To achieve low-frequency noise control, this study proposes a cascaded resonant structure with dual apertures and embedded tapered necks, realizing low-frequency broadband sound absorption through a hybrid series–parallel configuration. A theoretical analytical model for sound absorption characteristics was established based on the Transfer Matrix Method (TMM), and its reliability was verified by comparison with numerical simulations. Furthermore, the effects of structural parameters—including total thickness, cavity height ratio, perforation depth, aperture diameter, and taper—on sound absorption performance were analyzed. A broadband optimization design for multi-cell hybrid series–parallel coupling was conducted using a genetic algorithm, targeting the low-frequency range of 200–400 Hz and the broadband range of 400–1000 Hz, respectively. The results indicate that the hybrid series–parallel structure effectively overcomes the modal quantity limitations of traditional single-resonance systems, exciting denser resonant modes and significantly improving the flatness of the broadband absorption curve. With a total thickness not exceeding 50 mm, the optimized 3 × 3 unit structure achieves average sound absorption coefficients of 0.7638 in the 200–400 Hz band and 0.7991 in the 400–1000 Hz band. Experimental results show good agreement with both theoretical and simulation data, validating the feasibility of the composite system in achieving continuous broadband sound absorption within a compact space. This study promotes the miniaturization design of acoustic metamaterials and holds broad application prospects in the field of low-frequency noise control.
Acoustic vector sensors (AVSs) are widely employed in underwater surveillance, source localization, target tracking, seabed exploration, and autonomous underwater vehicle (AUV) applications owing to their ability to simultaneously measure acoustic pressure and particle velocity. However, when deployed on underwater platforms, AVSs are susceptible to structure-borne vibrational noise from onboard machinery and propulsion systems, which degrades acoustic measurements and direction-of-arrival (DOA) estimation performance. Existing vibration mitigation approaches primarily focus on sensor isolation or mechanical damping and provide limited compensation for vibration-induced interference during signal processing. To address this challenge, a vibrational noise cancellation (VNC) method is proposed that employs a secondary accelerometer as a reference sensor together with a multichannel adaptive filtering framework to suppress platform-induced vibration components. Simulation studies quantify the influence of vibrational noise on DOA estimation under varying vibration intensities and signal-to-noise ratios (SNRs). Performance is evaluated using the signal-to-interference-plus-noise ratio (SINR) and DOA root-mean-square error (RMSE). The results demonstrate improved SINR and reduced DOA estimation error across all tested conditions, with consistent performance even under spectral overlap between the desired signal and vibration components. Experimental validation using a secondary accelerometer-integrated AVS prototype under controlled laboratory and submerged water-tank conditions demonstrates the effectiveness of the proposed VNC framework. The results validate the proposed multichannel adaptive filtering approach as an effective means of mitigating platform-induced vibration, providing a practical signal-processing solution for improving the performance of AVS-based underwater sensing platforms.
Differential microphone arrays (DMAs) utilize signal subtraction between closely spaced sensors to approximate the spatial derivative of the acoustic pressure field, enabling frequency-invariant beampatterns within compact array geometries. This operating principle, however, makes DMAs inherently sensitive to positional inaccuracies, where even small deviations in sensor placement can significantly degrade array performance. In this work, the effects of position errors arising from array fabrication and mounting imprecision are systematically investigated. A geometric error model is proposed to quantify these effects in planar DMAs (PDMAs) by characterising sensor position inaccuracies into translational and rotational errors. An analytical formulation of the quantized beampattern for a general first-order PDMA is developed, and the impact of these errors is evaluated relative to the ideal case. Results indicate that even sub-millimeter translational errors significantly degrade null depth (ND) thereby reducing the array’s interference suppression capability while introducing minor beampattern variations at angles farther away from the nulls. In contrast, rotational errors cause more severe ND degradation, reduce the directivity index, and induce a uniform angular shift and distortion of the beampattern, displacing both mainlobe and nulls. Insights into how geometric errors vary with signal frequency, inter-sensor spacing, and array steering are presented. A method for compensating known position errors is also discussed. Experimental results validate the proposed error model for the representative TL and RL error cases.
