About this Journal Submit a Manuscript Table of Contents
Advances in Mechanical Engineering
Volume 2014 (2014), Article ID 210717, 10 pages
http://dx.doi.org/10.1155/2014/210717
Review Article

Condition Monitoring of Blade in Turbomachinery: A Review

1Institute of Noise and Vibration, Universiti Teknologi Malaysia, International Campus, Jalan Semarak, 54100 Kuala Lumpur, Malaysia
2Razak School of Engineering and Advanced Technology, Universiti Teknologi Malaysia, 54100 Kuala Lumpur, Malaysia

Received 9 July 2013; Revised 20 January 2014; Accepted 20 January 2014; Published 11 March 2014

Academic Editor: Mario L. Ferrari

Copyright © 2014 Ahmed M. Abdelrhman et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Blade faults and blade failures are ranked among the most frequent causes of failures in turbomachinery. This paper provides a review on the condition monitoring techniques and the most suitable signal analysis methods to detect and diagnose the health condition of blades in turbomachinery. In this paper, blade faults are categorised into five types in accordance with their nature and characteristics, namely, blade rubbing, blade fatigue failure, blade deformations (twisting, creeping, corrosion, and erosion), blade fouling, and loose blade. Reviews on characteristics and the specific diagnostic methods to detect each type of blade faults are also presented. This paper also aims to provide a reference in selecting the most suitable approaches to monitor the health condition of blades in turbomachinery.

1. Introduction

Blades are extensively used in power generation turbomachinery such as the compressors and gas turbines. In these machines, there could be more than a thousand blades located in both the turbine and compressor sections to transfer energy between the rotor and fluid. Thus, the operation of these machines is largely dependent on the condition of these blades. Blades often operate in hostile and high stress environment that can potentially lead to blade failures. The failure of a single blade can potentially compromise the total integrity of the machine. In such instances, even a blade failure rate of 1 in 1000 is not acceptable. According to Farrahi et al. [1] and Poursaeidi et al. [2], blade failures in gas turbines and compressors mostly originate from some form of initial damage or defect of the blades caused by Foreign Object Damage (FOD), ingested debris, or manufacturing defects. These minor defects or damage can propagate over time and eventually lead to total blade failures. Common types of blade faults and blade failures in turbomachinery include blade rubbing, low and high cycle fatigue failures, creep, fouled blade, loose blade, and blade induced FOD. A review on the most common failure mechanisms in gas turbine blades was provided by Carter [3].

This paper provides a review on the strategy for monitoring the condition of blades as well as the specific techniques applicable to monitoring different types of blade faults commonly found in turbomachinery.

2. Strategies for Condition Monitoring of Blade

Two different types of strategies can be employed to monitor the condition of a blade: monitoring machine’s operating parameters (e.g., vibration, pressure, temperature) and performing signal analysis (e.g., Fourier’s analysis, wavelet analysis, artificial intelligence). In most instances, both strategies are deployed concurrently to obtain the most important information from the signal at hand. In this paper, the literature review on each of these two strategies is presented.

2.1. Monitoring of Machine’s Operating Parameters
2.1.1. Vibration Measurements

Vibration analysis has been the most widely used blade condition monitoring method for decades. Time domain waveform signals are typically transformed into the frequency domain using Fourier’s transformations. After that, interpretation of the vibration spectrum is performed to extract useful information about the health of the machine. Detailed monitoring of a blade using vibration analysis is commonly achieved by monitoring the relative changes of its blade pass frequency (BPF) and its harmonics as presented by Mitchell [4]. The use of vibration analysis for blade faults diagnosis has also been studied by Simmons [5, 6], Parge et al. [7], and Parge et al. [8]. They have observed that the relative changes in the BPF and its harmonics amplitude can provide useful information about the occurrence of blade rubbing. Al-Badour et al. [9] studied the vibration signals caused by stator-to-blade rubbing during the machine runup and coast-down and they found that this method is effective in detecting blade rubbing. The authors Abdelrhman et al. [10] also investigated the effectiveness of vibration analysis for blade rubbing in multistage rotor system at different rotor stages and different rubbing severities. Satyam et al. [11] used Cepstrum analysis to analyze the vibration signal in order to detect blade faults. It was shown that the conventional frequency analysis in machinery vibration is incapable of accurately determining the defects in blades. On the other hand, Cepstrum analysis can accurately identify the harmonics and sideband families of BPF. It is therefore a better technique for blade faults detection especially in ship and submarine applications. In addition, Randall and Sawalhi [12] also employed the same technique to remove unwanted frequency components from the BPF components of interest. This allows for a better identification of blade fault based on the BPF monitoring approach. Mathioudakis et al. [13] conducted an experimental study to correlate between the compressor casing vibration and the pressure field around the compressor blades. They found that the casing vibration can be correlated with unsteady pressure field around blades and thus provide a clearer picture of the blade in the interior of the casing as compared to conventional vibration signals. Barragan [14] reported that through a detailed comparison of vibration spectrum against a vibration spectrum library of severe faults such as FOD, blade loss part, blade rub, and loose joints, blade fault could be detected. Chang and Chen [15] studied the detection of cracked blades by analyzing vibration signals with a spatial wavelet analysis method. They found that the position of the cracked blade can be identified based on the proposed method. Zielinski and Ziller [16] presented a noncontact blade vibration measurement method on axial flow turbine compressor blades. This method offered a better means of detecting cracks in rotor blades. Nevertheless, it is highlighted by Baines [17] that vibration analysis could only detect the blade faults if severe damage occurs at the blade. A minor deformation of a few blades, such as rubbing, will not be detected. Hee and Leong [18] used vibration signal to generate operational deflection shape (ODS) of rotor casing to detect blade rub. A comprehensive review on the use of vibration as a diagnostic tool for blade fault has been done by the authors and can be found in [19].

2.1.2. Strain Gauge Measurements

Strain gauges are typically used to measure the amount of deformation on the surface of a turbine blade. Scalzo et al. [20] employed the strain gauge method to monitor the stress profile of blades of an industrial gas turbine. It is reported that this method enables the monitoring of blade fatigue failures by studying the characteristic of flow-induced resonant vibration. It is well understood that a crack presence in a turbine blade will alter its natural frequency and a cracked blade may thus experience resonance conditions even at its normal operating condition. Mercadal et al. [21] proposed the use of noncontacting stress monitoring system (NSMS) to monitor blade’s resonance whereby a damaged blade is found to have shifted in its resonant frequencies. A method of using strain gauge to monitor low cycle fatigue faults in turbine blade was also proposed by Kumar et al. [22]. Barschdorff and Korthauer [23] commented that the deployment of this method in operational gas turbine may face some challenges as the strain gauge device may fail due to the extremely high temperature in the gas turbine.

