Vibration Signal Analysis Research Articles

Multitype chatter detection via multichannelinternal and external signals in robotic milling

Robot milling chatter monitoring is susceptible to interference from robot pose, feed direction, inertial excitation-dominated low-frequency flexible vibration (IELFFV), and spindle noise. Therefore, this paper proposes a multi-type chatter detection method utilizing multi-channel internal and external signals. First, the multichannel vibration signals and internal controller signals were acquired and aligned. The vibration signals were converted into an engagement coordinate system. Then, based on the milling vibration signal characterization and sensitivity analysis of vibration signals to chatter, a combination of three chatter indicators, calculated via filtered displacement, filtered acceleration, and internal signals, was utilized to detect structural mode-dominated low-frequency chatter, IELFFV, and tool mode-dominated high-frequency chatter. Finally, the effectiveness of the proposed method in detecting the multitype chatter was demonstrated via robot milling experiments. However, the process of the tool entering and exiting the workpiece could be misclassification as TMHFC, which would be addressed in subsequent research.

Blasting Vibration Control and Signal Analysis of Adjacent Existing Deterioration Tunnels

Building a new tunnel adjacent to an existing tunnel has become a common means of transformation in engineering. Existing tunnels are prone to some deterioration, such as cavities and cracks under long-term traffic load. This kind of deterioration tunnel is prone to collapsing under the action of blasting. Therefore, the vibration caused by blasting should be strictly controlled. Based on the reconstruction project of the Bo Jiling Tunnel, this paper puts forward the method of mechanical cutting in a central position combined with an ordinary detonator to reduce blasting vibrations. ANSYS/LS-DYNA version 19.2, was used to simulate two conditions of full-section blasting and central mechanical cutting blasting. By comparing the stress and velocity of the existing tunnel, the damping effect of mechanical cutting blasting is analyzed. Via field experiments, the superiority of the mechanical cutting method in reducing blasting vibration is further discussed. At the same time, the relationship between the main vibration frequency and the peak velocity of the existing deterioration tunnel is obtained by wavelet packet analysis of the field experimental data. The frequency band energy distribution in each direction of vibration velocity is also obtained. The results show that the central mechanical cutting increases the blasting free surface, and the mechanical cutting method reduces the vibration velocity by 36.3%. The third frequency band (31.25~46.875 Hz) is the most concentrated, which is the dominant frequency band of the signal. The novelty of this paper is to propose mechanical cutting of the central hole instead of traditional blasting for existing deterioration tunnels. The feasibility of this method is verified by numerical simulation and field tests. The relationship between peak vibration velocity, band energy, and tunnel frequency is clarified, which can better control blasting vibration and ensure the safety of existing deterioration tunnels.

A study of surface fatigue wear and influence of contaminated lubricant in journal bearing system using tribological and vibration analyses

In the present work, journal bearing performance was evaluated using condition monitoring processes namely wear debris, vibration signal, lubricant oil degradation and temperature analyses. The journal bearing system was operated at 800 rpm and 200 N constant speed and load over a period of 1200 h. The lubricant oil samples were collected for every 300 h of test duration for oil degradation and wear particle analyses by using Fourier transform infrared radiation (FTIR) and scanning electron microscope/energy dispersive X-ray spectroscopy (SEM/EDS) techniques, respectively. FTIR results provided specific chemical bonds and functional groups contained in the degraded oil. The vibration signals were acquired for every 100 h of operation, and a fast Fourier transform (FFT) frequency spectrum technique was employed for vibration signal analysis. The friction coefficient and specific wear rate parameters were considered to evaluate the tribological performance of the journal-bearing system. Statistical parameters, namely root-mean-square and kurtosis, were evaluated. Wear mechanisms such as abrasive, adhesive and surface fatigue wear were observed on the journal-bearing contact surfaces using a USB optical microscope. The scoring on the bearing element confirmed the fault development and resulted in an increase in vibration amplitudes. Higher mass loss in the bearing material was observed due to high friction and wear rates owing to its being a softer material than that of the journal material. The results obtained from the aforementioned parameter assessment techniques provided a reliable information on the defects which appeared on journal-bearing contact surfaces.

