Stochastic Noise Generation Research Articles

Aerodynamic noise and its reduction of the marine gas turbine air exhaust system.

The aerodynamic noise of pipelines is an important part of the noise of a ship's system. This paper conducted numerical investigations on the flow and acoustic characteristics of the marine gas turbine exhaust system. The near-field and far-field acoustic characteristics of the internal flow noise of the exhaust system are calculated by employing the Möhring's sound analogy method. In addition, the far-field acoustic characteristics of the external jet flow noise of the exhaust system are calculated by employing the stochastic noise generation and radiation (SNGR) method. Two kinds of protrusions are added to the main nozzle outlet to achieve noise reduction. The internal sound field of the marine exhaust system is dominated by low frequency sound sources, which are more obvious as the exhaust mass flow rate decreases. As for the external sound field of the marine exhaust system, the peak frequency of the far-field noise spectrum decreases with the decrease in the exhaust mass flow rate. The eight periodic protrusions perform better in reducing the internal aerodynamic noise of the exhaust system, while the five aperiodic protrusions perform better in reducing the external jet noise of the exhaust system.

Coupling a Neural Network with a Spatial Downscaling Procedure to Improve Probabilistic Nowcast for Urban Rain Radars

Rainfall forecasting plays a key role in mitigating environmental risks in urban areas, which are subject to increasing hydrogeological risk due to transformations in the urban landscape. We present a new technique for probabilistic precipitation nowcasting by generating an ensemble of equiprobable forecasts, which is especially useful for weather radars with limited spatial range, and that can be used operationally on devices with low computational capacity. The ensemble members are obtained by a novel stochastic noise generation process, consistent with the spatial scales not resolved by the prediction model, which allows continuous downscaling of the output of a deep generative neural network. Through an in-depth analysis of the results, for precipitation accumulated in the first hour, measured by all the most robust skill indicators, extended to an entire year of data at 5-min time resolution, we demonstrate that the proposed procedure is able to provide calibrated, reliable, and sharp ensemble rainfall forecasts with a quality comparable or superior to the state-of-the-art classical alternative optical flow technique. The ensemble generation procedure we propose is sufficiently general to be applied in principle to other deterministic architectures as well, thus enabling their generalization in probabilistic terms.

Numerical and Experimental Research on Single-Valve and Multi-Valve Resonant Systems

Multiple valves in the pipeline system belong to obvious periodic structure distribution types. When a high-speed airstream flows through the pipeline valve, it produces obvious aero-acoustic and acoustic resonance. Acoustic resonant systems with single and six-pipe valves were investigated to understand the flow and acoustic characteristics using a numerical simulation method and testing method. The strongest acoustic resonance occurred at a specific flow velocity with a corresponding Strouhal number of 0.47 corresponding to the geometric parameters in the paper. Moreover, acoustic resonance occurred in a certain velocity range, rather than increasing with the increase of the velocity of the pipeline. This regular increase provided an important theoretical basis for the prediction of the acoustic resonant and ultimate acoustic load of a single-valve system. When the pipeline was attached with multiple valves and the physical dimension was large, the conventional aero-acoustics calculation results were seriously attenuated at high frequency; the calculation method involving a cut-off frequency in this paper was presented and could be used to explain the excellent agreement of the sound pressure level (SPL) below the cut-off frequency and the poor agreement above the cut-off frequency. A new method involving steady flow and stochastic noise generation and radiation (SNGR) was proposed to obtain better results for the SPL at the middle and high frequencies. The comparison results indicated that the traditional method of Lighthill analogy and unsteady flow could accurately acquire aerodynamic noise below the cut-off frequency, while the new method involving steady flow and SNGR could quickly acquire aerodynamic noise above the cut-off frequency.

