A method and system for predicting the combined wear and fatigue failure of turbine blades

Abstract

This invention provides a method and system for predicting the combined wear and fatigue failure of turbine runner blades, relating to the field of hydropower equipment operation and maintenance technology. The method includes: establishing a fluid computational domain model and mesh generation; constructing a flow field based on single-phase flow characteristics; constructing a two-phase flow system using a discrete phase model; constructing a material wear model based on the characteristics of suspended matter passing through the machine; obtaining a geometrically varying wear prediction model and calculating the model; calculating the fluid-structure interaction strength characteristics of the initial structural field, and based on the calculation results, calculating the fatigue characteristics of the initial structural field; reconstructing the flow channel geometrically varying model based on fatigue characteristic data, and calculating the flow and fatigue characteristics during the reconstruction process; and evaluating the fatigue failure prediction of the runner under the influence of wear. This invention utilizes dynamic mesh smoothing technology to achieve real-time reconstruction of the geometric model and iterative coupling of the flow field and structural field during the wear process, improving the accuracy and scientific rigor of failure prediction.

Description

Technical Field

[0001] This invention relates to the field of hydropower equipment operation and maintenance technology, specifically to a method and system for predicting the combined wear and fatigue failure of turbine blades. Background Technology

[0002] Hydropower stations in high-altitude areas of western China face challenges such as turbine runner wear, fatigue, and failure due to harsh operating conditions, as well as flow instability caused by wear. It is essential to adopt a low-specific-speed turbine runner structure with long and short blades, ensuring sufficient strength and rigidity, a low cavitation coefficient, and good cavitation performance and efficiency. Statistics show that runner failure caused by silt abrasion results in an average efficiency reduction of 1.5%–6% for turbines in large and medium-sized power stations in my country, and a reduction of 3%–10% for small and medium-sized turbines. Silt abrasion and damage also shorten the unit maintenance cycle and increase the workload of unit overhaul. Some runners suffer severe damage and require emergency shutdown for maintenance during a single flood season, and the maintenance period for some power station units has been shortened to one year, far exceeding the three-year maintenance period for power stations in other regions. Frequent equipment maintenance causes inconvenience to users' production and daily life. Taking a hydropower station in a high-altitude area of ​​western China as an example, due to silt abrasion and subsequent fatigue damage, the unit's output load has decreased significantly, and vibration exceeds standards in some operating conditions, seriously affecting the safe and stable operation and economic benefits of the power station.

[0003] In high-altitude river basin hydroelectric generating units, the failure mechanism of turbine flow components exhibits a significant combined characteristic of sediment abrasion and fatigue. Rivers in this region are influenced by a hydro-wind-solar complementary energy system, topography, and climate conditions, resulting in high load fluctuations (load range fluctuation ≥70%), high suspended mass (annual average >1.0 kg / m³), and high flow velocities (local runner velocity ≥45 m / s). These factors collectively constitute the external driving conditions that accelerate runner failure. From the perspective of material damage mechanisms, suspended mass abrasion is essentially an abrasive wear process. Suspended mass containing a high proportion of hard particles (Mohs hardness ≥5) is carried by high-speed water flow, impacting the component surfaces at an impact velocity of 20-50 m / s, thus cutting and chiseling them. During sediment abrasion, high-speed moving sediment particles act like micro-abrasives, continuously impacting the surfaces of turbine blades, guide vanes, and bottom rings. According to the theory of elastoplastic deformation, this impact causes dislocation slip and plastic accumulation on the material surface, resulting in plastic deformation and spalling. This forms typical furrows and pits, damaging the integrity of the turbine runner surface, creating a rough and uneven microstructure and stress concentration areas. It also deteriorates the flow velocity and flow regime of the local flow field, leading to the preferential initiation of fatigue cracks on the material surface. This further weakens the material's load-bearing capacity, making the component surface more susceptible to impact and cutting by silt particles, accelerating the wear process and causing the wear depth to increase exponentially with operating time. These two factors reinforce each other, creating a vicious cycle.

[0004] Meanwhile, the complex dynamic loads induced by cyclic stresses during long-term, flexible operation of the turbine under a hydro-wind-solar complementary energy system, combined with the pressure pulsations induced by unsteady flow fields and their resulting mechanical vibrations, constitute a complex alternating stress environment. At the level of combined action mechanisms, sediment abrasion promotes fatigue failure through multiple pathways. First, the wear-induced deterioration of the turbine runner surface roughness significantly increases the stress concentration factor, resulting in a significant shortening of the fatigue crack initiation life. Second, the wear-induced stripping of the material surface layer weakens the material's fatigue resistance, and the resulting subsurface plastic deformation layer forms a high-density dislocation cell structure, altering the surface stress distribution and making the stress state in the wear area more complex, accelerating the material stripping rate. This positive feedback mechanism leads to nonlinear degradation of component material properties; the rate of decay of residual strength over operating time is several times higher than under single-effect conditions. Under this combined action, the material properties of the turbine runner continuously deteriorate, cracks propagate rapidly, and ultimately, the component fractures, seriously threatening the safe and stable operation of the turbine.

[0005] Currently, calculations of mud and sand wear and fatigue strength using a single runner are commonly used, which cannot accurately simulate and reproduce the actual fatigue damage process of turbine runners in high-altitude, muddy water turbines. Summary of the Invention

[0006] To address the technical problem that existing technologies, which use single runner blade wear calculations and fatigue strength calculations, cannot accurately calculate the failure time of the actual combined fatigue damage of the turbine runner, this invention proposes a method and system for predicting the combined failure of runner blade wear and fatigue.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for predicting the combined wear-fatigue failure of turbine blades, the method comprising: A fluid computational domain model and mesh generation are established; a flow field is constructed based on the fluid computational domain model, mesh generation, and single-phase flow characteristics; a two-phase flow system is constructed based on the constructed flow field using a discrete phase model; a material wear model is constructed based on the material wear model and the parameters of the suspended mass characteristics set for the two-phase flow system; a geometric morphology gradual wear prediction model is obtained based on the material wear model; and the gradual wear prediction model is calculated. The fluid-structure interaction strength characteristics of the initial structural field are calculated; based on the results of the fluid-structure interaction strength characteristics calculation, the fatigue characteristics of the initial structural field are calculated; based on the fatigue characteristic data of the initial structural field, the flow channel geometry gradient model is reconstructed; the flow characteristics and fatigue characteristics are calculated during the flow channel geometry gradient model reconstruction process; and the fatigue failure prediction of the impeller under the influence of wear is evaluated.

