Turbine systems play a crucial role in the production of wind energy. According to Allemang, De Clerck, Niezrecki, & Blough (2012), wind is currently the fastest growing renewable source of energy in the world. Nevertheless, the industry is still experiencing untimely turbine-related problems that result in increases in the cost of energy. With the increase in the size and installation of offshore turbines, these failures become considerably costly (Askeland, Fulay, & Bhattacharya, 2009). Thus, there is necessary for the industry to increase the reliability of turbines and to reduce downtime. In order to comprehend the premature failures of turbines and propose probable improvements, concerned bodies, such as the National Renewable Energy Laboratory, have initiated a conglomerate called the Gearbox Reliability Collaborative (GRC). Such consortiums aim at ensuring the reliability of turbine systems in the production of energy and other area of usage. It is through these groupings that the involved parties are able to come up with gearbox design, manufacture and maintenance techniques that have a common goal of improving reliability and extending the lifetime of turbines (Cho, Bode, & Kim, 2010). Condition Monitoring (CM) is one of the areas under research in the energy industry. It can substantially assist the industry in achieving the common objective of the improved turbine uptime. This is because it facilitates the best operation and maintenance practices. In other words, condition monitoring can target almost all its key subsystems, such as nacelle, blades, drivetrain, tower, and foundation. In this regard, the paper discusses condition-monitoring techniques used to detect on the turbine blades. The techniques discussed by the paper include: ultrasonic, thermo graphic, radiograph, and full matrix capture. Also, the paper discusses the blade characteristics that determine the appropriate technique in detecting defects. Notably, it will be of great significance to discuss the problems, such as corrosion and crack, which cause turbine malfunctions.
Condition Monitoring Techniques and their Characteristics
Condition Monitoring is mainly implemented using an integrated approach by coordinating various commercial equipment suppliers (Chou & Tu, 2009). It takes an integrated approach since no single technique can offer reliable and comprehensive solutions needed by the energy industry. The figure below shows the setup of the condition monitoring system. This system was designed in order to provide reliable diagnostics whenever a turbine drivetrain components start to fail. The stress represents the Acoustic Emission technique that covers the frequency range that lies above 20 kHz (East & Eisenmann, 2002). According to Star (2001), the inline particle counts represent the number of ferrous particles present in the main loop. The offline oil condition sensor measures the total ferrous debris in parts per million (ppm), oil quality in customized scale, and relative humidity in as a percentage. Various dynamometer tests have the same setup with very small and minor differences (Energy and Environmental Analysis, 2008).
Ultrasonic Flaw Detector
Theflaw detection is the most common and oldest application of ultrasonic testing (Faulstich, 2006). Since the 1940s, the principals of physics that explain the sound wave propagation through solids have been deployed in detecting hidden crack, porosity, voids, and internal discontinuities in metals, composites, plastics, and ceramics (Gallar, 2011). According to Star (2001), sound waves of high frequency reflect from defects in various predictable ways. Such sound waves produce characteristic echo patterns, which can be recorded and displayed by the portable devices. Ultrasonic testing is entirely safe and nondestructive, and it is a well-known detection method in various manufacturing, the service and processing industry, particularly in applications integrating structural and weld metals (Gallardo, Rodr%u0131guez, & Herrera, 2002).
The use of ultrasonic in detecting flaws in turbine blades borrows much from physics of sound waves. According to (Ginley & Cahen, 2011), sound waves refer to organized mechanical waves that travel through solid, liquid or gas substances that act as media. These waves travel through a certain medium at a particular speed and in a foreseeable direction. In addition, when the waves bump into a boundary having a different middling, they will undergo reflection according to certain principles of physics. It is these principles of physics that motivate the ultrasonic flaw detection. According to Star (2001), all sound waves oscillate at a given frequency that can be experienced as a pitch in an acquainted range of a noticeable sound. The normal frequency of human hearing spans to an extreme of 20 kHz; while the common of ultrasonic defect applications use frequencies that range from 500 kHz to 10 MHz (Donachie & Donachie, 2002). Sound waves do not travel effectively via other ordinary gases at frequencies in the range of megahertz, though it travels freely via most engineering materials. The speed at which sound waves travel changes based on the medium. The density of the medium and its elastic properties affect the speed of sound energy and various sound waves travel at the dissimilar velocities (Gipe, 2004). According to Star (2001), any form of waves travels has a related wavelength, which is the distance from one consistent point to another in the wave cycle.