Compressibility Modified RANS Simulations for Noise Prediction of Jet Exhausts with Chevron

The impact of compressibility modified RANS turbulence closures is investigated for high subsonic round and chevron jet flows with Mach = 0.9 and Re = 1.03×106, including the predicted acoustic noise generation. The well-documented chevron jet flow and noise cases, namely NASA SMC000 and SMC006 are selected as the simulation configurations. Two compressibility RANS closures are considered, which are based on the k-ε turbulence model. The first type only considers the compressibility dissipation rate, and the second type accounts for three modifications of compressibility dissipation rate, pressure dilation and production limiter. The acoustic noise is calculated employing the SNGR (Stochastic Noise Generation and Radiation) method using the flow prediction of the three-dimensional RANS simulations. The results show that both of the two types of compressibility modified RANS models improve the accuracy of the mean flow and turbulence quantities. This results in more accurate jet noise predictions than with the standard RANS model. The first type modification is found to be moderate and the second type is remarkable. The noise results by the second type model, i.e. Sarkar2 model, agree with the experimental data quite well. For the mean flow field, the compressibility modified model (Sarkar2 model) estimates a shorter potential jet core, and improved predictions of the velocity in the downstream region are observed. The study demonstrates the importance of considering the compressibility modified RANS closure for the noise prediction of high-speed jets via the comparison to experimental data. Hence, the SNGR method is found to be cost effective for jet noise prediction, when compared to other approaches.

An Efficient Method for Prediction of the Flow-induced Vehicle Interior Noise

Wind noise is an important attribute for vehicle interior quietness. To improve the efficiency of wind noise simulation, a method of Stochastic Noise Generation and Radiation (SNGR) is used to rebuild the flow field based on steady simulation result. For wind noise simulation, the flow field result of transient simulation is only applied to low frequency, and the result of the SNGR method is applied to middle and high frequency. Finally, correcting the SNGR result by correction coefficient and combining the two results. The experiment and simulation results have good agreement. The results shows that the hybrid method for predicting the wind noise in vehicle can not only obtain good results, but also improve computational efficiency.

Aeroacoustic topology optimization of noise barrier based on Lighthill's acoustic analogy

Topology optimization for the shape of noise barrier taking into account the aeroacoustic sound source is presented based on Lighthill's acoustic analogy approach. In the previous acoustic topology optimization based on the Helmholtz equation, only simple acoustic sound sources were taken into account. To overcome this limitation, the present aeroacoustic topology optimization is performed using Lighthill's equation that represents the sound propagation by an aeroacoustic sound source. In this study, aeroacoustic sound sources are randomly generated through the stochastic noise generation and radiation method. As an interpolation method of material properties by continuous design variables, the linear filter and the sigmoid penalty function are used. The top shape of noise barrier is designed by the optimization examples for various conditions, such as random aeroacoustic sound sources, interpolation methods, and target frequencies.

Applications of directional wavelets, Universal Multifractals and anisotropic scaling in ensemble nowcasting; a review of methods with case‐studies

AbstractThis article begins with a review of the scale decomposition and stochastic noise generation aspects of the ensemble nowcasting process, as used in the Short Term Ensemble Prediction System (STEPS). Alternative methods are then suggested, with case‐study examples, which could lead to enhanced performance and may also be applicable in other fields. The standard FFT‐band‐pass‐filtering scale decomposition, essential to this process, is shown to be practically the same as a redundant Continuous Wavelet Transform (CWT). This immediately makes accessible a number of more advanced methods from the wavelet literature. Particular attention is given to directional wavelet transforms and their use for analysing anisotropic scaling behaviour, common in weather images. With a view to numerically efficient operational usage, the focus is then shifted to Discrete Wavelet Transforms (DWTs). While the standard DWT has a number of disadvantages, including limited orientational selectivity in two dimensions, the relatively recent Dual Tree Complex WT (DTCWT) overcomes most of these. Examples are given that show how its six‐fold directionality can be exploited, to create synthetic features that replace the unpredictable small‐scale parts of a (weather‐radar‐based) extrapolation nowcast with more realistic spatially varying anisotropy. Further, the Bounded Log‐normal Cascade noise model, as used in STEPS, is reconsidered and a proposal to use Universal Multifractal realizations is presented. For the noise parameters, some operationally viable estimation methods are considered, such as the Double Trace Moment technique and a structure‐function‐based approach, which could work on readily available DTCWT coefficients.

