Session E02: Acoustics: Hydroacoustics and General (3:10pm - 3:55pm CST)

Deployment of a Mobile Four Sensor Infrasound Array for Severe Weather

Infrasound, sound at frequencies below 20 Hz has been observed to be emitted by tornado-producing storms up to two hours before tornadogenesis. Due to the low atmospheric attenuation of sound at these low frequencies, they may be detected several hundreds of kilometers away. If the received infrasound signals can be correlated with thermodynamic and flow field properties of the storms and/or tornadoes, passive infrasound monitoring has potential for the study and prediction of tornadoes and other severe weather. Previous work accomplished at Oklahoma State University has focused the observations from a single stationary array located in Stillwater, OK. However, this array is too far away from NEXRAD II radar stations to allow for low-level storm characterization. Therefore, a second array was designed to be deployable at various sites in Oklahoma within 50 km of the radar stations. Containing four microphones, the new array was deployed during the summer of 2020 at various sites affiliated with the Oklahoma Mesonet. Details of the array's design and deployment will be presented, as well as preliminary data collected during non-tornadic thunderstorms in 2020.   

Infrasound Propagation in the Atmospheric Conditions of Tornado Producing Storms

Tornado producing storms have been shown to emit sound below the threshold of human hearing (20 Hz), also known as infrasound. Infrasound can be detected over long ranges due to the low atmospheric attenuation at these frequencies. The current work computationally investigates the propagation of infrasonic signals through specific atmospheric conditions. This is done by utilizing a collection of atmospheric modeling codes known as AVO-G2S and a collection of numerical models for long range propagation of infrasound known as NCPAprop. These two open source codes are used to provide an estimation on how a particular infrasonic signal would behave on a specific date and time. Thus, this work will report findings on the propagation of infrasound utilizing historic meteorological data to analyze the atmospheric conditions during tornado producing storms that occurred in 2017 and 2019 in Oklahoma. Specifically, the likelihood is estimated that the emitted infrasound from a storm would be detected at an infrasound array located in Stillwater, OK.   

Scaling properties of the Ffowcs-Williams and Hawkings equation for complex acoustic source close to a free-surface

We investigate sources of discrepancy arising when using the acoustic results obtained at model scale, to deduce the acoustic field at full scale. Two aspects are pointed out: the classical similarity principle between a model and a prototype and the effects of a free-surface on the reflection of the acoustic signal. The analysis is carried out taking advantage of the Ffowcs-Williams and Hawkings equation. The free-surface is considered as a smooth plane, taking advantage of the method of images, thus considering the half-space Green function. As a test case we consider an open-sea ship propeller, tested in pulling condition at a model scale in the limit of Fn$=$0. The fluid dynamic data were obtained in a previous work, through the use of a Large-Eddy simulation. As a main result, application of similarity theory shows that perfect similarity can be obtained if the speed of sound at the laboratory scale is properly scaled. Overall, results show that imperfect similarity and absence of a free-surface in model tests, introduce errors when rescaling the model test data to full scale. In other words, acoustic data obtained at the model scale may be not significant for the acoustic characterization of the full scale propeller. In particular, the error becomes not negligible in the far field, where the quadrupole acoustic field, associated to the wake, predominates and noise spectrum may be substantially modified. The study may be considered of general importance and may be applied to a wide range of problems in hydroacoustics.   

A three-dimensional numerical model for the noise propagation in an inhomogeneous medium.

The wave equation for a viscoelastic fluid is solved to predict the acoustic response to a noise source in an inhomogeneous marine environment. The algorithm is able to handle the presence of the free-surface, stratification in the water column and interaction with a bottom layer which causes attenuation and reflection. The present model considers a monopole as a noise source. A second order central difference scheme is used for spatial and time derivatives on uniform grid. Spurious reflections at the boundaries are damped using the Perfectly Matched Layer. Validations are performed against significant benchmarks: the ideal ocean waveguide and the Pekeris waveguide. The results show that the method is able to reproduce correctly the propagation and reflection of the acoustic waves within the water column. This is true both in case of homogeneous medium and in the more interesting case of density variable medium with a bottom layer with attenuation. Advantages of the present method over the classical Helmholtz equation algorithm is the flexibility to handle complex noise sources. At the conference, the propagation of the noise generated by more complex sources, like a dipole and a quadrupole, will be presented and discussed.   

Sound radiated by a compliant wall in turbulent channel flow

Structural excitation due to turbulent flow generates sound. We simulate the far-field sound radiated by a compliant wall in a turbulent channel flow using Direct Numerical Simulation (DNS) at $Re_{\tau}=180$ and $400$. The fluid-structure-acoustic interaction is assumed to be one-way coupled. In the fluid domain, we solve the incompressible Navier-Stokes equations, and in the solid domain, we solve the linear elasticity equations using the time-domain finite element method. To compute the acoustic pressure radiated due to the structural excitation, we use half-space Green's function formulation. We validate the acoustics solver by comparing the sound computed from synthetic structural excitation to analytical results. We will discuss the effect of boundary conditions on the radiated sound and the structural excitation for the two Reynolds numbers. Further, we will study the fluid-structure-acoustic coupling by combining the fluid DNS data with the solid modal decomposition and the acoustic Green's function to identify the dominant wall-normal regions that contribute to the far-field sound as a function of frequency.   

Underwater Radiated Noise from Coastal Ferry Vessels: The Influence of Propellers and Operating Conditions

Cavitation-induced noise from propellers dominates the underwater radiated noise (URN) from ships – the largest source of underwater noise pollution worldwide. Vessel noise has often been related to travel speed, and slowdowns been found to be effective in reducing overall noise in high-traffic areas. However, radiated noise it not universally correlated with travel speed. We investigated field-measured radiated noise levels from eight coastal ferry vessels each operating at a range of speeds, and considered the relationships between noise at individual 1/3-octave band levels and vessel operating conditions in order to uncover the physical mechanisms and characteristic acoustic markers of the specific noise behaviors. We considered a gamut of propeller parameters including depth, pitch, and slip and determined that only speed was reliably related to URN. Among vessels for which speed and noise were anti-correlated at most frequencies, all of which used controllable pitch propellers, acoustic markers were found that indicated a change in the noise generation regime when the propellers were under-loaded. The acoustic signatures of these regimes appear to correspond to different types of propeller cavitation.