2.1.3. Pressure Measurements

The use of pressure field distortion around rotating blades was proposed by Mathioudakis et al. [13] to diagnose blade deformation faults. It was shown that the pressure transducer signals mounted in the inner casing of the industrial gas turbine provide better information about the condition of fouled and twisted blades than the vibration analysis. This method has also been studied by Barschdorf and Korthauer [23] and Valero and Esquiza [24]. However, it is reckoned that this method is difficult to employ for operational gas turbines due to the difficulty of mounting pressure sensors at the surface of turbine blades.

2.1.4. Acoustic and Acoustic Emission Measurements

Acoustic emission techniques have proven to be an effective tool for early fault detection in rotating machinery [25]. Leon and Trainor [26] presented an online blade fault diagnostic method based on acoustic signals. This method utilized internal mounted acoustic sensors to identify the occurrence and location of propagating blade cracks, blade row damage, blade flutter, and aerodynamic events such as condensation shock. Willsch et al. [27] introduced a new technique for online monitoring of blade cracks and spalling. This technique involves the use of acoustical waveguides, optical waveguides, and millimeter waves. Nonlinear acoustics measurements techniques were then employed to evaluate fatigue in turbine blades. Nonlinear acoustics measurements techniques were also applied by Hinton et al. [28] to monitor blades fatigue. Randall and Rocketdyne [29] and Graham et al. [30] introduced blade faults diagnostic method based on acoustic emission monitoring to detect faults such as blade-to-stator rubbing, loose turbine disks, and blade cracking. The use of acoustic emission (AE) in monitoring gas turbine operating parameters is also investigated by Douglas et al. [31]. An experimental study done by Banov et al. [32] and Urbach et al. [33] showed that when fatigue crack has become sufficiently long, AE techniques can be used to effectively detect cracks of blades at a much more early stage than vibration analysis.

2.1.5. Debris Monitoring

Debris monitoring can be used to monitor the electrostatic and debris caused by blade rubbing. This method has been studied by Cartwright and Fisher [34] and Fisher [35]. In this technique, electrostatic sensors mounted in the turbine gas path were used to monitor the electrostatic charge that grows with blade rubbing. Besides, Powrie and McNicholas [36] also used this method to monitor the electrostatic charge and debris in the exhaust gas stream in order to detect blade rub and combustion chamber deterioration.

2.1.6. Blade Tip Monitoring

Optical measurement of blade vibration has been used for blade tip clearance monitoring by Simmons et al. [37]. This is done by installing an optical blade tip sensor at a spot on the blade surface. The probe of the fiber optic bundle will then image the spot. Any change in the distance between the blades tip and probe would cause the images spot on the fiber optic to move across the face of the bundle, and the distance is directly related to the change in blade to the probe. This method has been tested on an experiment rig and could be used to monitor blade tip clearance. This method can therefore be used to provide early detection of blade rubbing in gas turbine. Von Flotow et al. [38] used capacitance blade tip sensor to detect parameters such as the time of arrival and angle of arrival and thus provide information to detect various types of blade faults such as loss of blade, crack, rubbing, and bend blade. A powerful and reliable technique for blade tip clearance measurements that can be used under extremely harsh environments is presented by Steiner [39]. This method utilizes a heavy duty blade tip probe to detect shaft eccentricity or blade oscillations and thus provides information on the condition of blade. This method has also been studied by Sheard [40] among others.

2.1.7. Temperature Monitoring

In this method, blades running hotter than other blades can be detected and thus provide useful information for the prediction of creep. A blade temperature measurement system was also developed by Land Infrared Inc. This system is claimed to be capable of giving early warning of potential blade failures caused by overheating, such as creep. Annerfeldt et al. [41] presented a thermocrystal technique for monitoring the temperature of turbine blades and vanes. The technique was shown to be reliable and to have good accuracy in temperature gradients measurements. The study of heat transfer and stress distributions on a gas turbine blade was investigated by Kim et al. [42]. It was reported that the highest temperature on a turbine blade is located at the stagnation point of the blade leading edge. The point of maximum stress on a turbine blade is at its root. Turbine blade failures could therefore be estimated by monitoring the total stress resulting from combination of thermal load and cooling effects. A commercial online Thermal Barrier Coating (TBC) blade monitoring system has also been developed by Siemens Westinghouse Power Corporation Engineering in collaboration with Siemens Corporate Research [43]. The technique combines innovative access port design, high-speed infrared imagery, a tailor-made overall control and image evaluation system, and related TBC lifting models. The presented online technique is able to capture two-dimensional quantitative infrared images of row-1 blades during full engine operation and can help in not only ensuring the safe blades operation but also extending the blade’s life.

2.1.8. Performance Monitoring

Performance monitoring involves the acquisition of a variety of data (temperature, pressure, and speed) located along the gas turbine and then calculating the performance of the parameters such as mass flow rate and compressor and turbine operating efficiency. Performance monitoring methods are a valuable tool to detect blade fouling and rotating stall. These faults usually will cause some aerodynamics distortion on blades and affect the overall performance of the machine. This method has been studied by Dundas [44], Dundas et al. [45], and Meher-Homji and Boyce [46] amongst others. An algorithm for blade fault diagnosis using hybrid method (performance and vibration monitoring) was developed by Kubiak et al. [47]. It is shown that by using this hybrid algorithm, blade faults such as the wearing-out effects of the blade and blade fouling could be detected. Table 1 provides a summary of the blade condition monitoring methods discussed in the proceeding section.

tab1
Table 1: Summary of blade condition monitoring methods.
2.2. Advanced Signal Analysis Techniques
2.2.1. Wavelet Analysis

Aretakis and Mathioudakis [48] proposed a blade fault diagnostic method based on wavelet analysis. Wavelet analysis is performed on the time signals of the casing vibration, unsteady pressure, and acoustic measurements taken from commercial gas turbines. They highlighted that a distinctive difference could be seen from the wavelet map of healthy and faulty signals around BPF level. They have also demonstrated that each type of blade faults (fouled blade, twisted blade, and mistuned stator blade) could generate a unique signature of wavelet pattern based on the pressure signals. The authors investigated the feasibility of wavelet analysis for multistage blade faults diagnosis [49]. In another research paper [50] the current authors have also explored the usage of wavelet analysis for loose blade detection. Loose blade was found to be only detectable during the rotor coast-down process. Application of wavelet analysis in monitoring the different types of blade faults such as rub due to rotor eccentricity or creep has also been presented by the authors of [51]. It was showed that wavelet analysis can be used to reveal useful information about the blade condition and also to diagnose root cause of blade faults. Yuan et al. [52] used the wavelet analysis method to detect cracked turbine blades. It was demonstrated that both the Short Time Fourier Transform (STFT) and Mallat’s wavelet transform could be used to obtain the characteristics of a cracked blade based on the high frequency impacts signals found on the STFT and Mallat’s wavelet. Peng et al. [53, 54] reported on application of wavelet scalogram and wavelet phase spectrum to detect blade rubbing and rubbing-caused impacts. They have found that these techniques can be used to better detect rub impact signal as compared to FFT analysis. A blade fault diagnosis approach is also proposed by Qing-Yang et al. [55] by using the combination of the Morlet wavelet transform and neural network for feature recognition purpose. Wang et al. [56] also applied lifting wavelet transformation to detect blade faults. The current authors also studied the various applications of wavelet analysis technique [57, 58] to analyze vibration signal for blade rubbing detection.