Influence of Turbulence Statistics on Stochastic Jet-Noise Prediction with Synthetic Eddy Method

The separation algorithms that solve a flowfield and a sound field separately are used to estimate the jet noise accurately while suppressing the increase of computational cost. In this study, the synthetic eddy method (SEM) was applied to a strong shear flow to simulate a major sound source of jet noise, and the influence of turbulence statistics used in the SEM on the predicted far-field jet noise was investigated. The divergence-free SEM was employed as a reconstruction method of the turbulent field, and the shear effect due to the background flow was provided to the generated velocity field. In addition, the dissipation effect of the turbulence in the shear layer was modeled by decorrelating the generated velocity fluctuations. The estimated noise using the linearized Euler simulations appeared larger than the experimental value when parameters of the SEM were set based on actual flow conditions. In contrast, the sound pressure spectra agreed well with the experiment by increasing the time scale of the noise source. With the present stochastic noise prediction framework based on the SEM, turbulence statistics and far-field sound pressure spectra were correlated, which would help to modify parameters in stochastic noise generation methods to achieve desired far-field sound pressure spectra.

Prediction of ducted diaphragm noise using a stochastic approach with adapted temporal filters

The noise production by ducted single- and double-diaphragm configurations is simulated using a stochastic noise generation and radiation numerical method. The importance of modeling correctly the anisotropy and temporal de-correlation is discussed, based on numerical results obtained by large eddy simulation. A new temporal filter is proposed, designed to provide the targeted spectral decay of energy in an Eulerian reference frame. An anisotropy correction is implemented using a non-linear model. The acoustic propagation problem is solved using Lighthill’s aeroacoustic analogy with a tailored Green’s function obtained analytically. Comparison with scale--resolved data indicates that accurate far-field noise predictions are obtained for both single and double diaphragm configurations, with computational costs significantly reduced with respect to the scale-resolved approach. A grouping scheme for the noise sources based on the octree structure is introduced to minimize the memory requirements and further reduce the computational cost.

Jet noise prediction via coupling large eddy simulation and stochastic modeling

A novel method for efficient and accurate jet noise prediction is developed and tested in the framework of coupled LES and stochastic noise modeling. In the proposed method, the low frequency range of the acoustic spectrum, which corresponds to large turbulent structures, is resolved with an implicit LES model by employing higher-order spatial and temporal discretizations. The high frequency range, which corresponds to the fine-scale turbulence, is modeled via a stochastic broadband noise generation model. The inputs to the stochastic model are represented by statistics from an axisymmetric RANS jet simulation. The smaller synthetic scales are convected by larger scales, accounting for the effects of sweeping from the larger LES-resolved turbulent scales. The farfield acoustic data is obtained using either the Linearized Euler Equations, or an acoustic analogy based on the Ffowcs-Williams Hawkings method. The method is evaluated on cold and heated jets at different Reynolds numbers.

Combined CFD-Stochastic Analysis of an Active Fluidic Injection System for Jet Noise Reduction

In the framework of DANTE project (Development of Aero-Vibroacoustics Numerical and Technical Expertise), funded under the Italian Aerospace Research Program (PRORA), the prediction and reduction of noise from subsonic jets through the reconstruction of turbulent fields from Reynolds Averaged Navier Stokes (RANS) calculations are addressed. This approach, known as Stochastic Noise Generation and Radiation (SNGR), reconstructs the turbulent velocity fluctuations by RANS fields and calculates the source terms of Vortex Sound acoustic analogy. In the first part of this work, numerical and experimental jet-noise test cases have been reproduced by means RANS simulations and with different turbulence models in order to validate the approach for its subsequent use as a design tool. The noise spectra, predicted with SNGR, are in good agreement with both the experimental data and the results of Large-Eddy Simulations (LES). In the last part of this work, an active fluid injection technique, based on extractions from turbine and injections of high-pressure gas into the main stream of exhausts, has been proposed and finally assessed with the aim of reducing the jet-noise through the mixing and breaking of the turbulent eddies. Some tests have been carried out in order to set the best design parameters in terms of mass flow rate and injection velocity and to design the system functionalities. The SNGR method is, therefore, suitable to be used for the early design phase of jet-noise reduction technologies and a right combination of the fluid injection design parameters allows for a reduction of the jet-noise to 3.5 dB, as compared to the baseline case without injections.

