Accident reason analysis method for turbine blade

Abstract

The invention discloses an accident cause analysis method for a turbine blade. Accidents of steam turbine blades, from premature cracking to catastrophic accidents of parts in use, are the biggest threats to planned maintenance and reliable operation of large-scale units. The method comprises the following steps: collecting accident evidence; carrying out engineering analysis; corrective measures are made, the design or materials are modified, the stress and the service life are recalculated, the modification range is from limited repair of a damaged area to complete replacement of the blade, the characteristics of the blade are physically changed, and the design modification is either provided by a manufacturer and independently verified or originated from independent engineering analysis. And detailed information of finite element analysis is converted into manufacturing specifications. The method is applied to accident cause analysis of the turbine blade.

Description

Technical field:

[0001] The invention relates to the technical field of steam turbine blades, in particular to a method for analyzing causes of steam turbine blade accidents. Background technology:

[0002] During the life of each steam turbine, tens of millions of dollars are spent on proper maintenance, efficient operation, and routine repairs. Steam turbine blade incidents, from premature cracking to catastrophic failures of components in service, are the greatest threat to planned maintenance and reliable operation of these large units. Therefore, correct and timely analysis of the root causes of blade incidents is critical to managing current issues and developing long-term strategies for similar units operating with the same blades. The present invention provides technicians with an analytical method for investigating the causes of steam turbine blade incidents, using a staged approach to determine the most common accident damage mechanisms and root causes of steam turbine blades. Summary of the invention:

[0003] The purpose of the present invention is to establish a method for analyzing the causes of turbine blade accidents, effectively solve the problems caused by blade accidents to the daily operation of the unit, and provide a method for analyzing the causes of turbine blade accidents.

[0004] The above purpose is achieved through the following technical solutions:

[0005] A method for analyzing the cause of an accident of a steam turbine blade comprises the following steps:

[0006] Step 1: Collect accident evidence;

[0007] Collect basic evidence of the accident and supplement it with material analysis and metallographic examination to determine the potential root cause and damage mechanism and eliminate other factors, while conducting detailed engineering analysis of the design to finally determine the root cause and design a permanent repair plan;

[0008] Step 2: Conduct engineering analysis;

[0009] Finite element analysis is performed at the location where the cracks are formed to obtain key details about the stresses. When conditions permit, the analysis results are verified with dynamic frequency data. Combined with the blade's material properties, stress and life prediction, and fracture mechanics models, the primary failure mechanism and its source are separated from the secondary failure mechanism. If the cracks are caused by excessive operating stress, the design must be changed.

[0010] Step 3: Develop corrective measures;

[0011] Modifications to the design or materials are made and stresses and life are recalculated. Modifications range from limited repairs to damaged areas to complete replacement of the blade, whose characteristics have been physically changed. Design modifications are either proposed by the manufacturer and independently verified, or they are derived from independent engineering analysis and the details of the finite element analysis are converted to manufacturing specifications.

[0012] In the accident cause analysis method for steam turbine blades, the step 1 of collecting accident evidence includes: collecting blade inspection data, collecting unit operation data, completing metallographic analysis, evaluating industry-wide data, and identifying accident mechanisms.

[0013] The method for analyzing the cause of an accident of a steam turbine blade, the engineering analysis in the second step includes: defining material properties, defining geometric shapes, establishing a finite element model, calculating static stress, calculating natural frequencies and modes, calculating dynamic stress, verifying the model with test data, estimating crack initiation life and formulating corrective measures.

[0014] In the accident cause analysis method for steam turbine blades, the corrective measures formulated in step three include: identifying design changes, model and finite element analysis, calibrating service life and selecting a final solution.