A failure diagnosis method of ball bearing based on optimized multiscale variational modal extraction with synchro squeezing transformation

When a compound failure occurs in a bearing, the failure information included in vibration signals is featured by being weak, complex and combined. This fact makes it difficult to identify a compound failure precisely. Variational mode extraction (VME) solves the problem of variational mode decomposition (VMD) in being difficult to determine the number of decomposition layers and has certain applications in the identification of bearing failures. However, the initial expected central frequency of VME and option of penalty factor is crucial to the extraction of expected mode. To precisely determine the central frequency and penalty factor of VME and fulfill the correct extraction of feature information of compound failures in bearing, the paper has proposed a method based on synchro squeezing transform (SST) and information entropy to adaptively determine the parameters in VME (SST-VME). Firstly, SST is used to make time-frequency analysis of original vibration signals and characteristic frequency bands with larger energy are chosen from time-frequency spectrum to adaptively determine the expected central frequency of VME. Secondly, as information entropy has representation capacity for information included in the signal, penalty factor of VME was adaptively determined according to information entropy. Thirdly, original signals were subjected to VME according to expected central frequency and penalty factor adaptively determined; the expectation mode obtained was denoised by singular value decomposition algorithm. Finally, the type of compound failures of bearing is determined according to the frequency spectrum of denoised signals. To verify the effectiveness of proposed method, a comparison with VMD algorithm is conducted. As indicated by the result, the proposed method is more precise and comprehensive than VMD algorithm to extract the feature information corresponding to compound faults of bearing, and thereby correctly determines the type of compound failures.

An adaptive reweighted-Kurtogram for bearing fault diagnosis under strong external impulse noise

In vibration-based fault diagnosis of rotating machinery, strong impulse noise constitutes a widely encountered source of external interference, thereby posing significant challenges to the effective extraction of diagnostic information. In this context, an adaptive reweighted-Kurtogram (ARKurtogram) method is proposed. The bearing transient fault signatures are accurately extracted through two aspects – that is, precise frequency band division and robust band selection indicator. Specifically, we use the dual-tree complex wavelet packet transform (DTCWPT) as the band division tool and propose the sub-bands rearranged and ensemble DTCWPT which addresses the band disorder problem, yields far richer band division patterns and maintains excellent performance of DTCWPT in signal decomposition. Meanwhile, we introduce a robust indicator characterizing periodic fault impulses to accurately select the fault band. In comparison with the classical kurtosis and the advanced reweighted-kurtosis, this indicator is insensitive to strong external impulse noise and its calculation process contains a sliding window averaging approach which makes the calculation entirely automated without human intervention. The validity of the ARKurtogram is investigated with the analysis of vibration signals measured from fault-injection experiments. The comparisons with the classical fast Kurtogram method and the advanced Autogram method highlight the advantages of the method in extracting bearing transient fault signatures from complex signals with strong external impulse noise. It shows high potential in the application of industrial bearing fault diagnosis.

Graph-Guided Higher-Order Attention Network for Industrial Rotating Machinery Intelligent Fault Diagnosis

Data-driven approaches have gained great success in the field of rotating machinery fault diagnosis for its powerful feature representation capability. However, in most of the current studies, model training process requires massive fault data which is costly to gather or even unavailable in some extreme operating conditions. At the same time, structural relationships between samples are not fully exploited to facilitate the model performance. In response to these problems, a novel graph-guided higher-order attention network (GHOAN) for rotating machinery fault diagnosis is proposed in this study. Specifically, the proposed approach incorporates the advantages of improved graph attention network (GAT) model and the multi-order neighborhood feature perception to achieve richer feature representation by aggregating features from multiple neighborhood domains. In this way, effective fault diagnosis may be achieved by using fewer training samples based on vibration signal analysis. The results of experiments conducted on two benchmarking datasets and a practical experimental platform show that proposed GHOAN achieve superior performance.

