The variation of the open-cell polyurethane foam thickness within the seating system can affect the dynamic response of the occupant-seat system and the riding discomfort. This study was aimed to develop and optimize a finite element model of the occupant-seat system incorporating the skeleton and muscle tissue for predicting the seat transmissibility and assessing the riding discomfort with different foam thicknesses. The key model parameters influencing the prediction of the seat transmissibility were investigated with the sensitivity analysis, and the correlation between the parameters and the seat transmissibilities was quantified by fitting the polynomial functions. The genetic algorithm was also utilized to iterate the squared error between the predicted and the measured seat transmissibilities, thereby obtaining the optimal parameter values. The best fit of the vertical inline and fore-and-aft cross-axis seat transmissibility predicted by the optimized model increased by 13.34 % and 14.41 %, respectively. When the dynamic stiffness of the foam decreased with the increase of the thickness, the peak frequency of seat transmissibilities, the weighted root-mean-square acceleration and the SEAT value exhibited the same downward trend.
For the current rolling bearing rolling degradation trend prediction problem, its general use of a single timefrequency domain features as the degradation characteristics, the effect is often not satisfactory. To improve the prediction accuracy, a rolling bearing degradation trend prediction method based on multi-scale Boltzmann-Shannon interaction entropy (abbreviated as MBSIE) and Transformer-TCN is proposed. The original rolling bearing signals are decomposed and noise-reduced by EEMD, the filtered IMF components are used for signal reconstruction, and the degradation indicators are constructed by fusing the filtered time-frequency domain features with MBSIE, the Transformer-TCN neural network is introduced for prediction. The data from PHM2012 is used for instance analysis. The results show that the prediction accuracy of the proposed method reaches 94%, which is better than the traditional entropy such as arrangement entropy and sample entropy as the degradation features, proving the feasibility and relative superiority of this paper in predicting the bearing degradation trend.
This study investigates the application of acoustic emission tomography (AET) using the algebraic reconstruction technique (ART) algorithm for damage detection and visualization in homogeneous and multilayered materials. The research employed a systematic four-stage experimental approach to evaluate how sensor configuration, acoustic signal (ray) coverage, and source signal frequency affect reconstruction accuracy. The first stage examined signal propagation along the surface of a homogeneous aluminum specimen. The second stage analyzed signals propagating through the material volume to provide comprehensive assessment of internal damage. The third stage extended this acoustic emission (AE) tomographic approach to multi-layered structures, revealing additional challenges including impedance mismatches and wave distortions that increased reconstruction errors. The final stage addressed those limitations by implementing higher-frequency source signals, which reduced reconstruction errors and improved damage detection precision due to shorter wavelengths providing finer resolution and greater sensitivity to the defects. The findings demonstrate the effectiveness of AET in identifying damage across diverse material and sensor configurations, ray coverage, and highlight its adaptability as a non-destructive evaluation (NDE) technique. This research provides elaborate and valuable insights into optimizing AET methodologies for structural health monitoring (SHM) and enhancing its applicability to complex material systems using ART algorithm.
Accurate impact localization remains challenging for structural health monitoring (SHM) systems due to wave propagation complexities and high measurement uncertainties. This study presents a novel baseline-free framework that combines simulated annealing (SA) and genetic algorithm (GA) to characterize low-velocity impacts in plate structures. The proposed framework provides robust impact localization without relying on baseline signals, material properties, or detailed geometric parameters. Experimental impact tests were conducted using a coarse network of piezoelectric sensors arranged in multiple configurations to evaluate the performance of the proposed framework for impact characterization, full-field group velocity mapping, and computational efficiency. The framework predicted impact locations with an average relative error of less than 5% and R-values of about 0.98 under the optimal sensor configuration. The analysis of sensor placement indicated asymmetric sensor layouts yield 50% higher performance compared to symmetric layouts. The combined SA-GA framework successfully identified various impacts both inside and outside the regions confined by active sensor network with probabilities exceeding 90%. Additionally, the results demonstrated that the proposed framework consistently outperformed standalone SA and GA algorithms in both impact identification accuracy and the consistency of the velocity field.