Sound waves travelling in solids take various modes of propagation defined by the motion involved (Glover, Mulukutla, Sarma, & Overbye, 2011). For instance, shear and longitudinal waves are the most popular modes used in the ultrasonic detection. Additionally, plate and surface waves are also used in some occasions. A motion of particle perpendicular to the direction of wave propagation describes longitudinal waves. A motion of particle vertical to the bearing of wave propagation characterizes a shear or transverse wave. Star (2001) views a plate wave as a multifaceted vibration mode in thin plates with a wavelength of less than one. Sound waves can undergo conversions from one state to another. Shear waves are commonly generated in test materials, including turbine blades, by bringing together longitudinal waves at the chosen points (Hahn, Durstewitz, & Rohrig, 2005). Star (2001) also points out that sound energy at ultrasonic frequencies can help detect flaws in the turbine blades. At such frequencies, the sound energy is guiding and the beams deployed for detection. The angle of incidence equals the angle of reflection in scenarios where sound reflects off a boundary (Jørgensen, et al., 2004).
An ultrasonic transducer uses the above principles of physics to detect flaws in the turbine blades (Star, 2001). In broad terms, a transducer is just a device for converting energy from one state to another. They transform electrical energy into very high frequency sound energy and vice versa. A typical ultrasonic flaw detector uses an active component of piezoelectric ceramic, polymer or ceramic (Katherine & Miller, 2008). Very high electrical voltage excites this element to vibrate at a known frequency, generating an electric pulse. A wear plate, which safeguards the element, covers its front part. The rear part of the element is merged to a supporting material, which involuntarily reduces the vibrations once the generation of sound is complete. Since sound waves travelling at ultrasonic frequencies do not travel efficiently through gases, an extremely thin layer of liquid is used between the test piece and the transducer. The figure below shows a typical ultrasonic transducer used in detecting flaws in turbine blades (Nelligan, 2009).
There are five kinds of transducers widely used in flaw detection applications, such as turbine blade (Leizerovich & Le%u012Dzerovich, 2005). The first kind is the contact transducer, as suggested by the name, are commonly used in a straight contact with the turbine blades. These transducers introduce a sound wave that is vertical to the surface of the turbine blade. Contact transducers are extremely important in detecting porosity, voids, and delamination or cracks that are corresponding to the external surface of the blade (Meetham & Van de Voorde, 2000). The second kind of transducer is the angle beam transducer. These transducers are deployed along with epoxy or plastic wedges to introduce longitudinal waves into the turbine blade at a chosen angle. The angle beam transducers are commonly applicable in the weld inspection. The third kind of ultrasonic transducer is the delay line transducers that integrate a short delay line or a plastic between the test piece and the active element. Additionally, they are deployed to enhance the near surface resolution. According to Star (2001), angle beam transducers are appropriate for high temperature testing because the delay line safeguards the active element from thermal effects. The other type of the ultrasonic transducer is the immersion transducer designed to direct sound waves into the turbine blade through a column of water (Singh & Lucas, 2011). They are of great significance in the situations that require sharply focused beams to enhance flaw resolution and programmed scanning applications. The last type of the ultrasonic transducer is the dual element transducers that use the separate receiver and transmitter elements in a solitary assembly (Sohre, 2000). They are appropriate for detecting flaws in the turbine blades that have rough surfaces. It is worth noting that they have a high tolerance for high temperatures, as well.
Contemporary ultrasonic flaw detectors, including the panametrics-NDT Epoch series, are portable, small, and micro-processor-based. Because of this features, they are appropriate for both field and shop use (Meetham & Van de Voorde, 2000). They produce and display an ultrasonic waveform, and use analysis software to detect and classify flaws in the turbine blades and other test pieces. According to Star (2001), they include an ultrasonic receiver, as well as software and hardware that are used for capturing and analyzing signals, and displaying waveforms.
Active thermography, such as inductively and optically heated thermography, offers a non-contact and rapid inspection method for detecting flaws present in a plane that is parallel to a surface (Allemang, De Clerck, Niezrecki, & Blough, 2012). Active thermography method has been deployed successfully to enhance sensitivity and image contrast of active thermography for flaws in the turbine blades that have irregular surface or nonhomogeneous emissivity. Nevertheless, the current approach of actively heating a wide area of an object using power flash lamps seems not to be effective for the detection of cracks developing perpendicular to the surface (Askeland, Fulay, & Bhattacharya, 2009). The detection of such cracks in the metallic components and turbine blades are not exceptional, as it is a major issue for the energy industry. As much as the non-destructive (NDT) techniques, such as eddy currents and ultrasonic, are widely used, there is an ever-increasing interest in studying the capability of techniques based on thermo graphics to offer more efficient, rapid and convenient inspection method.