Porous materials such as glass wool are widely used in aircraft fuselage insulation systems for their sound absorption performance. In operational aeronautical environments, their acoustic performance may differ from that measured under nominal laboratory conditions due to several factors such as protective coverings, installation procedures and moisture variations occurring during flight operations and throughout the aircraft service life. Despite their practical relevance, the effects of these non-nominal conditions on sound absorption variability remain insufficiently characterized. This study investigates the frequency-dependent sound absorption coefficient of aeronautical glass wools under controlled non-nominal conditions by combining impedance tube measurements, machine-learning techniques and Shapley additive explanations (SHAP). The investigated factors include material type, protective layers, relative humidity, humidity cycling and both controlled and operator-dependent mounting configurations. Results indicate that material, layer, relative humidity and humidity cycling significantly influence the acoustic response, with moisture-related effects exhibiting a strongly frequency-dependent behavior concentrated within three distinct frequency bands. Installation-related effects are also found to produce identifiable spectral variations, with operator-dependent mounting conditions generally associated with higher absorption levels and shifts of the peak response toward lower frequencies. The results further indicate that more controlled and uniformly distributed contact conditions improve measurement repeatability and reduce installation-induced variability.
A finite element-boundary element (FE-BE) framework is developed to investigate the vibroacoustic response of a functionally graded acoustic black hole (FG-ABH) panel subjected to thermal loading. The structural response of the panel is obtained using the FE method, while the radiated acoustic field is evaluated through a direct BE formulation. Before the main parametric study, the numerical framework is validated through benchmark comparisons available in the literature and an experimental investigation on a damped one-dimensional ABH beam and plate configurations. The experimental comparison shows that the model can reproduce the ABH-induced vibration localisation observed in the measured operational deflection patterns. The effects of the ABH configuration, damping treatment, thermal environment, and structural stiffening are examined in terms of averaged quadratic velocity (AQV), radiated sound power (RSP), and radiation efficiency. The results show that the FG-ABH panel exhibits a distinct vibroacoustic behaviour compared with the corresponding uniform FG panel. In the lower-frequency range, the local reduction in bending stiffness caused by the ABH profile produces resonance shifts and locally higher vibration amplitudes. With increasing frequency, however, the tapered region promotes flexural-wave localisation, leading to clear reductions in structural vibration and acoustic radiation. Beyond the estimated ABH cut-on frequency of nearly 940Hz, broadband suppression of AQV and RSP is observed. The viscoelastic damping layer further improves the reduction performance, with AQV suppression of nearly 10–15dB in the higher-frequency region. The study further reveals that the unstiffened FG-ABH panel becomes unstable under elevated thermal loading because of thermal softening and stiffness loss in the tapered region. To address this limitation, cross stiffeners are incorporated into the FG-ABH panel. The stiffened configuration preserves the broadband vibroacoustic suppression capability of the ABH while ensuring stable behaviour under thermal conditions. Although thermal loading shifts the structural resonances towards lower frequencies, the broadband sound radiation response remains comparatively less sensitive. The findings indicate that the combined FG-ABH and stiffener configuration can provide an effective passive strategy for reducing vibration and sound radiation in lightweight structures operating under thermal environments.
A novel acoustic metastructure composed of a graded helical duct and Helmholtz resonators (GHD-HR) for low-frequency broadband sound absorption is proposed and modeled using the transfer matrix method. The structure consists of graded helical duct (GHD) connected in series with multiple Helmholtz resonators (HRs). It achieves efficient, low-frequency, broadband sound absorption at a deep sub-wavelength scale in a compact geometry. The results demonstrate that the synergistic interaction between the graded helical duct and HRs yields multiple high-absorption peaks in the low-frequency and maintains good sound absorption performance in the mid-to-high frequency. Parametric analysis indicates that the neck and cavity dimensions, and the structural parameters of the helical duct, are key influencing factors. A genetic algorithm is employed to optimize these parameters, improving the average absorption coefficient from 0.802 to 0.919 within the target frequency band of 1–1000 Hz. The experimental results show good agreement with the theoretical predictions and finite element simulations, validating the proposed model. The metastructure offers a compact, deep-subwavelength solution for low-frequency broadband sound absorption and holds promise for engineering noise control applications.