[0015] Beneficial effects:

[0016] 1. The present invention provides practical help for the root cause analysis of steam turbine blades. During the life of each steam turbine, the corresponding maintenance, efficient operation and daily repairs cost tens of millions of yuan. Among them, premature cracking or breaking of blades is the biggest threat to the scheduled maintenance and reliable operation of these large units. Correct and timely analysis of the root cause can manage the current problem and establish a long-term strategic foundation for the use of the same blades in similar units. The invention aims to provide a blade accident analysis method to steam turbine technicians, which involves the coordination of multiple technical contents and professional engineering disciplines. The accident investigation process is relatively complicated. If the safety issues of the blades are not resolved, they will continue to crack or break. The downtime and power production losses caused by repeated accidents will far exceed the cost and time required for accident cause analysis and replacement strategy. Therefore, it is necessary to develop a method for accident cause analysis of steam turbine blades, which is standardized and universal, and is applicable to all blade accident analysis of steam turbine units. Description of the drawings:

[0017] Figure 1 : Flowchart of Cause Analysis of Steam Turbine Blade Accidents

[0018] Figure 2 :Critical Dimension Measurement of Steam Turbine Blades after Accident

[0019] Figure 3:Schematic diagram of steam turbine blade structure

[0020] Figure 4 :Finite Element Cyclic Symmetric Model of Steam Turbine Blades

[0021] Figure 5 : Campbell diagram of turbine blade frequency

[0022] Figure 6 : SAFE diagram of turbine blade frequency. Specific implementation method:

[0023] A method for analyzing the cause of an accident of a steam turbine blade comprises the following steps:

[0024] Step 1: Collect accident evidence;

[0025] Collect basic evidence of the accident and supplement it with material analysis and metallographic examination to determine the potential root cause and damage mechanism and eliminate other factors, while conducting detailed engineering analysis of the design to finally determine the root cause and design a permanent repair plan;

[0026] Step 2: Conduct engineering analysis;

[0027] Finite element analysis is performed at the location where the cracks are formed to obtain key details about the stresses. When conditions permit, the analysis results are verified with dynamic frequency data. Combined with the blade's material properties, stress and life prediction, and fracture mechanics models, the primary failure mechanism and its source are separated from the secondary failure mechanism. If the cracks are caused by excessive operating stress, the design must be changed.

[0028] Step 3: Develop corrective measures;

[0029] Modifications to the design or materials are made and stresses and life are recalculated. Modifications range from limited repairs to damaged areas to complete replacement of the blade, whose characteristics have been physically changed. Design modifications are either proposed by the manufacturer and independently verified, or they are derived from independent engineering analysis and the details of the finite element analysis are converted to manufacturing specifications.

[0030] Collect blade inspection data (such as Figure 1As shown, task 1 in stage 1). After a blade accident occurs, the first thing to do is to protect the scene, collect all debris, make relevant records as detailed as possible, and take photos of the damaged blades. The photos include the condition of the blades after the cylinder is opened, including whether there is scaling, corrosion, water brushing, etc. The accident blades are classified, grouped, photographed, and recorded, and the matching status of the blades and various components is photographed and recorded. The accident blades and parts are protected, including protecting the fracture, reducing the number of handling times, and maintaining the collection of parts. The basic dimensions of the accident blades are measured and recorded, and the maintenance data are checked to see if the dynamic and static clearances meet the standards. Combined with the attached Figure 2 Note: The dimensions of the accident blade that need to be measured after the cylinder is opened include but are not limited to the following: the tip clearance of the accident blade (D1 dimension), the axial width of the blade root (D2 dimension), the rotor radius on the steam inlet side of the blade (D3 dimension), the rotor radius on the steam inlet side of the blade (D4 dimension), the dimension from the root of the blade to the rotor axis (D5 dimension), the dimension from the top of the blade to the rotor axis (D6 dimension), etc.