2.2.2. Artificial Intelligence and Pattern Recognition Techniques

Probabilistic neural networks method was presented by Kyriazis et al. [59] to diagnose various types of blade faults. Angelakis et al. [60] proposed a method of using neural networks to diagnose blade faults in gas turbines. In his study, neural network-based fault diagnosis was used as pattern recognition tool to discriminate the patterns of faulty from healthy blades using signals measured from twelve different sensors or signals such as vibration and pressure signals. Loukis et al. [61] presented an automated blade faults detection method based on spectral pattern analysis technique. They have incorporated an innovative feature into a computer expert system to automatically detect and identify the type of fault in gas turbine by deriving the values of discriminants calculated from spectral patterns of fast response measurement signals such as casing vibration, internal pressure, and acoustic emission. It was found that blade fouling, blade twisting, and stator blade restaggering could be detected automatically based on this method. Aretakis et al. [62] presented a hybrid method to diagnose compressor blade fouling faults based on the combination of pattern recognition techniques known as the geometric and statistical pattern recognition technique. These pattern recognition techniques were shown to be viable to diagnose minor blade faults. Dedoussis et al. [63] presented a method based on the numerical simulation of blade fault signatures from unsteady wall pressure signals. This method produced the blade fault signatures of gas turbine using only theoretical computations. They employed a panel technique for the calculation of the flow field around blade cascades. From the calculation, time waveform of the pressure signal at a location on the casing wall facing the rotating blades is developed. By processing the faulty time, waveform signal will enable the construction of the faulty spectra signatures. This method gives rise to the possibility of establishing the blade fault signatures without having to perform any experiments at all.

2.2.3. Computational Fluid Dynamics (CFD)

Aretakis et al. [62] demonstrated the application of CFD to derive blade faults signatures. It was reported that this method can thus provide a basis for blade faults identification in gas turbine and compressors. Stamatis et al. [64] presented a study of blade faults diagnosis in gas turbine based on the CFD method. In this paper, a measured quantity was used as an input for physical modeling and the physical blade configuration is produced as the final output. This method provides a direct geometrical picture of the blade faults and does not rely on the interpretation of conventional fault signatures. Besides, Koubogiannis et al. [65] also used CFD method to create a database for gas turbine blade fault diagnosis solely based on signal modeling approach. They have shown that blade fault signature databases could be created without the need to conduct any costly experiments by using the method of unstructured and parallel CFD processing. CFD method was also investigated by Yokoyama et al. [66]. Hameed and Manarvi [67] also used CFD to predict crack locations at turbine blades surface.

2.2.4. Statistical Analysis

Romessis et al. [68] proposed the use of probabilistic reasoning to derive statistical information. It was showed that this method could be used to detect component faults (i.e., blade related faults) of jet engine. Loukis et al. [69] developed an automated method for gas turbine blade fault diagnosis based on the principals of statistical pattern recognition. The decision-making feature is based on the derivation of spectral patterns from dynamic measurement data. The calculations of discriminate coefficients with respect to the reference spectral patterns of the faults were done by taking into account their statistical properties. In this method, the success rate of automated decision making is further improved, and the need for intuitive discriminate selection is eliminated. Zhang et al. [70] presented a signal recognition method to recognize the blade fault based on the symmetry and similarity principal by making use of the means, variance, and means of variance to eliminate the effect caused by the parameter differences between blades. The plotting of means of variances showed that it can be used to distinguish the faulty from the healthy blade.

3. Condition Monitoring of Different Types of Blade Faults

In this section, the most frequently occurring blade faults in turbomachinery are categorized into five types, namely, blade rubbing, blade fatigue failure, blade deformation (twisting, creeping, corrosion, and erosion), blade fouling, and loose blade. A review of the most suitable method to detect each type of blade faults is presented in the following section.

3.1. Blade Rubbing

The occurrence of blade rubbing in turbomachinery has become more prevalent with the advent of high performance turbomachinery design. This is because the primary design consideration of these machines is to minimize the operational clearances between rotating blades and casing in order to increase cycle efficiencies and thus reduce the overall fuel consumption [71]. According to Barschdorff and Korthauer [23], blade rubbing and blade fatigue failure were reckoned to be the most prevalent blade faults in gas turbine with 23% of total blade failures in gas turbines contributed by rubbing whist 18.5% by blade fatigue failure (i.e., crack, loss of part, and FOD). The consequences of blade rubbing could be very serious as it can lead to other more destructive failures in machinery such as FOD due to broken blade parts. Figure 1 illustrates the effect of blade rubbing on the surface of compressor rotor.

210717.fig.001
Figure 1: Marks of blade rubbing on compressor rotor.

