Numerical models will be developed for the prediction of silencer transmission loss under the operating conditions present in gas turbine exhausts. In these systems the large diameter ducts and high operating temperatures produce a challenging acoustic environment due to the unverified behaviour of fibrous materials at high temperatures and the existence of complex sound fields. To understand the behaviour of fibrous materials at high temperatures their bulk acoustic properties are measured using a modified impedance tube which can heat material samples up to a temperature of 500 C. It will be demonstrated that the high temperature material properties can be extrapolated from room temperature measurements given knowledge of the temperature dependant flow resistivity. Finite element numerical models using point collocation and mode matching techniques to predict the transmission loss of silencers are developed and successfully validated. Dissipative silencer designs with various cross-sectional designs are explored numerically and experimentally according to common industry standards. It is demonstrated that transmission loss may be optimised by the arrangement of the fibrous material across the cross-section. The accurate numerical models allow for effe cient silencers to be designed reducing silencer size and cost. A new hybrid silencer is presented combining dissipative and reactive elements with the aim of increasing the low frequency attenuation of large silencers while maintaining an effective broadband spectrum. Measurements and predictions show this innovative design to be successfull. Application of the hybrid silencer allows for more flexible noise control solutions when design is limited by low frequency noise.

Large silencers used in the power generation industry usually have a very large cross section at the inlet and outlet. Higher-order modes will populate the inlet and outlet even at very low frequencies. Although the silencer itself is often modeled by a three-dimensional analysis tool such as the boundary element method (BEM) or finite element method (FEM), a direct computation of the transmission loss (TL) from the BEM or FEM model can be challenging without incorporating certain forms of modal expansion. A so-called “impedance-to-scattering matrix method” is proposed to extract the modes at the inlet and outlet from the BEM impedance matrix based on the point collocation method. The BEM impedance matrix relates the sound pressures at the inlet and outlet to the corresponding particle velocities, while the scattering matrix relates the modes at the inlet and outlet. Normally there are more boundary elements than the total number of modes at the inlet and outlet, and a least-squares procedure is used to condense the element-based impedance matrix to the mode-based scattering matrix. The TL computation will follow if a certain form of the incident wave is assumed and the outlet is non-reflective. Several commonly used inlet/outlet configurations are considered in this dissertation, which include axisymmetric, non-axisymmetric circular, and rectangular inlet/outlet shapes. In addition to the single inlet and outlet silencers, large multi-inlet and multi-outlet silencers are also investigated. Besides the collocation-based impedance-to-scattering matrix method, an integral-based impedance-to-scattering matrix method based on the reciprocal identity is also proposed for large silencer analysis. Although it may be more time-consuming to perform the additional numerical integration, an integral-based method is free of any uncertainties associated with collocation points. The computational efficiency, accuracy and stability are compared between two proposed methods. One bonus effect of producing the scattering matrix is that it can also be used to combine subsystems in series connection. The Redheffer’s star product is introduced to combine scattering matrices of subsystems. In the design stage, rapid assessment of the silencer performance is always preferred. However, the existing analytical approaches are only suitable for simple dissipative silencers such as straight lined ducts. A two-dimensional first-mode semi-analytical solution is developed to quickly evaluate the performance of tuned dissipative silencers below the cut-off frequency. The semi-analytical solution can also serve as a validation tool for the BEM.

It is difficult to predict the sound radiation from large factory roofs. The existing infinite panel theories of sound insulation are not sufficient when the sound radiates at grazing angles. It has been shown that the reason for the collapse of the theory is the well known result for the radiation efficiency. This research will present a simple analytic strip theory, which agrees reasonably well with numerical calculations for a rectangular panel. Simple analytic strip theory has lead to the conclusion that it is mainly the length of the panel in the direction of radiation, rather than its width that is important in determining its radiation efficiency. The findings of the current research also indicated that apart from the effect due to coincidence, a panel was non-directional compared to an opening.