[0008] On the other hand, the present invention also provides a combined wear-fatigue failure prediction system for turbine blades. The system includes a memory for storing computer program instructions and a processor for executing the program instructions. When the computer program instructions are executed by the processor, the system is triggered to execute the above-mentioned combined wear-fatigue failure prediction method for turbine blades.

[0009] Compared with the prior art, the beneficial effects of the present invention are: 1. Significantly Improved Prediction Accuracy and Scientific Rigor: This invention breaks through the traditional computational framework of single damage or fatigue modes, proposing a combined wear-fatigue failure prediction method for turbine blades based on fluid-structure interaction. By constructing a full-chain model of "initial flow field-structural field co-calculation → geometrical gradual wear prediction → dynamic fatigue damage assessment → mesh reconstruction iteration," it achieves bidirectional joint calculation of wear morphology evolution and fatigue crack propagation. The simulation capability of the full-chain model can simulate the entire process from the initial state of the turbine blade to wear-fatigue combined failure, including the gradual change of geometric morphology, dynamic changes in the flow field, and fatigue crack propagation, providing a powerful tool for the full life cycle management of turbine blades. The introduction of the quantitative relationship between sediment impact kinetic energy and material plastic deformation accurately characterizes the wear depth growth characteristics over time, making wear prediction more scientific and accurate. This innovation significantly improves the accuracy and scientific rigor of failure prediction.

[0010] 2. Enhanced Dynamic Simulation and Real-Time Reconstruction Capabilities: A dynamic mapping model between surface roughness and stress concentration factor is established to quantify the attenuation effect of wear on fatigue life, enabling the simulation process to dynamically reflect changes in actual working conditions. Furthermore, dynamic mesh smoothing technology is utilized to achieve real-time reconstruction of the geometric model and coupled iteration of the flow-structure field during the wear process. This technology improves the accuracy and efficiency of the simulation, making the prediction results closer to reality.

[0011] 3. Optimization of economic efficiency and operational stability: Accurate failure prediction allows for advance scheduling of maintenance plans, reducing unplanned downtime and thus improving the operational stability and economic benefits of the power station. It provides key technical support for the wear-resistant design, operational optimization, and preventive maintenance of high-altitude turbine runners, helping to extend equipment lifespan and reduce maintenance costs. Furthermore, the prediction method and system provided by this invention offer strong technical support for the operation and maintenance management of hydropower equipment, making operation and maintenance decisions more scientific and rational.

[0012] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0013] Figure 1 This is a diagram illustrating the wear-fatigue combined action mechanism according to the present invention; Figure 2This is a flowchart of the fatigue failure prediction method for a rotor under the influence of wear according to the present invention; Figure 3 This is a flowchart of the method for predicting the combined failure of wheel wear and fatigue according to the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings, so as to more clearly understand the purpose, features and advantages of this invention. It should be understood that the embodiments shown in the drawings are not intended to limit the scope of this invention, but are only for illustrating the essential spirit of the technical solutions of this invention. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0015] Unless the context requires otherwise, throughout the specification and claims, the word “comprising” and its variations, such as “including” and “having”, shall be understood to have an open, inclusive meaning, that is, to be interpreted as “including, but not limited to”.

[0016] Throughout this specification, references to "an embodiment" or "an embodiment" indicate that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Therefore, the appearance of "in an embodiment" or "an embodiment" in various places throughout the specification does not necessarily refer to the same embodiment. Furthermore, a particular feature, structure, or characteristic may be combined in any manner in one or more embodiments.

[0017] The singular forms “a” and “the” used in this specification and the appended claims include plural references unless otherwise expressly stated herein. It should be noted that the term “or” is generally used to mean “and / or” unless otherwise expressly stated herein.

[0018] In the following description, in order to clearly demonstrate the structure and working method of the present invention, a number of directional terms will be used. However, terms such as "front", "back", "left", "right", "outside", "inside", "outward", "inward", "up", and "down" should be understood as convenient terms and not as limiting terms.

[0019] The implementation details of the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The following content is only for the convenience of understanding the implementation details and is not necessary for implementing this solution.

[0020] In the actual operation of hydropower stations in high-altitude areas of western China, turbine runners have long faced complex and severe operating conditions. High suspended sediment content, extreme flow velocity fluctuations, and alternating loads resulting from flexible operation under a hydro-wind-solar complementary energy system constitute the dual driving factors for runner material failure. On the one hand, sediment containing hard particles (Mohs hardness ≥ 5) impacts the runner blades at high speeds of 20-50 m / s, forming typical groove wear morphology through plowing action, destroying surface integrity and inducing stress concentration. On the other hand, the pressure pulsation and mechanical vibration caused by unsteady flow fields superimpose to subject the runner to fatigue damage in a complex alternating stress environment. Existing technologies often analyze sediment wear or fatigue strength in isolation, neglecting the positive feedback mechanism between the two: the surface roughness deterioration caused by wear reduces fatigue crack initiation life by more than 40%, while the subsurface plastic deformation caused by fatigue damage accelerates the material spalling rate, forming a vicious cycle of "wear-fatigue," ultimately leading to a sharp drop in runner efficiency, excessive vibration, or even fracture failure.