[0031] Collect unit operation data (such as Figure 1 As shown in the figure, task 2 in stage 1). Check whether there are any abnormalities in the operating conditions before the blade accident, such as whether the operating parameters are normal, whether there is overload, overspeed and abnormal frequency operation, whether there is blade scaling, corrosion, water brushing, etc. If the appearance inspection or fracture analysis shows that the surface is related to corrosion, the corrosive medium should be identified, and the deposits on the blade surface can be taken for ion analysis; if there are signs that it is related to water hammer, the source of the water should be found. At the same time, pay attention to the water quality analysis and steam quality in the operation records. At the same time, check whether there have been atypical events, such as the unit approaching or exceeding the load, the back pressure exceeding the allowable range, and the power grid interference.

[0032] Complete metallographic analysis (such as Figure 1 As shown in Figure 1, task 3 in stage 1). The purpose of metallographic analysis is to determine whether the material is qualified and to find relevant information from the fracture surface. The specific implementation includes physical and chemical testing to verify whether the blade material meets the standard. The broken blade is removed and the residue can be used for analysis. If it is not removed, small pieces can be sawed off for spectral, hardness, composition and metallographic analysis. A further implementation is fracture analysis. From the fracture morphology, it can be roughly determined whether it is tangential vibration, axial vibration or torsion; from electron microscopy and electronic scanning, it can be determined whether it is mechanical fatigue, corrosion fatigue, stress corrosion or micro-vibration corrosion.

[0033] Evaluate industry-wide data (e.g. Figure 1 As shown in Figure 4, task 4 in stage 1) is mainly to conduct comparative analysis of blades of similar units or even the same unit in terms of design, manufacturing, assembly, operation, etc. to find the cause of the blade accident. If there has been a similar accident before and the current accident is similar enough, the same remedial measures can be accepted.

[0034] Initial identification of the failure mechanism (e.g. Figure 1 The goal of this task is to use all available evidence collected in Phase 1 to identify potential causes and to reach a preliminary conclusion, which may still change as more information becomes available.

[0035] If a temporary strategy has been formed in Phase 1, a blade repair or replacement strategy can be implemented. If a detailed analysis is required, the tasks of Phase 2 can be performed. The specific implementation method is as follows.

[0036] Define the blade material properties (such as Figure 1 As shown in Figure 1, Task 1 in Phase 2). The material properties here refer to the fatigue characteristic parameters for estimating the low-cycle fatigue and high-cycle fatigue of the blade, which are usually obtained from material fatigue tests and are used to estimate the number of cycles or hours before crack initiation of the blade. The fatigue characteristic parameters include fatigue strength coefficient, fatigue strength index, fatigue ductility coefficient, cyclic hardening coefficient, cyclic strain hardening index, etc.

[0037] Determine the blade geometry (e.g. Figure 1 The purpose of this task is to provide information about the geometry and dimensions of the blade design, as well as the connection details (root, tie rod, shroud, etc.) between the target blade and other blades and wheel grooves (combined with Figure 3 Description) in order to construct the finite element model of the entire blade (combined with Figure 3 illustrate).

[0038] Establish a finite element model (such as Figure 1 As shown in Figure 3, task 3 in stage 2). The finite element method is the main means of stage 2. Its main purpose is to calculate the life damage of the blade at the operating speed. In order to correctly calculate the natural frequency and mode of the entire blade, the calculation model needs to have cyclic symmetry, combined with Figure 4 Note that A1 and A2 are two cyclic symmetry planes of the blade shroud, B1 and B2 are two cyclic symmetry planes of the blade tie rod, and C1 and C2 are two cyclic symmetry planes of the blade rotor groove. Special mesh processing is required in the contact surface of the blade root, the contact surface between the shrouds, the contact surface between the tie rod and the tie rod hole, etc. These areas require consistent mesh nodes on the contact master surface and the contact slave surface, especially near the crack initiation position, to give a reasonable mesh density, which determines the accuracy of the finite element calculation.

