Examensarbeten för masterexamen

Browse

Senast publicerade

Visar 1 - 5 av 3405
  • Post
    Noise reduction of mine and tunnel ventilation; investigations on scale models and methods to determine noise reduction
    (2023) Arand, Sally; Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Kropp, Wolfgang; Lundberg, Oskar; Rantakokko, Renny
    Ventilation systems used in mines and tunnels contain big axial flow fans, which dominate the sound environment together with the noise generated from heavy machinery. To fulfil the demands regarding work-placement and environmental regulations and with increasing awareness of noise induced health issues, the reduction of the emitted sound is investigated in a thesis project with practical measurements and simulations. To do fast investigations a scale model of the ventilation system was designed. The fan scaling laws, which are used to scale fan properties between different fan sizes, showed, that by reducing the fan size and the rotational speed the fan performance will decrease highly. The resulting effect on the flow properties is rather small. The blade passage frequency will change, which leads to a change in the acoustic similarity. CFD simulations were performed in Comsol Multiphysics 6.1 to investigate the pressure and velocity field closely over a perforated plate and around a centrumbaffle, which are both part of the silencer of the ventilation system. The results showed a laminar-sublayer close to the wall and an increase in the fluctuations of the velocity and the pressure in the flow duct when adding the centrum-baffle. In the practical part of the thesis the silencer of the ventilation system was scaled and its acoustic properties were determined with measurements. The transmission loss measurement methods were compared by using the two room method with microphones and a sound intensity probe. Both methods showed similar results. Insertion loss and level difference were measured in another setup. Overall there was no big difference between the silencer characteristics measured from the different methods. It was found that the most effective reduction is brought from the absorption material. The perforated plate does not add to the noise reduction. At last an acoustic simulation for the silencer was set up in Comsol to investigate the transmission loss of the silencer further. Due to limitations in computational resources, the model is restricted to plane wave excitation, which limits the accuracy of the model and the possibility to directly compare the measurements with the simulation results. But both show the most effect due to the absorber and no added reduction by the perforated plate.
  • Post
    Speech enhancement for non-stationary noise around a machine cabin
    Zhou, Yiliang; Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Ahrens, Jens; Frenne, Nicklas
    This thesis is mainly concerned with the solution for speech enhancement in the presence of non-stationary noise around the machine cabin. This allows outside speech to enter the cabin and reduces unwanted noise. The application scenario of this work is a signal processing system using a microphone array outside the cabin to capture signals and using different algorithms to enhance the speech signals. With the help of this system, the machine operator is able to get speech information in a noisy environment. The noise sources in this situation are more complex and non-stationary. Examples of noise sources include engine noise, traffic noise, or other construction activities. Previous work done by Tomoya chose the microphone array configuration and developed the beamforming method. From the results of the valuable work, it is found that beamforming is able to increase the signal-to-noiseratio (SNR) in the current situation, but the sound quality and SNR are still limited due to the low input SNR and non-stationary noise environment. Therefore, modified beamforming and new methods are implemented in this work. noise cancellation(NC) predicts the transfer path for the noise signal and removes it by controlling the minimum error of the output. Noise suppression(NS) uses the scheme of spectral subtraction to subtract the noise spectrum from noisy speech spectrum. A combination of beamforming and noise cancellation and a combination of beamforming and noise suppression method are developed and evaluated. The result shows a better performance for this low input SNR and non-stationary noise case.
  • Post
    Enlargement of perceived size of sound source using diffusive geometries
    (2024) Xu, Zhiwen; Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Ahrens, Jens