[0021] To address this technical bottleneck, refer to Figure 1-Figure 3 As shown, this invention breaks through the traditional computational framework of single damage or fatigue modes and proposes a combined wear-fatigue failure prediction method for turbine blades based on fluid-structure interaction. This method achieves bidirectional joint calculation of wear morphology evolution and fatigue crack propagation by constructing a full-chain model of "initial flow field-structural field co-calculation → geometrical gradual wear prediction → dynamic fatigue damage assessment → mesh reconstruction iteration." The initial flow field-structural field co-calculation is fundamental, providing initial conditions for the flow field and structural field in subsequent steps. Geometrical gradual wear prediction relies on the results of the initial flow field calculation and simultaneously provides input for geometrical morphology changes in dynamic fatigue damage assessment. Dynamic fatigue damage assessment is performed based on geometrical gradual wear prediction, considering the influence of geometrical morphology changes on fatigue characteristics. Mesh reconstruction iteration runs throughout the entire process, ensuring the accuracy and stability of the computational model. The innovations are: 1) Introducing a quantitative relationship between the impact kinetic energy of sediment and the plastic deformation of materials to accurately characterize the growth characteristics of wear depth over time; 2) Establishing a dynamic mapping model between surface roughness and stress concentration coefficient to quantify the attenuation effect of wear on fatigue life; 3) Developing dynamic mesh smoothing technology to achieve real-time reconstruction of the geometric model and flow-structure field coupling iteration during the wear process. Verified by actual cases, the error between the predicted remaining life of the turbine runner and the on-site maintenance data is controlled within 8%, providing key technical support for the wear-resistant design, operation optimization, and preventive maintenance of turbine runners at high altitudes.

[0022] The fatigue failure prediction method for impeller blades under the influence of wear provided by this invention has the following specific process flow: S1. Establish the fluid computational domain model and mesh generation.

[0023] Based on the actual CAD model of the turbine runner, the fluid computation domain is extracted, including key components such as runner blades, guide vanes, and flow channels, to ensure that the geometric boundaries are consistent with actual operating conditions. For example, for turbines operating at high altitudes in western China, special consideration must be given to the geometric details of areas such as sediment inlets and outlets.

[0024] Define the inlet, outlet (pressure outlet or free outflow), and wall conditions (no-slip wall, roughness settings) of the fluid computation domain. For example, the inlet needs to set the suspended mass volume fraction and velocity distribution, and the outlet needs to define pressure or flow rate conditions, etc.

[0025] In some embodiments, the mesh generation strategy encompasses type selection, quantity optimization, and quality control: Regarding mesh type, structured meshes are suitable for regular regions (such as cylindrical flow channels), offering high computational efficiency but struggling to adapt to complex boundaries; unstructured meshes (such as a mixture of tetrahedral and hexahedral meshes) are suitable for complex geometries (such as twisted blades and guide vanes), allowing for localized refinement of key areas such as the leading and trailing edges of blades and sediment impact zones; hybrid meshes combine the advantages of both, using structured meshes in the main flow channel and unstructured meshes in locally complex areas. Mesh quantity optimization involves sensitivity analysis through parametric studies (e.g., 5 million to 15 million mesh numbers). When the mesh number is ≥10 million, computational efficiency and experimental error are <1%, and meshes are fined 2-3 times in sediment impact zones and flow separation zones to capture high-gradient physical quantities. For quality control, structured meshes must have an orthogonality >0.7, while unstructured meshes must have an interior angle >30°, an aspect ratio <5, and a twist <0.85 to avoid numerical errors and computational instability.

[0026] S2, based on the fluid computational domain model of S1, mesh generation, and single-phase flow characteristics to construct the flow field.

[0027] First, the steady flow characteristics of the entire flow channel were calculated, and the convergence results of the steady flow characteristics calculation were used as the initial conditions for the unsteady flow characteristics calculation. Then, the unsteady flow characteristics were calculated to obtain the internal and external flow characteristics of the turbine under wide load flexible operation conditions, the law of operation conditions, pressure pulsation and vortex flow distribution and evolution and other dynamic characteristics, which provided the flow field background conditions for subsequent multiphase flow simulation.

[0028] Before predicting the gradual wear of the flow channel geometry, it is necessary to use a hydraulic mechanical erosion testing system to experimentally measure and fit a wear quantity relationship model of the impeller material, which includes wear characteristic constants and velocity exponents. Then, through the study of the long-term wear characteristics of suspended matter, the wear rate acceleration factor is obtained. A long-term wear prediction model considering the coupled wear rate acceleration factor is established and written as a user-defined variable to characterize the wear depth. Finally, a dynamic mesh wear numerical simulation calculation of the gradual wear of the multiphase flow channel morphology is carried out to predict the wear morphology of the flow channel. The specific steps are as follows: S3. Construct a sediment-water two-phase flow system using a discrete phase model (DPM).

[0029] Based on the flow field constructed in S2, a Discrete Phase Model (DPM) is introduced to simulate the motion and wear process of sediment particles. DPM is a numerical method in computational fluid dynamics (CFD) used to simulate the behavior of discrete phases (such as particles, droplets, bubbles, etc.) in multiphase flow. In this embodiment, multiphase flow is the phenomenon of two different phases (or different components, different motion states) of matter flowing together simultaneously (i.e., sediment-water).

[0030] The volume fraction of suspended matter passing through the turbine runner at high altitudes is much less than 10%, so the interaction between particles and the influence of particle volume fraction on the liquid phase can be disregarded, making a sparse particle flow continuous phase model suitable. Based on the suspended matter sampling and testing results from high-altitude hydropower sites, the characteristic parameters of the suspended matter passing through the runner were set, including: particle density, median particle size, particle injection velocity, particle mass flow rate, drag force model, and particle shape characteristics. Using the DPM model, the motion trajectory, spatial distribution, and impact characteristics of sediment particles within the turbine runner at high altitudes under sparse flow conditions were obtained. The momentum exchange between particles and water flow (such as drag force) was quantified, and the wear location, local wear rate, and energy input of key runner components (such as blades and guide vanes) were predicted.

[0031] Furthermore, the wear rate acceleration factor was obtained through the study of the long-term wear characteristics of suspended matter. The wear rate acceleration factor is based on the flow field characteristics of S2 (such as shear stress distribution) to modify the particle-flow field coupling strength, so as to reflect the influence of long-term wear on the flow state.

[0032] In summary, by using the DPM model and setting the parameters of the suspended mass characteristics, the single-phase flow field is extended into a two-phase dynamic system of sediment-water flow, enabling the prediction of impeller wear based on the flow channel geometry.

[0033] S4. Construct a material wear model.

[0034] The suspended mass characteristic parameters set in S3 (such as particle density, median particle size, and injection velocity) are used as experimental variables for the material wear model to ensure consistency between the two-phase flow simulation and the material wear model experiment. The output of S3 can also be used as boundary conditions for the material wear behavior provided by S4 (e.g., which areas require focused wear analysis).

