Search

Contemporary fault diagnosis algorithms constitute advanced signal processing techniques integrated with the data-driven feature classification algorithms which make an effective fault diagnosis scheme for rotating machinery such as gearboxes and motors. Feature extraction is a prevalent task which is intended to assist the fault diagnosis process by eliciting a set of condition indicators (features) from the input raw signal. In actual scenario, the gearboxes may have multiple stages and are rather operating under fluctuating speeds. The feature extraction technique employed at medium and high ranges of operating speed may not be adequate during low operating speeds. In this present study, the feature extraction abilities of discrete wavelet transform (DWT) and empirical mode decomposition (EMD) in terms of their relative effectiveness while ascertaining the local gear tooth defects of a multistage gearbox are compared. Two local gear tooth defects, namely root crack and tooth chip with three severity levels, are seeded artificially. The experiments are carried out on a three-stage spur gearbox experiencing fluctuating operating speeds. Vibration analysis is performed, and the recorded raw vibration signatures are decomposed using DWT and EMD analyses separately. Mother wavelet selection is done using the criteria of energy-to-Shannon entropy ratio. The identification of intrinsic mode functions (IMFs) is made by examining the Pearson correlation coefficient. Various descriptive statistics are obtained from the wavelet coefficients and IMFs and the potential indices among them are chosen by implementing the decision tree algorithm. Finally, support vector machine (SVM) algorithm is executed to distinguish among the various defect severity levels. It has been observed that the SVM in conjunction with DWT has resulted in better classification than SVM in conjunction with EMD.

Previous research investigators have exploited machine-learning algorithms to diagnose the defects in rotating machinery. However, with increasing complexity in the design of rotating machinery, it is quite challenging to quantify the faults precisely. In this present study, an attempt has been made to predict the defect severity of the rotating machinery using Adaptive Neuro-Fuzzy Inference System (ANFIS). This ANFIS algorithm employs artificial neural networks to define the membership functions, rules and weights to construct the fuzzy inference system. Experiments are performed on a multi-stage spur gearbox model while it is subjected to fluctuating operating speeds. Two local defects on bearing race as well as on gear tooth with four different severity levels are seeded intentionally. Three condition monitoring (CM) strategies, namely, vibration, lubrication oil and acoustic signal analyses are executed, and the raw data is recorded synchronously. The raw vibration and acoustic waveforms are decomposed through discrete wavelet transform to extract the descriptive statistics from the wavelet coefficients. Among them, most discriminating features are selected and given as input to ANFIS classification tool to train the network for obtaining the Sugeno-type FIS, which in turn estimates the severity of the component. Later, the features from the individual CM strategies are combined to devise an integrated feature dataset which is further channelled as input to the ANFIS for predicting the defect severity levels. The investigation reveals that, the proposed integrated feature set in conjunction with ANFIS can discriminate between the defect severity conditions of the gears as well as bearings under fluctuating speeds.

Early detection of bearing defects may prevent the occurrence of catastrophic failures of the whole associated system. Condition monitoring strategies such as vibration and acoustic signal analyses are employed for incipient fault diagnosis of bearings. The current investigation attempts to compare the fault diagnostic capabilities in terms of their effectiveness in early detection of local bearing defects. Experiments are performed on a three-stage gearbox under constant and fluctuating operating conditions of speed. Wavelet coefficients are achieved from the acquired raw signals by discrete wavelet transform and various statistical features are obtained. Most contributing features among them are chosen by decision tree. Further, the extracted features are classified based on their fault severity levels using support vector machine algorithm. The experimental investigation revealed that vibration signal analysis outperformed the acoustic signal analysis under the experimental operating conditions.

Rolling element bearing faults of a laboratory scale wind turbine gearbox operating under nonstationary loads have been diagnosed using condition monitoring (CM) techniques such as vibration analysis, acoustic analysis, and lubrication oil analysis. Two local bearing faults, namely, bearing inner race fault and bearing outer fault are seeded in the gearbox. The raw data from these techniques are decomposed and wavelet approximation coefficients of level four (a4) are extracted using discrete wavelet transform (DWT). A plethora of statistical features is computed from the wavelet approximation coefficients and the most significant features are being identified by implementing the decision tree algorithm. The classification efficiencies of each of these CM techniques are compared by using the support-vector machine algorithm. Furthermore, an integrated CM scheme is developed by combining the individual CM techniques and the fault diagnosing ability of the integrated CM scheme is compared with the individual CM techniques. A principal component analysis-based approach is used as a feature classification algorithm and an input feature matrix is formed by combining the significant features extracted from vibration, acoustic, and lubrication oil analysis. It has been observed that the integrated CM scheme has provided better classification interpretations than the single CM techniques and it can be extended for real time fault diagnosis of a wind turbine gearbox.