[0039] Calculate static stress (such as Figure 1As shown in Figure 4, Task 4 in Phase 2). There are two main purposes for calculating the steady-state static stress of the blade under centrifugal force and steam bending loads. The first purpose is to identify accident mechanisms, such as low-cycle fatigue, stress corrosion cracking, thermal fatigue, etc. These calculated values ​​are ultimately used as inputs to the respective damage models. The second purpose is to determine the location and magnitude of the average stress affected by dynamic stress and high-cycle fatigue. The centrifugal tensile stress calculation formula for the blade under centrifugal force load is:

[0040]

[0041] σ 总 =σ1+σ2

[0042] Where σ 总 is the total stress of the blade; σ1 is the centrifugal tensile stress of the blade part; σ2 is the centrifugal tensile stress of the shroud part; ρ is the blade material density; ω is the angular velocity of rotation; R1 is the radius of the blade top section; r1 is the radius of the blade root section; b is the axial width of the shroud; t is the shroud pitch; h is the shroud thickness, R2 is the average radius of the shroud; A is the top cross-sectional area of ​​the shroud.

[0043] The calculation formula for the steam bending stress of the blade profile is:

[0044]

[0045] σ B =M B / W B

[0046] Where σ B is the blade steam bending stress; M B is the steam bending moment; W B F is the blade bending section coefficient; D is the steam power; L S is the blade height; a is the blade root and top chord length; D is the maximum cross-sectional thickness of the blade.

[0047] Calculate natural frequencies and modes (such as Figure 1 When analyzing the root causes of high-cycle fatigue on longer blades, it is most important to know whether the blade's natural frequency is detuned, slightly detuned, or potentially resonant. For shorter blades on HP and IP turbines, it is still important to know the blade's natural frequency because they are typically designed to avoid coincidence with the nozzle pass frequency generated by the upstream stators and to withstand the dynamic stresses that would be generated even under assumed resonant conditions. The results of the frequency analysis of the blade are usually presented in a Campbell plot (e.g. Figure 5 As shown, the horizontal axis is the speed and the vertical axis is the frequency) and the SAFE diagram (as shown Figure 6The Campbell diagram shows the dynamic frequency from 0 rpm to full speed, while the SAFE diagram shows the frequency of a single pitch diameter at full speed. The two diagrams represent the same data from different perspectives. The natural frequency of the blade is ultimately used to calculate the dynamic stress of the blade. The blade dynamic frequency calculation formula is:

[0048]

[0049] Where f 静 is the natural frequency of the blade when it rotates; f 静 is the natural frequency of the blade when it is stationary; K is the harmonic number; I is the blade height; E is the elastic modulus of the blade material; I0 is the moment of inertia of the blade section; g is the gravitational acceleration; A is the cross-sectional area of ​​the blade; ρ is the density of the blade material; B is the dynamic frequency coefficient; and n is the operating speed.

[0050] Calculate dynamic stress (such as Figure 1 As shown, Task 6 in Phase 2). The goal of this task is to estimate the stress that the blade may be subjected to at the crack initiation location based on the analysis of the natural frequency results and determine the vibration mode that is most likely to cause the crack. If the frequency analysis indicates the possibility of a vibration mode operating under resonant conditions, then this task is to determine whether the dynamic stress is high enough to pose a threat. If the problem is related to damage tolerance, the results of the dynamic stress analysis provide a computational basis for estimating the critical crack size of pits, erosion notches, foreign object damage, etc. When calculating dynamic stress, three important input parameters that determine the accuracy of the calculation are the blade natural frequency, the blade damping coefficient, and the steam excitation force on the blade surface obtained based on CFD calculations. The calculation formula for the blade dynamic stress caused by the steam excitation force is:

[0051] Δσ B =0.04σ B

[0052]

[0053] Where σ d is the blade dynamic stress; σ B is the blade steam bending stress; Δσ B For the exciting force; is the magnification factor

[0054] Validate the model with experimental data (e.g. Figure 1As shown, Task 7 in Phase 2). While the predicted values ​​of static stresses can be reasonably checked by simple formulas, the structural dynamics calculations based on the finite element method are best evaluated by comparison or correlation with frequency tests. There are two common test methods. The static frequency test of the blade usually refers to the test of the blade at 0 rpm. The test can be performed on a single blade (suspended or in a test fixture) or on the blade installed in the disk / rotor. The dynamic frequency test of the blade is a test to capture the natural frequency of the rotating blade in a controlled non-steam environment. The test can usually be performed using radio telemetry.