To date, abundance of studies have been conducted to understand the effects and mechanisms of blade rubbing in turbomachinery. Choy and Padovan [72] investigated the characteristics of the nonlinear dynamics of rubbing and have established the relationship of various parameters of rubbing excitation such as the relationship of rub force and energy levels with rubbing duration and incidence separation angles. Laverty [73] studied the mechanics of rubbing between a compressor blade tip seals and rotor casing. He found that the total energy of rubbing is mainly contributed by the incursion rate of the rubbing as compared to rubbing velocity and the thickness of the blade. He concluded that the overall rubbing energy increased in proportion to the quantity of blades that are involved in the process of rubbing. Sawicki et al. [74] studied the dynamic behavior of rotors rubbing and found that the vibration spectrum of rubbing is mainly dominated by subharmonic, quasiperiodic, and chaotic vibration components. Ahrens et al. [75] conducted an experimental study to investigate the resulting contact forces (in radial and tangential direction) during the process of rubbing. Roques et al. [76] formulated a mathematical rotor-stator model of a turbogenerator in order to study the speed transients and angular deceleration associated with rubbing. These reports, amongst others, have provided a deeper understanding of the mechanics and mechanisms of rubbing and thus enabled a better interpretation of rub related observations and signals in relation to their actual physical condition that occurred. Blade rubbing detection in turbomachinery is often accomplished by establishing the vibration symptoms of rubbing from time domain (i.e., vibration waveform, orbit plot), frequency domain (i.e., FFT), and also time-frequency domain (i.e., wavelet and STFT) signal analysis. A literature review by Muszynska [77] provided exhaustive information related to vibration, rotor dynamics, and the resulting rotor orbit during rubbing. This information could be used as a reference to detect blade rubbing in turbomachinery. Kubiak et al. [78] highlighted that blade rubbing could be detected if the blade passing frequency (BPF) amplitude is found to be exceptionally high in the vibration spectrum. Beside this, the presence of the abnormal frequency harmonics peaks (2x, 3x, 4x, etc.) of the operating speed in the vibration spectrum could also indicate the occurrence of blade rubbing. A drastic escalation in rotor subharmonic peaks in the vibration spectrum could also infer the presence of blade rubbing as reported by Meher-Homji [79]. Patel and Darpe [80] studied the early detection of rubbing in turbomachinery based on the vibration signal measured during a coast-up of a machine. They found that the Hilbert-Huang signal analysis could be applied to detect early rubbing in a rotor system based on the vibration signal measured during the coast-up of a machine. Wavelet analysis method was also widely used to detect rubbing in turbomachinery. For instance, Peng et al. [54] conducted a study to determine the effectiveness of using the conventional scalograms as compared to the reassigned scalograms for the detection of rubbing. They found that when rubbing occurred, its rubbing impacts could lead to an increase in vibration amplitude at high frequency region. They concluded that the vibration amplitude peaks of high frequencies region increase in correspondence with the increase of severity of rubbing. Wang and Chu [81] proposed a method to determine the location of rubbing in a rotor bearing system based on acoustics emission and wavelet analysis. Peng et al. [82] used wavelet analysis as a means for feature extraction of rubbing impact signal in a rotor system. Hee and Leong [83] conducted an experimental study to understand the dynamic responses of casing deflection profile due to different configurations of blade rubbing and found that the detailed analysis of amplitude and phase angle of the casing vibration could be used for blade rubbing classification. Leong and Lim [84] proposed a method that involved wavelet analysis technique to detect blade looseness and monitoring blade pass frequency signals for blade rubbing and cracks identification. Cong et al. [85] reported the use of an impact energy model for the evaluation of rotor rub impact fault where spectrum analysis can show a distinctive pattern for different severity states of the blade rub.

Besides vibration analysis, acoustic emission method was also proposed by Mba and Hall [86] to detect blade rubbing. They found that the stress waves and acoustic emission produced during the frictional rubbing process can propagate across the turbine surface and across the bearing interface allowing measurement to be made by the sensor attached to bearing housing. Meher-Homji [87] commented that the simple acoustic measurement could also be used to detect blade rubbing using stethoscopes during gas turbine startup and shutdown and this technique has already been adopted by gas turbine operator. Fisher [35] commented that many techniques such as the performance and vibration monitoring are dependent on a secondary effect being produced by the rubbing fault. For instance, blade rubbing can only be detected by these techniques provided sufficient material loss has occurred and results in an imbalance or loss of performance. He has adopted an engine distress monitoring system (EDMS) to monitor in real time the debris produced by the gas path component, such as blade rubbing and combustor faults. He has found that this method is capable of providing a more direct indicator of blade rubbing as the measurement made based on the actual debris produced by the fault probably before remedial action is required, thus giving a sufficient prognosis period. Simmons et al. [37] presented a method for blade tip clearance monitoring using an optical probe and found that this method could provide early indication of a blade rubbing if the measured blade tip clearance has reduced considerably.

3.2. Blade Fatigue Failure (Crack, Loss of Part, and FOD)

Blade fatigue failures in turbomachinery are generally caused by high and low cycle fatigue of the blades. High cycle fatigue in turbine blades is often caused by aerodynamic excitations or self-excited vibration (flutter). High cycle fatigue damage occurrs when stress levels are above its fatigue strength. In contrast, low cycle fatigue occurs as a result of frequent start-stop cycle of a machine which can lead to crack in bores and bolt hole areas of compressor and turbine disks that operate under high centrifugal stress. In this situation, a minute flaw could grow into crack which upon attaining critical size could result in total rupture of blade. In addition, crack can also be caused by resonant fatigue in blade. Resonant fatigue is an important failure mechanism which arises when a periodic force acts at a frequency corresponding to a blade natural frequency. If the damping is inadequate, crack will eventually develop and propagate and total blade failure could occur as commented by Meher-Homji [87]. A nonintrusive measurement method for measurements of torsional vibration signals for blade crack detection was also studied by Maynand and Trethewey [8890].

It is well understood that any damage in turbine blades, including cracks, can cause shift in blade’s resonance frequencies. Therefore the monitoring of the blade resonance condition is often adopted to detect a crack in blade. Lackner [91] presented a crack detection method for compressor blades in gas turbine engine using eddy current sensors. He observed that crack in blade was found to lower the resonant frequency of the first torsion mode of the blade. Mercadal et al. [21] employed a method using a noncontacting stress monitoring system to monitor blade resonance condition. Various methods of nondestructive testing such as eddy current and ultrasound were also often used to detect crack in turbine blades. Telemetry measurement method to monitor direct blade vibration was also used and studied by Scalzo et al. [20] and Mercadal et al. [21]. However, the methods of monitoring blade vibration using eddy current and strain gauge seem to be only applicable during machinery outage condition.

Vibration analysis is often employed in the field to monitor health condition of turbomachinery. However, vibration analysis is only capable of detecting fatigue blade failure if there is massive loss of blade part that could cause the changes in rotor dynamics characteristics of the machine. Meher-Homji [87] presented a case study whereby severely damaged first-stage turbine blades could not even show up in its vibration spectrum. Doftman and Moroslav [92] presented in their paper that a crack at the turbine blade root could be caused by the fatigue initiated by the rotor torsional vibration. Torsional vibration is generally a sporadic, transient phenomenon provoked by a sudden load change on the grid. Hence, the measurement and monitoring of the torsional vibration could be used to provide early warning for crack at turbine blade root. This method was also studied by Maynard and Trethewey [93].

3.3. Blade Deformation (Creep, Twisting, Erosion, and Corrosion)

Blade deformation in turbomachinery could occur due to creep, erosion, corrosion, and FOD induced blade twisting. Figure 2 depicts photographs of compressor blade deformation. Generally, blade deformation could cause obstruction in the flow of a machine and therefore could be detected by measuring the pressure field around the blades. This method has been proposed by Mathioudakis et al. [13] and Dedoussis et al. [63]. As shown by Aretakis and Mathioudakis [48], the distortion of pressure signals around the blades could indicate the occurrence of twisted blade. Beebe [94] also commented that the erosion of the blade due to the solid particles could also be detected using the performance monitoring method when gas turbine efficiency has dropped substantially. Beside this, a technology developed by GE power system known as taggant coating system can be used to detect any blade corrosion or erosion of blade. Any detection of the coated material at exhaust collector can be used as the indication of the presence of corrosion and erosion of the Mann [95].

fig2
Figure 2: Photographs of blade deformation.
3.4. Blade Fouling

Blade fouling is an important mechanism leading to performance deterioration in turbomachinery over time. Kurz et al. [96] explained that fouling is normally caused by the adherence of particles to airfoils and annulus surfaces. A case study of the causes and effects of fouling on gas turbine operation can be found in Meher-Homji [97, 98]. He found that compressor blade fouling could be detected based on performance monitoring method. The solids or condensing particles in the air and combusted gasses could precipitate on the rotating and stationary blade, thus causing some changes to the aerodynamic profile and invariably dropping the compressor mass flow rate and affecting the turbine flow coefficient. Therefore the dropping of the overall power output and thermal efficiency of the gas turbine could indicate the presence of the fouled blade. Mathioudakis et al. [13] presented that the measurement of the pressure field around rotating blades in turbomachinery is a more sensitive method as compared to performance monitoring to detect fouled blade. However, pressure monitoring method is difficult to be applied for gas turbine and compressor during actual operation.