Random vibration-based progressive fatigue damage monitoring in thermoplastic coupons: a preliminary investigation

In recent years there has been an increasing trend towards the utilization of composites, particularly thermoplastics, in various components of aerostructures. The prelusion of such materials has underscored the significance of investigating their fatigue behavior and developing reliable methods for detecting fatigue damage. In this context, vibration-based techniques hold significant potential as they leverage the inherent excitation provided by in-flight noise and turbulence. This study aims at assessing the progressively accumulated fatigue damage in thermoplastic coupons via random vibration signals while accounting for operational and inter-structural uncertainties. The experimental process consists of preliminary tension and fatigue tests, interrupted fatigue tests, C-Scan inspection tests, and non-contact random vibration tests. Consecutive fatigue states are obtained by performing fatigue tests at intervals of 10 000 cycles for a population of 7 coupons. At each interruption, ultrasonic C-Scan and vibration inspection tests are performed, allowing for the visualization of fatigue damage and random vibration signal analysis. Welch Power Spectral Density estimates are employed and are shown to have good potential for distinguishing among different fatigue states despite the inevitable population and experimental uncertainty. Furthermore, fatigue damage is found to progress symmetrically and laterally along the free edges of the test coupons, which is explained by the free edge effect.

Multi-layer convolutional dictionary learning network for signal denoising and its application to explainable rolling bearing fault diagnosis

As a vital mechanical sub-component, the health monitoring of rolling bearings is important. Vibration signal analysis is a commonly used approach for fault diagnosis of bearings. Nevertheless, the collected vibration signals cannot avoid interference from noises which has a negative influence on fault diagnosis. Thus, denoising needs to be utilized as an essential step of vibration signal processing. Traditional denoising methods need expert knowledge to select hyperparameters. And data-driven methods based on deep learning lack interpretability and a clear justification for the design of architecture in a “black-box” deep neural network. An approach to systematically design neural networks is by unrolling algorithms, such as learned iterative soft-thresholding (LISTA). In this paper, the multi-layer convolutional LISTA (ML-CLISTA) algorithm is derived by embedding a designed multi-layer sparse coder to the convolutional extension of LISTA. Then the multi-layer convolutional dictionary learning (ML-CDL) network for mechanical vibration signal denoising is proposed by unrolling ML-CLISTA. By combining ML-CDL network with a classifier, the proposed denoising method is applied to the explainable rolling bearing fault diagnosis. The experiments on two bearing datasets show the superiority of the ML-CDL network over other typical denoising methods.

A fault diagnosis method for motor vibration signals incorporating Swin transformer with locally sensitive hash attention

Identification of motor vibration signals is one of the important tasks in motor fault diagnosis and predictive maintenance, and wavelet time–frequency diagram is a commonly used signal analysis method to extract the frequency and time characteristics of signals. In this paper, a method based on local sensitive hashing (LSH)-Swin transformer network is proposed for identifying the wavelet time–frequency diagrams of motor vibration signals to analyze the fault types. The traditional Swin transformer model converges slowly due to the smoothing of the attention distribution when dealing with data with sparse features, while the method proposed in this paper reduces the smoothing of the computed attention and enables the network to learn the key features better by introducing locally-sensitive hash attention in the network model, dividing the sequences in the input attention into multiple hash buckets, calculating the attention weights of only some of the vectors with a high degree of hash similarity, and by sampling discrete samples with the use of the Gumbel Softmax. The experimental results show that the method proposed in this paper has better recognition accuracy and higher computational efficiency compared with the traditional network when processing wavelet time–frequency maps of motor vibration signals, and its validation accuracy reaches 99.7%, the number of parameters also has a decrease of about 13%, and the training network to reach converged epochs is also faster. The method in this paper can provide an effective solution for the analysis and processing of motor vibration signals, and has certain application value in practical engineering.

Vibration feature extraction and fault detection method for transmission towers

AbstractThis paper presents a novel bolt looseness detection method for power transmission towers based on vibration signal analysis. The proposed method utilizes pulse excitation to extract the vibration signal of the tower, which is then adaptively decomposed using the Variational Mode Decomposition of Spider Wasp optimizer (SWVMD). This overcomes limitations of traditional Variational Mode Decomposition methods by leveraging bio‐inspired optimization to improve signal decomposition. Simulated signals processed with different optimization methods verify the superiority of the SWO approach. Field tests on a 110‐kV transmission tower further demonstrate the effectiveness of the proposed SWVMD technique for analyzing on‐site vibration data. A new improved intrinsic multiscale sample entropy feature is also introduced for bolt state characterization. A Spider Wasp Support Vector Machine classifier is developed to realize accurate bolt loosening monitoring using the extracted features. Dynamic response tests under varying bolt conditions show that the method can identify early loosening and reduce tower damage risks compared to conventional techniques. This novel vibration‐based detection framework presents an innovative application of nature‐inspired computing for power infrastructure health monitoring.

