Acoustic characterization of absorbing materials using dynamic mode decomposition techniques
Title
Acoustic characterization of absorbing materials using dynamic mode decomposition techniquesAuthor(s)
Directors
Prieto Aneiros, AndrésNayak, Ashwin
Date
2020-07-21Center/Dept./Entity
Universidade da Coruña. Facultade de InformáticaDescription
Traballo fin de mestrado (UDC.INF). Matemática industrial. Curso 2019/2020Abstract
[Abstract] In general, the simulation of physical phenomena through numerical methods tends to be a computationally
intensive task, but this is particularly true in the eld of acoustics. Due to the fast changing
derivatives and the innately second order formulation, a ne mesh needs to be used, and in order for
the time discretization to be well behaved, a small time step needs to be chosen as well. In addition to
this, the testing of acoustic propagation in a single domain is rarely of interest, since most applications
involve the design of acoustic barriers or transmitters, which means that most problems solved in the
eld involve couplings. One of the coupled mediums is usually a
uid, and it is common for the other to
be a porous material since they are the most e ective sound absorbers.
It is in fact because of this absorption that the porous models can get very complex. Time convolutions
are usually needed for the modeling of high frequency noise, which makes the simulation process very
costly. It becomes apparent that a reduced order method that is able to cut the computation time down
is a worthwhile tool to have. Among the reduced order methods (ROMs) the chosen one is a method
that is able to make predictions into the future from a reduced amount of snapshots.
Dynamic Mode Decomposition (DMD) is a technique developed in 2010 by Peter Schmid [47]. It is
based on a Singular Value Decomposition (SVD) into which dynamics are added, making it able to
not only reconstruct available data using a reduced order representation, but also able to expand the
dimensionality in the time dimension in order to make predictions about the future. This means that
a simulation spanning a shorter time can be run and the remaining sector of the time domain can be
predicted by DMD, which adds up to a signi cantly faster process.
DMD is a data-driven method, which means that no information about the dynamic model is needed, only
a series of snapshots are used. It has been used in the
uid dynamics community, where it originated, and
a number of elds including video processing [21, 14], epidemiology [41] and neuroscience [5]. A number
of acoustic models are developed in this work, and then, they are used to test the capabilities of DMD
in acoustic problems and to nd its limitations.
The motivation of this problem arises from a collaboration between the Technological Institute for Industrial
Mathematics (ITMATI) and Micro
own Technologies through the ROMSOC project. Together
they started a project that became Ashwin Nayak's PhD thesis, in which the objective is to design a
multilayer windshield for an acoustic probe by modeling the acoustic eld both inside and outside of the
windshield considering acoustic and other physical phenomena such as
uid, thermal and poro-elastic
e ects in an unbounded domain. The present master thesis project reduces Nayak's problem to a 1D
simpli ed problem and attempts to develop a method that could be, in the future, generalized to his
problem and reduce the computation time needed.
Keywords
Acoustic
Acoustic propagation
Numerical methods
Acoustic propagation
Numerical methods
Rights
Os titulares dos dereitos de propiedade intelectual autorizan a visualización do contido deste traballo a través de Internet, así como a súa reproducción, gravación en soporte informático ou impresión para o seu uso privado e/ou con fins de estudo e de investigación. En nengún caso se permite o uso lucrativo deste documento. Estos dereitos afectan tanto ó resumo do traballo como o seu contido Los titulares de los derechos de propiedad intelectual autorizan la visualización del contenido de este trabajo a través de Internet, así como su repoducción, grabación en soporte informático o impresión para su uso privado o con fines de investigación. En ningún caso se permite el uso lucrativo de este documento. Estos derechos afectan tanto al resumen del trabajo como a su contenido