Urban courtyards serve as important semi-enclosed spaces that influence residents’ acoustic comfort and psychological wellbeing, yet their perceptual qualities remain under-researched compared to street canyons. This study investigates how height-to-width ratio (H/W) and natural design features (trees, vegetation, and water elements) affect perceived enclosure and pleasantness in urban courtyards through multisensory assessment. A laboratory experiment was conducted using virtual reality environments and acoustic simulations with 33 participants who evaluated courtyards varying in H/W (0.3 and 1.0), size (750 m2 and 3,000 m2), and natural features across visual-only, audio-visual, and audio-only conditions. Results showed that higher H/W consistently increased perceived enclosure while reducing pleasantness in both visual-only and audio-visual conditions. Under audio-only conditions, H/W had negligible effects on pleasantness in small courtyards, though a modest but significant effect was observed in large courtyards (η2 = 0.14). Natural features, particularly trees with birdsong, significantly enhanced pleasantness (η2 = 0.18–0.48) and produced modest, context-dependent reductions in perceived enclosure (η2 = 0.08–0.15 for scenario effects). Larger courtyards were perceived as more pleasant and open (less enclosed) than smaller ones in both visual-only and audio-visual conditions. Audio-visual presentation generally increased perceived enclosure (lower openness scores), while also increasing pleasantness relative to the visual-only condition. These findings offer preliminary evidence-based insights that may inform courtyard design practice, with the caveat that all participants were Japanese and stimuli reflected Western courtyard typologies; cross-cultural replication with more diverse samples is needed before broader generalisation.
The market for air purifiers has grown owing to the increasing attention on indoor air quality. However, consumers are annoyed by their noise characteristics. In this study, an algorithm for creating a masking sound was developed to effectively improve the subjective psychological and cognitive evaluation of the operating noise by validating through a jury test and reflecting individual noise sensitivity. Experiments were conducted to measure the operating noise of an air purifier under 16 conditions combining four levels of air purifier-to-microphone distance and four levels of wind strength. An algorithm for generating a masker sound was developed to adapt to the operating conditions. This masker sound was combined with the operating noise. To validate the effectiveness of this process, three combined sounds were selected and subjective evaluations were performed. The Weinstein Noise Sensitivity Scale (WNSS) questionnaire was also filled out by 25 subjects. The sound pressure level of the air purifier operating noise was low. However, it contained psychologically unpleasant tonal components that could be quantified based on the tonality. The algorithm developed for designing the masker sound was shown to be effective in reducing the tonality of the operating noise of the air purifier used in this study and the sound pressure level of the combined sound. The subjective evaluation and WNSS were used to calculate the response score and noise-sensitivity weighted response score, respectively, of the combined sounds of operating noise and masker sound. Through the jury test results, it was indicated that a tendency existed for the combined sound incorporating the proposed natural sound masker to receive higher comfort and satisfaction ratings. This was notwithstanding that its loudness increased marginally compared with the operating noise. The proposed approach shows potential for improving the subjective perception of air purifier operating noise while reducing tonality.
While the broad acoustic insulation has been considered as an important technique in underwater acoustics, there has been a challenge of achieving broadband underwater acoustic insulation with thin panels. For instance, conventional impedance-mismatch layers have suffered from Fabry-Perot transmission peaks, whereas locally resonant metamaterials generally provide attenuation only over a narrow frequency range. In this research, an elastic coiling-up metapanel for broadband underwater acoustic insulation based on Bragg-scattering-induced bandgap. Here, the proposed metapanel introduces a folded elastic load-transfer path inside a compact unit cell, where longitudinal, shear, and bending deformations are coupled to reduce the effective stiffness without increasing the physical thickness. A theoretical model based on a periodic diatomic mass-spring chain is developed to describe the Bragg bandgap formation and to estimate the band-edge frequencies from the equivalent masses and stiffnesses of the unit cell. Numerical simulations validate the predicted bandgap behavior, transmission reduction, hydrostatic-pressure response, and geometric design trends. For experimental validation, an ABS-based metapanel was fabricated and tested in an underwater acoustic tank. The measured insertion loss shows an attenuation-enhanced region that is consistent with the theoretical and numerical bandgap predictions. These results demonstrate that elastic coiling-up provides an effective design principle for compact broadband underwater acoustic insulation.