    Interaural cross correlation, which quantifies the similarity of binaural signals, is essential for immersive and rendered auditory environment. Previous research has proven that reducing interaural correlation leads to expansion of perceived size of a source. The Schroeder diffuser is a prevalently used acoustic design to create pleasant acoustic environment since it was firstly developed. It helps to eliminate many acoustic problems in many scenarios through scattering waves in multiply directions. In this work, the Schroeder diffuser is employed to manufacture natural sound field with reduced interaural correlation so that models of source enlarger can be studied further. To explore the properties of the sound field of the Schroeder diffuser, the produced sound field is simulated and a method of representing the sound field with circular harmonic expansion is applied. The characteristics of the sound field of the Schroeder diffuser are investigated to discover their connectivity with low interaural correlation in order that any other arbitrary sound field can be manipulated to achieve it. In light of such discovery, two models of simple sound fields which are impulse field and plane wave field are utilized to produce the sound field with low interaural correlation with proper modification engaged. These simple models are capable of saving computational cost as simulating the Schroeder diffuser is pretty resource and time consuming.
  • Post
    Unified Auralization of Room Simulations; Investigation of sampling grids and other parameters
    (2024) Xuanye, Liu; Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Ahrens, Jens
    With the development of computer science and the increase in computing power, the finite-difference time-domain method has gradually become a viable option for solving wideband, transient problems in acoustics simulation. In this method, the binaural response can be extracted by local sampling of the finite-difference grid using an array of receivers. This thesis focuses on the effects of the grid size, the spacing of grid points, and the distribution of grid points in the case of a volumetric sampling. This thesis also investigates the effect of the change in impulse response length and others. These effects include variations in numerical robustness and the frequency properties of the generated audio signal. In order to achieve this study, this thesis simulates the sound field formed by a single plane wave propagation. By sampling this sound field with a volumetric grid, the plane wave density function is calculated using the plane wave decomposition method and integrated with free-field head-related transfer functions in the spherical harmonics domain to get the binaural impulse response. The results of the valuable work show that the numerical stability of a volumetric grid with uniformly distributed sampling points is greatly affected at high frequencies due to spatial aliasing. In addition, the results also give some valuable conclusions for the selection of the impulse response length and the selection of the number of sampling points.
  • Post
    Modelling the effects of railway implemented low-height noise screens; an investigation of train track ballast impedance
    (2024) Hildén, Marius; Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Chalmers tekniska högskola / Institutionen för arkitektur och samhällsbyggnadsteknik (ACE); Forssén, Jens; Källman, Magnus
    Conventional tall noise barriers (measuring approximately 3 meters and above) are commonly employed to mitigate noise from railways in urban environments. They are effective and their noise reduction outcomes can be accurately estimated using existing, low-order geometrical ray-acoustic models, e.g. Pierce’s thin hard diffracting screen solution. However, tall noise barriers arguably have an adverse effect on surrounding landscape, as well as obscuring the sightlines of both train operators and passengers. In cases where noise levels can be adequately attenuated using a low-height noise screen (LHNS) it can be a preferable implementation in regard to aesthetic, cost, and maintenance aspects. The current problem with implementing LHNS is that their noise reduction outcomes are, due to fundamental design, difficult to accurately estimate. This is a problem in large-scale urban development projects where the margin of error is small, often leading to LHNS being disregarded in favor of conventional noise screens. To improve the accuracy of insertion loss (IL) estimations from LHNS, a previously imple mented 2.5D boundary element method (BEM) model used for calculating railway LHNS IL is revised. The main focus of the revision regards the surface impedance of the BEM-modeled train track. Measurements have been performed on ballasted train tracks to serve as validation data for an impedance parameter study of ballasted train track surfaces. The resulting set of impedance parameters have been used in 2.5D BEM-models simulating the sound pressure field of different train shapes with and without LHNS, in other words estimating the IL of LHNS for different railway applications. The IL results are compared with existing LHNS IL measurements from other projects. The simulated results demonstrate a generally accurate alignment with existing measurement data for IL in third-octave frequency bands for passenger trains, however results differ between different measurement comparisons. In the case of industrial trains, results are less promising. This is hypothesized to be a result of the source model used in the simulations being inaccurate for industrial trains. Further investigation/development of source models used for different train types is a recom mended starting point for improving the reliability of the 2.5D BEM simulations. Access to more LHNS IL measurement validation data is also considered necessary. Nonetheless, the yielded results indicate that the revised impedance parameters have been an effective step in improving LHNS IL estimation when compared with previous BEM-model results.