Noise generated by fluid power applications can be treated using bladder-style suppressors, and an optimal operating condition for these devices is sought in this thesis. Bladder-style suppressors employ a compliant nitrogen-charged bladder to create an impedance change within the system, reflecting the noise back to the source and preventing it from propagating downstream. The noise in a hydraulic system is created by a pump, the flow source in a hydraulic system, and can be separated into three categories: fluid-borne noise, structure-borne noise and airborne noise. Fluid-borne noise places addition stress on sealing surfaces, potentially causing leaks. Airborne noise can be uncomfortable, even hazardous depending on the level. Bladder-style suppressors primarily treat fluid-borne noise; however, it is seen in the literature that fluid-borne noise is the cause of structure-borne and airborne noise. This thesis presents an optimization method for finding the optimal charge pressure for implementation with a given system operating over a broad range of system pressures. The optimization weights suppressor performance by the spectral content of the fluid-borne noise as well as the duty cycle of the system. A single charge pressure works well over a small range of system pressures, though many fluid power applications operate over a larger range of system pressure than the usable range of a suppressor. For systems operating over an extremely broad pressure range, two suppressors charged to different pressures are used to treat the noise in the entire system pressure range. To determine suppressor performance experimental measurements were performed, and models developed, of the transmission loss of this type of device. A multi-microphone method using transfer function relationships between six sensors determines the transmission loss of the suppressor under test. An equivalent fluid model modeling the wave behavior both upstream and downstream, as well as within the suppressor, was created to predict suppressor transmission loss. Optimal configurations are found for a set of system pressures, charge pressures and duty cycles. Analysis of the results shows the time weighting has a more significant impact on the optimum charge pressure than the frequency weighting, as shown by duty cycles considered in this thesis. In addition, all charge pressures selected as optimal for either single suppressor optimizations or double suppressor optimizations, exhibit the highest transmission loss for a single system pressure in the pressure duty cycle for a simulated machine.

This study concerns with acoustical analysis of exhaust mufflers for earth-moving machinery. The study arises from the fact that there is a need for further noise reduction emitted by earth-moving machinery produced by Hidromek Inc. in order to be on the safe side of the limits stated in European Noise Directive 2000/14/EC. The acoustical performance of the muffler is investigated both experimental and numerical means. A three-dimensional finite element method is performed to calculate the transmission loss. An experimental setup is also developed to measure the transmission loss. Chung-Blaser, two-source and two-load methods are applied to measure the transmission loss of single expansion chamber with extended inlet/outlet. The experimental setup is verified by comparing the test results obtained by two-load method with one-dimensional analytical solution obtained by transfer matrix method. Transmission loss of the muffler of interest, calculated by finite element method is compared with the experimental results. Sound power level of earth-moving machinery and insertion loss of the muffler is measured to investigate the acoustical performance of the muffler. These results can then be served as guidance for the acoustical design of a muffler.

Compact fully anechoic chambers may gradually replace open area test sites as the preferred type of testing facility for the measurement of radiated emissions. This dissertation theoretically and experimentally investigates the suitability of a compact fully anechoic chamber with ferrite tiles being partially lined on the floor ground as a free-space environment at frequency range below 300MHz. The results show that, for vertical polarization, the practical chamber will behave like a free-space test site. However, for the horizontal polarization, it seems have relatively significant effect on the NSTL performance and will not behave well like a free-space test site. Based on FDTD modeling approach, a numerical tool with the measuring tunable dipole antennas being considered is established to simulate the NSTL performance of the compact EMC chamber with different layout design of ferrite tiles. Furthermore, several designs of reduced coverage of the ferrite tiles in the compact chamber are proposed and discussed. In addition, this dissertation also investigates the correlation test of a 3m fully anechoic chamber ¡]FAC¡^ and a 10m semi-anechoic chamber ¡]SAC¡^ at frequency range from 80MHz to 1000MHz. Based on correction factor between two test sites, the EMI behaviors of the EUT in a 10m SAC could be predicted rapidly at frequency from 200MHz to 900MHz. Thus, a 3m FAC has been expected to provide a low cost as a pre-compliance test site, if the predicted EMI behaviors of the EUT are close to the practical measured results in a 10m SAC.

