Julien Maillard

Advanced time domain wave‐number sensing for structural acoustic systems. I. Theory and design

This paper discusses new work concerned with developing structural sensors and associated signal processing techniques that provide time domain estimates of far‐field pressure or structural wave‐number information. The sensor arrangement consists of multiple accelerometers whose outputs are passed through an array of linear filters. The impulse response of each filter is constructed from the appropriate Green’s function for the elemental source area associated with each sensor. The outputs of the filter array are then summed in order to predict far‐field pressure or wave‐number information somewhat analogous to the well‐known boundary element technique. A major significance of the approach is that it provides time domain information and can thus be efficiently applied to active structural acoustic control approaches.

Advanced time domain wave‐number sensing for structural acoustic systems. Part III. Experiments on active broadband radiation control of a simply supported plate

The present work gives further developments and experimental testing of a new time domain structural sensing technique for predicting wave‐number information and acoustic radiation from vibrating structures. Most structure‐borne active sound control approaches now tend to eliminate the use of microphones located in the far field by developing sensors directly mounted on the structure. In order to reduce the control authority and complexity required to minimize sound radiation, these sensors should be designed to provide error information that is solely related to the radiating part of the structural vibrations, e.g., the supersonic wave‐number components in the case of planar radiators. The approach discussed in this paper is based on estimating supersonic wave‐number components coupled to acoustic radiation in prescribed directions. The spatial wave‐number transform is performed in real time using a set of point structural sensors with an array of filters and associated signal processing. The use of the ...

ADVANCED TIME DOMAIN WAVE-NUMBER SENSING FOR STRUCTURAL ACOUSTIC SYSTEMS. II: ACTIVE RADIATION CONTROL OF A SIMPLY SUPPORTED BEAM

A real time structural acoustic sensor and associated signal processing is developed and applied to the active control of sound radiated by a simply supported beam. The sensor consists of multiple accelerometers mounted on the structure. An array of FIR filters processes the measured structural information to provide an estimate of the structural wave‐number component coupled to acoustic radiation in a prescribed direction. This time domain signal is used as the error information in a feedforward adaptive control approach. The single channel filtered‐X LMS algorithm is implemented here. Computer simulations in the discrete time domain demonstrate the ability of the sensor to replace the use of error microphones in the far field. The described sensor represents a significant alternative to the use of distributive structural sensors (for example piezoelectric material) by providing accurate radiation information over a broadband frequency range.

Comparison of two structural sensing approaches for active structural acoustic control

A numerical study comparing the use of two structural sensing approaches for sound radiation control is performed on a baffled rectangular plate. The first sensing approach implements an array of accelerometers whose outputs are filtered to construct an estimate of the sound pressure radiated at given angles in the far field. The second method uses the same array of point sensors to estimate the net volume acceleration of the plate. Results show the improved performances of the sensor observing far-field sound radiation over a volume acceleration based on sensor.

Sample-based engine noise synthesis using an enhanced pitch-synchronous overlap-and-add method

An algorithm for the real time synthesis of internal combustion engine noise is presented. Through the analysis of a recorded engine noise signal of continuously varying engine speed, a dataset of sound samples is extracted allowing the real time synthesis of the noise induced by arbitrary evolutions of engine speed. The sound samples are extracted from a recording spanning the entire engine speed range. Each sample is delimitated such as to contain the sound emitted during one cycle of the engine plus the necessary overlap to ensure smooth transitions during the synthesis. The proposed approach, an extension of the PSOLA method introduced for speech processing, takes advantage of the specific periodicity of engine noise signals to locate the extraction instants of the sound samples. During the synthesis stage, the sound samples corresponding to the target engine speed evolution are concatenated with an overlap and add algorithm. It is shown that this method produces high quality audio restitution with a low computational load. It is therefore well suited for real time applications.

Innovative tools for urban soundscape quality: Real-time road traffic auralization and low height noise barriers

Although noise exposure has been unanimously recognized for its impacts on people’s health, noise is still a current problem in many cities across the world. Besides, most noise analysis tools are based on long term average levels, which were initially meant to be used for inter-city infrastructures and are therefore insufficient to quantify the soundscape quality in urban environments. In this paper, we present some of the recent developments and results regarding two classes of innovative tools dedicated to the improvement of urban soundscape quality. First, recent advances in the field of real-time auralization of road traffic noise are presented and the benefit of realistic sound field restitution as a new analysis tool for the evaluation of urban planning projects is discussed and demonstrated. Second, some promising results regarding innovative urban noise reducing devices, taken among other works from the European project HOSANNA, are presented. In particular, the potential of vegetation and low he...

