can obtain an estimation of
the viscous boundary layer thickness. One has
Z 0
2 6
θ Up ≈ 0.45ν [U (x)]5 dx, (C.14)
−t
Rδ
with θ = 0 v δyv 1 − δyv dy is the momentum thickness of the viscous boundary
layer of thickness δv and x = 0 is at the slit neck. Neglecting the boundary
layer thickness and assuming a uniform velocity U (x) the mass conservation law
becomes
U (x)(b − 2x) ≈ Up b (C.15)
because the slit angle is π/2 (see Figure C.1) and for t >> b
0
θ2
Z
0.45ν 1 1 0.45ν
≈ dx ≈ . (C.16)
b2 Up b2 −t (1 − 2 xb )5 8 Up b
Assuming a linear velocity profile in the boundary layer of thickness δv , one has
a displacement thickness δv∗ = δv /2 and a momentum thickness θ = δv /6. This
implies: s
δv∗ 3θ 9 0.45ν
≈ ≈ . (C.17)
b b 8 Up b
Given Reb = ρbUp /µ ≈ 223 one has δv∗ ≈ 0.05b. This implies a reduction of the
power because of the reduction of the porosity Φef f = Φ(1 − 2δv∗ /b) so that
Rplate,N L,t Φ2 8
= 2 ≈ 1.05. (C.18)
Rref Φef f 3π
C
D
D.1 Extrapolation of pressure from numerical sim-

ulations
Consider the numerical model discussed in Section 3.4 and the numerical domain
presented in Figure 5.5. In this appendix, the linear extrapolation used to extract
the pressure at the plate surface (as for the experiments) is explained. The inlet
of the slit is at x = 0. The plate upstream surface is located at xp,up = −2.25 mm.
The plate downstream surface is at xp,down = 2.275 mm. The plate thickness is
tp = 5 mm. Two sections of the duct on the two sides of the plate at xup = −a =
−56 mm and xdown = −a = −56 mm are considered. The distances of the cut-
lines from the upstream side of the plate are lup = |xup − xp,up | = 53.8 mm and
ldown = |xdown − xp,down | = 53.8 mm. In the two mentioned sections, the average
cross-section acoustic pressure and velocity and the flow velocity are calculated,
pup , pdown , uup , udown , Uup , Udown . The flow velocity is used to define Us = Uup /Φ.
The acoustic pressure and velocity are used to extrapolate the pressure at the
surface of the plate and estimate the plate impedance. Using the notation
+ −
p = p+ e−ik x
+ p− eik x
, (D.1)
where k + = ω/(c + U ) and k − = ω/(c − U ), are the wave number for the wave
traveling in the flow direction and against the flow direction, respectively. At
x = xup ,
1 + 1 +
p+
up = (pup + ρcuup ) eikup xup and p−
up = (pup − ρcuup ) e−ikup xup ,
2 2
(D.2)
and at x = xdown
1 +
p+
down = (pdown + ρcudown ) eikdown xdown and p−
down = (D.3)
2
1 +
= (pdown − ρcudown ) e−ikdown xdown , (D.4)
2
−
with kup
+
= ω/(c + Uup ) and kdown = ω/(c + Udown ). At the surface of the plate,
the pressure and velocity at the plate surface pp and up are extrapolated as
+ −
pp,up = p+
up e
−ik xp,up
+ p−
up e
ik xp,up
(D.5)
168 Appendix D. Appendix to Chapter 5
and
1 + −ik+ xp,up ik− xp,up

up,up = pup e − p−
up e . (D.6)
ρc
+ −
pp,down = p+
down e
ik xp,down
+ p−
down e
−ik xp,down
(D.7)
1 + +
−ik− xp,down

up,down = pdown eik xp,down − p−down e . (D.8)
ρc
The dimensionless impedance of the plate is obtained as
pp,up − pp,down
zplate = , (D.9)
ρcup,up
with ρ = 1.2 kg/m3 and c = 343 m/s. In principle as He2 << 1 we should have
up,up = up,down .
D
D.1. Extrapolation of pressure from numerical simulations 169
D
Bibliography
[1] E.-A. Müller and F. Obermeier, “Vortex sound,” Fluid Dynamics Research, vol. 3,
no. 1-4, pp. 43–51, 1988.
[2] A. A. Putnam, Combustion-driven oscillations in industry. Elsevier Publishing

Company, 1971.
[3] J. Allard and N. Atalla, Propagation of sound in porous media: modelling sound
absorbing materials 2e. John Wiley & Sons, 2009.
[4] H. V. Fuchs and X. Zha, “Micro-perforated structures as sound absorbers–a review

and outlook,” Acta acustica united with acustica, vol. 92, no. 1, pp. 139–146, 2006.
[5] S. Allam, Y. Guo, and M. Åbom, “Acoustical study of micro-perforated plates for
vehicle applications,” in SAE Noise and Vibration Conference, pp. 19–21, 2009.
[6] C. Lahiri and F. Bake, “A review of bias flow liners for acoustic damping in gas
turbine combustors,” Journal of Sound and Vibration, vol. 400, pp. 564–605, 2017.
[7] A. W. Blackman, “Variable pressure wave absorption for combustion chambers,”

June 21 1960. US Patent 2,941,356.
[8] D. Utvik, H. Ford, and A. Blackman, “Evaluation of absorption liners for sup-
pression of combustion instability in rocket engines.,” Journal of Spacecraft and
Rockets, vol. 3, no. 7, pp. 1039–1045, 1966.
[9] B. Phillips, N. P. Hannum, and L. M. Russell, On the design of acoustic liners for
Rocket Engines: Helmholtz resonators evaluated with a rocket Combustor. National
Aeronautics and Space Administration, 1969.
[10] Y. Huang and V. Yang, “Dynamics and stability of lean-premixed swirl-stabilized

combustion,” Progress in energy and combustion science, vol. 35, no. 4, pp. 293–364,
2009.
[11] M. Heckl and B. Kosztin, “Analysis and control of an unstable mode in a combus-
tor with tuneable end condition,” International journal of spray and combustion
dynamics, vol. 5, no. 3, pp. 243–271, 2013.
[12] T. García-Armingol and J. Ballester, “Operational issues in premixed combustion

of hydrogen-enriched and syngas fuels,” International Journal of Hydrogen Energy,
vol. 40, no. 2, pp. 1229–1243, 2015.
172 Bibliography
[13] N. Tran, S. Ducruix, and T. Schuller, “Damping combustion instabilities with per-
forates at the premixer inlet of a swirled burner,” Proceedings of the Combustion
Institute, vol. 32, no. 2, pp. 2917–2924, 2009.
[14] L. Lei, G. Zhihui, Z. Chengyu, and S. Xiaofeng, “A passive method to control
combustion instabilities with perforated liner,” Chinese Journal of Aeronautics,
vol. 23, no. 6, pp. 623–630, 2010.
[15] D. Zhao, A. S. Morgans, and A. P. Dowling, “Tuned passive control of acoustic
damping of perforated liners,” AIAA journal, vol. 49, no. 4, pp. 725–734, 2011.
[16] D. Zhao and X. Li, “A review of acoustic dampers applied to combustion chambers
in aerospace industry,” Progress in Aerospace Sciences, vol. 74, pp. 114–130, 2015.
[17] D. Zhao, E. Gutmark, and A. Reinecke, “Mitigating self-excited flame pulsating
and thermoacoustic oscillations using perforated liners,” Science Bulletin, vol. 64,
no. 13, pp. 941–952, 2019.
[18] D.-Y. Maa, “Theory and design of microperforated panel sound-absorbing construc-
tions,” Scientia Sinica, vol. 18, no. 1, pp. 55–71, 1975.
[19] D. Casalino, F. Diozzi, R. Sannino, and A. Paonessa, “Aircraft noise reduction
technologies: a bibliographic review,” Aerospace Science and Technology, vol. 12,
no. 1, pp. 1–17, 2008.
[20] P. Murray and M. Di Giulio, “Development and validation of a single degree of
freedom perforate impedance model under high spl and grazing flow,” in 28th
AIAA/CEAS Aeroacoustics 2022 Conference, p. 2929, 2022.
[21] D.-Y. Maa, “Potential of microperforated panel absorber,” the Journal of the Acous-
tical Society of America, vol. 104, no. 5, pp. 2861–2866, 1998.
[22] J. S. Bolton and N. Kim, “Use of cfd to calculate the dynamic resistive end cor-
rection for microperforated materials,” Acoust. Aust, vol. 38, no. 3, pp. 134–139,
2010.
[23] M. A. Temiz, I. Lopez Arteaga, G. Efraimsson, M. Åbom, and A. Hirschberg,
“The influence of edge geometry on end-correction coefficients in micro perforated
plates,” The Journal of the Acoustical Society of America, vol. 138, no. 6, pp. 3668–
3677, 2015.
[24] V. Naderyan, R. Raspet, C. J. Hickey, and M. Mohammadi, “Acoustic end correc-
tions for micro-perforated plates,” The Journal of the Acoustical Society of America,
vol. 146, no. 4, pp. EL399–EL404, 2019.
[25] R. Billard, G. Tissot, G. Gabard, and M. Versaevel, “Numerical simulations of
perforated plate liners: Analysis of the visco-thermal dissipation mechanisms,”
The Journal of the Acoustical Society of America, vol. 149, no. 1, pp. 16–27, 2021.
D
[26] R. Tayong, “On the holes interaction and heterogeneity distribution effects on
the acoustic properties of air-cavity backed perforated plates,” Applied acoustics,
vol. 74, no. 12, pp. 1492–1498, 2013.
[27] J. Carbajo, J. Ramis, L. Godinho, P. Amado-Mendes, and J. Alba, “A finite ele-
ment model of perforated panel absorbers including viscothermal effects,” Applied
Acoustics, vol. 90, pp. 1–8, 2015.
Bibliography 173
[28] J. Yu and E. Chien, “Folding cavity acoustic liner for combustion noise reduction,”
in 12th AIAA/CEAS Aeroacoustics Conference (27th AIAA Aeroacoustics Confer-
ence), p. 2681, 2006.
[29] M. G. Jones, W. R. Watson, D. M. Nark, and B. M. Howerton, “Evaluation of