3.5. Loose Blade

Loose blade in turbomachinery is a special condition whereby it usually involves not only the blade itself but also the mechanism of blade locking or blade attachment design. The authors Lim and Leong [50] have encountered a loose blade condition in one of the power stations in Malaysia. An experimental study of loose blade found that loose blade is not detectable under normal operation condition due to centrifugal force. Instead, it can only be detected during rotor coast-down stage by observing the impactive signals caused by the loose blade. Beside this, Kuo [99] also conducted a study on the diagnosis of the loose blade with Fourier’s analysis method on vibration signal. He used neural networks and fuzzy logic methods to develop a pattern recognition algorithm to enhance the detection of loose blade using vibration analysis.

4. Conclusion

Blade faults remain as one of the most elusive and frequently occurred faults in turbomachinery. The selection of blade monitoring method depends largely on the characteristics of blade faults as well as the practicality of the measurement itself. In summary, vibration analysis and BPF method can be used effectively for the detection of severe blade rubbing. For minor or early blade faults detection, pressure measurements, blade tip monitoring, acoustic emission, and debris monitoring should be used in lieu of vibration measurements. For the detection of blade fatigue failures, strain gauge measurements and blade vibration monitoring methods can be used to monitor any change in blade resonance frequency. For blade deformation and blade fouling detections, performance monitoring was found to be the most effective method available to date. This paper aims to provide an overview of the most feasible approaches for condition monitoring of blade in turbomachinery.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

This work is supported by the ERGS Grant (R.K130000.7840.4L089) financed by the Ministry of Education of Malaysia.