[0035] A hydraulic mechanical abrasion testing system was used to accurately simulate the turbulence state on the surface of the impeller blades through a disc-type planar flow abrasion test method, thereby obtaining the wear characteristics of the impeller blade material. Using these wear characteristics as input, a wear quantity relationship model considering the influence of the median particle size of different materials was established, and the wall wear rate was solved to obtain an accurate material wear model.

[0036] S5. Based on the material wear model, obtain a geometric morphology gradual wear prediction model.

[0037] Based on the long-cycle wear test results of the runner blades, a wear rate acceleration factor of S3 is introduced to modify the material wear model, so as to construct a gradual wear prediction model of the runner's flow wall geometry. The gradual wear prediction model of geometry considers the influence of the local wear process on the runner blade surface on the flow regime and wear, and obtains the true wear characteristics of the runner blade material.

[0038] During the long-term operation of turbine blades, the continuous impact of sediment particles leads to gradual wear of the blade surface material, causing dynamic evolution of the flow wall geometry. This evolution not only alters the original streamlined design of the blades but may also further exacerbate the wear process by affecting local flow regimes (such as velocity distribution, pressure gradient, and turbulence intensity), forming a combined effect of morphology change, flow regime alteration, and accelerated wear. To accurately predict this complex process, a predictive model considering gradual geometric changes needs to be constructed based on long-term wear experimental results.

[0039] Specifically, the wear rate acceleration factor in S3 is introduced to modify the material wear model constructed in S4, quantifying the nonlinear impact of geometric shape changes on local wear rates (e.g., the edges of wear pits experience accelerated wear due to concentrated flow velocity, forming a "pitting effect"). The modified model can dynamically reflect the gradual change process of blade surface morphology: on the one hand, it predicts the wear distribution of the next stage based on the current morphology parameters; on the other hand, it feeds back the morphology changes to the flow field calculation module to resolve the flow parameters, forming a closed-loop iteration of "flow field-wear-morphology". Through this method, the geometric shape gradual wear prediction model can not only obtain the wear characteristics of materials under real working conditions (such as the wear resistance decay law over time), but also accurately simulate the entire process of the blade surface changing from initial smoothness to local unevenness, and then to overall profile change, providing key technical support for runner maintenance cycle optimization, anti-wear coating design, and hydraulic performance evaluation. The final geometric morphology gradual wear prediction model has the core advantage of breaking through the assumption of fixed morphology or uniform wear in traditional models, and realizing high-precision prediction of the dynamic evolution of blade surface under complex multi-physics coupling.

[0040] S6. Calculation of the gradual wear prediction model.

[0041] The gradual wear depth is the mesh movement deformation value of the flow channel surface, and the accuracy of the wear morphology prediction is determined by the accuracy of the wear rate solution obtained by the custom wear model. To calculate the mesh movement deformation value of the turbine runner flow channel, the flow field, particles, and wear amount are calculated transiently at each time step in the numerical solution process, while the evolution of the wear morphology of the flow wall is updated based on the mesh movement deformation value calculated at each time step. Dynamic mesh technology transforms the static prediction model from the previous step into a prediction tool with spatiotemporal evolution capabilities, enabling simulation of the entire process from "local wear" to "global deformation".

[0042] In the wear prediction method based on solving the transient flow field, the mesh movement deformation value is related to the corresponding wear time interval. A sufficient number of flow field iterations needs to be set in the calculation to ensure that the flow field has converged after each mesh movement before the next mesh movement. First, the turbulent flow field is solved using the RNG k–ε model (an RNG k–ε model is a turbulence model derived from the Renormalization Group Theory, used to solve for the distribution of turbulent kinetic energy (k) and turbulent dissipation rate (ε) in a turbulent flow field). Simultaneously, the particle trajectory is calculated using a discrete particle model within the Lagrange framework (a Lagrange framework is a method for describing the motion of individual particles; its core idea is to track the changes in motion parameters such as position, velocity, and acceleration over time from the perspective of a single particle). The wear rate is calculated based on the particle-wall interaction using the McLaury wear prediction model (an engineering model for predicting particle erosion wear, whose core is to calculate the wear volume or depth caused by repeated particle impacts by quantifying the relationship between particle impact parameters (such as velocity, angle, and size) and material properties (such as hardness). The wear rate is further corrected using a wear rate acceleration factor determined by a rotating disk suspended mass wear test. The obtained wear depth is used as input to a dynamic mesh program, where the surface node displacement is iteratively updated according to a convergence criterion defined by a mesh node movement threshold. This allows for gradual deformation of the geometric boundaries throughout the simulation. Finally, the wear prediction results are compared with field measurement data to evaluate the model's accuracy and applicability.

[0043] S7. Perform fluid-structure interaction strength calculations for the initial structural field.

[0044] A turbine runner stiffness and strength calculation model was established. The dynamic pressure field obtained from the single-phase flow characteristics of S2 was coupled to the turbine runner stiffness and strength calculation model to perform initial structural field fluid-structure coupling strength characteristic calculation. The fluid-structure coupling strength characteristic calculation, by coupling the dynamic pressure field of the single-phase flow with the turbine runner stiffness and strength calculation model, obtained the dynamic stress distribution (including time-averaged stress, pulsating stress, and high stress concentration regions), overall deformation mode, and vibration characteristics (such as displacement field, natural frequencies, and mode shapes) of the runner under complex flow field conditions. It also revealed the influence of coupling mechanisms such as flow-induced vibration and cavitation-wear synergy on the structural response. Finally, a complete structural field result was formed, including stress time history, strain energy density, and boundary condition verification data, providing key inputs for subsequent fatigue life prediction and wear morphology evolution analysis. The initial structural field refers to the complete structural system composed of the geometry, material properties, and mechanical boundary conditions of the turbine runner's structure (such as blades, hub, and lower ring components) in the initial state of "no significant wear," serving as the basic structural model for subsequent coupling of flow field loads and analysis of strength characteristics.

[0045] S8. Calculation of fatigue characteristics of the initial structural field.