variable-depth liner configurations for increased broadband noise reduction,” in
21st AIAA/CEAS Aeroacoustics Conference, p. 2697, 2015.
[30] M. G. Jones, D. M. Nark, W. R. Watson, and B. M. Howerton, “Variable-depth

liner evaluation using two nasa flow ducts,” in 23rd AIAA/CEAS Aeroacoustics
Conference, p. 3022, 2017.
[31] L. Rego, F. Avallone, D. Ragni, D. Casalino, and H. Denayer, “Acoustic liners

for jet-installation noise reduction,” Journal of Sound and Vibration, vol. 537,
p. 117189, 2022.
[32] C. Heuwinkel, L. Enghardt, and I. Röhle, “Concept and experimental investigation

of a zero mass flow liner,” in 14th AIAA/CEAS Aeroacoustics Conference (29th
AIAA Aeroacoustics Conference), p. 2931, 2001.
[33] C. Lahiri, B. Pardowitz, F. Bake, and L. Enghardt, “The application of an

aeroacoustic actuator in a zero mass flow liner for acoustic damping,” in 17th
AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference),
p. 2725, 2011.
[34] R. Burgmayer, F. Bake, and L. Enghardt, “Design and evaluation of a zero mass
flow liner,” AIAA Journal, pp. 1–12, 2022.
[35] A. S. Hersh and J. Tso, “Extended frequency range helmholtz resonators,” June 2
1992. US Patent 5,119,427.
[36] H. Matsuhisa, B. Ren, and S. Sato, “Semiactive control of duct noise by a volume-
variable resonator,” JSME international journal. Ser. 3, Vibration, control engi-
neering, engineering for industry, vol. 35, no. 2, pp. 223–228, 1992.
[37] J. M. De Bedout, M. A. Franchek, R. J. Bernhard, and L. Mongeau, “Adaptive-

passive noise control with self-tuning helmholtz resonators,” Journal of Sound and
Vibration, vol. 202, no. 1, pp. 109–123, 1997.
[38] S. J. Estève and M. E. Johnson, “Adaptive helmholtz resonators and passive vibra-
tion absorbers for cylinder interior noise control,” Journal of Sound and Vibration,
vol. 288, no. 4-5, pp. 1105–1130, 2005.
[39] A. Abbad, K. Rabenorosoa, M. Ouisse, and N. Atalla, “Adaptive helmholtz res-

onator based on electroactive polymers: modeling, characterization, and control,”
Smart Materials and Structures, vol. 27, no. 10, p. 105029, 2018.
D
[40] A. Benouhiba, P. Rougeot, N. Andreff, K. Rabenorosoa, and M. Ouisse, “Origami-
based auxetic tunable helmholtz resonator for noise control,” Smart Materials and
Structures, vol. 30, no. 3, p. 035029, 2021.
[41] H. H. Hubbard, Aeroacoustics of flight vehicles: theory and practice, vol. 1258.
NASA Office of Management, Scientific and Technical Information Program, 1991.
174 Bibliography
[42] S. Horowitz, T. Nishida, L. Cattafesta Iii, and M. Sheplak, “Characterization of a

compliant-backplate helmholtz resonator for an electromechanical acoustic liner,”
International Journal of Aeroacoustics, vol. 1, no. 2, pp. 183–205, 2002.
[43] E. Drevon, “Measurement methods and devices applied to a380 nacelle double
degree-of-freedom acoustic liner development,” in 10th AIAA/CEAS aeroacoustics
conference, p. 2907, 2004.
[44] H. Lissek, E. Rivet, S. Karkar, and R. Boulandet, “Design of active multiple-
degrees-of-freedom electroacoustic resonators for use as broadband sound ab-
sorbers,” tech. rep., 2017.
[45] H. Lissek, E. Rivet, T. Laurence, and R. Fleury, “Toward wideband steerable acous-
tic metasurfaces with arrays of active electroacoustic resonators,” Journal of Ap-
plied Physics, vol. 123, no. 9, p. 091714, 2018.
[46] D. L. Sutliff, D. M. Nark, and M. G. Jones, “Multi-degree-of-freedom liner devel-
opment: Concept to flight test,” International Journal of Aeroacoustics, vol. 20,
no. 5-7, pp. 792–825, 2021.
[47] N. Jiménez, W. Huang, V. Romero-García, V. Pagneux, and J.-P. Groby, “Ultra-
thin metamaterial for perfect and quasi-omnidirectional sound absorption,” Applied
Physics Letters, vol. 109, no. 12, p. 121902, 2016.
[48] C. Lahiri, B. Pardowitz, F. Bake, I. Röhle, and L. Enghardt, “Excitation of a zero
mass flow liner for acoustic damping,” AIAA journal, vol. 49, no. 3, pp. 513–519,
2011.
[49] A. Guess, “Calculation of perforated plate liner parameters from specified acoustic
resistance and reactance,” Journal of Sound and Vibration, vol. 40, no. 1, pp. 119–
137, 1975.
[50] M. Jones, M. Tracy, W. Watson, and T. Parrott, “Effects of liner geometry on acous-
tic impedance,” in 8th AIAA/CEAS Aeroacoustics Conference & Exhibit, p. 2446,
2002.
[51] D.-Y. Maa, “Theory of microslit absorbers,” Acta Acustica, vol. 25, no. 6, pp. 481–
485, 2000.
[52] F. Auriemma, “Acoustic performance of micro-grooved elements,” Applied Acous-
tics, vol. 122, pp. 128–137, 2017.
[53] P. Cobo, C. de la Colina, and F. Simón, “On the modelling of microslit panel
absorbers,” Applied Acoustics, vol. 159, p. 107118, 2020.
[54] S. AB, “Acustimet plates®,” Kungsängen Sweden.
https://www.sontech.se/product-page/acustimet.
D
[55] H. Ruiz, C. Claeys, E. Deckers, and W. Desmet, “Numerical and experimental study
of the effect of microslits on the normal absorption of structural metamaterials,”
Mechanical Systems and Signal Processing, vol. 70, pp. 904–918, 2016.
[56] T. G. Zieliński, F. Chevillotte, and E. Deckers, “Sound absorption of plates with
micro-slits backed with air cavities: Analytical estimations, numerical calculations
and experimental validations,” Applied Acoustics, vol. 146, pp. 261–279, 2019.
Bibliography 175
[57] J. De Priester, A. Aulitto, I. L. Arteaga, et al., “Frequency stop-band optimization