The infrared thermography techniques (IRT) have been developed to detect surface cracks (Glover, Mulukutla, Sarma, & Overbye, 2011). The IRT combines the use of thermography and laser scans. The introduction of the new IRT technique called the singular method was a major step towards identifying cracks on the turbine blades. In this method, the singular electro-thermal fields generated around crack tips identify cracks. The application of the periodically modulated electric current to a crack causes the singular current field to oscillate in the same frequency (Leizerovich & Le%u012Dzerovich, 2005). The cyclic change of the singular electro-thermal field leads to the cyclical changing of the distribution of temperature that can be imaged by the lock-in mode with the reference signal of the modulated electric current. Star (2001) calculated the current distribution prompted by the external coil to investigate the flow of current that is near the surface of the turbine blade. According to Star (2001), the cracks are noticeable through a direct observation of the heating process because of the concentrated density of current and modification of diffusion. In addition, Star (2001) also claims that one can visualize the surface cracks by deploying vibration-induced friction heating of the crack in titanium.
Another technique that is also an improvement of the traditional thermography is the laser scanning approach (Ginley & Cahen, 2011). This technique proved to be an extremely successful way of detecting cracks on the surface of turbine blades. The technique deploys a laser heat source for the excitation of the specimen where the resulting temperature of the surface is recorded with an infrared camera. The beam is concentrated on the test sample by using the optical scanner that generates the required lateral heat flow (Katherine & Miller, 2008). A high-speed infrared camera records the resolved time distribution of temperature. Based on the data collected, anisotropies that cause cracks can be noticed.
Forced diffusion thermography is also another approach used to detect cracks on the turbine blades. This technique projects a pattern of dynamic heat in order to force the flow across the crack, therefore optimizing measurable thermal gradient (Meetham & Van de Voorde, 2000). The elementary idea is that a crack impedes the heat flow, creating a gradient in the thermal image that clearly defines the crack. As gradients resulting from the variance of emissivity from the surface of the turbine blade can be misinterpreted as cracks, the deployment of the spatial crack derivative through applying two opposing heat flows to the surface can be useful (Katherine & Miller, 2008). This is because it enables the separation of the flaw image from the emissivity crack or noise. Nevertheless, the derivative of the thermal image magnifies the noise so that the gradient image does not provide better information than the raw thermal image.
Radiography is among the traditional and still effective non-destructive techniques used for detecting flaws. The most significant application of radiography is in the gas industry. It is used as a tool for inspecting welds in the pipelines (Sohre, 2000). With the development of online digital radiography that deploys sensitive fluorescent plates, offline analog radiography that deploys films is still beneficial. The rapid development of wind energy systems reveals the necessity for an automatic and reliable weld inspection system. The checking of low and poor quality films of welds is a time-consuming process for certified experts and reduces their eye sensitivity and inspection accuracy (Sherma & Fried, 2003). It is worth stressing that more than 60 per cent of radiographic images in industries are not defective, but require being isolated in visual check, the design of radiographic system should act as an efficient tool in reducing the mass work of the process of film checking. The process of extracting welds from the background image can be achieved manually through user’s selection or automatically by a computer (East & Eisenmann, 2002).
The inspection of welded turbine blades is necessary to ensure that the quality of the weld meets the provisions of the operations and design that, in turn, assures reliability and safety. The visual interpretation of radiographic films is extremely hard when a lot of flaws are to be detected and calibrated. In addition, the quality assessment by the involvement of humans raises the costs of energy produced by wind turbines (Askeland, Fulay, & Bhattacharya, 2009). Various experts also do not have a similar opinion about the given radiographic film since they depend on their previous experiences. In order to eliminate the slowness and inconsistency of the evaluation by humans in meeting operations with increased reliability, radiography is useful in inspecting welds in the turbine blades.