In this study, the ungrounded and grounded virtual-inerter-based dynamic vibration absorbers (VIDVAs) are proposed by using an inertial actuator with relative or absolute acceleration feedback. The relative acceleration feedback signal is created by using two accelerometers (on the proof-mass and the base of the inertial actuator), whereas the absolute acceleration feedback is generated by an accelerometer on the proof-mass alone. The natural frequencies of the proposed VIDVAs can be tuned by using a virtual inerter without changing its physical design. The tunable VIDVAs in this study aim to reduce resonant vibrations as well as vibrations at a single forcing frequency. The tuning capabilities and control performances of the proposed VIDVAs are investigated numerically and experimentally. The results show that both VIDVAs can reduce resonant vibration significantly, while the grounded VIDVA exhibits broader mode splitting. The experimental results also show that a high level of vibration reduction was achieved by using grounded VIDVA in a wide range of forcing frequencies. Compared to ungrounded VIDVA, the study indicates that the grounded VIDVA is easier to implement and can achieve better control performance.
This study develops a weighted cost function-based optimization framework for tuning the time delay parameter in a nonlinear vehicle suspension with time-delayed acceleration feedback control. A quarter-vehicle model is first established to characterize suspension dynamics. The steady-state response is derived analytically using the harmonic balance method, while a stability analysis of the linearized system delineates the feasible ranges of time delay and feedback gain coefficient. A composite cost function is then formulated by integrating weighted contributions from the sprung mass acceleration, suspension dynamic deflection, and tire dynamic load. Optimization of the time delay parameter is carried out for both positive and negative feedback control configurations. A comparative performance evaluation between constant and frequency-dependent time delay strategies is conducted across the human-sensitive frequency band (4-8 Hz). Numerical simulations validate the theoretical predictions, demonstrating average reductions of 45.97% in the cost function, 35.54% in the sprung mass acceleration, and 30.65% in the tire dynamic load. Although a slight increase in the suspension dynamic deflection is observed, the results confirm that optimized time delay can simultaneously improve ride comfort and driving safety.
This paper proposes a computational model for analyzing an acoustic scattering from an underwater elastic shell structure constrained by acoustic boundaries. The method derives the underwater acoustic field expression using the wave superposition method (WSM). It incorporates the image source principle to establish equivalent source configurations for acoustic boundary modeling, and couples the shell's vibrational response with the acoustic field through acoustic-structure interaction conditions on the shell surface. The dynamic response of the shell under acoustic excitation is obtained via the precise transfer matrix method. The proposed approach is not constrained by the geometry of the object or the frequency band of interest, while effectively avoiding the singularity issues inherent in conventional WSM. The method's accuracy has been verified through comparisons with numerical solutions. Furthermore, this work presents an experimental study of underwater target scattering under laboratory conditions, introducing a systematic experimental methodology and data processing protocol specifically designed for theoretical validation. Additionally, a detailed analysis of acoustic field patterns at varying immersion depths provides further insights into the effects of acoustic boundaries.
Variational Mode Decomposition (VMD) requires the manual setting of decomposition algorithm parameters, leading to low fault identification accuracy. To address this issue, this paper proposes a fault diagnosis method based on the Whale Optimisation Algorithm (WOA)-optimized VMD, combined with Multi-scale Permutation Entropy (MPE) and Support Vector Machine (SVM). The specific steps are as follows: Firstly, the Whale Optimisation Algorithm is introduced to optimize the number of modal decompositions (K) and the quadratic penalty factor (alpha) of Variational Mode Decomposition, obtaining the optimal VMD parameters. After decomposing the vibration signal, the multi-scale permutation entropy of each Intrinsic Mode Function (IMF) component is calculated separately. Finally, the extracted feature data are input into the SVM classifier, and the Whale Optimisation Algorithm is used to optimize the parameters of SVM to obtain the WOA-SVM optimization model. Meanwhile, ten-fold cross-validation is introduced to train the WOA-SVM, ultimately achieving accurate classification and diagnosis of fault signals. Analyses are conducted using datasets from the Bearing Data Centers of Jiangnan University and the University of Western Ontario (Canada) respectively. The results show that the fault identification accuracy of this method reaches 98.36% and 99.13%, respectively, and its performance is superior to that of methods such as VMD-PE, EMD-MPE, VMD-SVM, and VMD-MPE.
