Advances in Noise and Vibrations for Machines

Vibration analysis and monitoring are currently required in various fields of industry, from automotive and aeronautics to manufacturing and quality control, and from machining and maintenance to civil engineering. They are also a key aspect in the case of validating Finite Element Analysis (FEA) models and prototypes in cutting-edge fields of aerospace technologies and technologies for future accelerators and detectors.

Machine vibrations are a normal and typically unavoidable result of moving and rotating parts or the effects of the ground-motion transfer on an object. To mitigate the negative effects of vibration, in many cases, they need to be assessed and monitored. Vibrations, as well as noise measurements, are also a good indication of systems’ behavior or the degrading quality of the product in the case of manufacturing. Thus, they are a perfect way to perform systems and machines diagnostics [1] or quality control processes for the products [2,3,4]. Vibration, especially modal analysis of structures and machines, is also a good way to validate a finite element model [5,6].

The aspects of measurement, assessment, and mitigation of noise and vibration are wide in range. In the case of industrial application, the prominent usage is in the field of diagnostics of the machines. Vibration analysis has become one of the most widely applied techniques in machine diagnostics due to its sensitivity to any, even very small, changes in dynamic behavior [7]. Mechanical components such as bearings, gears, and shafts generate characteristic vibration signatures under normal operating conditions, which can be used as a reference for detecting any deviations from normal operation [5,8]. Any potential deviations from these patterns often indicate the presence of sometimes serious faults such as imbalance, misalignment, wear, or, in worst-case scenarios, cracks and other critical problems [9]. The systematic, real-time measurements and interpretation of vibration signals enable early fault detection, thereby supporting predictive maintenance strategies, which are also a key issue in Industry 4.0. This approach minimizes unplanned downtime (cost reduction), extends machine lifetime (alignment with green strategies), and enhances overall operational safety (social impact) [10]. As a non-invasive method, e.g., using laser vibrometers or microphones, vibration-based diagnostics provides continuous monitoring without interrupting machine performance, making it a practical and effective tool in industrial applications [11,12]. Those methods, due to their flexibility and wide range of applications, are also typically used for structural health monitoring and structural dynamics. They provide sensitive indicators of material degradation and damage evolution when needed [13,14]. Acoustic emissions and vibration responses reveal changes in structures’ behavior, such as stiffness, cracks, delamination of composites [15], or fatigue, well before visible signs occur. By analyzing frequency shifts, damping variations, or transient acoustic events, it is possible to detect, localize, and assess the severity of defects in real time [16]. These techniques enable continuous, non-invasive monitoring of infrastructure and mechanical systems [17], supporting preventive maintenance and improving safety and reliability across civil [18], aerospace [19], and mechanical engineering applications [20,21]. Moreover, similar methods are used to monitor and evaluate vibration impact on new and existing buildings, tunnels, and, in the case of underground facilities, experimental chambers [22]. Furthermore, the effects, both in real time [23] and statistically over more extended periods [24,25], of using different heavy-duty equipment or the impact of natural seismic events on equipment or structures can be calculated in an easy way. Another aspect worth mentioning is vibrations and dynamic stability of structural elements of machines, such as plates and beams, which can be described, modeled, or tested with vibration-based techniques [26,27]. Another important issue is the use of vibration for accurate identification of system dynamics. This is often connected to predicting the dynamic systems [28] and modeling of passive or active vibration suppression systems [29,30]. When considering moving or transporting structural elements, machines, or sensitive devices, that are part of a vibration suppression system, the aspect of performing analysis of permissible vibration levels is an important one [31]. Depending on the means of transportation, timeframe, and access to power, among other factors, the choice of measurement equipment is the most problematic element [32]. This topic is closely connected to road traffic-induced vibration and noise, which are a modern type of contaminant of urban environments [33,34].

Those are just some of the examples of research areas. Independent of the application, the crucial practical consideration should focus on proper signal processing and the choice of measurement equipment (both the excitation source, if needed, and the sensors). Typically, modal hammers and electrodynamic shakers are used to input the energy to the system and accelerometers, seismic sensors, or geophones to measure the response. This approach is not possible if measurement is performed on light and ultralight structures and micromachines for aerospace and future detector technology applications [35]. For fragile low-mass objects, a fully non-contact approach must be used, such as using 2D or 3D laser Doppler vibrometers [6].

Taking into account the above examples, this Special Issue, “Advances in Noises and Vibrations for Machines”, brings together diverse studies addressing those topics. Especially the topics of noise and vibration challenges in modern machinery are explored: from electrical motors and machine tools to barrier design and system diagnostics. Each contribution advances the field with innovative methodologies, rigorous modeling, and real-world validation, showing that this is valid and applicable field of study.

The following authors, from well-known scientific centers, have contributed to those issues.