AIRFRAME NOISE PREDICTION USING THE SNGR METHOD

In this paper, the Stochastic Noise Generation and Radiation method (SNGR) is used to predict airframe noise. The SNGR method combines a stochastic model with Computational Fluid Dynamics (CFD), and it can give acceptable noise results while the computation cost is relatively low. In the method, the time-averaged mean flow field is firstly obtained by solving Reynolds Averaged Navier-Stokes equations (RANS), and a stochastic velocity is generated based on the obtained information. Then the turbulent field is used to generate the source for the Acoustic Perturbation Equations (APEs) that simulate the noise propagation. For numerical methods, timeaveraged RANS equations are solved by finite volume method, and the turbulent model is K – ε model; APEs are solved by finite difference method, and the numerical scheme is the Dispersion-Relation-Preserving (DRP) scheme, with explicit optimized 5-stage Rung-Kutta scheme time step. In order to test the APE solver, propagation of a Gaussian pulse in a uniform mean flow is firstly simulated and compared with the analytical solution. Then, using the method, the trailing edge noise of NACA0012 airfoil is calculated. The results are compared with reference data, and good agreements are demonstrated.

Experimental and numerical investigation of noise generated by rotor blade passing an exhaust grille

This paper considers the noise generated by interaction between the rotor blade and the exhaust grille. Acoustic measurements were first performed in a semi-anechoic chamber to investigate the characteristic of the noise. An approach that combines a Reynolds averaged Navier-Stokes calculation with the stochastic noise generation and radiation method is used to identify noise sources. With the computed flow solutions, the broadband noise radiated by the grille is evaluated based on a semi-empirical noise model. The numerical results are compared with the measured noise levels and good agreement is obtained.

Prediction of noise radiation from basic configurations of landing gears by means of computational aeroacoustics

Noise radiation from aircraft during the takeoff and landing has become a major issue for inhabitants living in the vicinity of airports and thus for regulation authorities and aircraft developers. However the numerical simulation of aeroacoustic noise, especially for complex geometries like a landing gear, remains one of the most difficult challenges in aeroacoustics. The present study, aiming at predicting noise radiation from basic geometries as well as the noise radiation of a simplified landing gear, employs a hybrid approach that combines a CFD simulation with the decoupled computational aeroacoustics (CAA) simulation. Flow-induced noise is assumed to originate from turbulence. Reynolds-averaged Navier–Stokes equations with different closure approaches can be employed to gain the required turbulent quantities. Subsequently, quantities as the mean flow velocities, pressure, density, turbulent kinetic energy and dissipation rate of the CFD simulation are the starting point for the generation of the transient acoustic sources by the stochastic noise generation and radiation (SNGR) method. It is assumed that the acoustic phenomena do not provide feedback to the mean flow field and turbulence and thus a recalculation of the flow field is not required. Since the propagation of sound is insignificantly influenced by turbulent and viscous effects, it can be described by the Euler equations in the near field. The CAA simulation is extended with a Ffowcs Williams Hawkings (FWH) module that calculates the noise levels in the far field upon integrating the surface source terms on a porous FWH surface within the CAA domain. The results of the simulations are compared with experimental data, obtained by measurements in an acoustic wind tunnel.

Acoustic Source Terms for the Linearized Euler Equations in Conservative Form

A rather novel approach to predict jet noise is the Stochastic Noise Generation and Radiation (SNGR) method. The SNGR method uses the linear Euler equations as an acoustic analogy together with source terms which are modeled. In other studies the Euler equations on primitive form are used. In the present work the linear Euler equations on conservative form are used. Due to this, new source terms have to be derived for the conservative set of equations. A formal derivation of the correct source terms for the linear Euler equations on conservative form is presented. Simplified versions of the derived source terms are also developed. To validate the derived source terms a direct simulation of a forced 2D mixing layer is carried out. The solutions to the linearized Euler equations with source terms are compared to the solution of the direct simulation and show a good agreement. All simulations are performed using Tam and Webb's fourth order DRP scheme and a four step fourth order Runge-Kutta time marching technique. Artificial selective damping introduced through the numerical scheme is used to avoid spurious waves. Absorbing boundary conditions based on characteristic variables, Engquist and Majda, are used at the free boundaries and a buffer layer is added at the outflow.