The rapid growth of electric-vehicle (EV) technology makes in-cabin vibration and noise control critical to ride comfort. The high operating frequency of traction motors, combined with the absence of a conventional engine block, intensifies mid-frequency (approximate to 500-2000 Hz) vibro-acoustic coupling. Lightweight body structures further complicate control and call for more accurate hybrid finite-element/statistical-energy-analysis (FE-SEA) models. This work integrates an edge-based smoothed finite-element technique (ES-FEM) into a hybrid FE-SEA framework (ES-FE-SEA) to improve boundary continuity and smoothing-domain construction, thereby increasing the accuracy and stability of mid-frequency acoustic-vibration predictions. A reduced-scale cabin model is built in VA One to conduct parametric studies on mesh density, element aspect ratio, mesh type, and damping-loss factor; predictions are assessed with Monte-Carlo analyses. Results show that mesh density strongly affects vibration response, smaller aspect ratios shift peak energy toward higher frequencies, and damping substantially changes both energy levels and peak velocity. The ES-FE-SEA model consistently improves mid-frequency predictions over a conventional FE-SEA formulation and is suitable for acoustic-package optimization and EV-cabin NVH management. Future work will validate model parameters experimentally and extend the framework to multi-physics coupling for broader engineering applicability.
Two-wheeled gyroscopic robots are challenging to balance due to their intrinsic instability and sensitivity to external disturbances. This study proposes a hybrid control approach that combines a linear quadratic regulator (LQR) and a regression-based machine learning model to increase stability and adaptability. The regression model is trained with data collected from an LQR-controlled system to predict the optimal control responses under different conditions. The proposed approach utilizes state-space representation and the Jacobian linearization technique to design and simulate the robot's performance under both controlled and random disturbance profiles. Simulation results indicate that the proposed method ensures steady motion even in dynamic, uncertain environments and increases resistance to external disturbances. Compared to the conventional LQR control method, the regression-enhanced model dynamically adjusts the control parameters, effectively reducing oscillations throughout the robot's motion and achieving an average energy saving of approximately 80 %. These results demonstrate that robust, adaptive control in autonomous two-wheeled robotic systems can be achieved by fusing data-driven models with classical control techniques.
Bearings in electric locomotives serve as core components of the transmission system. Their periodic impulsive fault signals are easily overwhelmed by intense noise, rendering traditional methods ineffective at extracting weak fault features under low signal-to-noise ratio (SNR) conditions, thereby limiting early diagnostic capabilities. To address this challenge, this paper proposes a bearing fault detection method that integrates the dq transformation, kurtosis statistics, and the Duffing chaotic oscillator: the dq transformation separates the periodic impulsive fault components, kurtosis statistics adaptively suppress noise, and the Duffing system identifies weak signal features. Comparative analyses using simulations and experiments demonstrate that the proposed method achieves higher fault detection accuracy in low-SNR environments, along with superior timeliness and stability in feature extraction.
Modern vehicle suspension systems play a crucial role in ensuring ride comfort, road handling, and overall vehicle stability. Active suspension systems, unlike passive ones, utilize control strategies to dynamically adjust suspension forces, improving vehicle performance under varying road conditions in real time. This research paper focuses on the analytical modeling and performance analysis of a Fuzzy Logic Controller (FLC) and a Proportional-Integral-Derivative (PID) controller for a full-car active suspension system using a 7-degree of freedom (DOF) model. The developed model considers the vertical motion of the vehicle body, pitch, and roll dynamics, along with the suspension dynamics of all four wheels. A hydraulic actuator is incorporated at each suspension to provide active control forces. The control strategies are designed to minimize vehicle body displacement and acceleration, thereby improving ride comfort and stability. The FLC is developed using fuzzy inference rules and triangular membership functions, while the PID controller follows conventional control logic. Simulation results indicate that the FLC significantly reduces suspension travel, body displacement, and acceleration when compared to both the passive and PID-controlled suspensions. The FLC achieves superior performance by mitigating pitch and roll motions, which are crucial for overall ride quality. Root mean square (RMS) values of suspension parameters confirm that the proposed fuzzy controller enhances vehicle dynamics, providing better vibration isolation and improved ride comfort. The study demonstrates the effectiveness of the FLC in handling nonlinearities in active suspension systems, making it a promising approach for real-time automotive applications.