Moien Masoumi and Berker Bilgin have contributed with the paper “Experimental Acoustic Noise and Sound Quality Characterisation of a Switched Reluctance Motor Drive with Hysteresis and PWM Current Control” [36]. They evaluate the acoustic emissions of a switched reluctance motor under PWM control. The study correlates motor operating regimes with sound quality metrics and identifies design parameters that influence the performance. Their findings provide valuable insights into the contributions of these elements to acoustic noise characteristics and offer a foundation for improving the motor’s acoustic behavior during most of the life cycle stages (from design to control stages).

Shih Ming Wang et al. have contributed with the paper “Optimization of Machining Parameters for Reducing Drum Shape Error Phenomenon in Wire Electrical Discharge Machining Processes” [37]. This study addresses the “drum shape error” that occurs in Wire Electrical Discharge Machining (WEDM) of thick workpieces, caused by wire vibration and debris accumulation. Experiments were conducted to evaluate the influence of machining parameters such as open voltage, pulse ON/OFF time, and servo voltage on both shape error and material removal rate (MRR). Full factorial and Taguchi methods were applied to analyze parameter effects, and regression models were developed to build optimization rules. A C#-based human–machine interface was created to recommend optimal parameter settings. Verification experiments confirmed prediction accuracies of 94.7% for shape error and 99.2% for MRR, with drum shape error reduced by nearly 40%. Results showed that while optimization effectively minimized error, MRR decreased significantly, highlighting the trade-off between precision and efficiency. The findings provide a practical system for process optimization and suggest that future work should examine material variation, workpiece thickness, and flushing pressure adjustments to further mitigate error.

Xinyu Li, Riliang Liu, and Zhiying Zhu presented the topic of “Data Acquisition and Chatter Recognition Based on Multi Sensor Signal Processing in Machining” [38]. This study, among others, examines chatter behavior in blade whirling milling, which is an efficient process for blade manufacturing. Multisensor signals, including vibration and acoustic emissions, were collected during both roughing and finishing operations. Time-, frequency-, and time–frequency-domain features were extracted and reduced using principal component analysis (PCA) to ensure relevant information was retained. These features were then applied to an MLGRU-based chatter recognition model enhanced with a self-attention mechanism. Experimental validation, presented by the authors, showed an average recognition accuracy of more than 89% with a response time below 0.4 s. This confirms the model’s reliability and timeliness. Moreover, the results also revealed that blade edge regions, especially during roughing, are most susceptible to chatter. This is due to reduced stiffness. The approach provides a solid basis for real-time chatter monitoring and future suppression strategies to improve machining stability.

Hao Wu, Lingshan He, Ziyu Tao, Duo Zhang, and Yunke Luo contributed with a study titled “An Integrated SEA–Deep Learning Approach for the Optimal Geometry Performance of Noise Barrier” [39], which presents a topic of urban rail transit systems, which are increasingly challenged by environmental noise pollution, particularly along densely populated corridors. The presented study evaluates the performance of vertical and fully enclosed noise barriers using a multiaspect combination of field measurements, statistical energy analysis, and optimization techniques. Train pass-by noise data revealed strong near-track amplification, most pronounced at 1 m from the barrier and 2 m above the rail, with increases up to 1.47 dB for VB and 4.13 dB for FB in the 400–2000 Hz range. To address this issue, micro-perforated panels (MPPs) were introduced and optimized using Maa’s theory, effectively eliminating amplification and achieving 4–8 dB reductions at sensitive locations. A neural network surrogate model (R² = 0.9094, MSE = 0.8711) was further developed to guide barrier geometric design, demonstrating that barrier height and installation distance jointly govern insertion loss, with optimal configurations achieving up to 12 dB. The combined use of MPP absorbers and surrogate optimization provides a practical, high-performance strategy for urban rail noise mitigation, with future work directed toward integrating cost and sustainability analyses.

Kamran Foroutan and Farshid Torabi submitted the paper “Nonlinear Vibration of Oblique-Stiffened Multilayer Functionally Graded Cylindrical Shells Under External Excitation with Internal and Superharmonic Resonances” [40]. This study develops a semi-analytical method to analyze nonlinear vibrations of oblique-stiffened multilayer functionally graded cylindrical shells under external excitation. The influence of material gradation, geometry, and other properties on vibration behavior is systematically evaluated. Findings of the authors can support the design and optimization of aerospace and automotive structures by, e.g., improving resistance to dynamic loads, reducing weight, and enhancing durability, among others. The approach enables accurate vibration prediction across varied potential materials and configurations, offering safer, more efficient, and application-specific engineering solutions and future designs.

The work “Determining Vibration Characteristics and FE Model Updating of Friction-Welded Beams” [41] by Murat Şen examines the vibration behavior of friction-welded SAE 304 steel beams produced at rotational speeds of 1200, 1500, and 1800 rpm. Using Experimental Modal Analysis, the first five bending mode frequencies, damping ratios, and mode shapes were identified. Results showed a slight reduction in natural frequencies with increasing welding speed, with the largest variation of 3.28% in the first mode. Damping ratios exhibited mode-dependent trends, while mode shapes remained highly consistent across speeds, with similarities above 93.8%. The findings demonstrate how welding speed affects dynamic properties and provide a validated modeling framework for predicting and optimizing the vibration performance of welded structures.