Numerical Solution of Acoustic Propagation Problems Using Linearized Euler Equations

The goal of this work is to study some numerical solutions of acoustic propagation problems using linearized Eider's equations. The two-dimensional Euler's equations are linearized around a stationary mean flow. The solution is obtained by using a dispersion-relation-preserving scheme in space, combined with a fourth-order Runge-Kutta algorithm in time. This numerical integration leads to very good results in terms of accuracy, stability and low storage. The radiation of a source hi a subsonic and supersonic uniform mean flow is investigated. The numerical estimates are shown to be in excellent agreement with the analytical solutions. Next, a typical problem in jet noise is considered, the propagation of acoustic waves in a sheared mean flow, and the numerical solution compares favorably with ray tracing. The final goal of this work is to improve and to validate the Stochastic Noise Generation and Radiation (SNGR) model. In this model, the turbulent velocity field is modeled by a sum of random Fourier modes through a source term in the linearized Euler's equations. The implementation of acoustic sources in the linearized Euler's equations is thus an important point. This is discussed with emphasis on the ability of the method to describe correctly the multipolar structure of aeroacoustic sources. Finally, a nonlinear formulation of Euler's equations is solved hi order to limit the growth of instability waves excited by the acoustic source terms.

Aeroacoustic simulations and stochastic approach using linearized Euler’s equations

Sound generation and propagation in a turbulent flow is a very difficult numerical problem. Indeed, acoustic fluctuations are very small by comparison to the mean and turbulent fields. Linearized Euler’s equations provide accurate numerical solutions only in working on perturbations. In the stochastic noise generation and radiation model, the mean flow is calculated by solving the Reynolds-averaged Navier–Stokes equations. Then, Euler’s equations are linearized around this mean flow field. A turbulent source term is introduced in the equations, and the turbulent velocity field is modeled by a sum of random Fourier modes. The numerical solution is obtained by using a dispersion-relation-preserving scheme in space, combined with a fourth-order Runge–Kutta algorithm in time. In order to validate the numerical method, radiation of multipolar sources in a uniform and sheared mean flow has been investigated in a two-dimensional case. The calculated solutions favorably compare with analytical solutions or with ray tracing. Then, the SNGR model is applied to a subsonic jet noise in a 3-D geometry. Numerical results and future developments will be discussed.

Stochastic approach to noise modeling for free turbulent flows

A new approach to noise modeling for free turbulent flows is presented. The equations governing the sound field are obtained in two steps. The first step consists of treating the mean and turbulent components of the flow while the acoustic perturbations are neglected. In the second step, a set of equations is derived for the acoustic variables. On the left-hand side of this system, one finds the linearized Euler equations, whereas the right-hand side exhibits source terms related to the turbulent fluctuations and their interactions with the mean flow. These terms are modeled using a stochastic description of the three-dimensional turbulent motion. This is achieved by synthesizing the velocity field at each point in space and for all times with a collection of discrete Fourier modes. The synthesized field posesses the suitable one- and two-point statistical moments and a reasonable temporal power spectral density. The linearized Euler equations including a stochastic description of noise sources are solved numerically with a scheme based on a fractional step treatment. Each one-dimensional problem is solved with a weak formulation. A set of calculations are carried out for a simple freejet. Comparisons between calculations and experiments indicate that a spatial filtering of the source terms is required to obtain the expected level in the far field. Realistic pressure signals, power spectral densities, and sound field patterns are obtained. It is indicated that the stochastic noise generation and radiation (SNGR) approach may be applied to more complex flows because the numerical codes used to calculate the mean flowfield and the wave propagation are not specific of jet configurations. The limitations of the present model lie in the statistical properties of the synthetic turbulent field and in the use of an axisymmetric modeling of the acoustic propagation.