Weikang Jiang

In acoustic testing for aircraft engines, duct-mounted microphone arrays are widely used for sound source localization, mode identification, and acoustic liner experiments. Phase calibration of arrays is considered equally critical as amplitude calibration; however, such calibration is often constrained by duct geometry and necessitates repeated disassembly, which poses a risk of mechanical damage. An in-situ microphone array calibration method based on a rotating sound source has been proposed in this paper. Because a rotating source enables frequency-shift of distinct azimuthal modes, the resulting mode-resolved sound fields can be used to distinguish positional phase differences from inter-channel phase errors in single-ring microphone arrays. By combining the proposed axial and single-ring phase calibration techniques, calibration of arrays with a large number of microphones is achieved and validated through Monte Carlo simulations and test rig experiments. Results demonstrate that the proposed method offers the advantages of in-situ calibration within acoustic ducts, compatibility with high-channel-count arrays, and robust, efficient operation.

The impedance eduction method is applied in rectangular ducts, with the focus being placed on flat-panel acoustic liner samples traditionally. To measure the accurate acoustic impedance of the acoustic liner in a real engine, it is necessary to extend the impedance eduction method to cylindrical ducts. For the impedance eduction method in circular ducts with large transversal dimensions, consideration must be given to the termination impedance of higher-order cut-on modes. So, an extended impedance eduction method incorporating termination impedance is developed for multimodal acoustic propagation in cylindrical ducts. To simultaneously search for more variables, including the higher-order termination impedance, the multiple measurements with different incident modes are included in the proposed impedance eduction procedure. Numerical simulations and experiments are employed to validate the proposed method. The Monte Carlo Method is used to evaluate the uncertainty from the sound pressure propagating nonlinearly to the impedance.

When the wake of a marine propeller moves through the geomagnetic field, a low-frequency electric field is induced in the surrounding seawater. The spatial distribution of this kind of field has been well analyzed in conventional studies. However, the equivalent source that controls the far-field response has not been identified from conservation principles. A charge-neutral dipole framework for the wake-induced electric field is proposed in this study. Under the low-magnetic-Reynolds-number quasi-static approximation, the field satisfies a motional-induction Poisson equation. Far-field multipole analysis shows that the monopole term vanishes for a localized wake with no net boundary current. The leading far-field response is therefore a dipole. The dipole moment is further related to measurable wake parameters using a double-Gaussian model of the two-mode near wake. The model is validated by a one-way coupled computational fluid dynamics (CFD)--electromagnetic simulation of a scaled DTMB 5415 propeller. The CFD volume integral and the closed-form expression agree within 0.55\% in magnitude and 0.1° in direction. The resulting model provides a compact tool for electromagnetic-signature estimation and parameter inversion.

A resonant extremely low-frequency (ELF) mechanical antenna prototype with an adjustable resonant frequency is developed for long-range cross-media communication. Magnetostrictive actuation is studied to achieve high energy conversion rate. The resonance mechanism is utilized to significantly enhance radiation efficiency compared to a conventional amplifying lever. This prototype employs a high-performance electret thin film as the source of electromagnetic radiation, offering advantages such as miniaturization and high radiation intensity. Additionally, the radiation mechanism of the electret with a single charge is studied, and a vibration-radiation model of the electromagnetic dipole is established. The vibration and the magnetic field simulations and tests are conducted to evaluate the prototype’s performance. A tenfold vibration amplification is obtained at resonance with the designed lever compared to the off-resonance state. As a result, the designed mechanical antenna can achieve a radiation distance of up to 92.2 km at 1 fT. This study demonstrates a promising path for developing miniaturized and efficient ELF transmitters.

The Spinning Mode Synthesizer (SMS) is an effective test rig for studying noise propagation and radiation in aeroengines and gas turbines for investigating the mechanism of combustion oscillation, where specific acoustic modes are generated through precise amplitude modulation and phase shifting of an artificial sound source system. Conventional SMSs neglect the variation of cross-sections and work at room temperature, and fail to correctly characterize the propagation of acoustic modes in combustors. A methodology is proposed to excite duct modes in cylindrical ducts with small axial gradients of diameter and temperature. The variation of axial wavenumber and sound velocity separately influenced by varying diameter and temperature gradients is theoretically investigated and considered in the process of mode excitation. The coefficients of the observation matrix for acoustic field propagation are derived as functions corresponding to the axial coordinate by solving the three-dimensional wave equations using the Wentzel–Kramers– Brillouin (WKB) approximation. Moreover, the transition of cut-on modes into cut-off modes at some special wavenumber dependent on the diameter and temperature of the duct is theoretically investigated. Through the input of target modes and inversion of the obtained observation matrix, the amplitude and phase angles of the loudspeaker array required by mode excitation are determined. A spinning mode synthesizer is established on a cylindrical duct with a varying diameter to validate the proposed method of mode excitation, equipped with 24 loudspeakers distributed in 3 rings, and both azimuthal modes and radial modes are successfully generated in the experiments. The amplitude of target modes stimulated by the facility is 10 dB higher than any other cut-on modes across the operating frequency range, demonstrating a great signal-tonoise ratio and robustness of the proposed method.

To compute the acoustic wave scattered by local flow inhomogeneities like vortices, an integro-differential scheme is developed to numerically solve the Lippmann-Schwinger formulation of the linearized Euler equations in the frequency domain, achieving spectral-order accuracy and a computational complexity of [ ]. This formulation converts the aeroacoustic scattering equation into a second-order integro-differential system using the Green’s function of the Helmholtz equation. The singular convolution integrals are evaluated with the kernel truncation method, and high-order derivatives are computed using Fourier series. Validated against several aeroacoustic benchmarks, our method shows significant advantages in efficiency and accuracy, while avoiding special treatments for radiating boundary conditions. Additionally, we extend this formulation to unsteady flows, interpret the spectral broadening effect as the convolution of incident waves with the medium’s motion in the frequency domain, and offer an example for its numerical simulation.

This paper presents a semi-analytical method, referred to as the linear-velocity-profile fast field program (LFFP), for predicting two-dimensional sound fields in ambient parallel mean flows. The proposed method incorporates the linear velocity layering method into the fundamental framework of fast field program (FFP) to achieve reduced computational costs and enhanced precision, particularly under high-velocity gradient conditions. The accuracy of LFFP is validated through a two-dimensional jet case by comparison with the linearized Euler equation in frequency-domain. In shear flow cases, results obtained from various combinations of Mach number difference and shear layer thickness suggest that the reason for the higher precision of LFFP, compared to traditional FFP, primarily arises from its consideration of the second velocity gradient term in the Pridmore-Brown operator within each single layer. To systematically and mathematically explain this observation, residual analysis is introduced. Furthermore, based on the residual analysis, the multi-staircase layering model is subsequently developed to improve computational efficiency and adapt to sound field calculations in environments with multiple vertically variable ambient quantities.