References

  1. G. H. Farrahi, M. Tirehdast, E. M. Khalil Abad, S. Parsa, and M. Motakefpoor, “Failure analysis of a gas turbine compressor,” Engineering Failure Analysis, vol. 18, no. 1, pp. 474–484, 2011. View at Publisher · View at Google Scholar · View at Scopus
  2. E. Poursaeidi, A. Babaei, F. Behrouzshad, and M. R. Mohammadi Arhani, “Failure analysis of an axial compressor first row rotating blades,” Engineering Failure Analysis, vol. 28, pp. 25–33, 2013.
  3. T. J. Carter, “Common failures in gas turbine blades,” Engineering Failure Analysis, vol. 12, no. 2, pp. 237–247, 2005. View at Publisher · View at Google Scholar · View at Scopus
  4. J. S. Mitchell, “examination of pump cavitation, gear mesh and blade performance using external vibration characteristics,” in Proceedings of the 4th Turbomachinery Symposium, Texas A & M University, 1975.
  5. H. Simmons, “A non-intrusive method for detecting HP turbine blade resonance,” ASME Paper 86-JPGC-Pwr-36, 1986.
  6. H. Simmons, “Nonintrusive detection of turbine blade resonance,” in Proceedings of the 3rd EPRI Conference on Incipient Failure Detection in Power Plants, Philadelphia, Pa, USA, 1987.
  7. P. Parge, B. Trevillion, and P. Carle, “Machinery interactive display and analysis system description and applications,” in Proceedings of the 1st International Machinery Monitoring and Diagnostic Conference, pp. 176–182, Las Vegas, Nev, USA, September 1989.
  8. P. Parge, B. Trevillion, and P. Carle, “Non-intrusive vibration monitoring for turbine blade reliability,” in Proceedings of the 2nd International Machinery Monitoring and Diagnostic Conference, pp. 435–446, Los Angeles, Calif, USA, October 1990.
  9. F. Al-Badour, M. Sunar, and L. Cheded, “Vibration analysis of rotating machinery using time-frequency analysis and wavelet techniques,” Mechanical Systems and Signal Processing, vol. 25, no. 6, pp. 2083–2101, 2011. View at Publisher · View at Google Scholar · View at Scopus
  10. A. M. Abdelrhman, M. S. Leong, L. M. Hee, and K. H. Hui, “Vibration analysis of multi stages rotor for blade faults diagnosis,” Advanced Materials Research, vol. 845, pp. 133–137, 2014.
  11. M. Satyam, V. S. Rao, and C. G. Devy, “Cepstrum analysis—an advanced technique in vibration analysis of defects in rotating machinery,” Defence Science Journal, vol. 44, pp. 53–60, 1994.
  12. R. Randall and N. Sawalhi, “Use of the cepstrum to remove selected discrete frequency components from a time signal,” Rotating Machinery, Structural Health Monitoring, Shock and Vibration, vol. 5, pp. 451–461, 2011.
  13. K. Mathioudakis, A. Papathanasiou, E. Loukis, and K. Papailiou, “Fast response wall pressure measurement as a means of gas turbine blade fault identification,” Journal of Engineering for Gas Turbines and Power, vol. 113, no. 2, pp. 269–275, 1991. View at Scopus
  14. J. M. Barragan, “Engine vibration monitoring and diagnosis based on on-board captured data,” DTIC Document, 2003.
  15. C. Chang and L. Chen, “Damage detection of cracked thick rotating blades by a spatial wavelet based approach,” Applied Acoustics, vol. 65, no. 11, pp. 1095–1111, 2004. View at Publisher · View at Google Scholar · View at Scopus
  16. M. Zielinski and G. Ziller, “Noncontact blade vibration measurement system for aero engine application,” in Proceedings of the International Symposium of Air Breathing Engines (ISABE '05), 2005.
  17. N. Baines, “Modern vibration analysis in condition monitoring,” in Proceedings of the Noise and Vibration Control Worldwide, May 1987.
  18. L. M. Hee and M. Leong, “Experimental study of dynamic responses of casing deflection profile for blade rubbing classification,” Journal of Vibroengineering, vol. 14, no. 4, pp. 1668–1680, 2012.
  19. A. M. Abdelrhman, M. S. Leong, S. A. M. Saeed, and S. M. Al Obiadi, “A review of vibration monitoring as a diagnostic tool for turbine blade faults,” Applied Mechanics and Materials, vol. 229, pp. 1459–1463, 2012.
  20. A. J. Scalzo, J. M. Allen, and R. J. Antos, “Analysis and solution of a nonsynchronous vibration problem in the last row turbine blade of a large industrial combustion turbine,” Journal of Engineering for Gas Turbines and Power, vol. 108, no. 4, pp. 591–598, 1986. View at Scopus
  21. M. Mercadal, A. von Flotow, and P. Tappert, “Damage identification by NSMS blade resonance tracking in mistuned rotors,” in Proceedings of the 2001 IEEE Aerospace Conference, pp. 73263–73277, March 2001. View at Scopus
  22. S. Kumar, N. Roy, and R. Ganguli, “Monitoring low cycle fatigue damage in turbine blade using vibration characteristics,” Mechanical Systems and Signal Processing, vol. 21, no. 1, pp. 480–501, 2007. View at Publisher · View at Google Scholar · View at Scopus
  23. D. Barschdorff and R. Korthauer, “Aspects of failure diagnosis on rotating parts of turbomachinerys using computer simulation and pattern recognition methods,” in Proceedings of the International Conference on Condition Monitoring, Brighton, UK, May 1986.
  24. M. Valero and E. Esquiza, “Detection of incipient damage in rotating machinery by noise analysis,” in Proceedings of the Noise and Vibration Control Worldwide, 1988.
  25. S. M. A. Al-Obaidi, M. S. Leong, R. Hamzah, and A. M. Abdelrhman, “A review of acoustic emission technique for machinery condition monitoring: defects detection and diagnostic,” Applied Mechanics and Materials, vol. 229, pp. 1476–1480, 2012.
  26. R. L. Leon and K. Trainor, “Monitoring systems for steam turbine blade faults,” Sound and Vibration, vol. 24, no. 2, pp. 12–15, 1990. View at Scopus
  27. M. Willsch, T. Bosselmann, and N. M. Theune, “New approaches for the monitoring of gas turbine blades and vanes,” in Proceedings of the 2004 IEEE Sensors, pp. 20–23, October 2004. View at Scopus
  28. Y. L. Hinton, J. K. Na, W. T. Yost, and G. L. Kessel, Field Measurement of the Acoustic Nonlinearity Parameter in Turbine Blades, NASA Langley Technical Report Server, 2000.
  29. R. L. Randall and Rocketdyne, “Acoustic emission monitoring of steam turbines: final report,” in Proceedings of the Joint ASME/IEEE Power Generation Conference, Louis, Mich, USA, 1981.
  30. L. J. Graham, R. L. Randall, and C. Hong, Acoustic-Emission Monitoring of Steam Turbines, Final Report Rockwell International Corp, Energy System Group, Canoga Park, Calif, USA, 1982.
  31. R. Douglas, S. Beugné, M. Jenkins et al., “Monitoring of gas turbine operating parameters using acoustic emission,” in Proceedings of the 2004 EWAGE Conference, pp. 389–396, 2004.
  32. M. D. Banov, E. A. Konyaev, and D. A. Troenkin, “Procedure for estimating the fatigue strength of gas turbine blades by method of acoustic emission,” The Soviet Journal of Nondestructive Testing, vol. 18, no. 8, pp. 600–603, 1982. View at Scopus
  33. A. Urbach, M. Banov, V. Turko, and Y. Feshchuk, Diagnostics of Fatigue Damage of Gas Turbine Engine Blades by Acoustic Emission Method, pp. 906–911, 59, World Academy of Science, Engineering and Technology, Venēcija, Italy, 2011.
  34. R. A. Cartwright and C. Fisher, “Marine gas turbine condition monitoring by gas path electrostatic detection techniques,” in Proceedings of the International Gas Turbine and Aeroengine Conference, Orlando, Fla, USA, June 1991. View at Scopus
  35. C. Fisher, “Gas turbine condition monitoring systems: an intergrated approach,” in Proceedings of the 2000 IEEE Aerospace Conference, 2000.
