An improved test system for acoustical rating of air-movement devices was installed and evaluated at the Riverside Energy Efficiency Laboratory at Texas A&M University where measurements of sound pressure levels were carried out using an array of six microphones instead of the existing rotating boom- microphone setup. The new array setup did not generate any inherent transient noise peaks, which provided adequate signal-to-noise ratios suitable for low sone fan testing. The reverberation chamber was qualified for broad-band testing in the frequency range 50 Hz to 10 kHz. Important acoustical parameters, namely, reverberation time and natural modes of the chamber, were determined. The purpose of this study was to identify potential background noise sources by computing the coherence functions between microphones placed outside the chamber and a microphone placed within the chamber. No strong coherence was observed, thus indicating adequate sound attenuation characteristics of the chamber walls. The effect of background noise levels on the loudness rating of fans was evaluated. A low sone fan and a louder fan (loudness greater than one sone) were tested during night time when the background noise is the least and during daytime and with the air conditioners running (high background noise level). While both fan types showed no significant change in loudness when tested during daytime and during the night, accurate ratings were not obtained with the air-conditioners running due to inconsistent spectrum. Finally, it was observed that with the six decibels separation requirement between the fan and background noise spectra for a low sone fan, at very low frequencies (below 63 Hz), despite inadequate fan- background separation, the loudness rating of the fan does not change as the minimum perceived loudness at these frequencies is very high. At very high frequencies (greater than 5 kHz), the fan does not generate any noise and hence the fan and the background noise sound pressure levels are very close to each other.

This work was concerned with the aerodynamic and acoustic performance of small axial flow refrigerator fans. Relevant literature focusing on the principles of fan noise generation and fan system elements that contribute to noise generation was identified and discussed. A plenum chamber test rig was designed and constructed following ISO 10302-1. A range of 200 mm diameter pressed aluminium Air-Drive fans, commonly used in commercial refrigerator systems, were evaluated using the test rig. It was found that the performance of these fans was highly dependent upon the impedance of the system in which they are installed. Indications of rotating stall under high load were observed and shown to significantly increase the noise output of the fan without greatly affecting the air moving capability. A series of novel colour map plots are presented, which allow for a visual interpretation of fan performance over a wide range of operating conditions. A fanpack developed by Wellington Drive Technologies Ltd was evaluated. This fanpack was found to generate significant tonal noise. The design of the fanpack was evaluated and improvements are suggested. Prototype fans were developed from the fanpack based on promising concepts presented in the literature. The performance of four prototype fans was evaluated. The results were disappointing, but proof of concept was demonstrated.

An advanced model of active control of fan noise for ultra high bypass turbofan engines has been developed. This model is based on a boundary integral equation method and simulates the propagation, radiation and control of the noise generated by an engine fan surrounded by a duct of finite length and cylindrical shape, placed in a uniform flow. Control sources, modeled by point monopoles placed along the wall of the engine inlet or outlet duct, inject anti-noise into the duct to destructively interfere with the sound field generated by the fan. The duct inner wall can be lined or rigid. Unlike current methods, reflection from the duct openings is taken into account, as well as the presence of the evanescent modes. Forward, as well as backward (i.e., from the rear of the engine), external radiation is computed. The development of analytical expressions for the sound field resulting from both the fan loading noise and the control sources is presented. Two fan models are described. The first model uses spinning line sources with radially distributed strength to model the loading force that the fan blades exert on the medium. The second model uses radial arrays of spinning point dipoles to simulate the generation of fan modes of specific modal amplitudes. It is shown that these fan models can provide a reasonable approximation of actual engine fan noise in the instance when the modal amplitude of the propagating modes or the loading force distribution on the fan blades, is known. Sample cases of active noise control are performed to demonstrate the feasibility of the model. The results from these tests indicate that this model 1) is conducive to more realistic studies of active control of fan noise on ultra high bypass turbofan engines because it accounts for the presence of evanescent modes and for interference between inlet and outlet radiation, which were shown to have some impact on the performance of the active control system; 2) is very useful because it allows monitoring of any region of the acoustic field; 3) is computationally fast, and therefore suitable to conduct parametric studies. Finally, the potential that active noise control techniques have for reducing fan noise on an ultra high bypass turbofan engine is investigated. Feedforward control algorithms are simulated. Pure active control techniques, as well as hybrid (active/passive) control techniques, are studied. It is demonstrated that active noise control has the potential to reduce substantially, and over a relatively large far field sector, the fan noise radiated by an ultra high bypass turbofan engine. It is also shown that a hybrid control system can achieve significantly better levels of noise reduction than a pure passive or pure active control system, and that its optimum solution is more robust than the one achieved with a pure active control system. The model has shown to realistically predict engine acoustic behavior and is thus likely to be a very useful tool for designing active noise control systems for ultra high bypass turbofan engines. / Ph. D.