The initial development of a hybrid method for modeling outdoor sound propagation in urban areas

Accurately modeling urban outdoor sound propagation is a difficult problem. Using a frequency-domain or time-domain method is too expensive, and using an engineering, geometrical, or statistical method is too restrictive. For example, the ISO 9613-2, NMPB-Roads, and CNOSSOS-EU engineering methods only model diffraction for straight barriers. The Nord2000 and Harmonoise engineering methods can also model diffraction from wedge-shaped screens, but these modeling capabilities are still insufficient to model many potential noise mitigation solutions such as a gamma or T-shaped barrier. To extend the applicability of engineering methods, a detailed model like the boundary element method could characterize a complicated noise mitigation device versus a simple reference device. This presentation discusses the initial development of this hybrid method.

Active control of sound radiation from cylinders with piezoceramic actuators and structural acoustic sensing

In this paper, analytical and experimental results of an investigation of active control of sound radiated from cylinders are presented. The aluminum cylinder is 1 m in length, 25 cm in diameter, and 2.4 mm in thickness with two rigid endcaps at both ends. The excitation is a bandlimited random noise encompassing the first five modes of the cylinder and the control actuators are surface mounted piezoceramic transducers. Since it is desired to integrate the error sensors into the structure, the recently developed structural acoustic sensing (SAS) approach is extended to cylindrical coordinates and implemented using 12 accelerometers mounted on the cylinder. The SAS approach provides time‐domain estimates of far‐field radiated sound at predetermined radiation angles. The controller is a 3×3 filtered‐x LMS paradigm implemented on a TMS320C30 DSP. The results show excellent global control of the radiated sound over the frequency bandwidth of excitation. The SAS approach is shown to yield similar performances ...

Optimizing calculation points using Gaussian process regression

Many applications in computational acoustics calculate a value (e.g., sound pressure level) at many different locations to approximate the value throughout a region using interpolation. Often, the points are uniformly or exponentially spaced without a rigorous procedure to optimize the locations because directly minimizing the interpolation error is too computationally expensive. Instead, the interpolation error can be indirectly reduced by minimizing the maximum variance estimated using Gaussian process regression. This approach is less expensive because the variance at a point is not a function of the value at that point. Thus, each evaluation of the objective function (i.e., the maximum variance) does not require additional acoustical computations, which mitigates the cost of the objective function. As an example, this procedure is applied to an outdoor sound propagation case to approximate the insertion loss of a 3 m tall T-shaped barrier compared to 3 m tall straight barrier, which are modeled using the boundary element method. In this case, the optimized locations have a smaller interpolation error than uniformly and exponentially distributed points.Many applications in computational acoustics calculate a value (e.g., sound pressure level) at many different locations to approximate the value throughout a region using interpolation. Often, the points are uniformly or exponentially spaced without a rigorous procedure to optimize the locations because directly minimizing the interpolation error is too computationally expensive. Instead, the interpolation error can be indirectly reduced by minimizing the maximum variance estimated using Gaussian process regression. This approach is less expensive because the variance at a point is not a function of the value at that point. Thus, each evaluation of the objective function (i.e., the maximum variance) does not require additional acoustical computations, which mitigates the cost of the objective function. As an example, this procedure is applied to an outdoor sound propagation case to approximate the insertion loss of a 3 m tall T-shaped barrier compared to 3 m tall straight barrier, which are modeled using ...

Extending standard urban outdoor noise propagation models to complex geometriesa)

A hybrid method that combines a noise engineering method and the 2.5D boundary element method approximates outdoor sound propagation in large domains with complex objects more accurately than noise engineering methods alone and more efficiently than reference methods alone. Noise engineering methods (e.g., ISO 9613-2 or CNOSSOS-EU) efficiently approximate sound levels from roads, railways, and industrial sources in cities for simple, box-shaped geometries by first finding the propagation paths between the source and receiver, then applying attenuations (e.g., geometrical divergence and atmospheric absorption) to each path, and finally incoherently summing all of the path contributions. Standard engineering methods cannot model more complicated geometries but introducing an additional attenuation term quantifies the influence of complex objects. Calculating this extra attenuation term requires reference calculations but performing reference computations for each path is too computationally expensive. Thus, the extra attenuation term is linearly interpolated from a data table containing the corrections for many source/receiver positions and frequencies. The 2.5D boundary element method produces the levels for the real and simplified geometries and subtracting them yields a table of corrections. For a T-shaped barrier with two buildings, this approach reduces the mean error by approximately 2 dBA compared to a standard engineering method.