Monitoring of drill bit wear using sound and vibration signals analysis recorded during rock drilling operations

Monitoring of drill bit wear using sound and vibration signals analysis recorded during rock drilling operations

An analysis of vibration signals is conducted to monitor the condition of the clutch of a centrifugal pump, employing ISO 10816 standards

Vibration in the shaft of a centrifugal pump arises from various factors, including the spacing between clutches, belt thickness, width, and their positioning within the clutch assembly. Vibration serves as a vital diagnostic tool for assessing machinery, offering insights into both motion and translation. Understanding vibrations and analyzing corresponding data are essential for effective maintenance and troubleshooting protocols. By leveraging this knowledge, companies can mitigate downtime, enhance productivity, and extend the operational lifespan of their machinery. Vibration stems from cyclic forces acting upon machine components, triggering energy release within the structure and resulting in oscillations. Excessive vibration not only generates noise but also compromises pump performance and jeopardizes critical components, notably the shaft and bearings. The objective of this study is to monitor and analyze vibration in centrifugal pumps equipped with modified belt connections. Various configurations of belt thickness, width, distance, and positioning were explored. Real-time data was gathered using accelerometer sensors to capture X, Y, and Z-axis values, subsequently converted from analog to digital format. Among the configurations tested, the Outside-Inside belt positioning, with belt dimensions of 6.5mm thickness, 150mm width, and a 40mm distance, exhibited the lowest average vibration value at 8.72mm/s. However, according to ISO 10816 standards, this measurement falls within the unsatisfactory category, indicating the need for further refinement in pump design and maintenance practices.

Intelligent fault detection of spark plugs using vibration signal analysis and artificial neural networks

Intelligent fault detection of spark plugs using vibration signal analysis and artificial neural networks

A Comprehensive Investigation of Fault Signatures and Spectrum Analysis of Vibration Signals in Distributed Bearing Faults

A Comprehensive Investigation of Fault Signatures and Spectrum Analysis of Vibration Signals in Distributed Bearing Faults

ОЦЕНКА ВЛИЯНИЯ КОЭФФИЦИЕНТА МАСШТАБИРОВАНИЯ БАЗОВОЙ ФУНКЦИИ ВЕЙВЛЕТ-ПРЕОБРАЗОВАНИЯ НА СТЕПЕНЬ ПРОЯВЛЕНИЯ ОСНОВНЫХ НЕИСПРАВНОСТЕЙ ЦЕНТРОБЕЖНЫХ НАСОСНЫХ АГРЕГАТОВ В ВЕЙВЛЕТ-КАРТИНАХ