[0055] Estimate the crack initiation life (e.g. Figure 1 This represents the final task of the engineering analysis. The evidence gathered is integrated into a comprehensive report in which a consistent explanation of the incident is given from all aspects. A life estimation model is essentially any deterministic formula in which key parameters (such as stress-strain characteristics obtained from material testing) are applied to estimate life loss, and the resulting predictions are then compared to the number of operating hours and an assessment is made as to whether the mechanism can explain the formation of cracks over a known service life. If dealing with a pure fatigue problem, this preliminary estimate is usually sufficient to confirm the root cause. If the problem involves the auxiliary effects of erosion or corrosion, an estimate of the crack growth life or damage tolerance can also be made using a direct deterministic formula.

[0056] In order to estimate the initiation life of low-cycle fatigue cracks in blades, it is first necessary to estimate the plastic strain of blades based on stress according to Manson's life prediction model. The plastic strain is calculated as follows:

[0057]

[0058] Where ε p is the plastic strain; Δσ is the stress amplitude; K is the hardening coefficient; n is the strain hardening exponent.

[0059] With the plastic strain, the low cycle fatigue life at any location can be calculated using the following equation:

[0060]

[0061] Where N f is the cycle life; p is the plastic strain; ε f is the extension coefficient; c is the extension index.

[0062] In order to estimate the initiation life of high-cycle fatigue cracks in blades, the high-cycle fatigue life of blades can be estimated by the following formula:

[0063]

[0064] Where N f is the cycle life; d is the dynamic stress; f is the fatigue strength coefficient; σ mean is the average stress; b is the fatigue strength index.

[0065] If the failure is determined to be the result of premature crack formation due to excessive stress in Phase 2, then Phase 3 is entered where the initial design is changed and the finite element analysis is repeated to determine if the corrective actions provide a permanent solution. The specific implementation is as follows.

[0066] Identify design changes (such as Figure 1 As shown, Task 1 in Phase 3). The new design truly represents any modification to the original blade where significant modifications have been made to structures such as the root, shrouds or ribs, or to the blade material itself. Particularly for long blades, any adjustments to the mass or stiffness of the original blade disk / rotor system will have an impact on the natural frequencies. The changes provided in the replacement design are either corrective or preventive, and if it is concluded that all blade failures are due to excessive stress leading to premature crack formation, then the changes to the original design are evaluated as the third and final phase to determine if the corrective actions will provide a permanent fix. Therefore, based on the results of the blade root cause analysis, it is necessary to identify whether the blade requires a new design, whether a repair can be used to repair a surface damaged blade, or whether alternative inspection standards can be applied to manage and guide the remaining life of the blade.

[0067] Model and finite element analysis (such as Figure 1 As shown, Task 2 in Phase 3). With the advent of more powerful computers, model size (number of elements and mesh density) is no longer a big issue. However, it is possible to include enough detail at the location where cracks form to ensure reasonable and correct predicted stresses. The main limiting factor of any analysis depends on the final input, assumptions and boundary conditions specified. In particular, when calculating the natural frequencies of blades and the dynamic stresses generated by high-cycle fatigue, the dynamic models involved require a lot of practice and experience. The challenges of modeling the blade-disk system are further complicated when links or contact features between components are involved (such as bosses, shrouds and ties). If these settings are not entered correctly, they will introduce artificial stiffness or fixed contact points, resulting in errors in stresses and / or predicted frequencies. Therefore, it is necessary to establish a reasonable calculation model when designing new blades, and to give correct and reasonable boundary loads when performing finite element analysis.