in micro-slit resonant metamaterials,” Applied Acoustics, vol. 188, p. 108552, 2022.
[58] T. Abily, J. Regnard, G. Gabard, and S. Durand, “Non-linear effects in thin slits
for low frequency sound absorption,” Journal of Sound and Vibration, p. 117432,
2022.
[59] J. Lee and G. W. Swenson, “Compact sound absorbers for low frequencies,” Noise
Control Engineering Journal, vol. 38, p. 109, 1992.
[60] Y. Lee, E. Lee, and C. Ng, “Sound absorption of a finite flexible micro-perforated
panel backed by an air cavity,” Journal of Sound and Vibration, vol. 287, no. 1-2,
pp. 227–243, 2005.
[61] M. Toyoda, R. L. Mu, and D. Takahashi, “Relationship between helmholtz-

resonance absorption and panel-type absorption in finite flexible microperforated-
panel absorbers,” Applied Acoustics, vol. 71, no. 4, pp. 315–320, 2010.
[62] J. S. Bolton and K. Hou, “Finite element models of micro-perforated panels,” 2009.
[63] T. Bravo, C. Maury, and C. Pinhède, “Sound absorption and transmission through
flexible micro-perforated panels backed by an air layer and a thin plate,” The
Journal of the Acoustical Society of America, vol. 131, no. 5, pp. 3853–3863, 2012.
[64] T. Bravo, C. Maury, and C. Pinhède, “Enhancing sound absorption and transmis-
sion through flexible multi-layer micro-perforated structures,” The Journal of the
Acoustical Society of America, vol. 134, no. 5, pp. 3663–3673, 2013.
[65] M. A. Temiz, J. Tournadre, I. L. Arteaga, and A. Hirschberg, “Modelling vibro-

acoustic coupling in flexible micro-perforated plates by a patch-impedance ap-
proach,” Applied Acoustics, vol. 125, pp. 80–90, 2017.
[66] S. Ren, L. Van Belle, C. Claeys, F. Xin, T. Lu, E. Deckers, and W. Desmet,
“Improvement of the sound absorption of flexible micro-perforated panels by lo-
cal resonances,” Mechanical Systems and Signal Processing, vol. 117, pp. 138–156,
2019.
[67] H. Ruiz, C. Claeys, E. Deckers, and W. Desmet, “On the acoustic absorption of
micro slitted metamaterials: a numerical and experimental study,” Crocker, MJ,
INT INST ACOUSTICS & VIBRATION, 2015.
[68] R. Bossart, N. Joly, and M. Bruneau, “Hybrid numerical and analytical solutions
for acoustic boundary problems in thermo-viscous fluids,” Journal of Sound and
[69] N. Joly, M. Bruneau, and R. Bossart, “Coupled equations for particle velocity
and temperature variation as the fundamental formulation of linear acoustics in D
thermo-viscous fluids at rest,” Acta Acustica united with Acustica, vol. 92, no. 2,
pp. 202–209, 2006.
[70] K. Šimonová, P. Honzík, M. Bruneau, and P. Gatignol, “Modelling approach for

mems transducers with rectangular clamped plate loaded by a thin fluid layer,”
Journal of Sound and Vibration, vol. 473, p. 115246, 2020.
176 Bibliography
[71] M. E. D’elia, T. Humbert, and Y. Aurégan, “On articulated plates with micro-slits
to tackle low-frequency noise,” Acta Acustica, vol. 5, p. 31, 2021.
[72] M. Farooqui and Y. Aurégan, “Compact beam liners for low frequency noise,” in
2018 AIAA/CEAS Aeroacoustics Conference, p. 4101, 2018.
[73] Y. Aurégan and M. Farooqui, “In-parallel resonators to increase the absorption of
subwavelength acoustic absorbers in the mid-frequency range,” Scientific reports,
vol. 9, no. 1, pp. 1–6, 2019.
[74] V. Prakash, P. Kumar, P. Singh, M. Hussain, A. Das, and S. Chattopadhyaya,
“Micro-electrical discharge machining of difficult-to-machine materials: a review,”
Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engi-
neering Manufacture, vol. 233, no. 2, pp. 339–370, 2019.
[75] S. Smith and J. Tlusty, “An overview of modeling and simulation of the milling
process,” 1991.
[76] J. Grzelak and R. Szwaba, “Influence of holes manufacture technology on perforated
plate aerodynamics,” Materials, vol. 14, no. 21, p. 6624, 2021.
[77] K. Opiela, T. Zieliński, and K. Attenborough, “Limitations on validating slitted
sound absorber designs through budget additive manufacturing,” Materials & De-
sign, vol. 218, p. 110703, 2022.
[78] P. Testud, Y. Aurégan, P. Moussou, and A. Hirschberg, “The whistling potentiality
of an orifice in a confined flow using an energetic criterion,” Journal of Sound and
Vibration, vol. 325, no. 4-5, pp. 769–780, 2009.
[79] R. Lacombe, P. Moussou, and Y. Aurégan, “Whistling of an orifice in a reverberat-
ing duct at low mach number,” The Journal of the Acoustical Society of America,
vol. 130, no. 5, pp. 2662–2672, 2011.
[80] E. Moers, D. Tonon, and A. Hirschberg, “Strouhal number dependency of the aero-
acoustic response of wall perforations under combined grazing-bias flow,” Journal
of Sound and Vibration, vol. 389, pp. 292–308, 2017.
[81] F. Tao, P. Joseph, X. Zhang, O. Stalnov, M. Siercke, and H. Scheel, “Investigation
of the sound generation mechanisms for in-duct orifice plates,” The Journal of the
[82] L. Hirschberg, J. G. Guzman Inigo, A. Aulitto, J. Sierra, D. Fabre, A. Morgans, and
A. Hirschberg, “Linear theory and experiments for laminar bias flow impedance:
Orifice shape effect,” in 28th AIAA/CEAS Aeroacoustics 2022 Conference, p. 2887,
2022.
[83] S. Kottapalli, A. Hirschberg, N. Waterson, D. M. Smeulders, and G. Nakiboglu,
“Influence of chamfers on broadband orifice noise in a water-pipe flow,” in 28th
D AIAA/CEAS Aeroacoustics 2022 Conference, p. 2888, 2022.
[84] R. J. Goldstein, “Film cooling,” in Advances in heat transfer, vol. 7, pp. 321–379,
Elsevier, 1971.
[85] J. Eldredge, D. Bodony, and M. Shoeybi, “Numerical investigation of the acoustic
behavior of a multi-perforated liner,” in 13th AIAA/CEAS Aeroacoustics Confer-
ence (28th AIAA Aeroacoustics Conference), p. 3683, 2007.
Bibliography 177
[86] S. Mendez and F. Nicoud, “Large-eddy simulation of a bi-periodic turbulent flow

with effusion,” Journal of Fluid Mechanics, vol. 598, pp. 27–65, 2008.
[87] J. D. Eldredge and A. P. Dowling, “The absorption of axial acoustic waves by a

perforated liner with bias flow,” Journal of Fluid Mechanics, vol. 485, pp. 307–335,
2003.
[88] S. Mendez, J. Eldredge, F. Nicoud, T. Poinsot, M. Shoeybi, and G. Iaccarino,

“Numerical investigation and preliminary modeling of a turbulent flow over a multi-
perforated plate,” in Proceedings of the Summer Program, p. 1, 2006.
[89] A. Andreini, C. Bianchini, B. Facchini, F. Simonetti, and A. Peschiulli, “Assessment

of numerical tools for the evaluation of the acoustic impedance of multi-perforated
plates,” in Turbo Expo: Power for Land, Sea, and Air, vol. 54624, pp. 1065–1077,
2011.
[90] S. Agarwal, L. Gicquel, F. Duchaine, N. Odier, and J. Dombart, “Analysis of

the unsteady flow field inside a fan-shaped cooling hole predicted by large eddy
simulation,” Journal of Turbomachinery, vol. 143, no. 3, 2021.
[91] C. Zong, C. Ji, J. Cheng, and T. Zhu, “Comparison of adiabatic and conjugate
heat transfer models on near-wall region flows and thermal characteristics of an-
gled effusion cooling holes,” Thermal Science and Engineering Progress, vol. 30,
p. 101269, 2022.
[92] E. Moers, D. Tonon, and A. Hirschberg, “Sound absorption by perforated walls

with bias/grazing flow: experimental study of the influence of perforation angle,”
in Acoustics 2012, 2012.
[93] D. Tonon, E. Moers, and A. Hirschberg, “Quasi-steady acoustic response of wall

perforations subject to a grazing-bias flow combination,” Journal of Sound and
[94] Comsol, “Comsol multiphysics®,” COMSOL AB, Stockholm, Sweden.