The article “End-of-Line Quality Control Based on Mel-Frequency Spectrogram Analysis and Deep Learning” [42] by Jernej Mlinarič, Boštjan Pregelj, and Gregor Dolanc presents a deep learning framework for end-of-line inspection of brushless DC motors, integrating Mel-frequency spectrograms with a CNN–BiGRU architecture. Vibration and acoustic signals, recorded directly during production tests, were converted into spectrograms without pre-processing. A six-step data reduction strategy improved feature relevance, and Bayesian optimization with oversampling addressed dataset imbalance. Moreover, the fault detection extended beyond conventional threshold-based methods enables early identification of defects, which are typically detectable only in later testing stages. Overall, the presented method reduces reliance on typical manual inspection, enhances consistency, and improves efficiency in industrial diagnostics and inspections.

The work [43] by Guoyin Mo, Chengyu Liu, Guimian Liu, and Fuhao Liu titled “Improved Nonlinear Dynamic Model of Helical Gears Considering Frictional Excitation and Fractal Effects in Backlash” develops a nonlinear dynamic model of helical cylindrical gear systems under flexible support, incorporating the coupled effects of sliding friction, backlash, and fractal surface roughness. Using the Lagrange method and Runge–Kutta solutions, the model examines how driving speed, surface roughness, and fractal dimension influence dynamic responses. Results show that, compared with traditional models, the proposed 7-DOF formulation more accurately reflects gear behavior in highly flexible support conditions. The findings of the authors provide theoretical guidance for vibration suppression systems and noise reduction in helical gears.

The paper “Influence of Distributor Structure on Through-Sea Valve Vibration Characteristics and Improvement Design” [44] by Qingchao Yang, Zebin Li, Aimin Diao, and Zhaozhao Ma proposes a structural improvement to the electromagnetic hydraulic distributor in order to suppress transient noise generated during the opening and closing of sea valves. Performing theoretical calculations and simulations, the flow characteristics of the distributor were obtained and analyzed. A buffering slot was introduced into the flow port of the valve sleeve but without altering the stable working flow rate. Furthermore, a triangular slot design was selected to mitigate sharp pressure fluctuations and extend the distributor’s response time. Authors have performed the prototype testing, which confirmed that the modification significantly reduces transient noise, particularly under higher operating pressures, while preserving operational flow performance. These findings demonstrate that slotting the flow hole can effectively stabilize pressure changes during valve actuation and also enhance acoustic performance. Apart from the sea valves, the proposed design strategy offers a practical framework for improving noise control in other electro-hydraulic valve systems, providing both theoretical insight and engineering guidance for marine applications.

Finally, the work [45] by Cunrui Shen and Chihua Lu, titled “A Method to Obtain Frequency Response Functions of Operating Mechanical Systems Based on Experimental Modal Analysis and Operational Modal Analysis”, introduces a hybrid approach for identifying operational frequency response functions (OFRFs), addressing the limitations of operational modal analysis (OMA), which cannot provide mass-normalized mode shapes. The method supplements OMA with experimental modal analysis (EMA) conducted in the static state. Operational mode shapes are classified as either unchanged or altered relative to static conditions, with scaling performed via least squares for the former and the principal square root method for the latter. EMA data supply condensed mass information needed for normalization. Special attention is given to ensuring numerical stability. This includes strategies to mitigate matrix inversion issues. A methodical approach to validation through simulation and measurement demonstrates the method’s robustness against noise, incomplete data, and dynamic disturbances. Finally, results confirm that the approach reliably reconstructs operational FRFs with good accuracy. Limitations remain in cases involving high damping, high-frequency modes, or structural pose changes, which suggest directions for future refinement.

The contributions collected in this Special Issues share the following unifying motifs:

• Insight into Vibration Sources: from electromagnetic forces in motors to chatter dynamics in machining, the studies elucidate root causes and modeling strategies and algorithms.
• Advanced Signal Processing and Machine Learning: deploying autoencoders, Kriging models, and deep learning for predictive fault detection and design optimization.
• Noise Mitigation via Design Optimization: contributions demonstrate how parameter tuning—notably barrier geometry or machining conditions—can suppress unwanted vibrations and acoustic emissions.
• Bridging Simulation and Experimentation: all articles validate their computational solutions through experimental data, reinforcing practical significance.

This collection reflects a still-developing research landscape where data-driven diagnostics, AI-enhanced modeling, and advanced simulation frameworks converge to deliver quieter and more reliable machines or predict problems before they occur. The impact spans industrial automation, transportation, energy systems, and tool design. We anticipate these studies will inspire cross-disciplinary collaboration and inform future machine-design standards focused on noise and vibration control.

We express our gratitude to all authors, reviewers, and the editorial team whose efforts have made this Special Issue, “Advances in Noises and Vibrations for Machines”, possible.