Relevance Rotary-type machines, which include centrifugal pumping units, are widely used in the oil refining and petrochemical industries and represent an integral part of technological processes. Therefore, preventing their emergency failure is the most important task, which is solved using vibroacoustic diagnostics. Vibration diagnostic services of technical supervision departments process signals mainly using the spectral method of vibration signal analysis based on the Fourier transform. Along with the undeniable advantages, which include the simplicity of software implementation and the admissibility of optimization for the analysis of many functions of physical processes, this transformation has certain disadvantages. For example, due to the loss of information about the time factor, the idea of the dynamics of changes in the spectral composition of the signal is distorted, which does not allow for a comprehensive analysis of the vibration signal. In this regard, there is a need for additional methods of digital signal processing. In recent years, in the process of research carried out at the Ufa State Petroleum Technological University, it has been proven that a method based on wavelet transformation of vibration signals can be used to assess the technical condition of pumping and compressor equipment. This type of transformation is capable of recognizing not only periodic components of the signal, but also local ones, determining their time and frequency characteristics, and establishing the relationship between the levels of wavelet decomposition and the frequency ranges of wavelet filters. It has been determined that the level of decomposition during the wavelet transform of the signal can be taken as a criterion for clearing the vibration signal of the digital pump from the noise component in the presence of faults such as «imbalance» and «misalignment». Discrete wavelet transform of vibration signals makes it possible to determine amplitude modulation caused by the manifestation of faults of electromagnetic origin and inhomogeneity of the air flow in the gap between the rotor and stator of an electric motor, as well as to identify slowly developing (trend) signals associated with lubrication defects in the motor bearings of centrifugal pumping units. Wavelet patterns store complex information about the signal, however, when several types of faults or defects are combined, it is very difficult to identify the leading one and the degree of its development, especially in the early stages. In this case, the value of the scaling factor can serve as a characteristic of the degree of development of the fault, since it was found that wavelet patterns begin to differ comprehensively when setting the scaling factor of the wavelet transform of signals, starting from a value of 192, confirming the presence of an anomalous signal. Aim of research Increasing the accuracy of assessing the technical condition of centrifugal pumping units by selecting the scaling factor of the basic function of the wavelet transform of vibration signals. Research methods To practically solve the problems, methods and algorithms for digital processing of vibration signals based on the Fourier transform and wavelet transform were used. Results It has been established that for a more accurate assessment of the stage of development of the main faults of the digital pump using the continuous wavelet transform, it is necessary to use a scaling factor of the basic wavelet function equal to 512.

Application of advanced signal processing techniques in vibration analysis and fault diagnosis of multi-dimensional anthropomorphic wind turbines

Abstract Wind turbine operating conditions are complex. To ensure the turbine’s safe operation, it is essential to carry out condition monitoring and fault diagnosis of its vibration. In this paper, from the structure of wind turbines, fault types, and fault formation mechanisms, a wind turbine vibration condition monitoring system is established by designing different vibration condition monitoring sensors and combining them with the Internet of Things technology. The discrete Fourier transform is employed to preprocess the time-frequency data before extracting the specific features of the vibration signal by combining the Hilbert-Huang transform after obtaining the wind turbine vibration signal. The SC-TSFN model with spatio-temporal deep fusion is established to realize the fault diagnosis of wind turbines by combining the replaceable null convolution module, BiLSTM module and the self-attention mechanism. It has been found that when the tertiary meshing frequency fluctuates around 506.98 Hz at a fault characteristic frequency of 16.14 Hz, it indicates a fault in the tertiary high-speed shaft gear. The SC-TSFN model has a fault identification time of approximately 52 days before the actual fault downtime, and the model has a 92.05% accuracy rate for wind turbine fault identification. Relying on the signal processing technology to carry out the wind turbine vibration signal analysis and then input it into the fault identification model can realize the accurate identification of the fault state of the unit and provide technical support for the stable operation of wind turbines.

INVESTIGATION OF THE WEAR PROCESS OF MIXER ELEMENTS USING INFORMATION TECHNOLOGY

Today, mechanical engineering is one of the dynamically developing sectors of the economy of Kazakhstan. Any production process involving machines imposes requirements for reliability and durability to ensure stable and trouble-free operation of technological processes. In this regard, ensuring the operational characteristics of machines for a long time is a key factor when choosing a manufacturer. The quality of the materials used in the production of machines significantly affects the service life and operational characteristics of machines. The wear of machine parts is one of the main causes of breakdowns and accidents. The financial costs of car repair items are quite large. To date, the wear control of the mixer elements, removable elements of the blades is carried out only visually and replacement occurs without a system, after complete abrasion or breakage. Since these elements are not always available in the warehouse of the enterprise, downtime of equipment can take a considerable time. Modeling of the wear process of mixer elements in the matlab program allowed us to develop a simulation model using the vibroacoustic control method. Analysis and processing of vibration signals is carried out using standard programs. The developed algorithm of the process of monitoring the wear of the mixer elements allows you to form a library of signals and noises, comparing them with the reference values of the signals, to determine the degree of wear of the mixer elements. The algorithm can be implemented on standard hardware and software devices. Keywords: Mixer, wear, model, control, matlab, management.