The transient statistical energy analysis (TSEA) method is suitable for solving high-frequency structural dynamic response problems, and has been widely used in aerospace and marine fields. However, for complex structures such as hull shells and multi-compartment submersibles, there are problems such as a cumbersome computational process and huge computational volume. This paper proposes the Iterative TSEA (ITSEA) method, which utilizes an iterative approach to solve the energy balance equation combined with initial conditions. Validation shows that compared to TSEA, the ITSEA algorithm saves 40% of computational parameters and processes, while the relative error of the energy decay curve after 20 iterations is no more than 2.1%. ITSEA offers significant advantages for large-scale shock response calculations for complex structures, and for structures with more than 30 subsystems, the improvement in computational efficiency becomes significant.
The elastic wave bandgap characteristics of origami-based periodic strip-like waveguide structures are studied computationally. The computational models of waveguide structures are constructed by unidirectionally connecting unit cells composed of elastic plates according to Miura-ori or Jabara-ori (accordion fold) patterns, with the crease lines replaced by elastic interphase regions of finite width and thickness. The dispersion relations of these periodic structures are obtained by the finite element analysis based on the plate theory, and the elastic wave bandgaps lying at relatively low frequencies are identified. The results show that the location and width of the bandgaps depend significantly on the folding state of the structures, confirming their potential applicability as tunable acoustic metamaterials. The folding-state dependence of the bandgaps is shown for different parameters of the crease interphase regions and is discussed in light of unit-cell vibration modes at the bandgap edges. Furthermore, the wave transmission characteristics in the structures consisting of a finite number of unit cells are examined in relation to the bandgaps and the vibration modes of the corresponding periodic structures.
To identify the eigenfrequencies of a time-varying mass system and investigate the effect of mass change on the eigenfrequency, the experimental modal tests under different masses and the operational modal tests under time-varying mass are conducted. Recognizing that the structures within distinct time periods remain time-invariant, a short-time enhanced frequency domain decomposition (ST-EFDD) algorithm is proposed. The eigensystem realization algorithm (ERA) method is employed for the identification of eigenfrequencies in beams subjected to added masses. Concurrently, the ST-EFDD algorithm is employed to discern the eigenfrequencies of beam experiencing time-varying mass. The eigenfrequency identification results obtained from various methods are compared in this study. The findings indicate that the ERA method identifies the second-order eigenfrequency. Additionally, this study explores the distribution intervals of the second-order eigenfrequency identified by the ERA method under different excitations, including rubber head, nylon head, and aluminum head. The ST-EFDD algorithm is capable of identifying the third-orders eigenfrequency. This study investigates the distribution intervals of the eigenfrequency for each order under both flowing sand excitation and flowing sand conditions. The linear fitting of the ST-EFDD identification result reveals means squared deviations of 1.076, 28.268, and 39.508 under flowing sand excitation. Similarly, for flowing sand with exciter co-excitation, the corresponding values are 5.714, 8.308 and 22.712. Therefore, the ST-EFDD method can identify more eigenfrequencies and is more suitable for vibration systems with wide frequency and large excitation.
The ongoing electrification of the automotive sector and the increasing reliance on virtual prototyping pose new challenges for the accurate prediction of structural dynamics in electric machines. While digital twins are advancing, experimental validation remains indispensable due to the complexity and cost of physical testing. Among the most critical environmental influences is temperature, which significantly affects electromagnetic losses, material properties, and dynamic behavior. This study presents a comparative experimental modal analysis of laminated stators-both unwound and wound-under controlled thermal conditions ranging from 25 degrees C to 120 degrees C. The investigation reveals that modes with axial nodal lines and torsional characteristics are particularly sensitive to temperature-induced stiffness degradation and viscoelastic damping. In contrast, radially dominated modes exhibit greater thermal stability. A key contribution of this work is the experimental identification of temperature-induced mode fusion phenomena, characterized by frequency convergence and increased modal interaction. These effects are especially pronounced in wound stators due to the combined influence of winding mass, anisotropic stiffness, and resin softening. The findings underscore the necessity of incorporating temperature-dependent, anisotropic material models into simulation frameworks to ensure accurate prediction of modal behavior in electric machines. Future research should explore alternative stacking and winding techniques to further enhance thermal robustness.