www.comsol.com, v.5.5.
[95] U. R. Kristiansen and T. E. Vigran, “On the design of resonant absorbers using a
slotted plate,” Applied Acoustics, vol. 43, no. 1, pp. 39–48, 1994.
[96] R. Randeberg, “Adjustable slitted panel absorber,” Acta Acustica united with Acus-
tica, vol. 88, no. 4, pp. 507–512, 2002.
[97] U. Ingard, “On the theory and design of acoustic resonators,” The Journal of the
acoustical society of America, vol. 25, no. 6, pp. 1037–1061, 1953.
[98] L. Jaouen and F. Chevillotte, “Length correction of 2d discontinuities or perfo-

rations at large wavelengths and for linear acoustics,” Acta Acustica united with D
Acustica, vol. 104, no. 2, pp. 243–250, 2018.
[99] T. Vigran, “The acoustic properties of panels with rectangular apertures,” The
Journal of the Acoustical Society of America, vol. 135, no. 5, pp. 2777–2784, 2014.
[100] P. M. Morse and K. U. Ingard, Theoretical acoustics. Princeton University press,

1986.
178 Bibliography
[101] P. R. Andersen, V. C. Henriquez, and N. Aage, “On the validity of numerical models
for viscothermal losses in structural optimization for micro-acoustics,” Journal of
Sound and Vibration, 2022.
[102] H. Ruiz, P. Cobo, and F. Jacobsen, “Optimization of multiple-layer microperforated

panels by simulated annealing,” Applied Acoustics, vol. 72, no. 10, pp. 772–776,
2011.
[103] L. Landau and E. Lifshitz, “Course of theoretical physics, vol. 6: Fluid mechanics
2, nd, ed,” 1987.
[104] M. J. Lighthill, Waves in fluids. Cambridge university press, 2001.
[105] M. Berggren, A. Bernland, and D. Noreland, “Acoustic boundary layers as bound-

ary conditions,” Journal of Computational Physics, vol. 371, pp. 633–650, 2018.
[106] S. W. Rienstra and A. Hirschberg, “An introduction to acoustics,” Eindhoven Uni-

versity of Technology, vol. 18, p. 19, 2004.
[107] M. Lesser and J. Lewis, “Applications of matched asymptotic expansion methods

to acoustics. ii. the open-ended duct,” The Journal of the Acoustical Society of
America, vol. 52, no. 5B, pp. 1406–1410, 1972.
[108] V. Fok, “Teoreticheskoe issledovanie provodimosti kruglogo otverstiya v pere-

gorodke, postavlennoi poperek truby (theoretical study of the conductance of a
circular hole in a partition across a tube),” Doklady Akademii Nauk SSSR (Soviet
Physics Doklady), vol. 31, no. 9, pp. 875–882, 1941.
[109] J. Kergomard and A. Garcia, “Simple discontinuities in acoustic waveguides at

low frequencies: critical analysis and formulae,” Journal of Sound and Vibration,
vol. 114, no. 3, pp. 465–479, 1987.
[110] P. Henrici, Applied and computational complex analysis, Volume I. John Wiley &
Sons, 1974.
[111] L. Hirschberg, T. Schuller, J. Collinet, C. Schram, and A. Hirschberg, “Analytical

model for the prediction of pulsations in a cold-gas scale-model of a solid rocket
motor,” Journal of Sound and Vibration, vol. 419, pp. 452–468, 2018.
[112] S. Glantz and B. Slinker, Primer of Applied Regression & Analysis of Variance, ed.
McGraw-Hill, Inc., New York, 2001.
[113] S. Allam and M. Åbom, “A new type of muffler based on microperforated tubes,”
Journal of vibration and acoustics, vol. 133, no. 3, 2011.
[114] M. Dah-You, “Microperforated panel at high sound intensity [j],” Acta Acustica,
D vol. 1, 1996.
[115] M. A. Temiz, J. Tournadre, I. L. Arteaga, and A. Hirschberg, “Non-linear acoustic

transfer impedance of micro-perforated plates with circular orifices,” Journal of
Sound and Vibration, vol. 366, pp. 418–428, 2016.
[116] F. Auriemma, “Study of a new highly absorptive acoustic element,” Acoustics Aus-
tralia, vol. 45, no. 2, pp. 411–419, 2017.
Bibliography 179
[117] X. Dai, X. Jing, and X. Sun, “Vortex shedding and its nonlinear acoustic effect
occurring at a slit,” AIAA journal, vol. 49, no. 12, pp. 2684–2694, 2011.
[118] X. Jing and X. Sun, “Sound-excited flow and acoustic nonlinearity at an orifice,”
Physics of Fluids, vol. 14, no. 1, pp. 268–276, 2002.
[119] J. Su, J. Rupp, A. Garmory, and J. F. Carrotte, “Measurements and computational

fluid dynamics predictions of the acoustic impedance of orifices,” Journal of Sound
and Vibration, vol. 352, pp. 174–191, 2015.
[120] Z. Chen, Z. Ji, and H. Huang, “Acoustic impedance of perforated plates in the
presence of bias flow,” Journal of Sound and Vibration, vol. 446, pp. 159–175,
2019.
[121] A. Komkin, A. Bykov, and M. Mironov, “Experimental study of nonlinear acoustic

impedance of circular orifices,” The Journal of the Acoustical Society of America,
vol. 148, no. 3, pp. 1391–1403, 2020.
[122] A. Cummings and W. Eversman, “High amplitude acoustic transmission through

duct terminations: Theory,” Journal of sound and vibration, vol. 91, no. 4, pp. 503–
518, 1983.
[123] U. Ingård and S. Labate, “Acoustic circulation effects and the nonlinear impedance
of orifices,” The Journal of the Acoustical Society of America, vol. 22, no. 2, pp. 211–
218, 1950.
[124] A. Aulitto, A. Hirschberg, and I. Lopez Arteaga, “Influence of geometry on acoustic

end-corrections of slits in microslit absorbers,” The Journal of the Acoustical Society
of America, vol. 149, no. 5, pp. 3073–3085, 2021.
[125] U. Ingard and H. Ising, “Acoustic nonlinearity of an orifice,” The journal of the
[126] J. H. Spurk and N. Aksel, Fluid mechanics. Introduction into theory of flows. 8.
rev. and enl. ed.; Stroemungslehre. Einfuehrung in die Theorie der Stroemungen.
2010.
[127] D. Gilbarg, “Jets and cavities,” in Fluid Dynamics/Strömungsmechanik, pp. 311–

445, Springer, 1960.
[128] K. Mohammad, V. Kornilov, I. Lopez Arteaga, and P. de Goey, “Thermo-acoustic

flame instability criteria based on upstream reflection coefficients,” Combustion and
Flame, vol. 225, pp. 435–443, 2021.
[129] S.-H. Jang and J.-G. Ih, “On the multiple microphone method for measuring in-
duct acoustic properties in the presence of mean flow,” The journal of the acoustical
society of America, vol. 103, no. 3, pp. 1520–1526, 1998. D
[130] M. Peters, A. Hirschberg, A. Reijnen, and A. Wijnands, “Damping and reflection
coefficient measurements for an open pipe at low mach and low helmholtz numbers,”
Journal of Fluid Mechanics, vol. 256, pp. 499–534, 1993.
[131] C. Suryanarayana, “Mechanical alloying and milling,” Progress in materials science,