Multi-stage gearboxes are vulnerable to failures often due to the extreme operating conditions, which may result in long downtimes. The current investigation is intended to examine the fault diagnostic capabilities of the integrated vibro-acoustic condition monitoring scheme while diagnosing the local/lumped defects exist at different speed stages of a multi-stage gearbox subjected to fluctuating/varying speeds. Experiments are performed, and the raw vibration and acoustic signatures are acquired simultaneously from the three-stage spur gearbox. Later, the raw data signatures are processed individually through discrete wavelet transform, and various descriptive statistics are extracted. Further, feature-level fusion is executed to obtain the integrated vibro-acoustic feature vector set for various speed stages of the gearbox. Finally, the obtained integrated feature vector set is classified using principal component analysis (PCA). It is observed that PCA performed using the integrated vibro-acoustic scheme clearly distinguishes among the various damage severity levels of pinion tooth exist at different speed stages of the gearbox.

Gears are positioned to transmit the power/torque and they are subjected to severe fluctuating speeds. The repetitive loading on the gear tooth may lead to failure which can cause plastic deformation and/or removal of contacting surfaces of tooth or even the initiation of fatigue crack. These tooth failures affect the transmission of gear and always accompanied by a variation in the gear mesh stiffness (GMS). In this work, both analytical and finite element (FE)-based numerical models are presented to estimate the GMS and also to extract the vibration response. A low-contact ratio, single-stage spur gear transmission consisting of eight degrees of freedom has been simulated. Two gear tooth failures, namely, tooth crack and breakage with two different severity levels are simulated. The effects of gear tooth failures on the evolution of GMS are compared with the case of ideal gear tooth. Further, frequency spectrum analysis is carried out to examine the effects of gear tooth failures on the amplitude of vibration of gear. Finally, the results of the developed analytical model and FE-based numerical model are validated with the experimental observations. It is observed that, the damaged gear tooth produces a reduction in the GMS values and lead to sidebands in the frequency spectrum.

This paper proposes a wavelet and artificial intelligence-enabled rapid and efficient testing procedure for patients with Severe Acute Respiratory Coronavirus Syndrome (SARS-nCoV) through a deep learning approach from thoracic X-ray images. Presently, the virus infection is diagnosed primarily by a process called the real-time Reverse Transcriptase-Polymerase Chain Reaction (rRT-PCR) based on its genetic prints. This whole procedure takes a substantial amount of time to identify and diagnose the patients infected by the virus. The proposed research uses a wavelet-based convolution neural network architectures to detect SARS-nCoV. CNN is pre-trained on the ImageNet and trained end-to-end using thoracic X-ray images. To execute Discrete Wavelet Transforms (DWT), the available mother wavelet functions from different families, namely Haar, Daubechies, Symlet, Biorthogonal, Coiflet, and Discrete Meyer, were considered. Two-level decomposition via DWT is adopted to extract prominent features peripheral and subpleural ground-glass opacities, often in the lower lobes explicitly from thoracic X-ray images to suppress noise effect, further enhancing the signal to noise ratio. The proposed wavelet-based deep learning models of both, two-class instances (COVID vs. Normal) and four-class instances (COVID-19 vs. PNA bacterial vs. PNA viral vs. Normal) were validated from publicly available databases using k-Fold Cross Validation (k-Fold CV) technique. In addition to these X-ray images, images of recent COVID-19 patients were further used to examine the model's practicality and real-time feasibility in combating the current pandemic situation. It was observed that the Symlet 7 approximation component with two-level manifested the highest test accuracy of 98.87%, followed by Biorthogonal 2.6 with an efficiency of 98.73%. While the test accuracy for Symlet 7 and Biorthogonal 2.6 is high, Haar and Daubechies with two levels have demonstrated excellent validation accuracy on unseen data. It was also observed that the precision, the recall rate, and the dice similarity coefficient for four-class instances were 98%, 98%, and 99%, respectively, using the proposed algorithm.