Modern methods and means of technical diagnostics of artillery weapons and prospects for their development

The analysis of full-scale military operations in the east of Ukraine indicates an increase in the variety of firing tasks for artillery weapons and the intensity of their execution. This, in turn, requires the maintenance of artillery weapons and systems that ensure their proper operation in a serviceable and ready-to-use condition, as well as the prevention of sudden breakdowns and failures during combat missions.
 Taking into account the influence of the technical condition on the effectiveness and safety of the combat use of artillery weapons, there is a need to improve the methods and means of its diagnosis.
 This article analyzes the available domestic and foreign studies devoted to the development and improvement of methods for diagnosing the technical condition of artillery weapons. Existing devices for measuring, inspecting, evaluating and controlling the technical condition are considered. The existing systematization was expanded, generalization and addition of known methods and tools was carried out. The main possibilities, advantages and disadvantages of known methods, as well as promising directions for further research are determined.
 It has been established that the majority of studies of possible methods of diagnosing the technical condition of artillery weapons are focused mainly on the analysis of the parameters of only the gun barrel and do not consider the ability of the studied methods to assess the technical condition of artillery weapons comprehensively.
 The analysis also showed that despite the large number of existing techniques for diagnosing the technical condition (determining the degree of wear), most of them are not suitable for use directly in field conditions, morally outdated or very expensive.
 In addition, it was determined that the application of the analysis of acoustic and vibration signals that occur during a shot is a promising direction for diagnosing the technical condition of artillery weapons, which potentially has the ability to detect more malfunctions, breakdowns and deviations of technical parameters.

Research on Micro-Turbine Operating State Characterization Based on Bearing Vibration Signals Analysis

An air flow-driven micro-turbine, widely used in air-condition control systems in aircraft cabins, train coaches, etc., exhibits complex vibration behaviors under stable or unstable inlet flow conditions, and especially has a certain correlation with speeds. In this paper, the vibration responses of the micro-turbine undergoing stable and unstable inlet flows measured on a test rig are analyzed and compared by using different signal processing methods, which include time and frequency domain methods, and statistical and nonlinear methods. First, the test rig system of the airflow-driven micro-turbine, the instrument system and four typical experimental cases are introduced. Then the measured vibration signals are analyzed and compared by time domain characteristic parameters (peak-to-peak value, RMS value and kurtosis value), statistical parameters (auto-correlation and BDS), and amplitude spectra in the frequency domain, statistic spectrum indicators (SSI) based on Welch's periodogram of power spectra, and the spectra of selected IMF components based on Hilbert-Huang Transform (HHT). In particular, some nonlinear feature analyzing methods, including Pseudo-Poincare mapping diagrams and Lempel-Ziv(LZ) complexity, are also used for analyzing measured vibration responses. The obtained results using the above multiple methods are compared and show that, when the inlet flow of the turbine fluctuates significantly, the nonlinear characteristics of the turbine bearings are significantly higher than those of the relatively stable inlet flow and speed conditions. Under these circumstances, commonly used time-frequency analysis methods cannot characterize the different speed operating state of the turbine, and LZ-Complexity and other nonlinear characterization methods should be used to better understand the characteristics of different speeds under unstable conditions.This study provides references for the aerodynamic stability monitoring of the micro-turbine and its design improvement.

Multi-Rate Vibration Signal Analysis for Bearing Fault Detection in Induction Machines Using Supervised Learning Classifiers

Vibration signals carry important information about the health state of a ball bearing and have proven their efficiency in training machine learning models for fault diagnosis. However, the sampling rate and frequency resolution of these acquired signals play a key role in the detection analysis. Industrial organizations often seek cost-effective and qualitative measurements, while reducing sensor resolution to optimize their resource allocation. This paper compares the performance of supervised learning classifiers for the fault detection of bearing faults in induction machines using vibration signals sampled at various frequencies. Three classes of algorithms are tested: linear models, tree-based models, and neural networks. These algorithms are trained and evaluated on vibration data collected experimentally and then downsampled to various intermediate levels of sampling, from 48 kHz to 1 kHz, using a fractional downsampling method. The study highlights the trade-off between fault detection accuracy and sampling frequency. It shows that, depending on the machine learning algorithm used, better training accuracies are not systematically achieved when training with vibration signals sampled at a relatively high frequency.