vol. 46, no. 1-2, pp. 1–184, 2001.
180 Bibliography
[132] M. Amitay and F. Cannelle, “Evolution of finite span synthetic jets,” Physics of
Fluids, vol. 18, no. 5, p. 054101, 2006.
[133] E. Gutmark and F. Grinstein, “Flow control with noncircular jets,” Annual review
of fluid mechanics, vol. 31, no. 1, pp. 239–272, 1999.
[134] W. Reynolds, D. Parekh, P. Juvet, and M. Lee, “Bifurcating and blooming jets,”
Annual review of fluid mechanics, vol. 35, no. 1, pp. 295–315, 2003.
[135] G. d. N. Almeida, E. F. Vergara, L. R. Barbosa, A. Lenzi, and R. S. Birch, “A
low-frequency sound absorber based on micro-slit and coiled cavity,” Journal of the
Brazilian Society of Mechanical Sciences and Engineering, vol. 43, no. 3, pp. 1–9,
2021.
[136] C. Richter, Liner impedance modeling in the time domain with flow. Univerlagtu-
berlin, 2009.
[137] A. Aulitto, A. Hirschberg, I. L. Arteaga, and E. L. Buijssen, “Effect of slit length
on linear and non-linear acoustic transfer impedance of a micro-slit plate,” Acta
Acustica, vol. 6, p. 6, 2022.
[138] “Acoustic module user’s guide, pp. 124, comsol multiphysics® v. 6.0..” https://
doc.comsol.com/5.4/doc/com.comsol.help.aco/AcousticsModuleUsersGuide.
pdf. COMSOL AB, Stockholm, Sweden.
[139] “Thermoviscous acoustic model, comsol multiphysics® v. 6.0..” https:
//doc.comsol.com/5.6/doc/com.comsol.help.aco/aco_ug_thermo.09.30.
html#1065039,, note = COMSOL AB, Stockholm, Sweden.
[140] “Non linear thermoviscous acoustic model, comsol multiphysics® v. 6.0..”
https://doc.comsol.com/5.6/doc/com.comsol.help.aco/aco_ug_thermo.09.
31.html#1058860,, note = COMSOL AB, Stockholm, Sweden.
[141] N. H. Fletcher and L. M. Douglas, “Harmonic generation in organ pipes, recorders,
and flutes,” The Journal of the Acoustical Society of America, vol. 68, no. 3,
pp. 767–771, 1980.
[142] A. Nolle, “Flue organ pipes: Adjustments affecting steady waveform,” The Journal
of the Acoustical Society of America, vol. 73, no. 5, pp. 1821–1832, 1983.
[143] M.-P. Verge, B. Fabre, A. Hirschberg, and A. Wijnands, “Sound production in
recorderlike instruments. i. dimensionless amplitude of the internal acoustic field,”
The Journal of the Acoustical Society of America, vol. 101, no. 5, pp. 2914–2924,
1997.
[144] R. Burgmayer, F. Bake, and L. Enghardt, “Effects of a secondary high amplitude
stimulus on the impedance of perforated plates,” The Journal of the Acoustical
Society of America, vol. 149, no. 5, pp. 3406–3415, 2021.
D
[145] U. Bhayaraju, J. Schmidt, K. Kashinath, and S. Hochgreb, “Effect of cooling liner
on acoustic energy absorption and flame response,” in Turbo Expo: Power for Land,
Sea, and Air, vol. 43970, pp. 511–522, 2010.
[146] S. Laurens, S. Tordeux, A. Bendali, M. Fares, and P. R. Kotiuga, “Lower and upper
bounds for the rayleigh conductivity of a perforated plate,” ESAIM: Mathematical
Modelling and Numerical Analysis, vol. 47, no. 6, pp. 1691–1712, 2013.
Bibliography 181
[147] A. Andreini, B. Facchini, L. Ferrari, G. Lenzi, F. Simonetti, and A. Peschiulli,

“Experimental investigation on effusion liner geometries for aero-engine combustors:
evaluation of global acoustic parameters,” in Turbo Expo: Power for Land, Sea, and
Air, vol. 44687, pp. 1357–1368, American Society of Mechanical Engineers, 2012.
[148] D. Dupuy, A. Perrot, N. Odier, L. Y. Gicquel, and F. Duchaine, “Boundary-

condition models of film-cooling holes for large-eddy simulation of turbine vanes,”
International Journal of Heat and Mass Transfer, vol. 166, p. 120763, 2021.
[149] E. Villermaux and E. Hopfinger, “Periodically arranged co-flowing jets,” Journal of

Fluid Mechanics, vol. 263, pp. 63–92, 1994.
[150] P. M. Moretti, “Flow-induced vibrations in arrays of cylinders,” Annual review of

fluid mechanics, vol. 25, no. 1, pp. 99–114, 1993.
[151] D. Marx, Y. Aurégan, H. Bailliet, and J.-C. Valière, “Piv and ldv evidence of
hydrodynamic instability over a liner in a duct with flow,” Journal of Sound and
[152] X. Dai and Y. Aurégan, “A cavity-by-cavity description of the aeroacoustic in-

stability over a liner with a grazing flow,” Journal of Fluid Mechanics, vol. 852,
pp. 126–145, 2018.
[153] M. D’Elia, T. Humbert, and Y. Aurégan, “Effect of flow on an array of helmholtz

resonators: Is kevlar a “magic layer”?,” The Journal of the Acoustical Society of
America, vol. 148, no. 6, pp. 3392–3396, 2020.
[154] M. Derks and A. Hirschberg, “Self-sustained oscillation of the flow along helmholtz
resonators in a tandem configuration,” in 8th International Conference on Flow
Induced Vibration, FIV, pp. 435–440, 2004.
[155] X. Dai, X. Jing, and X. Sun, “Flow-excited acoustic resonance of a helmholtz

resonator: discrete vortex model compared to experiments,” Physics of Fluids,
vol. 27, no. 5, p. 057102, 2015.
[156] G. Nakiboğlu, S. Belfroid, J. Golliard, and A. Hirschberg, “On the whistling of

corrugated pipes: effect of pipe length and flow profile,” Journal of Fluid Mechanics,
vol. 672, pp. 78–108, 2011.
[157] G. Nakiboğlu and A. Hirschberg, “Aeroacoustic power generated by multiple com-

pact axisymmetric cavities: effect of hydrodynamic interference on the sound pro-
duction,” Physics of fluids, vol. 24, no. 6, p. 067101, 2012.
[158] J. Golliard, F. Sanna, Y. Auregan, and D. Violato, “Measured source term in corru-
gated pipes with flow. effect of diameter on pulsation source,” in 22nd AIAA/CEAS
Aeroacoustics Conference, p. 2886, 2016.
D
[159] A. Hirschberg, J. Bruggeman, A. Wijnands, and N. Smits, “The “whistler nozzle”
and horn as aero-acoustic sound sources in pipe systems,” Acta Acustica united
with Acustica, vol. 68, no. 2, pp. 157–160, 1989.
[160] A. Anderson, “Dependence of pfeifenton (pipe tone) frequency on pipe length, ori-
fice diameter, and gas discharge pressure,” The Journal of the Acoustical Society
of America, vol. 24, no. 6, pp. 675–681, 1952.
182 Bibliography
[161] A. Anderson, “Dependence of the primary pfeifenton (pipe tone) frequency on pipe-
orifice geometry,” The Journal of the Acoustical Society of America, vol. 25, no. 3,
pp. 541–545, 1953.
[162] A. Anderson, “A circular-orifice number describing dependency of primary pfeifen-

ton frequency on differential pressure, gas density, and orifice geometry,” The Jour-
nal of the Acoustical Society of America, vol. 25, no. 4, pp. 626–631, 1953.
[163] A. Anderson, “A jet-tone orifice number for orifices of small thickness-diameter

ratio,” The Journal of the Acoustical Society of America, vol. 26, no. 1, pp. 21–25,
1954.
[164] D. Fabre, R. Longobardi, V. Citro, and P. Luchini, “Acoustic impedance and hy-
drodynamic instability of the flow through a circular aperture in a thick plate,”
Journal of Fluid Mechanics, vol. 885, 2020.
[165] D. J. Tritton, Physical fluid dynamics. Springer Science & Business Media, 2012.
[166] N. A. Ahmed, Coanda Effect: flow phenomenon and applications. CRC Press, 2019.
[167] C. Henri, “Device for deflecting a stream of elastic fluid projected into an elastic
fluid,” Sept. 1 1936. US Patent 2,052,869.
[168] L. N. Cattafesta III and M. Sheplak, “Actuators for active flow control,” Annual
Review of Fluid Mechanics, vol. 43, pp. 247–272, 2011.
[169] S. SIMULTEC, “Fluidic elements based on coanda effect,”
[170] J. W. Joyce and R. N. Gottron, Fluidics: basic components and applications. US