Ce travail doctoral porte sur la synthèse expérimentale de champs de pression acoustique aléatoires sur des surfaces planes. L'objectif principal de cette reproduction est de pouvoir tester le comportement vibroacoustique de panneaux plans sous l'action de deux champs de pression aléatoires : le champ acoustique diffus et la couche limite turbulente. Une étude théorique et numérique est tout d'abord présentée, qui aborde la formulation de ce problème de reproduction sur la base de la théorie de l'holographie acoustique plane de champ proche, et qui permet de vérifier que celle-ci permet une résolution de ce problème. L'influence de nombreux paramètres est également étudiée et des points critiques dans la mise en oeuvre expérimentale de l'approche sont identifiés, parmi eux la densité élevée de sources nécessaire pour la reproduction d'une couche limite turbulente. Une première mise en oeuvre expérimentale est détaillée. Trois approches, celle basée sur l'holographie, une seconde sur la synthèse de champs sonores (ou Wave Field Synthesis ) ainsi qu'une troisième basée sur les moindres carrés sont utilisées pour le calcul des amplitudes complexes des sources de reproduction. Le concept d'antenne synthétique, pour lequel un petit élément d'un réseau est déplacé pour créer un réseau de grandes dimensions en post-traitement, sera utilisé pour résoudre une majorité des points critiques dans la mise en oeuvre. Des indicateurs vibroacoustiques fréquentiels tels que le Transmission Loss (TL) sont obtenus pour un panneau d'aluminium simplement supporté, et la comparaison avec des résultats de calculs numériques montrent que ces approches permettent l'estimation de ce TL sous les deux champs aléatoires d'intérêt. La possibilité d'estimer le TL dans le domaine des nombres d'onde est également illustrée. Une seconde mise en oeuvre expérimentale est réalisée sous la forme d'une comparaison entre la méthode décrite ci-dessus et la méthode des chambres couplées, couramment utilisée pour l'estimation du TL. Pour un panneau composite aéronautique représentatif et pour un champ acoustique diffus, les écarts entre les résultats obtenus par les deux méthodes sont jugés acceptables, et permettent une seconde validation. Parallèlement, une autre application a concerné la reproduction d'un champ acoustique diffus sur la surface d'un matériau poreux, afin d'en estimer le coefficient d'absorption sous ce champ excitateur. L'approche théorique de cette estimation est décrite et validée expérimentalement sur une mousse de mélamine, prouvant qu'elle est réalisable sous des champs acoustiques synthétisés et montrant un potentiel pour des mesures in situ. Suite à la conclusion de ce mémoire qui rappelle les résultats importants obtenus dans le cadre de ce travail doctoral, une possibilité de mesure des fluctuations de pression pariétale liées à la couche limite turbulente est finalement présentée en annexe.

With increased exposure to transportation noise, there have been continued efforts to help insulate homes from aircraft noise. Current aircraft noise guidelines are based primarily on outdoor sound levels. As people spend the majority of their time indoors, however, human perception is evidently more related to indoor sound levels. Investigations are being made to provide further insight into how typical residential constructions affect indoor response. A pilot study has built a single-room "test house", according to typical construction for mixed-humid climate regions, and has directly measured outdoor-to-indoor transmission of sound - with specific focus on continuous commercial aircraft signatures. The results of this study are being used to validate and improve modelling software that simulates a wide range of construction types and configurations for other US climate regions. The improved models will allow for increased flexibility in simulating the impacts of acoustic and energy retrofits. Overall, the project intends to improve the ability to predict acoustic performance for typical US construction types as well as for any possible design alterations for sound insulation.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The research presents the sound insulation in masonry walls coated with plaster. Three walls were built in laboratory, studied in specific tests. Measurements were carried out varying some components of the wall, such as the type of the block and the combinations of coarse mortar, plaster and stucco on the faces of the wall. The mortars used in the tests were industrialized and had an additive in their conglomerate that aimed to its structural and commercial improvement. Also, paints and textures were used alternately in the tests to verify the importance of this kind of coating on the sound transmission loss of the wall. The laboratory findings showed up the location of the coincidence frequency and the acoustic accomplishment of the coated masonry walls. Variation of these factors had a direct influence at the system hardness. The increase of the variation in the set caused by the addition of the mortar layers does vary, further of frequency of coincidence, the frequency of resonance wall thickness. It was still checked that three days of drying of the mortar was sufficient to have satisfactory results on the essays. The laboratory tests brought more agility with the evidence and the validation that the cure of the aggregate in the wall needs three days. Another important verification was the increase of 5 dB on the Weighted Noise Reduction Index (Rw) when the wall is coated in both faces. The using of paint and texture didn t contribute to the acoustic performance of the wall. Just, the utilization of texture shown a raise of acoustic performance of 1 dB on the Weighted Noise Reduction Index (Rw). That didn t happend with the duplication of the layer of plaster, that obtained acoustic insulation indicators lower than 1dB in the set. / A pesquisa investigou a maneira como acontece a isolação sonora em paredes de alvenaria estrutural com revestimento de argamassa. Foram construídas em laboratório três paredes, estudadas em ensaios específicos. O tipo do bloco e as combinações de chapisco, emboço e reboco foram avaliados alternadamente em cada bateria de ensaios. As argamassas utilizadas nos ensaios eram industrializadas e possuíam aditivos em seu conglomerado que visavam a sua melhora estrutural e comercial. Também, foram utilizadas tintas e texturas alternadamente nos ensaios para verificar a influência deste tipo de acabamento na perda de transmissão sonora da parede. As constatações laboratoriais mostraram a localização da frequência de coincidência e o desempenho acústico das paredes de alvenaria estrutural revestidas. A variação desses fatores possui influência direta conforme a rigidez do sistema. O seu aumento no conjunto causado pelo acréscimo de camadas de argamassa faz variar, além da frequência de coincidência, também a frequência de ressonância de espessura da parede. Aferiu-se, ainda, a possibilidade de diminuição dos dias de secagem da argamassa entre os ensaios realizados. Os testes em laboratório ganharam em agilidade, com a comprovação e validação do tempo de 3 dias para a cura da massa usada na parede. Outra importante verificação foi o incremento de 5 dB no Índice de Redução Sonora Ponderado (Rw) quando há o revestimento da parede em ambas as faces. A utilização de tinta e de textura não acrescentou ganhos acústicos substanciais para o conjunto da parede. Apenas, a utilização da textura apresentou um aumento de performance na ordem de 1 dB no índice de redução sonora ponderado (Rw). O que não aconteceu com a duplicação da camada de reboco, que obteve indicadores de isolação sonora inferiores a 1 dB no conjunto.