[0068] Calibration life (eg Figure 1As shown, Task 3 in Phase 3). When a blade failure occurs, high cycle fatigue is always an obvious suspect and the challenge is to determine whether this is the primary or secondary cause. Based on the design analysis, the magnitude of the dynamic stress is used in conjunction with the static stress at the crack location to determine the magnitude of the strain that will occur during continued operation. Appropriate material fatigue properties are derived from cycled specimens under controlled conditions. The high cycle fatigue initiation life is usually equivalent to the operating time before failure. At low stress values ​​below the material yield strength, high cycle fatigue failures usually occur in 10 8 or less cycles. To estimate the initiation life of high cycle fatigue, it is necessary to know the vibration frequency of the blade mode involved, and in turn, the magnitude of the dynamic stress can be reasonably estimated. Usually, if the blade is well designed, its vibration stress is generally on the order of several MPa, which leads to the estimated high cycle fatigue life exceeding enough cycles to adapt to the vibration damage of 20-30 years of continuous operation. When the stress approaches the order of 50MPa, especially when the dynamic stress and static stress coincide in a zone, the stress will pose a serious threat to the integrity of the designed blade structure.

[0069] Select the final solution (such as Figure 1 As shown, Task 4 in Phase 3) begins with the systematic collection of key evidence from Phases 1 and 2, which is used to decide whether to proceed or execute the available repair / replace strategies, depending on the level of confidence in the information on which they are based. However, since a full investigation takes several weeks to execute, the strategies executed in Phase 1 or Phase 2 are not permanent, but are used to buy time to return the equipment to service and plan the implementation of the final corrective actions and select the final solution.

[0070] The present invention adopts a three-stage processing flow to analyze the causes of turbine blade accidents, and through the causes of these blade accidents, to identify the damage mechanism and root cause. It effectively solves the problems caused by turbine blade accidents to the daily operation of the unit and prevents the recurrence of blade accidents.

Claims

1. A method for analyzing the cause of an accident of a steam turbine blade, characterized in that: The method comprises the following steps: Step 1: Collect accident evidence; Collect basic evidence of the accident and supplement it with material analysis and metallographic examination to determine the potential root cause and damage mechanism and eliminate other factors, while conducting detailed engineering analysis of the design to finally determine the root cause and design a permanent repair plan; Step 2: Conduct engineering analysis; Finite element analysis is performed at the location where the cracks are formed to obtain key details about the stresses. When conditions permit, the analysis results are verified with dynamic frequency data. Combined with the blade's material properties, stress and life prediction, and fracture mechanics models, the primary failure mechanism and its source are separated from the secondary failure mechanism. If the cracks are caused by excessive operating stress, the design must be changed. Step 3: Develop corrective measures; Modifications to the design or materials, ranging from limited repairs to damaged areas to complete replacement of the blade, and recalculation of stresses and life. Its features have been physically changed and the design modifications are either proposed by the manufacturer and independently verified, or they are derived from independent engineering analysis and the conversion of finite element analysis details into manufacturing specifications.

2. The method for analyzing the cause of an accident of a steam turbine blade according to claim 1, characterized in that: The collection of accident evidence in step one includes: collecting blade inspection data, collecting unit operation data, completing metallographic analysis, evaluating industry-wide data, and identifying the accident mechanism.

3. The method for analyzing the cause of an accident of a steam turbine blade according to claim 2, characterized in that: The engineering analysis in step 2 includes: defining material properties, defining geometric shapes, establishing finite element models, calculating static stresses, calculating natural frequencies and modes, calculating dynamic stresses, verifying models with test data, estimating crack initiation life, and formulating corrective measures.

4. The method for analyzing the cause of an accident of a steam turbine blade according to claim 3 is characterized in that: The corrective actions in step three include: identifying design changes, modeling and finite element analysis, calibrating service life and selecting a final solution.

Images