[0046] Based on the results of material fatigue tests and calculations of the initial structural field's fluid-structure interaction strength characteristics, fatigue prediction of the dynamic strain energy density amplitude of the coupled residual stress factor in the turbine is carried out. This method of predicting fatigue life using the dynamic strain energy density amplitude of the coupled stress factor is a fatigue life prediction method that integrates multi-physics coupling effects and energy density criteria. Its core lies in obtaining the dynamic stress field through fluid-structure interaction calculations, combining it with the dynamic strain energy density amplitude as a damage parameter, quantifying the accelerating effect of the coupled stress factor on fatigue damage, and ultimately achieving high-precision fatigue life prediction of the turbine under complex load conditions.

[0047] Next, the wear morphology evolution prediction method based on gradually varying geometric boundaries simulates the time-varying characteristics and evolution process of the wall geometry by moving the discrete wall mesh nodes. It relies on the dynamic mesh method (a technique in computational fluid dynamics (CFD) used to handle the dynamic changes of the fluid domain over time; its core is to adjust the mesh node positions in real time so that the mesh can adapt to the movement of the boundary or internal region, thereby accurately simulating unsteady flow problems) to dynamically update the mesh node positions of the discrete wall elements. Combined with topology optimization technology (a mathematical method that optimizes the material distribution within a given region based on given load conditions, constraints, and performance indicators, aiming to improve performance and achieve efficient resource utilization by adjusting the internal material layout of the structure), it eliminates geometric distortion caused by wear. Throughout the simulation, the wear time step is dynamically adjusted, with the maximum time step limited to within 0.02% of the total simulation time. To ensure mesh quality and avoid excessive local mesh distortion, each deformation threshold is also limited by a convergence criterion, requiring the maximum node displacement to be less than 2 micrometers. When both the deformation and time step meet the set requirements, the current time step ends and the next iteration step is executed until the wear depth calculation for the set total wear time is completed. This method can achieve accurate long-term wear prediction with reasonable computational cost. The reconstructed flow channel geometry model is then used for flow field and structural strength calculations. The specific steps are as follows: S9. Reconstruct the flow channel geometry gradient model based on the fatigue characteristic data of the initial structural field.

[0048] Based on the dynamic mesh change threshold calculated by wear, the flow channel geometry gradual model is reconstructed. The number of mesh smoothing steps is set using the dynamic mesh smoothing criterion. Mesh smoothing alone cannot provide a high-quality mesh, so the local reconstruction Remeshing function in the automatic dynamic mesh module needs to be activated. Automatic dynamic mesh is an advanced functional module in computational fluid dynamics software used to automatically manage the dynamic update of the mesh. Its core is to integrate three methods, Smoothing, Layering, and Remeshing, to achieve adaptive adjustment of the mesh when the boundary moves or deforms, thereby ensuring the accuracy and stability of the calculation. Simultaneously, it follows the principle of combining the spring approximation smoothing method (a dynamic mesh optimization technique based on the principle of spring mechanics, whose core is to abstract the connection between mesh nodes as a virtual spring, and realize the dynamic adjustment of mesh nodes by simulating the compression and stretching process of the spring, ultimately achieving a smooth and uniform mesh distribution) with local mesh reconstruction. By reasonably setting the spring stiffness coefficient and mesh quality threshold, while maintaining the mesh topology, it optimizes key indicators such as the aspect ratio and twist of the mesh elements. By dynamically controlling the movement trajectory of the mesh nodes and combining it with the real-time mesh quality evaluation mechanism, it effectively avoids computational instability caused by mesh distortion while simulating the time-varying characteristics of the wall geometry, providing a reliable mesh foundation for fluid-structure interaction calculations involving wear processes.

[0049] Based on the flow channel geometry gradual model reconstruction technology and the dynamic strain energy density amplitude fatigue prediction method coupled with residual stress factor, a calculation process for wear-fatigue combined damage failure of water turbine is constructed. The dynamic interaction between wear damage in the fluid domain and structural fatigue damage in the solid domain is realized through a two-way data transmission mechanism. The time-varying influence of sediment wear on the runner geometry is comprehensively considered, providing an analytical framework for the assessment of wear-fatigue combined damage failure.

[0050] S10, Calculation of flow characteristics during the reconstruction process of the flow channel geometry gradient model.

[0051] During the reconstruction of the flow channel geometry gradient model, the dynamically updated flow characteristic iterative algorithm cyclically interacts with the unsteady flow characteristics obtained from the numerical calculation of the flow channel after the reconstruction of the flow channel geometry gradient model and the flow field boundary correction caused by the deformation of the structural mesh, and performs flow characteristic calculations after the gradual wear model of the impeller is geometrically reconstructed.

[0052] S11, Fatigue characteristics calculation during the reconstruction process of the flow channel geometry gradient model.

[0053] During the reconstruction of the flow channel geometric gradient model, a dynamically updated iterative algorithm for fluid-structure interaction fatigue characteristics is employed. This algorithm utilizes flow channel geometric gradient wear and impeller fluid-structure interaction numerical calculations. It iteratively interacts with the unsteady flow characteristics obtained from the numerical calculations of the reconstructed flow channel after updating the flow field boundary, and with the flow field boundary correction caused by mesh deformation after impeller wear. Furthermore, after considering corrections for surface coefficient, size coefficient, and fatigue safety factor in the fatigue characteristic calculations, the impeller fluid-structure interaction strength characteristics and the dynamic strain energy density amplitude fatigue characteristics of coupled residual stress are calculated. The time step and deformation threshold limitation of the flow channel geometric gradient model reconstruction determine the time step for calculating the impeller's coupling damage threshold once during the entire calculation process.

[0054] S12. Fatigue failure prediction and assessment of the impeller under the influence of wear.

[0055] The fatigue failure prediction results of the rotor under wear effects (S1-S6 together form a complete chain for wear damage prediction, and their results directly affect the prediction of wear damage) and fatigue damage prediction results are used in the fatigue failure calculation of the rotor under wear effects. Based on Miner's linear damage accumulation theory, and according to the wear degradation and material mechanical property degradation models, combined with the fatigue failure prediction method of the rotor under wear effects, a fatigue failure prediction assessment of the rotor under wear effects is carried out. The safety factor of the wear degradation and material mechanical property degradation models is set to 1, and the safety factor of the fatigue failure prediction assessment model of the rotor under wear effects is also set to 1. When either the safety factor of the wear degradation and material mechanical property degradation models is >1, and the safety factor of the linearly accumulated damage prediction assessment model is >1, the rotor can be considered to have failed due to fatigue.