Radiographic testing has been one of the commonly used NDT technique used for detecting internal weld flaws (Totten, Westbrook, & Shah, 2003). The technique is based on the capability of X-rays to penetrate through metals that range from opaque to transparent and produce photographic images or records. According to Cho, Bode, and Kim (2010), the technique is also based on the fluctuating intensity of absorption of the radiation that produces the image of the object under the examination. The latent images undergo transformations into permanent shadow images of both external and internal condition of the object. The common weld flaws in the turbine blades include: the lack of fusion, gas cavities, and lack of penetration, cracks, slag inclusions, lamellar tearing, and shrinking cavities (Energy and Environmental Analysis, 2008).
The lack of fusion arises from very little heat input or very rapid transverse of welding torch (Allemang, De Clerck, Niezrecki, & Blough, 2012). The excess or lack of penetration results because of very high heat input or very slow transverse moment of the welding torch. The gas cavities results from the trapping of the gases by the solidified welded metal (Askeland, Fulay, & Bhattacharya, 2009). During the process of solidification, the trapped gases results in porosity in the turbine blade. Slag inclusion is of different types. It can be of any direction or shape just like slag lines or weaving faults. Cracks frequently arise from phosphorus and sulfur that occur at faults. In relation to this, Cho, Bode, and Kim (2010) classify cracks in two distinct types: longitudinal and transverse. Longitudinal cracks are usually straight lines running along the centerline of the weld bead. On the other hand, transverse cracks are usually straight lines running perpendicularly to the centerline (Ginley & Cahen, 2011). The lamellar tearing is a common flaw in the turbine blades made of low quality steels. It usually arises in the blades that have low ductility. The shrinkage cavities are because of the combination of steam or thermal shrinkage.
The processing of images of the materials having the defects can play a substantial role in extracting some significant information in the radiographic images of blades that might possess one or more kinds of defects (Totten, Westbrook, & Shah, 2003). During this process, the image segmentation plays a crucial role in isolating the desired object from the image so that measurements can be made on it afterwards. According to Cho, Bode, and Kim (2010), the major reason for the refusal of weldments is the presence of discontinuities or defects. It is termed as incomplete type of flaw (Cho, Bode, & Kim, 2010). The introduction or generation of these defects in a welded turbine blade can interfere with the welding process. Presently, there has been a huge deal of research and work on the development of the automated system for analysis, inspection and detection of incomplete type of flaw.
Recently, Cho, Bode, and Kim (2010) have proposed another method of detecting flaws in the radiographic weldment images by using the segmentation of morphological watershed. This technique proved its ability to detect flaws, such as warm holes and slag inclusions, in blades of turbines. However, Cho, Bode, and Kim's (2010) technique was incapable of detecting defects in the blades where the information about the defect is missing. Cho, Bode, and Kim (2010) also proposed another technique for classifying and detecting flaws in weld radiographs of turbine blades. This technique has been used to discriminate and detect defects, such as porosity, wormholes, linear slag inclusions, lack of fusion, and gas pores (Allemang, De Clerck, Niezrecki, & Blough, 2012).
Full Matrix Capture
The Full Matrix Capture (FMC) is a data acquisition method, which allows for the complete time domain signal to be captured from each element of a linear array probe (Allemang, De Clerck, Niezrecki, & Blough, 2012). Holmes first introduced this technique to the non-destructive techniques in 2005. It involves acquiring data concerning the flaws on a turbine blade by using a “transmit on one and receive on all” approach. The first element initially acts as a transmitter that generates a complete domain signals. As energy in any material at any given moment is generated from a single element, the following technique is frequently known as the sequential data acquisition method (Ginley & Cahen, 2011).
The imaging of the Full Matrix Capture is usually attained via the Total Focusing Method (TFM). The TFM method involves a representation of grid pixels of the region of interest. This region is defined by applicable amplitude information from the full matrix of data being extracted. The technique used to image the data is arrived at through a standard beam-forming and sum approach (Askeland, Fulay, & Bhattacharya, 2009).