  36. H. Powrie and K. McNicholas, “Gas path condition monitoring during accelerated mission testing of a demonstrator engine,” in Proceedings of the 33rd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, Paper AIAA97-2904, Seattle, Wash, USA, 1997.
  37. H. R. Simmons, D. L. Michalsky, K. E. Brewer, and A. J. Smalley, “Measuring rotor and blade dynamics using an optical blade tip sensor,” in Proceedings of the 35th ASME International Gas Turbine and Aeroengine Congress and Exposition, Brussels, Belgium, June 1990. View at Scopus
  38. A. von Flotow, M. Mercadal, and P. Tappert, “Health monitoring and prognostics of blades and disks with blade tip sensors,” in Proceedings of the 2000 IEEE Aerospace Conference, vol. 6, pp. 433–440, March 2000. View at Scopus
  39. A. Steiner, “Techniques for blade tip clearance measurements with capacitive probes,” Measurement Science and Technology, vol. 11, no. 7, pp. 865–869, 2000. View at Publisher · View at Google Scholar · View at Scopus
  40. A. G. Sheard, “Blade by blade tip clearance measurement,” International Journal of Rotating Machinery, vol. 2011, Article ID 516128, 13 pages, 2011. View at Publisher · View at Google Scholar · View at Scopus
  41. M. Annerfeldt, S. Shukin, M. Björkman, A. Karlsson, A. Jönsson, and E. Svistounova, “GTX100 Turbine section measurement using a temperature sensitive crystal technique. A comparison with 3D thermal and aerodynamic analysis”.
  42. K. M. Kim, J. S. Park, D. H. Lee, T. W. Lee, and H. H. Cho, “Analysis of conjugated heat transfer, stress and failure in a gas turbine blade with circular cooling passages,” Engineering Failure Analysis, vol. 18, no. 4, pp. 1212–1222, 2011. View at Publisher · View at Google Scholar · View at Scopus
  43. D. H. LeMieux, On-Line Thermal Barrier Coating Monitoring for Real-Time Failure Protection and Life Maximization, National Energy Technology Laboratory (US), 2003.
  44. R. E. Dundas, “The use of performance-monitoring to prevent compressor and turbine failures,” in Proceedings of the ASME Gas Turbine Congress, 1982.
  45. R. E. Dundas, D. A. Sullivan, and A. Abegg, “Performance monitoring of gas turbines for failure prevention,” in Proceedings of the ASME International Gas Turbine and Aeroengine Congress, Cologne, Germany, June 1992.
  46. C. B. Meher-Homji and M. P. Boyce, “Aerothermodynamic gas path analysis for health diagnostics of combustion gas turbines,” in Proceeding of the 1982 MFPG Conference on Detection, Diagnosis and Prognosis, National Bureau of Standards, 1983.
  47. J. A. Kubiak, J. M. Franco, A. Carnero, A. Rothhirsch, and L. J. Aguirre, “Diagnosis of failure of a compressor blade,” in Proceedings of the ASME Gas Turbine Conference and Exhibition, ASME Paper, 1987.
  48. N. Aretakis and K. Mathioudakis, “Wavelet analysis for gas turbine fault diagnostics,” Journal of Engineering for Gas Turbines and Power, vol. 119, no. 4, pp. 870–875, 1997. View at Scopus
  49. A. M. Abdelrhman, M. S. Leong, L. M. Hee, and W. K. Ngui, “Application of wavelet analysis in blade faults diagnosis for multi-stages rotor system,” Advances in Manufacturing and Mechanical Engineering, vol. 393, pp. 959–964, 2013.
  50. M. H. Lim and M. S. Leong, “Diagnosis for loose blades in gas turbines using wavelet analysis,” Journal of Engineering for Gas Turbines and Power, vol. 127, no. 2, pp. 314–322, 2005. View at Publisher · View at Google Scholar · View at Scopus
  51. L. M. Hee and M. S. Leong, “Improved blade fault diagnosis using discrete Blade Passing Energy Packet and rotor dynamics wavelet analysis,” in Proceedings of the 2010 ASME Turbo Expo, pp. 31–37, Glasgow, UK, June 2010. View at Scopus
  52. Q. Yuan, H. Xie, and Q. J. Meng, “Fault diagnosis of turbine blade by wavelet,” in Proceedings of the Asia Pacific Visitors Conference (APVC '99), Nanyang Technological University, Singapore, 1999.
  53. Z. Peng, Y. He, Q. Lu, and F. Chu, “Feature extraction of the rub-impact rotor system by means of wavelet analysis,” Journal of Sound and Vibration, vol. 259, no. 4, pp. 1000–1010, 2003. View at Publisher · View at Google Scholar · View at Scopus
  54. Z. K. Peng, F. L. Chu, and P. W. Tse, “Detection of the rubbing-caused impacts for rotor-stator fault diagnosis using reassigned scalogram,” Mechanical Systems and Signal Processing, vol. 19, no. 2, pp. 391–409, 2005. View at Publisher · View at Google Scholar · View at Scopus
  55. X. Qing-Yang, M. Xian-yao, H. Xin-Jie, and M. Song, “Gas turbine fault diagnosis based on wavelet neural network,” in Proceedings of the International Conference on Wavelet Analysis and Pattern Recognition (ICWAPR '07), pp. 738–741, November 2007. View at Publisher · View at Google Scholar · View at Scopus
  56. L. Wang, L. Ma, and Y. Huang, “The application of lifting wavelet transform in the fault diagnosis of reciprocating air compressor,” in Proceedings of the International Conference on Intelligent System Design and Engineering Application (ISDEA '10), pp. 574–576, October 2010. View at Publisher · View at Google Scholar · View at Scopus
  57. M. H. Lim and M. S. Leong, “Detection of early faults in rotating machinery based on wavelet analysis,” Advances in Mechanical Engineering, vol. 2013, Article ID 625863, 8 pages, 2013. View at Publisher · View at Google Scholar
  58. M. H. Leem and M. S. Leong, “Reconstruction of vital blade signal from unsteady casing vibration,” Advances in Mechanical Engineering, vol. 2013, Article ID 146983, 9 pages, 2013. View at Publisher · View at Google Scholar
  59. A. Kyriazis, N. Aretakis, and K. Mathioudakis, “Gas turbine fault diagnosis from fast response data using probabilistic methods and information fusion,” in Proceedings of the ASME 51st Turbo Expo: Power for Land, Sea & Air (GT '06), pp. 571–579, Barcelona, Spain, May 2006. View at Scopus
  60. C. Angelakis, E. N. Loukis, A. D. Pouliezos, and G. S. Stavrakakis, “A neural network-based method for gas turbine blading fault diagnosis,” International Journal of Modelling and Simulation, vol. 21, no. 1, pp. 51–60, 2001. View at Scopus
  61. E. Loukis, P. Wetta, K. Mathioudakis, A. Papathanasiou, and A. Papailiou, “Combination of different unsteady quantity measurements for gas turbine blade fault diagnosis,” in Proceedings of the ASME 36th International Gas Turbine and Aeroengine Congress and Exposition, Orlando, Fla, USA, June 1991. View at Scopus
  62. N. Aretakis, K. Mathioudakis, and V. Dedoussis, “Derivation of signatures for faults in gas turbine compressor blading,” Control Engineering Practice, vol. 6, no. 8, pp. 969–974, 1998. View at Scopus
  63. V. Dedoussis, K. Mathioudakis, and K. D. Papailiou, “Numerical simulation of blade fault signatures from unsteady wall pressure signals,” Journal of Engineering for Gas Turbines and Power, vol. 119, no. 2, pp. 362–369, 1997. View at Scopus
  64. A. Stamatis and K. Papailiou, “Using HPC in gas turbines blade fault diagnosis,” in Proceedings of the High-Performance Computing and Networking (HPCN Europe '97) Conference, pp. 11–20, Vienna, Austria, April 1997.
  65. D. G. Koubogiannis, V. P. Iliadis, and K. C. Giannakoglou, “Parallel CFD tool to produce faulty blade signatures for diagnostic purposes,” in Proceedings of the 1998 International Gas Turbine & Aeroengine Congress & Exhibition, June 1998. View at Scopus