[0056] In some embodiments, the fatigue characteristic calculation in S11 employs a fatigue prediction method that considers the effect of wear on strain energy density amplitude. Strain energy density, a key parameter in materials science and mechanics, characterizes the energy stored in a material per unit volume due to elastic and plastic deformation under external loads, effectively reflecting the distribution and changes of energy within the material. Under cyclic stress, after a material undergoes a high number of cycles, fatigue failure often first occurs in regions with high strain energy density. This is because these regions, due to the accumulation of higher energy, are more prone to inducing microstructural damage, thereby promoting crack initiation and accelerating crack propagation. Compared to the local stress method, which relies on peak equivalent stress (which loses the multiaxial stress direction characteristics and cannot directly measure the total amount and spatial range of energy accumulation driving fatigue damage), the strain energy density method directly quantifies local energy accumulation (naturally including stress gradient and critical volume effect), which is more in line with the essence of high-cycle fatigue as a micro-damage process driven by local energy accumulation. Therefore, it can more accurately describe and locate the actual fatigue initiation point (often located in the second highest stress zone near the peak stress point at the notch root), rather than strictly corresponding to the peak point that may be insufficient in energy accumulation due to excessively steep gradient or small critical volume, as the local stress method does.

[0057] The study of the service conditions and structural form of high-altitude hydro turbine runners revealed that the cyclic loads borne by the runners were significantly lower than the yield strength of the materials. According to the theory of continuum mechanics, when the amplitude of the cyclic load meets the following conditions... At this stage, the material is in the linear elastic deformation phase, and the impeller only experiences recoverable elastic strain under high-cycle fatigue cyclic loading. And without plastic strain Accumulation. Based on elastic strain energy density. As a core parameter of the energy method, it characterizes the reversible mechanical energy stored per unit volume of material during elastic deformation. According to the first law of thermodynamics, when a load does work on a material... All of it is converted into elastic potential energy At that time, strain energy density Represented as: (1) For linear elastic materials, the stress-strain relationship is linear, and the integral result is: (2) The Hooke's law tensor for isotropic linear elastic materials is: (3) in, It is the stress tensor. and It is Lamé's constant. For volumetric strain, It is the Kronecker symbol (when) , ;when , ), It is the strain tensor.

[0058] Lamé's constant is: (4) (5) in, It is the elastic modulus of the material. It is the Poisson's ratio of the material.

[0059] Based on the above formulas (1) and (3), the strain energy density of the linear elastic material is obtained. ： (6) because With integral variables Irrelevant, and , We can obtain: (7) double dot product of strain tensor Expand into component form: (8) The formula for strain energy density is: (9) in, e 11 , e 22 , e 33 These are the normal strain components, corresponding to directions 1, 2, and 3 of the coordinate system, respectively. x , y , z Linear strain (tensile / compressive deformation) of the axis. e 12 , e 23 , e 31 The shear strain components correspond to the coordinate systems "1-2", "2-3", and "3-1" respectively. xyz noodle, yz noodle, zx Shear strain (shape distortion) of the surface.

[0060] From the perspective of micromechanical mechanisms, strain energy density decomposition theory constructs an important framework for characterizing the deformation energy of materials. Among them, volumetric strain... With skew strain The decoupling essentially realizes the separation of the volume change energy and the shape change energy in the material deformation energy. This decomposition provides a key theoretical basis for fatigue damage analysis because the initiation and propagation of fatigue cracks are essentially a process of energy accumulation to dissipation, and the component characteristics of the strain energy density are directly related to the damage evolution under different deformation modes.

[0061] In some embodiments, the starting positions of the damage and failure of the research object, the runner, are all at the welded joints of the runner blades and the lower ring near the water outlet side. In order to deeply explore the fatigue failure mechanism of the runner, to evaluate the fatigue damage degree and life of the runner based on the strain energy density method, it is necessary to construct the quantitative relationship between the damage evolution equation and the fatigue life, couple the macroscopic stress response and the microscopic damage accumulation mechanism, and use the unified modeling of continuous damage mechanics.

[0062] The Von Mises multiaxial stress is equivalently processed into an equivalent uniaxial stress ; where, Von Mises ( Mises) is a mechanical criterion used to equivalent a complex multiaxial stress state to a uniaxial stress state, aiming to simplify the evaluation of the fatigue life of materials under multiaxial stress - to convert the complex three - dimensional stress (such as the superposition of tensile, shear, and bending stresses) endured by the runner blades under fluid - structure interaction into an equivalent uniaxial stress value, and then calculate the fatigue life in combination with the uniaxial fatigue test data.

[0063] (10) Where, s 1, s 2, s 3 are the principal stress components in the three - dimensional stress state; In the elastic strain range, the change range of the dynamic strain energy density under a single - cycle load is: (11) Where, is the amplitude of the dynamic strain energy density; E is the elastic modulus; e a is the alternating strain amplitude, which quantifies the "elastic strain fluctuation amplitude" of the material per unit volume of the runner blade under cyclic loading; According to the high - cycle fatigue Basquin equation, establish the relationship between the amplitude of the strain energy density and the fatigue life: (12) Where, is the material fatigue characteristic constant, is the fatigue life, b is the fatigue strength index.

[0064] When applying material fatigue characteristics to turbine components, the effects of turbine size factor, surface finish factor, and fatigue safety factor must be comprehensively considered. For the surface condition of a high-altitude hydroelectric turbine turbine blade, the influence of a correction factor on material fatigue life can be quantitatively characterized by introducing a correction factor. The turbine blade underwent surface polishing during the manufacturing stage. Based on the definition of the surface finish factor, the formula is as follows: (13) in, The surface finish coefficient of the blade. This is a size factor. This is the fatigue safety factor.