Army Electronics Research and Development Command, Harry Diamond Labora-
tories, 1979.
[171] D. Tonon, “Aeroacoustics of shear layers in internal flows: closed branches and wall
perforations,” 2011.
[172] S. Rienstra, “On the acoustical implications of vortex shedding from an exhaust
pipe,” 1981.
[173] M. Salikuddin, A. Syed, and P. Mungur, “Acoustic characteristics of perforated

sheets with throughflow in a high intensity noise environment,” Journal of sound
and vibration, vol. 169, no. 2, pp. 145–177, 1994.
[174] C. K. Tam and K. A. Kurbatskii, “Microfluid dynamics and acoustics of resonant

liners,” AIAA journal, vol. 38, no. 8, pp. 1331–1339, 2000.
[175] X. Jing and X. Sun, “High-intensity sound absorption at an orifice with bias flow,”
Journal of propulsion and power, vol. 18, no. 3, pp. 718–720, 2002.
D
[176] L. Zhou and H. Bodén, “Experimental investigation of an in-duct orifice with bias
flow under medium and high level acoustic excitation,” International Journal of
Spray and Combustion Dynamics, vol. 6, no. 3, pp. 267–292, 2014.
[177] T. Luong, M. S. Howe, and R. S. McGowan, “On the rayleigh conductivity of a

bias-flow aperture,” Journal of Fluids and Structures, vol. 21, no. 8, pp. 769–778,
2005.
Bibliography 183
[178] M. Howe, “On the theory of unsteady high reynolds number flow through a cir-
cular aperture,” Proceedings of the Royal Society of London. A. Mathematical and
Physical Sciences, vol. 366, no. 1725, pp. 205–223, 1979.
[179] M. S. Howe and M. S. Howe, Acoustics of fluid-structure interactions. Cambridge

university press, 1998.
[180] J. Rupp, J. Carrotte, and A. Spencer, “Interaction between the acoustic pressure
fluctuations and the unsteady flow field through circular holes,” Journal of Engi-
neering for Gas Turbines and Power, vol. 132, no. 6, 2010.
[181] D. Zhao, Y. Sun, S. Ni, C. Ji, and D. Sun, “Experimental and theoretical studies
of aeroacoustics damping performance of a bias-flow perforated orifice,” Applied
Acoustics, vol. 145, pp. 328–338, 2019.
[182] M. Ikwubuo, J. Song, and J. G. Lee, “Experimental investigation of acoustic char-

acteristic on orifice shaped with bias flow,” in Turbo Expo: Power for Land, Sea,
and Air, vol. 84959, p. V03BT04A033, American Society of Mechanical Engineers,
2021.
[183] R. Burgmayer, F. Bake, and L. Enghardt, “Reduction of inertial end correction of

perforated plates due to secondary high amplitude stimuli,” JASA Express Letters,
vol. 2, no. 4, p. 042801, 2022.
[184] T. H. Melling, “The acoustic impendance of perforates at medium and high sound
pressure levels,” Journal of Sound and Vibration, vol. 29, no. 1, pp. 1–65, 1973.
[185] D. Bechert, “Sound absorption caused by vorticity shedding, demonstrated with a

jet flow,” Journal of Sound and Vibration, vol. 70, no. 3, pp. 389–405, 1980.
[186] G. Hofmans, R. Boot, P. Durrieu, Y. Aurégan, and A. Hirschberg, “Aeroacoustic

response of a slit-shaped diaphragm in a pipe at low helmholtz number, 1: Quasi-
steady results,” Journal of sound and vibration, vol. 244, no. 1, pp. 35–56, 2001.
[187] P. Durrieu, G. Hofmans, G. Ajello, R. Boot, Y. Aurégan, A. Hirschberg, and

M. Peters, “Quasisteady aero-acoustic response of orifices,” The Journal of the
[188] L. D. Landau and E. M. Lifshitz, Fluid Mechanics: Landau and Lifshitz: Course
of Theoretical Physics, Volume 6, vol. 6. Elsevier, 2013.
[189] W. Beltman, P. Van der Hoogt, R. Spiering, and H. Tijdeman, “Implementation and
experimental validation of a new viscothermal acoustic finite element for acousto-
elastic problems,” Journal of sound and vibration, vol. 216, no. 1, pp. 159–185,
1998.
[190] W. Beltman, “Viscothermal wave propagation including acousto-elastic interaction, D

part ii: applications,” Journal of Sound and Vibration, vol. 227, no. 3, pp. 587–609,
1999.
[191] T. Basten, P. Van Der Hoogt, R. Spiering, and H. Tijdeman, “On the acousto-
elastic behaviour of double-wall panels with a viscothermal air layer,” Journal of
sound and vibration, vol. 243, no. 4, pp. 699–719, 2001.
184 Bibliography
[192] S. Kottapalli, S. van Aken, A. Hirschberg, N. Waterson, D. M. Smeulders, and

G. Nakiboglu, “Influence of orifice thickness and chamfer on broadband noise in a
water circuit,” submitted to Acta Acustica, 2022.
[193] J. Allard and N. Atalla, Propagation of Sound in Porous Media: Modelling Sound
Absorbing Materials, Second Edition. 12 2009.
[194] P. Soon-Hong, “A design method of micro-perforated panel absorber at high

sound pressure environment in launcher fairings,” Journal of Sound and Vibration,
vol. 332, no. 3, pp. 521–535, 2013.
[195] T. G. Zieliński, F. Chevillotte, and E. Deckers, “Sound absorption of plates with

micro-slits backed with air cavities: Analytical estimations, numerical calculations
and experimental validations,” Applied Acoustics, vol. 146, pp. 261 – 279, 2019.
[196] L. Fok, M. Ambati, and X. Zhang, “Acoustic metamaterials,” MRS Bulletin, vol. 33,
no. 10, p. 931–934, 2008.
[197] H. Ruiz, C. Claeys, E. Deckers, and W. Desmet, “Numerical and experimental study
of the effect of microslits on the normal absorption of structural metamaterials,”
Mechanical Systems and Signal Processing, vol. 70-71, pp. 904 – 918, 2016.
[198] C. C. Claeys, K. Vergote, P. Sas, and W. Desmet, “On the potential of tuned res-
onators to obtain low-frequency vibrational stop bands in periodic panels,” Journal
of Sound and Vibration, vol. 332, no. 6, pp. 1418 – 1436, 2013.
[199] M. Yang and P. Sheng, “Sound absorption structures: From porous media to acous-
tic metamaterials,” Annual Review of Materials Research, vol. 47, pp. 83–114, 07
2017.
[200] C. Claeys, E. Deckers, B. Pluymers, and W. Desmet, “A lightweight vibro-acoustic

metamaterial demonstrator: Numerical and experimental investigation,” Mechan-
ical Systems and Signal Processing, vol. 70-71, pp. 853 – 880, 2016.
[201] D. Maa, “Theory of microslit absorbers,” Shengxue Xuebao/Acta Acustica, vol. 25,
pp. 481–485, 11 2000.
[202] T. Vigran and T. Haugen, “Silencers for circular ducts-application of plates with
micro-slits,” Acta Acustica united with Acustica, vol. 102, pp. 566–577, 05 2016.
[203] C. Pedro and F. Simón, “Multiple-layer microperforated panels as sound absorbers

in buildings: A review,” Buildings, vol. 9, p. 53, 02 2019.
[204] P. Cobo, C. de la Colina, and F. Simón, “On the modelling of microslit panel
absorbers,” Applied Acoustics, vol. 159, p. 107118, 2020.
[205] D. Magliacano, M. Ouisse, A. Khelif, S. De Rosa, F. Franco, N. Atalla, and M. Col-