Traditional acoustic materials typically exhibit poor low-frequency sound insulation performance and weak dynamic tunability. To address this challenge, a light-tunable membrane-type acoustic metamaterial based on liquid crystal elastomer (LCE) is experimentally investigated in this study. The proposed metamaterial consists of a multi-cell LCE membrane, attached mass blocks, and a rigid supporting frame. Low-frequency sound insulation is achieved through local resonance of the membrane–mass units, while illumination provides a non-contact stimulus to modify the mechanical state of the LCE membrane and thereby modulate the sound transmission loss (STL) response. Impedance-tube measurements are conducted to characterize the STL response of the LCE metamaterial. The effects of illumination, RM257 excess ratios, and mechanical stretch ratios on the sound-insulation properties of the LCE metamaterial are experimentally investigated. The results show that the mechanical stretch ratio and RM257 excess ratio can serve as fabrication-stage parameters for tailoring the measured STL response. In addition, illumination induces non-contact modulation of the STL within the investigated frequency band, demonstrating the experimental feasibility of material-level tuning of low-frequency sound insulation in membrane-type acoustic metamaterials.
The transition from internal combustion engine vehicles (ICEVs) to electric vehicles (EVs) is a key component of urban decarbonization, yet its acoustic implications remain not fully captured by conventional noise assessment. This study presents a time-resolved noise mapping framework which is based on the CNOSSOS-EU guideline and driven by a microscopic traffic simulation, enabling the detailed analysis of transient noise events. The method is applied to the district of Gross Flottbek in Hamburg, Germany, to assess the impact of fleet electrification on noise pollution and human health. To model EV noise emissions, two recently proposed modifications of the CNOSSOS-EU emission model are implemented and compared. Besides the equivalent continuous night-time sound level LNight, the impact on wake-up reactions triggered by isolated pass-by events is studied. The results indicate that the overall impact on sleep quality from the transition to EVs remains moderate. The greatest impact on wake-up reactions is found along major and intermediate-traffic roads, while reductions of LNight are most pronounced in quieter residential areas. These findings demonstrate that time-resolved noise mapping is a valuable tool for a comprehensive evaluation of noise exposure, as it reveals health-relevant insights, such as the acute impact of transient events, which conventional average-based methods tend to miss.
The restorative benefits of urban bird songs are widely recognized, yet the mechanisms linking acoustic diversity to human perception remain unclear. This study investigates how bird song acoustic structures influence perceptual benefits in urban soundscapes, using voiceprint types as the analytical unit. In Shanghai, monitoring sites were established along a noise gradient. Through passive acoustic monitoring and species recognition, 21 common bird songs were identified and classified into five voiceprint types: SH, BVF, CF, BP, and FM. In a controlled laboratory experiment, 54 participants provided psychological ratings (liking, pleasantness, richness, comfort) and physiological measures (heart rate variability, respiration rate, skin conductance), enabling analysis of dose–response relationships between acoustic parameters and perceptual outcomes. Repeated-measures analyses showed that psychological evaluations differed across voiceprint types, with FM and SH calls receiving higher preference ratings and BVF calls receiving lower ratings PCA further showed that the first two components explained 55.62% of acoustic variance, separating frequency-related and temporal acoustic dimensions. Physiological responses showed descriptive, non-significant divergence from subjective preference, indicating that autonomic patterns should be interpreted cautiously. These findings link acoustic phenotypes to perception and provide preliminary, laboratory-based evidence for future health-oriented urban soundscape design.
Achieving efficient sound attenuation within limited space remains a common challenge in engineering noise control. While topology optimization can precisely tailor material distributions within a given design domain, its iterative nature is computationally expensive. To address this limitation, this study proposes a topology-optimization guided neural network framework for rapid inverse design of attenuation structures to achieve target transmission loss (TL). A fixed-size expansion chamber is used as a representative case study. First, a dataset consisting of internal topologies optimized for attenuation at different frequencies is generated, along with their corresponding TL spectra. A fully connected neural network is then trained to map the acoustic responses to their corresponding optimized topologies, thereby facilitating an efficient inverse design process. The trained model can accurately reconstruct these topologies, and verifications confirm that the predicted structures preserve the required attenuation characteristics. Comparisons with randomly generated structures show that the better performance mainly arises from the target-driven and physically meaningful training data, rather than from the neural network alone. Principal component analysis further shows that the optimized structures can be low-dimensionally represented and evolve continuously with the target frequency, while also identifying a local topology branch transition from a single-cavity-like configuration to a double-cavity-like configuration. Furthermore, sparse frequency training is conducted, and the results show that a reduced reference set can effectively generate attenuation structures for other frequencies. Finally, impedance tube experiments confirm good agreement with numerical predictions and demonstrate the practical effectiveness of the proposed framework for rapid design of passive noise control devices.