Condition-based maintenance methods are being used in expensive and crucial equipment thereby provides the maintenance team with the information required without actually dismantling or shutting down the machine. For example, wind turbines and large stationary gearboxes use this method to predict the health of the system, thereby avoiding the possibility of a catastrophic failure. Studies related to fixed axis gearbox are more when compared to studies related to planetary gearbox. Multiple studies related to single component fault diagnosis have been carried out over the past years. However, studies on multicomponent fault detection are comparatively less. Nevertheless, the practical application requires multicomponent fault diagnosis technics to reliably predict failures as the machinery used would have more than one point of failure. This paper reports a study on multi-fault diagnosis conducted in a scaled-down model of a wind turbine planetary gearbox subjected to stationary loading. A uniaxial piezoelectric accelerometer was used to acquire vibration data which was then processed with an algorithm after the features were extracted using statistical feature extraction. The power spectral density was analysed to interpret the results obtained.

Exposure of a building to a tornado often proves fatal, resulting in massive destruction of property and structures. The effect of disasters can be minimized by understanding the nature of fluid-structure interactions when a tornado hits a building on its path. Earlier researchers have investigated extensively on building models exposed to stationary type vortex generated in a laboratory type tornado simulator; however studies using translating type vortex are few. In the present investigation, the external and internal pressures experienced by a building model are discussed based on experiments conducted using a translating tornado-like flow simulator at Tokyo Polytechnic University, Japan. The swirl ratio which characterizes the strength of vortices generated is kept constant. The investigation attempts to explore the opening locations which can result in higher internal pressures and net roof forces in building and the vulnerability of roof structures of buildings when exposed to tornado-like flow. The effect of translating speed on internal pressure fluctuations when compared to those of external pressures are investigated using aerodynamic admittance functions. Results indicate that there is an increased amount of internal pressure fluctuations at higher translating speeds.

Remaining Useful Life (RUL) studies generate large volumes of data which demand complex computational resources to process. Removal of anomalies from the acquired raw data is a significant paradigm which makes the condition monitoring process robust. In this investigation, a fingerprinting based data abstraction technique is proposed to identify the prominent data points from the acquired data. Run-to-failure experiments are performed on a scaled gearbox to acquire vibration signatures. Continuous Wavelet Transform (CWT) is performed and the most prominent data points (fingerprints) are abstracted from the CWT coefficients. Descriptive statistics are computed for these fingerprints. Cumulative energy is computed from the fingerprint to build health index for predicting the RUL for different speed stages of gearbox. Zone demarcation points are estimated for gearbox stages to determine individual stage’s health. A comparison between different classification algorithms yielded RNNs (long short-term memory networks) as the best in conjunction with the proposed algorithm.

With the advent of tall and complex shaped structures combined with urbanization, the structural designers are posed with a greater challenge to design the structures sustainable to rapid environmental changes. During the design of tall structures, wind loads acting on them are a major factor that needs to be considered. As a pre-requisite, the designer should have the information regarding wind environment of the region, wind forces on the structures and the response of the structure under these forces.

The paper discusses on the design and development of a test-rig for estimating the vibration characteristics of a beam. The salient features of the test rig, which makes it possible to conduct a set of experimental cases to determine the nature of vibration under different variations, and the test cases envisaged are discussed. Using the test-rig developed, the main objective is to analyse the transverse vibrations of flexible rectangular beams of aluminium alloy and mild steel for different cross sections and to study the importance of excitation on vibration response. Comparative study is performed based on the effect of variations in the characteristics of the beam on frequency and amplitude of vibration for different end conditions. An effort is made to estimate the relation between exciting centrifugal force due to rotating eccentric mass and amplitude of vibration for various cases under consideration.

Fluid flow studies play a vital role in proper assessing of wind loadson buildings and other civil engineering structures with various shapes. The presence of adjacent structure may alter wind loads on the structure of interest, by altering the fluid dynamics surrounding the structure. This effect is termed as interference effect and it depends on various factors such as terrain category, wind angle, geometry & orientation of the structure, spacing between the structures etc. Thus, proper assessment of interference effect with above factors, can lead to efficient design of structures that can withstand the load variations due to the presence of interfering structures. The existing literature on wind interference effects, focuses more on experimental analysis of the problem, which is both expensive and time consuming. In this regard, computational fluid dynamics (CFD) approach can serve as a quicker and efficient tool. In an attempt to study the effect of geometry &orientation of structures on interference phenomenon, turbulent flows around the buildings, cooling towers and chimneys have been simulated through CFD approach by performing Large Eddy Simulation (LES). In the present study, interference factors of a typical chimney structure has been evaluated by considering the presence of an identical chimney at varying distances, under different wind angles.