An innovative implementation of the Herschel-Quincke tubes concept for the reduction of noise from turbofan engines is proposed here. The approach consists of installing circumferential arrays of Herschel-Quincke (HQ) tubes or waveguides in the inlet of the turbofan engine. An analytical technique was developed to predict the effects of HQ tubes applied to circular inlets. The modeling technique involves modeling the tubes-inlet interfaces as finite piston sources that couple the acoustic field inside the inlet with the acoustic field within the HQ tubes. An optimization technique based on genetic algorithms was also developed to be able to design and optimize the system parameters. The accuracy of the model was validated with experimental data obtained from two types of turbofan engines. Analytical predictions are shown to correlate well with experimental data. The analytical model is then used to provide insight into the noise control mechanisms involved in the system. It is shown that the energy in an incident mode is in part reflected back to the fan and that some energy is also scattered into other higher-order modes. Thus, the suppression of a particular mode is due to the combination of the scattered contributions from the various incident modes. The effects of the system parameters were analyzed and parametric studies were conducted. Different configurations for the arrays of HQ tubes such as helical patterns or tubes at an angle with respect to the inlet axis were also investigated. The results show the great potential of the HQ tubes system to reduce noise from turbofan engines. / Master of Science

A new active-passive approach for the control of noise radiated from a small axial fan was investigated. The approach involved the installation of an axial fan into a short duct with both passive and active noise control functions. First, a systematic methodology for the analytical modeling of finite-length ducts with multiple discontinuities was formulated. The procedure involved the modeling of a duct as a collection of simple duct sections, which were interconnected at multiple junctions. Analytical studies have shown that a short lined duct provides passive noise reduction effects through the mass-loading effect of the duct air volume at low frequencies and the sound absorption by a passive liner at high frequencies. It was also shown that active control can provide further noise attenuations at low-to-mid frequencies, thereby enhancing the overall noise control performance. Two alternate designs of active-passive noise control fan duct were considered. One was a simple non- segmented duct with a 2x2 active control and the other was an internally segmented duct with an 8x8 active control. It was indicated that the latter design possesses a significantly higher global noise control potential than the former with respect to both bandwidth and attenuation level. This was attributed to the reduction of the unwanted pressure contributions from the duct cross modes through the high frequency shifting of the associated cut-on frequencies. The experimental validation of the noise control approach was also carried out. An active-passive noise control fan duct incorporating the segmented duct design with 8x8 active control was constructed in conjunction with a hybrid feedforward-feedback control system. Experimental results have shown significant reductions in the total fan noise power associated with the first four BPF tones by the feedforward control and the broadband fan noise power by the feedback control. The overall active-passive noise control characteristics were observed to be in accordance with the analytical results. / Ph. D.

In this study a centrifugal fan is studied for noise characteristics and measurements in a semi- anechoic room. A semi-anechoic room is constructed inside Fluid Mechanics Laboratory of Mechanical Engineering Department has been qualified with respect to ISO 3745 standard. The fan characteristic is obtained as proposed in AMCA standards 210-75, by simply measuring the voltage and current of the motor during operation and calculating the power consumption of the assembly. Noise measurements are taken using two microphones attached to a multi-channel data acquisition and processing system in the semi anechoic room. Several different configurations of the vacuum cleaner with some parts removed or replaced systematically are considered during the noise measurements. Some of the results showed that the damping material placed inside the motor cover is proved to be very effective in noise reduction. Two different damping materials are examined for comparative evaluation.

[Truncated abstract] The subject of fan noise generating mechanisms and its control has been studied intensively over the past few decades as a result of the ever-increasing demand for more powerful fans. A unique feature of fan noise is that it consists of high-level discrete frequency noise related to the blade passing frequency, and low-level broadband noise due mostly to turbulent airflow around the fan. Of the two types of fan noise, the discrete frequency noise is the more psychologically annoying component. Past research into fan noise has shown that the discrete frequency fan noise are dipole in nature and are caused predominantly by the fluctuating lift acting on the surfaces of the fan blades. Based on this, several theoretical models have been established to correlate these fluctuating lift forces to the far-field sound pressure. However, one general assumption in these models is that the fan blades are assumed rigid, and the consequence of such an assumption is that it is unclear if the far-field sound pressure is caused solely by the aerodynamic lift force, or whether the blade vibration also plays a substantial role in the generation of the far-field fan noise. One of the goal of this thesis was thus to experimentally quantify the contribution of blade vibration to far-field fan noise and it was found that blade vibration, whilst coherent with the far-field fan noise, did not contribute significantly. Aside of this, several experiments aimed at filling knowledge gaps in the understanding of fan noise characteristics were also be conducted, in particular, to understand the relationship between far-field sound pressure level to blade lengths as well as the number of blades on the fan. The experiments showed that for fans with many blades, the dependency of the far-field sound pressure on blade length is stronger than fans with less blades. Furthermore, dipole measurements showed that the dipole characteristics of fan noise does not occur only at the discrete frequencies, but also within a range of broadband frequencies, implying that the source for both discrete and broadband is the same. The second section of this thesis deals with the study of vortex shedding and its active control. When a circular cylinder (or any object) is placed in a flow within a specified Reynolds number range, flow separation and periodical wake motion is formed behind the cylinder, which is known as vortex shedding. It has been found in previous research that this wake motion is affected by acoustic field imposed on it via loudspeakers. This suggests that there is a strong acoustic-vortex relationship. However, little of this relationship is understood as conventional methods of studying vortex centre around the use of hot-wire anemometry, which effectively measures the velocity fluctuation in the flow. This thesis is the first in using a microphone to study the acoustic characteristic of the vortex wake, and experimental results shows that the two parallel shear layers of the wake carry the strongest pressure signals at the vortex shedding frequency, whilst the entrapped region between the layers carries the strongest pressure signals at the first harmonic.