Causes of Turbine Malfunction
Frequently, when blade failures occur in a turbine, many experts would consider resonance as the main cause of the failure (East & Eisenmann, 2002). This results in lengthy and costly exploration of the probable modes resonant vibrations. Apparently, the experts are well knowledgeable about the exceedingly harmful effects of resonance; though, it might be the best idea for the design analysis. The operation on resonant frequencies is a fact with many mechanical-drive turbines. According to Cho, Bode, and Kim (2010), it is the major reason why the construction of machines, such as HP that have 30 in. blade, has some troubles. Apparently, certain blades in a turbine need to be strong in order to operate under the resonant conditions. If a non-defective blade encounters a failure under these circumstances, then it is obviously too weak, and thus, stress needs to be reduced to guarantee its survival. Most importantly, huge amounts of money and time are spent on examining the failed turbine blade (Askeland, Fulay, & Bhattacharya, 2009). However, whenever time comes to make the hardware robust, many experts settle for the sparsest form of improvement. This is because of cost reasons. Nevertheless, according to Cho, Bode, and Kim (2010), it should be firmly kept in mind that the strength of the blade is a linear function of cost. Consequently, this includes turbine length, wheels, and critical speeds. It is illogical to ask for a strong turbine blade unless the concerned parties are willing to face the cost. In light of this, the paper examines the various causes of turbine malfunctions, which includes corrosion, crack, centrifugal force, steam-induced stress, impact stress, low-cycle fatigue, and mode vibration.
Corrosion refers to a steady destruction of materials, particularly metals, through chemical reaction with the environment (Askeland, Fulay, & Bhattacharya, 2009). The common example is rusting, which is the materialization of iron oxides (Askeland, Fulay, & Bhattacharya, 2009). This form of corrosion characteristically produces salts or oxides of the original metal. Corrosion can also take place in other materials, such as polymers or ceramics, but in such scenarios, the term degradation is very common. Structural alloys corrode just from the exposure to the moisturized air, though the process can be interfered with certain substances. The concentration of corrosion in a single place in a turbine blade can result in pits or cracks. Since corrosion is a process controlled by diffusion, it takes place on the exposed surfaces of turbine blades. Entirely, the surface of turbine blades is open to the environment, which makes the whole turbine vulnerable to corrosion. Consequently, techniques, such as passivation and chromate conversion that attempt to overcome corrosion of the turbine blade, can increase the resistance if the blade is exposed to corrosion. Various forms of corrosion can lead to flaws in turbine blades: stress corrosion, corrosion-fatigue, and standby corrosion (Askeland, Fulay, & Bhattacharya, 2009).
Stress corrosion is rare in the turbine blades; though, it can be observed in rotors where it has resulted in some detrimental failures. A fracture can be branched or inter-granular. Corrosion is usually caused by very corrosive steam or a corrosive environment in the stressed parts of the turbine blade (Sherma & Fried, 2003). Usually, the corrosive environment arises from the presence of chlorine in the moist atmosphere or when the blade is left open without protection. According to Cho, Bode, and Kim (2010), the required static stress component is residual stress, majorly near regions where the local yield has taken place during operation.
Corrosion-fatigue is perhaps the most popular cause of failures in blade turbines in the wet regions of turbines. This results in detrimental effects on the dry or wet region near the saturation line. The following process is well described by Cho, Bode, and Kim (2010). It is a popularly acknowledged engineering habit to consider that the endurance limit of steel is about half of its tensile capacity. In fact, this is reduced by even slight imperfections on the surface and by other factors that might affect the following surface. In concurrence, Cho, Bode, and Kim (2010) also point out that it is a generally accepted practice that there is no endurance limit when corrosive actions are present. This implies that the level of stress that the turbine blade can withstand continues falling with the increasing number of cycles. To provide some perceptions of the involved magnitudes, a steel of about 200 thousand psi has an endurance limit of about 100 thousand psi (Meetham & Van de Voorde, 2000). Cycling stressing and corrosion acting together are more severe than they might be anticipated cumulatively from each other. One variable that shapes this could be the incapability of the metal to either develop a defensive oxide film. Indeed, it appears probably that corrosion products can function in a wedging fashion to strengthen stress at the advancing front.
The presence of silica in the steam that propels turbines forms deposits that can trap the corrosive chemicals (Allemang, De Clerck, Niezrecki, & Blough, 2012). These chemicals become progressively very sensitive to the corrosion fatigue. This makes the situation much worse. Other most common destructive chemicals are chlorides, caustics, carbonates, and carbon dioxide from the treatment of boiler feed water or raw water. These chemicals get into the turbine through priming or carry-over. They might also be injected into the steam desuperheater if boiler contaminated condensate or feed water is used (Glover, Sarma, & Overbye, 2011). Boiler feed is frequently injected during the start-up of the plant. It is very bad because when resonances are common in many stages. In addition, the resulting contaminated condensate is undergoing recirculation into the desuperheater. The leakage of seawater into the boiler condensate can also be another cause of contamination. Sulfides, especially hydrogen sulfide, and syngas through leakages into heat exchangers cause fatigue corrosion (Askeland, Fulay, & Bhattacharya, 2009).