To address the issue that a single coiled-channel acoustic metamaterial exhibits discrete absorption peaks and pronounced inter-peak valleys, this paper proposes a composite sound-absorbing structure that couples a coiled structure (CS) having stepped arithmetically decreasing channels with porous material (PM). The acoustic response is predicted using a theoretical model combining thermo-viscous acoustic theory and the transfer matrix method, and its accuracy is confirmed through both numerical simulation and impedance tube measurements. The internal channel width of the CS narrows with an arithmetic decrement along the direction of sound propagation. This configuration extends the effective propagation path while simultaneously adjusting the thermo-viscous losses and the surface impedance within the channels. The resulting multi-order resonances can manipulate the impedance at the lower end of the target frequency band and in the locally weak absorption regions of the composite structure. Meanwhile, the PM supplies a continuous background of viscous and thermal dissipation, which fills the absorption valleys between adjacent resonance peaks. When the two are coupled, the composite structure yields a more balanced absorption spectrum. The parameters of structure are jointly optimized by a particle swarm optimization (PSO) algorithm. The optimized ADCPs with a total thickness of 62 mm (λ/23) achieve an average absorption coefficient exceeding 0.88 over the range of 383–2000 Hz. Their broadband response originates from the complementary interplay between the staggered multi-order resonances of the CS and the porous dissipation provided by the PM. This structure offers a novel solution for low-frequency noise control in compact spaces.
Outdoor environments are increasingly used for self-directed learning, yet little is known about how learners actively construct and adapt acoustic conditions to support cognitive work. While soundscape research has documented both the adverse effects of noise and the restorative benefits of sound, existing studies have largely examined learning in fixed or prescribed environments, offering limited insight into how individuals actively shape their acoustic surroundings. Drawing on semi-structured interviews with 51 university students, this study examines how learners construct soundscapes while studying outdoors. Using a grounded theory approach, the analysis identifies a soundscape optimization process through which learners anticipate, perceive, and modify acoustic conditions in relation to task demands, mood, and spatial affordances. The findings further reveal two mechanisms through which sound supports self-directed learning: the utility of sound, including masking distractions, priming focus, and setting pace; and the effects of sound, including fostering presence, security, and restoration during ongoing cognitive engagement. Notably, these mechanisms extend existing theory by identifying how outdoor soundscapes mask not only external noise but also internal mental distractions, and how restoration operates during ongoing cognitive engagement. Together, these findings reframe soundscape perception as an active construction process rather than passive exposure, positioning acoustic environments as infrastructure that learners dynamically optimize in relation to their goals and needs.
The soundscape framework defined in ISO 12913 relies on survey-based assessments to characterise human perception through Perceptual Attributes (PA). While these attributes have been validated across multiple languages via the Soundscape Attribute Translation Project (SATP), these results were obtained through surveys, not through neurophysiological readings. This exploratory study addresses this potential neurophysiology of PA, specifically the relationship between Spanish SATP descriptors (language) and objective neurophysiological responses to sound recorded via electroencephalography (EEG).Twelve participants were exposed to six contrasting audio excerpts from the SATP database, representing extreme positions within the pleasantness–eventfulness space. Subjective perception was recorded using the Spanish SATP protocol, while objective brain activity was measured through multi-channel EEG to compute Power Spectral Density (PSD) across the Delta, Theta, Alpha, and Beta bands. Given the multidimensional nature of the data, Self-Organising Maps (SOM) were employed to determine if these relationships could be organised into a neurophysiology of these Perceptual Attributes (PA) linking sound and language as perception. The results explore the relationship between soundscape as defined by ISO 12913 and the multidimensional modulation of brain rhythms, primarily in the Alpha and Delta bands.Calm, non-anthropogenic sounds (e.g., birdsong) were associated with increased Alpha-band power, a marker for relaxation and comfort, whereas eventful or chaotic sounds elicited stronger Delta-band activity, consistent with pre-attentive tracking responses. SOM analysis revealed three well-defined clusters (Calm, Annoying, and Vibrant) that closely align with the PA constructs and mirror the spatial organisation of the ISO 12913 circumplex model.This study shows that data obtained via EEG capture perceptual structures consistent with translated PA attributes. By framing these findings within the SATP framework, this methodology can provide a reproducible protocol for future cross-linguistic harmonisation and evidence-based urban acoustic design.