[0065] Hydropower turbine runners are subjected to alternating stresses under various operating conditions for extended periods. This invention analyzes the fatigue characteristics of turbine runner blades based on Miner's linear cumulative damage criterion. Miner's theory assumes that fatigue damage under different stress levels can be linearly superimposed, providing a simple and effective analytical method for estimating fatigue life under variable amplitude loads in engineering applications. The formula is as follows: (14) in, D This represents the total cumulative damage value, where n i and N i The ratio represents the "single damage contribution" under the i-th working condition. The total damage D is obtained by summing the damage contributions of all working conditions. n i For the i-th working condition, the number of cyclic loads actually borne by the impeller; N i Let represent the fatigue life (limit number of cycles) that the wheel material can withstand under the i-th working condition.

[0066] According to the Basquin equation for high-cycle fatigue, combined with equations (13) and (14), the total cumulative damage value under each working condition is obtained. D as follows: (15) The failure prediction of the impeller under the combined effect of wear and fatigue is based on the linear cumulative damage theory of Miner's criterion. The fatigue damage in the wear process is quantified, and after considering safety factors such as size factor and surface factor, failure occurs when the total damage reaches a critical value ≥1.

[0067] Formula (15) takes into account the influence of wear on fatigue characteristics. By introducing strain energy density as a key parameter, it more accurately reflects the fatigue damage process of materials under complex load conditions. By predicting the fatigue life under different working conditions, maintenance plans can be arranged in advance, unplanned downtime can be reduced, and the operational stability and economic benefits of equipment can be improved.

[0068] The present invention also provides a combined wear-fatigue failure prediction system for turbine blades. The system includes a memory for storing computer program instructions and a processor for executing the program instructions. When the computer program instructions are executed by the processor, the system is triggered to execute the aforementioned combined wear-fatigue failure prediction method for turbine blades.

[0069] This invention proposes a method and system for predicting the combined wear and fatigue failure of turbine runners in hydropower stations located in high-altitude areas of western China. Addressing the complex and severe wear and fatigue failure problem faced by turbine runners under these conditions, the method constructs a full-chain model based on fluid-structure interaction (FSI), realizing a complete process from initial flow field-structure field co-calculation (FSI) to geometric gradual wear prediction, dynamic fatigue damage assessment, and iterative mesh reconstruction. This method introduces the quantitative relationship between sediment impact kinetic energy and material plastic deformation to accurately characterize the wear depth growth over time. Simultaneously, it establishes a dynamic mapping model between surface roughness and stress concentration coefficient to quantify the attenuation effect of wear on fatigue life. Furthermore, it utilizes dynamic mesh smoothing technology to achieve real-time reconstruction of the geometric model and iterative flow field-structure field coupling during the wear process, significantly improving the accuracy and scientific rigor of failure prediction. This provides strong technical support for the operation and maintenance management of hydropower equipment, helping to plan maintenance schedules in advance, reduce unplanned downtime, extend equipment service life, and optimize operation and maintenance management.

[0070] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the invention by those skilled in the art without departing from the spirit and essence of the invention. Such modifications or substitutions should all fall within the scope of the invention, or any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the invention should be covered within the protection scope of the invention. Therefore, the protection scope of the invention should be determined by the scope of the claims.

Claims

1. A method for predicting the combined wear-fatigue failure of a turbine blade, characterized in that, The method includes: A fluid computational domain model and mesh generation are established; a flow field is constructed based on the fluid computational domain model, mesh generation, and single-phase flow characteristics; a two-phase flow system is constructed based on the constructed flow field using a discrete phase model; a material wear model is constructed based on the material wear model and the parameters of the suspended mass characteristics set for the two-phase flow system; a geometric morphology gradual wear prediction model is obtained based on the material wear model; and the gradual wear prediction model is calculated. The fluid-structure interaction strength characteristics of the initial structural field are calculated; based on the results of the fluid-structure interaction strength characteristics calculation, the fatigue characteristics of the initial structural field are calculated; based on the fatigue characteristic data of the initial structural field, the flow channel geometry gradient model is reconstructed; the flow characteristics and fatigue characteristics are calculated during the flow channel geometry gradient model reconstruction process; and the fatigue failure prediction of the impeller under the influence of wear is evaluated.

2. The method according to claim 1, characterized in that, The construction of the flow field based on the fluid computational domain model, mesh generation, and single-phase flow characteristics includes: First, the steady flow characteristics of the entire flow channel are calculated, and the convergence result of the steady flow characteristics calculation is used as the initial condition for the unsteady flow characteristics calculation. Then, the unsteady flow characteristics are calculated to obtain the internal and external flow characteristics of the turbine under wide load flexible operation conditions, the law of operation conditions, pressure pulsation and vortex flow distribution and evolution dynamic characteristics.

3. The method according to claim 2, characterized in that, The construction of a two-phase flow system based on the established flow field using a discrete phase model includes: introducing a discrete phase model to simulate the motion and wear process of sediment particles based on the established flow field; the volume fraction of suspended mass passing through the turbine runner at high altitudes is less than 10%, and the interaction between particles and the influence of particle volume fraction on the liquid phase are not considered, so a sparse particle flow continuous phase model is applicable; setting the characteristic parameters of suspended mass passing through the turbine runner based on the sampling and testing results of suspended mass at high-altitude hydropower sites, including: particle material density, median particle size, particle injection velocity, particle mass flow rate, drag force model, and particle shape characteristics; and obtaining the motion trajectory, spatial distribution, and impact characteristics of sediment particles in the turbine runner at high altitudes under sparse flow conditions using the DPM model. Furthermore, the wear rate acceleration factor was obtained through the study of the long-cycle wear characteristics of suspended matter. The wear rate acceleration factor is based on the flow field characteristics to correct the particle-flow field coupling strength.

4. The method according to claim 3, characterized in that, The construction of a material wear model based on the suspended mass characteristic parameters set in the two-phase flow system includes: using the suspended mass characteristic parameters set in the two-phase flow system as experimental variables of the material wear model; using a hydraulic mechanical abrasion testing system to accurately simulate the turbulence state on the surface of the impeller blades through a disc-type planar flow abrasion test method to obtain the wear characteristics of the impeller blade material; using the wear characteristics as input to establish a wear amount relationship model for different materials considering the influence of the median particle size; and solving for the wall wear rate to obtain an accurate material wear model.