Recent experimental achievements in active noise control (ANC) for cooling fans have used near-field error sensors whose locations are determined according to a theoretical condition of minimized sound power. A theoretical point source model, based on the condition previously stated, reveals the location of near-field pressure nulls that may be used to optimize error sensor placement. The actual locations of these near-field pressure nulls for both an axial cooling fan and a monopole loudspeaker were measured over a two-dimensional grid with a linear array of microphones. The achieved global attenuation for each case is measured over a hemisphere located in the acoustic far field of the ANC system. The experimental results are compared to the theoretical pressure null locations in order to determine the efficacy of the point source model. The results closely matched the point source model with a loudspeaker as the primary source, and the sound power reduction was greatly reduced when error sensors were placed in non-ideal locations. A weakness of the current near-field modeling process is that a point monopole source is used to characterize the acoustic noise from an axial cooling fan, which may have multipole characteristics. A more complete characterization of fan noise may be obtained using a procedure based on the work of Martin and Roure [J. Sound Vib. 201 (5), 577--593 (1997)]. Pressure values are obtained over a hemisphere in the far field of a primary source and the contributions from point source distributions up to the second order, centered at the primary source, may be calculated using a multipole expansion. The source information is then used in the aforementioned theoretical near-field calculation of pressure. The error sensors are positioned using the complete fan characterization. The global far-field attenuation for the multipole expansion model of fan noise is compared to that of previous experiments. Results show that the multipole expansion model yields a more accurate representation the near field, but is not successful in achieving greater sound power reductions in the far field.

Active noise control (ANC) has been implemented using analog filters to reduce broadband noise from a small axial cooling fan. Previous work successfully attenuated narrow-band, tonal portions of the noise using a digital controller. The practical performance limits of this system were reached and it was desirable to attenuate the noise further. Additional research, therefore, sought to attenuate broadband noise from the fan using a digital controller, but performance was limited by the group delay inherent in the digital signal processor (DSP). Current research attempts to further attenuate broadband noise and improve performance of the system by combining the tonal controller with an analog feedback controller. An analog controller is implemented in parallel with the digital controller without degrading the performance of either individual controller. Broadband noise is attenuated in a certain frequency region, but at the expense of increasing noise in adjacent frequency regions. Results show that a single-input single-output (SISO) controller is preferable to a multiple-input multiple-output (MIMO) controller for this system.

The phase and amplitude gradient estimator (PAGE) method has proven successful in improving the accuracy of measured energy quantities over the p-p method, which has traditionally been used, in several applications. One advantage of the PAGE method is the use of phase unwrapping, which allows for increased measurement bandwidth above the spatial Nyquist frequency. However, phase unwrapping works best for broadband sources in free-field environments with high coherence. Narrowband sources often do not have coherent phase information over a sufficient bandwidth for a phase unwrapping algorithm to unwrap properly. In fact, phase unwrapping processing can cause significant error when there is no coherent signal near and above the spatial Nyquist frequency. However, for signals at any frequencies up to the spatial Nyquist frequency, the PAGE method provides correct intensity measurements regardless of the bandwidth of the signal. This is an improved bandwidth over the traditional method. For narrowband sources above the spatial Nyquist frequency, additional information is necessary for the PAGE method to provide accurate acoustic intensity. With sufficient bandwidth and a coherence of at least 0.1 at the spatial Nyquist frequency, a relatively narrowband source above the spatial Nyquist frequency can be unwrapped accurately. One way of using extra information, called the extrapolated PAGE method, uses the phase of a tone below the spatial Nyquist frequency and an assumption of a propagating field, and therefore linear phase, to extrapolate the phase above the spatial Nyquist frequency. Also, within certain angular and amplitude constraints, low-level broadband noise can be added to the field near a source emitting a narrowband signal above the spatial Nyquist frequency. The low-level additive broadband noise can then provide enough phase information for the phase to be correct at the frequencies of the narrowband signal. All of these methods have been shown to work in a free-field environment.