The underwater electrostatic hydraulic actuator (EHA) has high precision, fast response time, and low noise level, etc. It is widely used for ocean exploration and deep-sea operations. However, the tough underwater conditions could cause considerable problems, particularly with the hydraulic system of EHA, which is prone to leakage failures. To address this issue, this paper presents a methodology for diagnosing hydraulic cylinder leakage faults based on Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ICEEMDAN) combined with CNN, which has been optimized by using the improved Dung Beetle Optimization (IDBO). By studying the various properties of leakage signals from underwater EHA hydraulic cylinders, the denoised output from the ICEEMDAN approach serves as the input vector for the CNN model. The hyperparameters of the Convolutional Neural Network (CNN) model are fine-tuned and optimized by using the enhanced IDBO algorithm. The testing results reveal that the hydraulic cylinder fault diagnosis model combined with the ICEEMDAN-CNN optimized by the IDBO has an accurate rate of 97.73%; and it has a much superior diagnostic impact than other models. This strategy can significantly increase the accuracy of the underwater EHA defect diagnosis model.
This paper aims to mitigate the transmission of the vehicle structure-borne noise. An analytical model of the suspension vibration system is established and equations for force and velocity between subsystems are derived. A finite element model of the automobile chassis suspension is developed, which is further calibrated through modal testing. The power flow analysis method is employed to investigate the transmission characteristics of power within the suspension system by applying a six-component force at the wheel center. The power distribution along each path in both front and rear suspensions is calculated, identifying the path with the highest power transmission. Additionally, an evaluation is conducted on how shock absorber mass influences transmitted power. The results demonstrate a reduction in structure-borne road noise by 1.09 dB(A), thereby validating our analytical approach with precision.
Practical evidence has shown that professional musicians are often exposed to high sound levels, which can lead to noise-induced hearing loss. This risk is particularly noticeable in drummers. Percussion instruments produce impulsive noise that is known to be more harmful than continuous noise of the same energy. Additionally, metrological constraints limit the accurate assessment of the actual noise exposure that drummers experience during their practice sessions. The primary objective of this article is to present the results of noise exposure measurements taken from a group of 21 drummers during a controlled rehearsal session in a recording studio. Specialized equipment, including high-pressure microphones positioned at the entrance of the drummers' ear canals, was used for these measurements. The results showed that the overall noise level during a typical rehearsal reached 103 dBA, while the corresponding kurtosis-adjusted level was 108.3 dBA. Notably, 90.5% of participants exceeded the NIOSH recommended exposure limit, and 62% experienced impulsive peak levels above 140 dBC during the rehearsal session. These findings indicate a significant risk of hearing impairment for drummers who play for extended periods, highlighting the need for preventive measures to protect this group.
The effect of aerodynamic control and active suspension systems on vehicle handling and stability during dynamic maneuvers is examined in this study. The paper shows how vehicle behavior changes from oversteer to under-steer under different load conditions by analyzing the understeer coefficient in response to both longitudinal and lateral load transfers. The importance of suspension tuning is highlighted by the examination of the impact of roll stiffness distribution on handling characteristics. To increase stability, a control method based on departures from ideal yaw behavior is implemented, modifying the normal forces on each axle. The advantages of active systems in dampening oscillations, decreasing overshoot, and achieving faster stabilization is demonstrated by the comparison of passive and active systems across several performance parameters, including yaw rate, sideslip angle, and lateral acceleration. Actuator force and power profiles further confirm the control system's responsiveness and effectiveness. The improved trajectory tracking and decreased deviation made possible by active control are validated by high-speed maneuver data collected on both dry and wet roads. With a focus on the trade-offs between drag and downforce, the research also investigates aerodynamic force changes depending on spoiler features and dynamic indications. Overall, the results confirm that combining aerodynamic controls and active suspension greatly improves vehicle safety, cornering stability, and responsiveness.