For the measurement of sound transmission loss of double walls, acoustical measurements in accordance with international standards have been widely adopted, although sound transmission loss is closely related to the vibrations of each single panel separated by an air cavity. However, direct estimation of sound transmission loss from wall vibration measurements, without any acoustic instrumentation, offers a practical and deployable alternative. In this paper, sound transmission loss of double walls consisting of two panels separated by an air cavity and not mechanically connected to each other is theoretically formulated as a function of wall vibrations, based on a simple prediction model. A step-by-step derivation of the wall vibration ratio is presented, with detailed physical interpretation. The proposed formulation is validated experimentally on a representative ship partition panel in accordance with ISO 10140–2:2021. The estimated sound transmission loss from vibration measurements agrees well with acoustic measurements across the measured 1/3-octave bands up to 800 Hz. Numerical validation using the transfer matrix method across six additional panel and cavity configurations further confirms that this accuracy is not specific to the tested specimen. A statistical analysis of measurement point selection shows that one accelerometer per 1 m2 is sufficient to achieve a standard deviation below 1 dB, providing a practical deployment guideline.
To address the issue of incipient valve internal leakage faults, this study proposes an acoustic emission(AE) detection method based on multi-representation image encoding and a lightweight convolutional network. Time-frequency domain features are first extracted, and PCA was compared with three nonlinear KPCA kernels, including Gaussian, polynomial, and sigmoid kernels. PCA was selected as the final dimensionality-reduction method because it provided the best balance between variance retention, classification performance, and robustness under the present small-sample setting. The reduced feature sequences are subsequently encoded into two-dimensional images using the gramian angular summation field (GASF) to capture temporal dependencies. Classification is performed using MobileNetV3, with improved generalization capability enhanced through Dropout regularization, data augmentation, and AdamW optimization, achieving a 5% accuracy improvement compared to baseline methods on the same dataset. Experimental results demonstrate that the proposed four-class model achieved a leakage recognition accuracy of 94.96% on the mixed validation dataset when the three leakage-severity classes were merged into a single leakage category. On the independent cross-time test set, the corresponding leakage recognition accuracy reached 96.98%, suggesting potential temporal robustness within the studied operating range. Model interpretability is further validated through Grad-CAM and t-SNE visualizations. By balancing detection accuracy and computational efficiency, this method provides preliminary methodological support for developing portable intelligent diagnostic systems.
Exposure to background speech is known to impair working memory performance, a concern particularly relevant in open-plan office environments. This study aimed to examine how specific characteristics of speech, including semantic content and sound pressure level, affect cognitive performance and perceived workload. Controlled laboratory listening experiments were conducted within a common experimental framework, allowing different acoustic characteristics of speech to be compared directly. Participants completed Operation Span tasks (OSPAN) while being exposed to binaural speech signals. Perceived workload was assessed using the NASA Task Load Index (NASA-TLX). In Listening Experiment A, the role of semantic content was investigated using conditions consisting of intelligible speech, locally time-reversed speech, continuous noise, and silence. Listening Experiments B1 and B2 focused on the effects of varying sound pressure levels in speech conditions. Overall, exposure to speech resulted in decreased cognitive performance relative to silence, whereas continuous noise and low-level speech yielded performance outcomes comparable to silent conditions. No significant effect of semantic content was observed in Listening Experiment A. In Listening Experiments B1 and B2, performance declined as sound levels increased, with diminishing effects observed at lower levels. Subjective workload ratings varied across conditions, even where objective performance differences were not detected. These findings suggest that background speech introduces a cognitive cost that may be compensated for in short term through increased mental effort, potentially contributing to long-term fatigue.