  66. T. Yokoyama, Y. Hayashi, K. Osako, and Y. Komatsu, “Numerical and experimental study of turbine blade vibration in variable geometry turbochargers,” SAE Technical Paper 2005-01-1855, 2005.
  67. M. S. Hameed and I. A. Manarvi, “Using FEM and CFD for engine turbine blades to localize critical areas for non destructive inspections,” International Journal of Multidisciplinary Sciences and Engineering, vol. 2, no. 3, pp. 29–37, 2011.
  68. C. Romessis, P. Kamboukos, and K. Mathioudakis, “The use of probabilistic reasoning to improve least squares based gas path diagnostics,” Journal of Engineering for Gas Turbines and Power, vol. 129, no. 4, pp. 970–976, 2007. View at Publisher · View at Google Scholar · View at Scopus
  69. E. Loukis, K. Mathioudakis, and K. Papailiou, “Optimizing automated gas turbine fault detection using statistical pattern recognition,” Journal of Engineering for Gas Turbines and Power, vol. 116, no. 1, pp. 165–171, 1994. View at Scopus
  70. J. Zhang, G. Zhang, M. Xu, and B. Sun, “Signal recognition of blade faults,” Modelling, Simulation & Control B, vol. 54, no. 1-2, pp. 59–63, 1994. View at Scopus
  71. K. E. Turner, M. Dunn, and C. Padova, “Airfoil deflection characteristics during rub events,” ASME Journal of Turbomachinery, vol. 134, no. 1, Article ID 011018, 8 pages, 2012. View at Scopus
  72. F. K. Choy and J. Padovan, “Non-linear transient analysis of rotor-casing rub events,” Journal of Sound and Vibration, vol. 113, no. 3, pp. 529–545, 1987. View at Scopus
  73. W. F. Laverty, “Rub energetics of compressor blade tip seals,” Wear, vol. 75, no. 1, pp. 1–20, 1982. View at Scopus
  74. J. T. Sawicki, J. Padovan, and R. Al-Khatib, “The dynamics of rotor with rubbing,” International Journal of Rotating Machinery, vol. 5, no. 4, pp. 295–304, 1999. View at Scopus
  75. J. Ahrens, H. Ulbrich, and G. Ahaus, “Measurement of contact forces during blade rubbing,” in Proceedings of the Institution of Mechanical Engineers (IMechE '00), Paper No. C576/083/2000, pp. 259–263, 2000.
  76. S. Roques, M. Legrand, P. Cartraud, C. Stoisser, and C. Pierre, “Modeling of a rotor speed transient response with radial rubbing,” Journal of Sound and Vibration, vol. 329, no. 5, pp. 527–546, 2010. View at Publisher · View at Google Scholar · View at Scopus
  77. A. Muszynska, “Vibrational diagnostics of rotating machinery malfunctions,” International Journal of Rotating Machinery, vol. 1, pp. 237–266, 1995.
  78. S. J. Kubiak, R. G. González, G. A. García, J. G. Gómez Mancilla, and B. G. Urquiza, “Hybrid fault patterns for the diagnosis of gas turbine component degradation,” in Proceeding of the 2001 International Joint Power Generation Conference, New Orleans, La, USA, June 2005.
  79. C. B. Meher-Homji, “Blading vibration and failures in gas turbines: part B—compressor and turbine airfoil distress,” in Proceedings of the ASME Turbo Expo, Paper no. 95-GT-419, June 1995. View at Scopus
  80. T. H. Patel and A. K. Darpe, “Study of coast-up vibration response for rub detection,” Mechanism and Machine Theory, vol. 44, no. 8, pp. 1570–1579, 2009. View at Publisher · View at Google Scholar · View at Scopus
  81. Q. Wang and F. Chu, “Experimental determination of the rubbing location by means of acoustic emission and wavelet transform,” Journal of Sound and Vibration, vol. 248, no. 1, pp. 91–103, 2001. View at Publisher · View at Google Scholar · View at Scopus
  82. Z. Peng, Y. He, Q. Lu, and F. Chu, “Feature extraction of the rub-impact rotor system by means of wavelet analysis,” Journal of Sound and Vibration, vol. 259, no. 4, pp. 1000–1010, 2003. View at Publisher · View at Google Scholar · View at Scopus
  83. L. M. Hee and M. S. Leong, “Experimental study of dynamic responses of casing deflection profile for blade rubbing classification,” Journal of Vibroengineering, vol. 14, no. 4, pp. 1668–1680, 2012.
  84. M. S. Leong and M. H. Lim, “Blades rubs and looseness detection in gas turbines—operational field experience and laboratory study,” Journal of Vibroengineering, vol. 15, pp. 1311–1321, 2013.
  85. F. Cong, J. Chen, G. Dong, and K. Huang, “Experimental validation of impact energy model for the rubimpact assessment in a rotor system,” Mechanical Systems and Signal Processing, vol. 25, no. 7, pp. 2549–2558, 2011. View at Publisher · View at Google Scholar · View at Scopus
  86. D. Mba and L. D. Hall, “The transmission of acoustic emission across large-scale turbine rotors,” NDT and E International, vol. 35, no. 8, pp. 529–539, 2002. View at Publisher · View at Google Scholar · View at Scopus
  87. C. B. Meher-Homji, “Blading vibration and failures in gas turbines: part C—detection and troubleshooting,” ASME Paper 95-GT-420, 1995.
  88. K. P. Maynard and M. Trethewey, “Blade and shaft crack detection using torsional vibration measurements, part 1: feasibility studies,” Noise and Vibration Worldwide, vol. 31, no. 11, pp. 9–15, 2000. View at Publisher · View at Google Scholar · View at Scopus
  89. K. P. Maynard and M. Trethewey, “Blade and shaft crack detection using torsional vibration measurements, part 2: resampling to improve effective dynamic range,” Noise and Vibration Worldwide, vol. 32, no. 2, pp. 23–26, 2001. View at Scopus
  90. K. Maynard and M. Trethewey, “Blade and shaft crack detection using torsional vibration measurements, part 3: field application demonstrations,” Noise and Vibration Worldwide, vol. 32, no. 11, pp. 16–23, 2001. View at Publisher · View at Google Scholar · View at Scopus
  91. M. Lackner, Vibration and Crack Detection in Gas Turbine Engine Compressor Blades Using Eddy Current Sensors, Massachusetts Institute of Technology, 2003.
  92. L. S. Dorfman and M. Trubelja, “Torsional monitoring of turbine-generators for incipient failure detection,” in Proceedings of the 6th EPRI Steam Turbine/Generator Workshop, pp. 1–6, St. Louis, Mo, USA, August 1999.
  93. K. Maynard and M. Trethewey, “Blade and shaft crack detection using torsional vibration measurements, part 3: field application demonstrations,” Noise and Vibration Worldwide, vol. 32, no. 11, pp. 16–23, 2001. View at Publisher · View at Google Scholar · View at Scopus
  94. R. Beebe, “Condition monitoring of steam turbines by performance analysis,” Journal of Quality in Maintenance Engineering, vol. 9, no. 2, pp. 102–112, 2003. View at Publisher · View at Google Scholar · View at Scopus
  95. B. S. Mann, “Solid-particle erosion and protective layers for steam turbine blading,” Wear, vol. 224, no. 1, pp. 8–12, 1999. View at Scopus
  96. R. Kurz, K. Brun, and S. Mokhatab, “Compression roundup: gas turbine compressor blade fouling mechanisms,” Pipeline and Gas Journal, vol. 238, no. 9, pp. 18–21, 2011. View at Scopus
  97. C. B. Meher-Homji, “Gas turbine axial compressor fouling—a unified treatment of its effects, detection, and troubleshooting,” in Proceedings of the 1990 ASME Cogen Turbo IV, pp. 179–190, New Orleans, La, USA, August 1990. View at Scopus
  98. C. B. Meher-Homji, “Blading vibration and failures in gas turbines: part B—compressor and turbine airfoil distress,” ASME Paper 95-GT-419, 1995.
  99. R. J. Kuo, “Intelligent diagnosis for turbine blade faults using artificial neural networks and fuzzy logic,” Engineering Applications of Artificial Intelligence, vol. 8, no. 1, pp. 25–34, 1995. View at Scopus