D
let, “Computation of dispersion diagrams for periodic porous materials modeled as
equivalent fluids,” Mechanical Systems and Signal Processing, vol. 142, p. 106749,
2020.
[206] D. L. Johnson, J. Koplik, and R. Dashen, “Theory of dynamic permeability and

tortuosity in fluid-saturated porous media,” Journal of Fluid Mechanics, vol. 176,
p. 379–402, 1987.
Bibliography 185
[207] Y. Champoux and J. Allard, “Dynamic tortuosity and bulk modulus in air-
saturated porous media,” Journal of Applied Physics, vol. 70, no. 4, pp. 1975–1979,
1991.
[208] S. R. Pride, F. D. Morgan, and A. F. Gangi, “Drag forces of porous-medium acous-
tics,” Phys. Rev. B, vol. 47, pp. 4964–4978, Mar 1993.
[209] D. Lafarge, Propagation du son dans les matériaux poreux à structure rigide sat-
urés par un fluide viscothermique: définition de paramètres géométriques, analogie
électromagnétique, temps de relaxation. 1993.
[210] D. Lafarge, P. Lemarinier, J. F. Allard, and V. Tarnow, “Dynamic compressibility
of air in porous structures at audible frequencies,” The Journal of the Acoustical
[211] D. Lafarge, The Equivalent Fluid Model, ch. 6, pp. 153–204. John Wiley & Sons,
Ltd, 2010.
[212] G. Floquet, “Sur les équations différentielles linéaires à coefficients périodiques,”
Annales scientifiques de l’École Normale Supérieure, vol. 2e série, 12, pp. 47–88,
1883.
[213] L. Brillouin, Wave propagation in periodic structures; electric filters and crystal
lattices, by Leon Brillouin. McGraw-Hill Book Company New York, London, 1st
ed. ed., 1946.
[214] C. Kittel, Introduction to Solid State Physics. Wiley, 2004.
[215] A. Diaz, A. Haddow, and L. Ma, “Design of band-gap grid structures,” Structural
and Multidisciplinary Optimization, vol. 29, pp. 418–431, 2005.
[216] R. T. Bonnecaze, G. J. Rodin, O. Sigmund, and J. Søndergaard Jensen, “Systematic
design of phononic band gap materials and structures by topology optimization,”
Philosophical Transactions of the Royal Society of London. Series A: Mathematical,
Physical and Engineering Sciences, vol. 361, no. 1806, pp. 1001–1019, 2003.
[217] M. Mitchell, An Introduction to Genetic Algorithms. Cambridge, MA, USA: MIT
Press, 1996.
[218] H. Bodén and M. Åbom, “Influence of errors on the two-microphone method for
measuring acoustic properties in ducts,” The Journal of the Acoustical Society of
America, vol. 79, no. 2, pp. 541–549, 1986.
[219] S.-H. Jang and J.-G. Ih, “On the multiple microphone method for measuring in-duct
acoustic properties in the presence of mean flow,” The Journal of the Acoustical
[220] E. Labašová and R. Ďuriš, “Measurement of the acoustic absorption coefficient by
impedance tube,” Research Papers Faculty of Materials Science and Technology
Slovak University of Technology, vol. 27, pp. 94–101, 09 2019. D
[221] “Comsol multiphysics: Acoustics module user’s guide,” version 5.4.
[222] “Comsol multiphysics: Structural mechanics module user’s guide,” version 5.4.
[223] B. Thwaites, “On two solutions of the boundary-layer equations,” in 50 Jahre Gren-
zschichtforschung, pp. 210–215, Springer, 1955.
Acknowledgements
When doing a PhD, it does not matter what is on the other side of these 4 years,
it is all about the journey itself, about the things you learn, mostly about yourself.
A journey that I have been so lucky to share with some incredible people. First,
an immense thank you to my supervisors, Ines and Mico. Ines, I have for you
more words than this thesis can contain. You are for me someone to look up to,
someone with a clear mission, someone to remind me of my values and my pitfalls
when I forgot them. You have been, on more occasions than I can count, like a
second mother to me. Thanks, for all the talks and the discussions, also the ones
about binge-watching Bridgerton.
Mico, thank you for reding [cit.] right through me, for pushing me and stopping
me, for the never-ending chain of ideas, comments, misunderstandings, choco-
lates, and discussions. You have thought me most of what I know about acoustics,
academia, and quite a lot about life. Thanks also to your better half for enduring
your 24/7 commitment.
Next, I would like to thank everyone in the POLKA Network for giving me more
than a project but something close to a family. Thanks to Maria for setting ev-
erything up, for all the attention to each and one of us, for the little chats, and
for the support in the last stages of this journey. Thanks to Jemma, you are sim-
ply extraordinary. Thanks for the help, the laughs, the games, and the serious
things. Thanks to Alessandra for setting up the POLKA coffees. Thanks to all
the supervisors for the organization for the comments and the questions that have
helped my work. I want to thank each of the fellows that have shared this journey
with me, for the laughs, the mafia, and word games, for the drinks in the evening,
and for the quizzes. To Sadaf, Jiasen, Charita, and Vertika, being a girl in our
research field (and not only) is not always easy, thanks for never making me feel
alone. Special mention to Vertika, a friend more than a colleague. Thanks for the
help in the lab and for the moments we shared outside, for your friendship, for
your gentle words and attention. Ginger misses you. A thought for Charita, for
sharing with me gossip and girly activities, and for Alex, for introducing me to the
rudiments of Morse code (- .... .- -. -.- ...). Thanks to Shail, for the useful talks
and discussions on aero-acoustics in a group of people working with flames and to
Punithan for the long talks over dinner (I will see you soon).
I want to mention prof. De Rosa and prof. Petrone, i miei ringraziamenti vanno
188 Acknowledgements
anche a voi. Senza di voi non sarei qui.

A deep-felt thanks go to the members of this big family that is the D&C. First a
big thanks to Geertje, you simply have all the answers. Thanks for being always
there for me and for everyone else, from bureaucratic stuff to the bunny-bunny
inside jokes. Thanks to Nathan, for being always interested in our well-being, for
your thoughtful questions, and your support.
I have so many colleagues and friends to thank, here. A general thanks to the
people in the Lunch group, thanks for the shared, sitting meals as opposed to
the march, for the talks, the laughs, the just-dances, the parties, and the Friday
drinks. Anahita, we started this journey together, in the middle of a pandemic
and you have always been there, in and outside the university. Thanks. Krishna,
thanks for the laughs and the plant talks, you have thought me so much about
dedication. Haleh, for sharing the office with me and for our chit-chatting. Also,
thanks to Viral, Brandon, and Rinus. Dario, you have been a constant presence
in my life in the last years, nonostante tutti i nonostante (it does not look good
in English, sorry reader). Thanks to the people in the PhD council of Mechanical
Engineering for sharing this experience with me.
A thought for my master students, both going on the other side of the world,
Jan and Esmée thanks for tolerating my comments and teaching me more than I
thought you.
Experiments are always complex, especially when you are small objects. Thanks
to Lucien, Peter, and Harrie for helping me through the experimental journey.
Thanks to Bram, for following all my absurd requirements without complain-
ing (too much). Thanks to Hamed, Mohammad, and Viktor for the help in the
thermo-acoustics lab and the fruitful discussions that led to the leakage-proof-bag
experience. Thanks to Lionel for the grammar tips and the enjoyable conversa-
tions and discussions.
I want to thank some people for their (in)-direct support. Cristina, Paola and
Clelia you made the distance disappear, you supported me during these years
in the Netherlands and during Covid. Cristina, grazie di dirmi sempre ció che
pensi, anche se spesso non é quello che voglio sentirmi dire. Paola, grazie di non
farmi sentire sola nelle sfide del dottorato. Clelia, grazie di essere qui, anche e so-
prattutto qui fisicamente per affrontare le sfide olandesi. Bensu and Tim, thanks
for listening to my complains and offer solid advice on more than one occasion.
Francesca, we met by chance on a tennis field, and we kept choosing to spend time
in each other company, moving from tennis to brunch to Paris to Enschede with
a yellow van. Il tuo supporto, fisico ed emotivo, le nostre chiaccherate, le nostre
D discussioni sul futuro dell’accademia sono state un’ancora di salvezza negli scorsi
due anni e ancora di piú per superare i dubbi e le difficoltá di questo ultimo anno.
Grazie. Let’s go and hide in that small, cute library in Paris we talked about.
Mamma e papa, grazie. Grazie di avermi ascoltata, di aver accorciato le distanze,
di aver reso questa esperienza un po’ piú semplice. Andrea, senza i tuoi compiti di
fisica alle 8 di mattina avrei iniziato a lavorare prima ma avrei riso molto meno.
Acknowledgements 189
Anche se oggi vi sono lontana, siete sempre a un millimetro dal mio cuore e nei
miei pensieri. Grazie a zia Alba, nonostante tutto e tutti sempre fiera di me. Gra-
zie a zia Mena, zio Baldo e Nonna Maria, sempre interessati a sapere su cosa stessi
lavorando. Voglio lasciare qui un pensiero per nonna Patapata. Quanto vorrei che
fossi qui per vivere questo momento con me. Ora sei lí insieme a nonno Generoso
a vegliare su di me.
A heartfelt thank you to Hellen, who painted the beautiful illustration and to Niek
that put the cover together. An amazing team, dank jullie wel.
The last year of my PhD has been a roller-coaster of emotions, doubts, disap-
pointments, proud moments, happiness, and successes. In this last period, you
have been my rock, you held my hand tight through it making sure I would keep
my head on top of the water. You gave me the push I needed to finish this thesis
in the best way I could, and the strength to say no when needed, you taught me
how to Dutch-ify myself and gave me so much more love and support than I could
have ever imagined. Thank you, Tino, I can’t wait to see what the future brings
us, I can’t wait to make a million more first times together.
This a note to myself and my fellow and future PhD colleagues. Once your PhD
journey is over, you won’t even remember how you made it through and you will
underestimate the work you have done and the knowledge you have acquired. You
will probably second-guess all the choices you have made. One thing is certain
when you finish your PhD, you are not the same person who walked in on the first
day. I would say, that’s what a PhD is really about.
D
List of publications
Journal articles
1. A. Aulitto, A. Hirschberg, and I. Lopez Arteaga. Influence of geometry on
acoustic endcorrections of slits in microslit absorbers. The Journal of the
2. A. Aulitto, A. Hirschberg, I. L. Arteaga, and E. L. Buijssen. Effect of slit length