5. The method according to claim 4, characterized in that, The method for obtaining a geometrical wear prediction model based on a material wear model includes: modifying the constructed material wear model using the wear rate acceleration factor to quantify the nonlinear influence of geometrical changes on the local wear rate; the modified model can dynamically reflect the gradual change process of the blade surface morphology.

6. The method according to claim 5, characterized in that, The calculation of the gradual wear prediction model includes: The calculation of the gradual wear prediction model includes calculating the gradual wear depth, which is the mesh movement deformation value of the flow channel surface. In order to realize the calculation of the mesh movement deformation value of the turbine runner flow channel, the flow field, particles and wear amount are calculated transiently at each time step in the numerical solution process, and the evolution of the wear morphology of the flow wall is updated based on the mesh movement deformation value calculated at each time step. In the numerical solution process, the flow field, particles, and wear are calculated transiently at each time step. A sufficient number of flow field iterations are required to ensure that the flow field converges before the next mesh movement, after each mesh movement. First, the turbulent flow field is solved using the RNG k–ε model, while particle trajectories are calculated using a discrete particle model within a Lagrange framework. The wear rate is calculated based on the particle-wall interaction using the McLaury wear prediction model, and further corrected using the wear rate acceleration factor determined by the rotating disk suspended mass wear test. The obtained wear depth is used as input to the dynamic mesh program, where the surface node displacement is iteratively updated according to the convergence criterion defined by the mesh node movement threshold to obtain the mesh movement deformation value.

7. The method according to claim 6, characterized in that, The reconstruction of the flow channel geometry gradient model based on fatigue characteristic data of the initial structural field includes: Based on the dynamic mesh change threshold calculated by wear, the flow channel geometry gradual model is reconstructed. The number of mesh smoothing steps is set using the dynamic mesh smoothing criterion, and the local reconstruction function of the automatic dynamic mesh module in the computational fluid dynamics software is activated. At the same time, following the principle of combining the spring approximate smoothing method with local mesh reconstruction, by setting the spring stiffness coefficient and mesh quality threshold, the aspect ratio and twist of the mesh cells are optimized while maintaining the mesh topology, and the movement trajectory of the mesh nodes is dynamically controlled.

8. The method according to claim 7, characterized in that, The fatigue characteristic calculation of the process of reconstructing the geometric gradient model of the flow channel includes: The service conditions and structural form of high-altitude turbine runners mean that the amplitude of the cyclic loads borne by the runners is lower than the yield strength of the materials. According to the theory of continuum mechanics, when the amplitude of the cyclic loads meets the following conditions... At this stage, the material is in the linear elastic deformation phase, and the impeller only experiences recoverable elastic strain under high-cycle fatigue cyclic loading. And without plastic strain According to the first law of thermodynamics, when all the work done by the load on the material is converted into elastic potential energy, the elastic strain energy density is... Represented as: （1）； For linear elastic materials, the stress-strain relationship is linear, and the integral result is: （2）； The Hooke's law tensor form for isotropic linear elastic materials is: （3）； in, It is the stress tensor. and It is Lamé's constant. For volumetric strain; It is the Kronecker symbol, when , ;when , ; It is the strain tensor; Based on formulas (1) and (3), the strain energy density of linear elastic materials is obtained. : （6）； because With integral variables Irrelevant, and , ,get: （7）； double dot product of strain tensor Expand into component form: （8）； The formula for strain energy density is: （9）； in, ε 11 , ε 22 , ε 33 These are the normal strain components, corresponding to the coordinate systems respectively. x , y , z Linear strain of the shaft; ε 12 , ε 23 , ε 31 The shear strain components correspond to the coordinate systems respectively. xy noodle, yz noodle, zx Shear strain of the surface.

9. The method according to claim 8, characterized in that, Based on the strain energy density method for assessing the fatigue damage degree and life of the turbine runner, a quantitative relationship between the damage evolution equation and fatigue life is constructed. The Von Mises multiaxial stress is then used to treat the stress as an equivalent uniaxial stress. : （10）； in, σ 1. σ 2. σ 3 represents the principal stress components under three-dimensional stress conditions; Within the elastic strain range, the variation range of dynamic strain energy density under a single cyclic load is: （11）； in, For dynamic strain energy density amplitude, E It is the elastic modulus; ε a It is the alternating strain amplitude, which quantifies the "elastic strain fluctuation amplitude" of a unit volume of material under cyclic loading on the turbine blade; Based on the Basquin equation for high-cycle fatigue, the relationship between dynamic strain energy density amplitude and fatigue life is established: （12）； in, It is the material fatigue characteristic constant. It is fatigue life. b It is the fatigue strength index; When applying material fatigue characteristics to turbine runner components, the effects of runner size factor, surface finish factor, and fatigue safety factor must be comprehensively considered. For the surface condition of the turbine runner blades, a correction factor is introduced to quantitatively characterize its impact on material fatigue life. The runner blades have undergone surface polishing during the manufacturing stage; based on the definition of the surface finish factor, the formula is as follows: （13）； in, The surface finish coefficient of the blade. This is a size factor. The fatigue safety factor; During operation, the turbine runner is subjected to alternating stress under multiple working conditions for a long period of time. Based on Miner's linear cumulative damage criterion, assuming that fatigue damage under different stress levels is linearly superimposed, the formula is as follows: （14）； in, D This represents the total cumulative damage value, where n i and N i The ratio represents the "single damage contribution" under the i-th working condition. The total damage D is obtained by summing the damage contributions of all working conditions. n i For the i-th working condition, the number of cyclic loads actually borne by the impeller; N i Let be the fatigue life (limit number of cycles) that the wheel material can withstand under the i-th working condition. According to the Basquin equation for high-cycle fatigue, combined with equations (13) and (14), the total cumulative damage value under each working condition is obtained. D as follows: （15）； The failure prediction of the impeller under the combined effect of wear and fatigue is based on the linear cumulative damage theory of Miner's criterion. The fatigue damage during the wear process is quantified, and after considering the safety factor, failure occurs when the total damage reaches a critical value ≥1.

10. A combined wear-fatigue failure prediction system for turbine blades, the system comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein, When the computer program instructions are executed by the processor, the system is triggered to execute the method for predicting the combined wear-fatigue failure of the impeller blade as described in any one of claims 1 to 9.

Images