Tornadoes pose serious threat to buildings and structures. Studies related to tornado-structure interaction are often restricted to post-damage investigations. Detailed tornado-structure interaction studies are required to understand the basic flow mechanisms that inflict damage. Terrain roughness is often a criterion that determines the strength and characteristics of tornado and consequently its effect on structures. In the present investigation, the effect of ground roughness on the internal pressures developed inside a building model exposed to a stationary vortex is analyzed using the Ward-type tornado simulator at Tokyo Polytechnic University. The effects of ground roughness on net local roof wind forces are also investigated. Results show that the internal pressures decrease in magnitude with introduction of roughness, whereas the resulting net local roof forces increase in magnitude with introduction of roughness.

Violent wind vortices and their influence on structures have been an area of research during the past few decades. Effects of these vortices on building structures depend on vortex size relative to building size, relative distance between vortex center and building location, terrain condition and swirl ratio. The present investigation takes into account factors such as building model location with respect to vortex center and terrain condition that influence the surface pressures on a building model cube exposed to a swirl flow by performing a systematic set of laboratory experiments using a Ward-type tornado-like-flow simulator developed at Tokyo Polytechnic University, Japan. The developed vortex was stationary and the effect of translation was not taken into account. The scalar component of velocity was determined at different heights and radii across the simulator floor. Mean and peak pressures acting on the building model cube considering the effect of the above factors is discussed. Salient features of peak pressure coefficients on a building surface occurring under a tornado-like flow regime are also discussed.

Induction motors, which are used worldwide as the ''workhorse'' in industrial applications, are intermittently subjected to faults, mainly the stator faults. In this paper, fault diagnostics of induction motor using current signature analysis, with wavelet transform, is treated as a pattern classification problem. The major steps in pattern classification are feature extraction, feature selection and classification. The feature extraction is done by wavelet transforms, using different wavelets which allow the use of long time intervals where there is precise low-frequency information, and shorter regions where there is precise high-frequency information. The extracted features are classified using the new generation pattern classification technique of Support Vector Machine (SVM) identification. Then the relative capability of the different wavelets, in performing the stator winding fault identification is analyzed and the best wavelet is selected.

Roller bearing is one of the most widely used rotary elements in a rotary machine. The roller bearing's nature of vibration reveals its condition and the features that show the nature are to be extracted through some indirect means. Statistical parameters like kurtosis, standard deviation, maximum value, etc. form a set of features, which are widely used in fault diagnostics. Finding out good features that discriminate the different fault conditions of the bearing is often a problem. Selection of good features is an important phase in pattern recognition and requires detailed domain knowledge. This paper addresses the feature selection process using decision tree and uses kernel based neighborhood score multi-class support vector machine (MSVM) for classification. The vibration signal from a piezoelectric transducer is captured for the following conditions: good bearing, bearing with inner race fault, bearing with outer race fault, and inner and outer race faults. The statistical features are extracted therefrom and classified successfully using MSVM. The results of MSVM are compared with and binary support vector machine (SVM).

Wind energy is an emerging, clean and renewable source of energy. It is estimated that by year 2035, wind energy will be generating more than 25% of the world's electricity according to International Energy Agency (IEA). With the increase in demand for wind energy, its maintenance issues are becoming more prominent. The scheduled maintenance is more economical than unscheduled repair resulting from failure. So a continuous condition monitoring of various critical components like bearings, gearbox, and shafts of wind turbine is essential in order to enable predictive maintenance. 10% of the total failure is contributed by the bearings, shaft and gear box failures, but the downtime is more than 50% of the total downtime. This paper discusses the development of a bench-top test rig which is designed to mimic the operating condition of an actual wind turbine and use it for monitoring its condition so as to diagnose the incipient faults in its critical components using latest machine learning algorithms such as Artificial Neural Network (ANN).

R K Goliya, N K Samaiya, G R Sabareesh, Abhay Gupta Sr. Lecturer of Civil Engineering, Jaypee University of Engineering & Technology, Guna, MP, India, ravindra.goliya@juet.ac.in Asst. Professor of Civil Engineering, Jaypee University of Engineering & Technology, Guna, MP, India, nitin.samaiya@juet.ac.in Asst. Professor, WEAC, Jaypee University of Engineering & Technology, Guna, MP, India gr.sabareesh@juet.ac.in Director, ESCOM Consultants Pvt. Ltd., NOIDA, UP, India, abhay@escom.in abhay@escom.in ABSTRACT

Tornadoes are deadliest of all winds causing damage to life and property in many parts of the world. Their uncertainty and associated danger makes real time investigations difficult. In the past three decades attempts are made to simulate tornado-like flow in the laboratory and expose scaled building models to it. Many earlier attempts did not consider the translating effect of vortices. The present investigation explores the characteristics of the tornado like vortex developed using the translating tornado-like flow simulator at Tokyo Polytechnic University and the generated flow field surrounding it.