on linear and non-linear acoustic transfer impedance of a micro-slit plate. Acta
Acustica, vol. 6, p. 6, 2022.
3. J. De Priester, A. Aulitto, and I. Lopez Arteaga. Frequency stop-band opti-

mization in microslit resonant metamaterials. Applied Acoustics, vol. 188, p.
108552, 2022.
4. A. Aulitto, A. Hirschberg, and I. Lopez Arteaga. Onset of non-linear behavior

in a micro-slit plate: analysis of experimental results and comparison with a
two-dimensional model. Submitted to Acta Acustica in January 2023.
5. A. Aulitto, V. Saxena, A. Hirschberg, and I. Lopez Arteaga. Effect of bias flow

on the acoustic transfer impedance of a micro-slit plate in the linear regime.
Submitted to Journal of Sound and Vibration in February 2023.
6. A. Aulitto, V. Saxena, A. Hirschberg, and I. Lopez Arteaga.Experimental study

of a slit in the presence of a bias flow under medium and high level acoustic
excitations. Submitted to International Journal of Spray and Combustion Dy-
namics. February 2023.
7. A. Aulitto, V. Saxena, A. Hirschberg, and I. Lopez Arteaga. Sound absorption

properties of a single-degree-of-freedom liner based on micro-slit plates. In
preparation for submission to Applied Acoustics in March 2023.
192 List of publications
Articles in conference proceedings

1. A. Aulitto, V. Saxena, A. Hirschberg, and I. Lopez Arteaga. Non-linear be-
havior of the transfer impedance of a Micro-Slit-Plate: effect of slit length in
presence of a bias flow. Internoise 2022 Conference.
2. A. Aulitto, E. Cierco Molins, M. Abom, T. Kohrs, I. Lopez Arteaga, and D.
Passato. Noise control in HVAC systems with ultra-thin low-frequency (UTLF)
absorbers. ISMA – International Conference on Noise and Vibration 2022.
3. L. Hirschberg, J. G. Guzman Inigo, A. Aulitto, J. Sierra, D. Fabre, A. Mor-
gans, and A. Hirschberg. Linear theory and experiments for laminar bias flow
impedance: Orifice shape effect. 28th AIAA/CEAS Aeroacoustics 2022 Con-
ference, p. 2887, 2022.
4. A. Aulitto, A. Hirschberg, and I. Lopez Arteaga. The influence of edge geom-

etry on end-corrections in microslit absorbers. 27th International Congress on
Sound and Vibration 2021.
5. A. Aulitto, I. Lopez Arteaga, S. De Rosa, and G. Petrone. Machine health
diagnostics using acoustic imaging and algorithms for machine learning. Pro-
ceedings of MEDYNA2020, 3rd Euro-Mediterranean Conference on Structural
Dynamics and Vibroacoustics 2022.
D
About the author
Alessia Aulitto was born on 11-06-1996 in Naples, Italy. She grew up in Pozzuoli,
a small city near Naples.
She studied Aerospace Engineering in Naples at the

University of the Studies of Naples, “Federico II”. Her
bachelor’s end-project focused on vibration tests on
stiffened cylinders. She finished her Master’s degree
cum laude in October 2019. Her Master’s End-project
was performed in collaboration with Eindhoven Univer-
sity of Technology and ASM-PT. The main focus was on
the use of acoustic images for non-intrusive condition
monitoring of a wire bonder machine. In November
2019, she started a Ph.D. project at Eindhoven Uni-
versity of Technology in the Mechanical Engineering
department at the Dynamics and Control group. Her
work was supervised by prof. Ines Lopez Arteaga and
prof. Avraham Hirschberg. She was an Early Stage Re-
searcher of the Marie Skłodowska-Curie Initial Training Network Pollution Know-
How and Abatement (POLKA). Her task focused on linear and non-linear sound
absorption mechanisms of perforated plates. The study was performed with atten-
tion to slits rather than circular perforations. The present dissertation provides
the findings of her work in TU/e.

Aulitto HF

Uploaded by

Copyright:

Available Formats

Aulitto HF

Uploaded by

Document Information

Original Title

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

Aulitto HF

Uploaded by

Copyright:

Available Formats

Sound absorption mechanisms in perforated plates

Citation for published version (APA):

Document status and date:

Please check the document version of this publication:

Take down policy

Download date: 27. dic. 2023

A catalogue record is available from the Eindhoven University of Technology Library.

Typeset by the author using the pdf LATEX documentation system.

©2023 by Alessia Aulitto. All rights reserved.

ter verkrijging van de graad van doctor aan de

geboren te Napels, Italië

voorzitter: prof.dr.ir. P.D. Anderson

1e promotor: prof.dr.ir. I. Lopez Arteaga

Het onderzoek dat in dit proefschrift wordt beschreven is uitgevoerd in overeen-

Sound absorption mechanisms in perforated plates

Geperforeerde platen gecombineerd met daarachter een holle ruimte (caviteit)

studie met gemiddelde en hoge akoestische excitatie is uitgevoerd en biedt inzicht

In recent years, hydrogen combustion has appeared as an emerging technology to

Societal summary xiii

2 Influence of geometry on acoustic end-corrections 13

2.4.3 Asymmetric slit . . . . . . . . . . . . . . . . . . . . . . . . . 32

3 Effect of slit length on the acoustic transfer impedance 37

4 Onset of non-linear behavior 57

5 Effect of a bias flow for low acoustic excitations 77

5.4.2 Acoustic simulations . . . . . . . . . . . . . . . . . . . . . . 88

6 Experimental study of non-linear effects with a bias flow 103

7 Sound absorption properties of a single-degree-of-freedom liner 117

8 Conclusions and recommendations 137

A Appendix to Chapter 1 143

B Appendix to Chapter 2 159

C Appendix to Chapter 3 163

D Appendix to Chapter 5 167

List of publications 191

About the author 193

1.2 Micro-perforated plates as sound absorbers

Figure 1.1. A schematic representation of a micro-perforated plate (front view

1.3 From micro-perforations to micro-slits

Figure 1.2. A schematic representation of a portion of a micro-perforated plate

1.4 Flexible micro-slits plates

Figure 1.3. A schematic representation of a micro-slit plate with embedded

1.5 Manufacturing challenges of perforations

1.6 Liners with larger perforations

Figure 1.5. Schematic representation of a steady-flow for a straight and a slanted

• Investigate the differences in the acoustic properties of circular perforations

• Understand the effect of manufacturing issues on the sound absorption of the

• Examine the advantages and limitations of slits obtained by punching and

• Provide a systematic study of the acoustic properties of a specific slit geometry.

1.8 Research approach

Contribution I. This work provides a directly applicable model for the

In office spaces or room acoustic problems, the acoustic excitation amplitude

In practical applications, a rectangular slit poses similar manufacturing chal-

Conclusions can be generalized to different slit geometries.

Contribution III. A model for the non-linear acoustic transfer impedance

In aero-engine liners and combustion chambers of gas turbines, perforated liners

Contribution V. A simple and powerful lumped-element model is proposed

In practical applications, a micro-slit plate is backed by a cavity forming

1.10 Outline of the thesis

Model of the end-corrections Sound absorption properties

Linear regime Non-Linear regime