Orientador: Luiz Carlos Pereira da Silva / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-05T12:14:14Z (GMT). No. of bitstreams: 1 Menezes_TacianadeVasconcelos_D.pdf: 6420793 bytes, checksum: f382ddde0cefab51526d7f41efc2ff66 (MD5) Previous issue date: 2005 / Resumo: Este trabalho apresenta um método para a alocação das perdas ativas de transmissão entre os agentes de um mercado de energia elétrica. No Brasil, a metodologia a ser implementada para a alocação das perdas baseia-se em métodos incrementais e na definição de fatores de perda tanto para a carga quanto para a geração. Esses fatores são calculados por meio de fluxo de carga CC, que considera algumas aproximações nas equações do fluxo de potência convencional. Devido a isto, acredita-se que este método possa fornecer uma alocação das perdas injusta aos participantes do mercado. Esta tese propõe uma metodologia para o cálculo dos fatores de perda, baseada em métodos incrementais, mas considerando a não-linearidade das equações do fluxo de carga, utilizando para isso um modelo de fluxo de carga CA. Os métodos CC e CA são aplicados a um sistema composto de 5 barras e a uma versão do sistema norte-nordeste brasileiro. Os resultados obtidos comprovam as limitações do método incremental CC, justificando a utilização de uma metodologia CA para a alocação das perdas. O método CA proposto mostrou-se eficiente e preciso no cálculo dos fatores de perda, além de permitir, diferentemente da metodologia CC, a consideração da sensibilidade da perda ativa total do sistema com a variação da potência reativa, o que a maioria dos métodos negligencia / Abstract: This work presents a method for active power transmission loss allocation among electrical market participants. In Brazil, the methodology to be implemented for loss allocation is based on incremental methods and defines loss factors for generators and loads. This technique uses a DC load flow to calculate the loss factors, and so it does not consider the complete power flow equations. Consequently, this method may provide unfair loss allocation depending on the system operation condition. This thesis proposes a methodology to calculate the loss factors, based on incremental methods but using an AC load flow taking into account system non-linearity. The DC and AC methods are applied to a small system composed by five buses and to a version of the Brazilian North-Northwest system. The results prove the limitation of DC incremental method, justifying the use of a methodology based on AC load flow equation for losses allocation. The AC method proposed shows to be efficient and accurate to calculate the loss factors as well permits to include directly, differently from the DC method, the total active power loss sensitivities with respect to reactive power changes, which most of loss allocation techniques neglect / Doutorado / Energia Eletrica / Doutor em Engenharia Elétrica