Various modes of failure in blades made from super alloys might be observed in the turbines, especially gas turbines (Totten, Westbrook, & Shah, 2003). The long-term turbine operation results in the microstructural dreadful conditions of super alloy blades. In several cases, the dreadful conditions lead to a significant change in the mechanical properties that can result in the blade failures. According to Cho, Bode, and Kim (2010), thermal fatigue cracks are the typical form of edge failures in the gas turbine blades. In the long-term operation, cracks also form on the blades made of wrought high alloys. Coating cracking is motivated by the local corrosion failure of the turbine blade under the coating (Askeland, Fulay, & Bhattacharya, 2009).
On the other hand, industrial manufacture of blades by different manufacturers might sometimes result in the displacement of the mass at the center (Allemang, De Clerck, Niezrecki, & Blough, 2012). Such a displacement causes static failures of blades. Such blade failures have been evident in the marine, aircraft and stationery gas turbine plants. Another contributing factor to static failure of the blade is overheating. It results from the deviation of temperature from the normal operating conditions. The increased metal temperature causes a drastic deterioration of mechanical properties and reduced metal fatigue resistances that causes blade cracking. Various scholars, such as Cho, Bode, and Kim (2010), have also reported that high-temperature corrosion and surface de-alloying may cause cracking. The analysis of Cho, Bode, and Kim's (2010) research also indicates that local elevation of the content of sulfur related causes sulfur-oxide corrosion attack.
A study conducted by Cho, Bode, and Kim (2010) showed that the maintenance team occasionally reported about the cracking in the turbine blades. The study also showed that corrosion fatigue cracks are evident at the inlet of the air-cooling holes. From this perspective, Cho, Bode, and Kim (2010) assert that the only stresses that are present at the inlet of air-cooling holes are the thermal stresses. This thermal stresses apparently arise from the incoming cooling air, since this region does not experience mechanical stress (Allemang, De Clerck, Niezrecki, & Blough, 2012). Therefore, it is rational to believe that the intensity of thermal stresses would be the highest at the interior surface near the entrance of the air inlet hole.
Condition Monitoring is mainly implemented using an integrated approach by coordinating various commercial equipment suppliers. It takes an integrated approach since no single technique can offer reliable and comprehensive solutions needed by the energy industry. The flaw detection is the most common and oldest application of ultrasonic testing. Since the 1940s, the principals of physics that explain the sound wave propagation through solids have been deployed in detecting hidden crack, porosity, voids, and internal discontinuities in metals, composites, plastics, and ceramics. An ultrasonic transducer uses the above principles of physics to detect flaws in the turbine blades. In broad terms, a transducer is just a device for converting energy from one state to another. It transforms electrical energy into very high frequency sound energy and vice versa. Contemporary ultrasonic flaw detectors, including the panametrics-NDT Epoch series, are portable, small, and micro-processor-based. Thus, they are appropriate for both field and shop use. Active thermography, such as inductively and optically heated thermography, offers a non-contact and rapid inspection method for detecting flaws present in a plane that is parallel to a surface. Radiography is among the traditional and still effective non-destructive techniques used for detecting flaws. The most significant application of radiography is in the gas industry as a tool for inspecting welds in pipelines. Radiography involves four steps: image enhancement, edge identification, segmentation, and feature extraction. The Full Matrix Capture (FMC) is a data acquisition method, which allows for the complete time domain signal to be captured from each element of a linear array probe. Frequently, when blade failures is detected in a turbine, many experts will conclude on resonance as the main cause of the failure. This results in lengthy and costly exploration of the probable modes resonant vibrations. Steam-induced stress refers to the bending stress resulting from a driving force or steam load. There are two forms of steam-induced stress, which are steady state and alternating. Impact stresses also arise from blade rubs or thrust bearing failures and water slugging. Extremely high impact strength of the material is very appropriate in minimizing the resulting damage if such accidents take place. Low-cycle fatigue may be one of causes of blade failure; though, many people do not always consider it. This implies that failures resulting from low-cycle fatigues that occur within few hundred to a few thousand stress cycles.