Efficient anti-cavitation mixed-flow water turbine runner under karst cave water quality and optimization method

Abstract

The invention provides an efficient anti-cavitation mixed-flow water turbine runner under karst cave water quality and an optimization method. The method comprises the steps that dynamic environment data of a target power station are obtained, a parameter model of a turbine runner is constructed, an initial runner scheme is generated based on the parameter model, and the parameter model is used for designing variables; establishing a simulation model based on the dynamic environment data; under various working conditions, performing flow characteristic calculation on the initial runner through the simulation model to generate a performance prediction result; constructing a mapping relation based on the design variable and the performance prediction result; and based on the mapping relationship, determining a target combination of the design variables to generate a runner design scheme. According to the method, multi-working-condition simulation and optimization are driven through dynamic environment data, a runner design scheme capable of adapting to dynamic changes of karst cave water quality is generated, and the problem that a traditional design environment is poor in adaptability is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydraulic machinery, and in particular to a high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality and an optimization method. BACKGROUND

[0002] China is rich in water energy resources, and hydropower stations are distributed throughout the country. Due to the influence of local climate and terrain, the water quality conditions of each hydropower station are different. Liulangdong Hydropower Station is the first hydropower station directly using underground rivers in karst areas for power generation. The runner type of Liulangdong Hydropower Station is old, and the cavitation cracks are serious. In addition, through maintenance, it is found that the bottom ring design of the unit has defects, which increases the maintenance period and causes the current unit to have low water energy efficiency, which seriously affects the economic benefit of the power station.

[0003] When the hydraulic turbine runner operates in the underground river water quality in the karst area, it faces extremely harsh working conditions. The water quality in the karst area is rich in calcium carbonate and carries angular and sharp rock particle inclusions, which causes the runner to be subjected to the synergistic effect of cavitation, abrasive wear, and chemical corrosion. More seriously, due to the influence of seasons and weather, the water flow characteristics change dynamically: high sediment content in the wet season leads to wear, and high salinity in the dry season leads to cavitation and scaling. This dynamic composite damage mechanism is a problem that has not been fully considered in existing runner design.

[0004] The existing runner design (such as the conventional HL runner) can be optimized for stable and single water quality conditions, and the blade profile (especially the outflow edge) is difficult to adapt to such a multi-mode, dynamically changing environment. SUMMARY

[0005] The present application provides a high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality and an optimization method to solve the problem that the runner cannot adapt to a multi-mode, dynamically changing environment.

[0006] In a first aspect, the present application provides a high-efficiency anti-cavitation mixed-flow hydraulic turbine runner optimization method under karst water quality, comprising: obtaining dynamic environment data of a target power station, the dynamic environment data including hydro-meteorological data, water quality characteristic data, and cavitation nucleus data under different seasonal typical working conditions; constructing a parameter model of the hydraulic turbine runner to generate an initial runner scheme based on the parameter model, the parameter model being used for design variables; establishing a simulation model based on the dynamic environment data; performing flow characteristic calculation on the initial runner under multiple working conditions through the simulation model to generate performance prediction results; constructing a mapping relationship based on the design variables and the performance prediction results; Based on the mapping relationship, a target combination of the design variables is determined to generate a runner design scheme.

[0007] By driving multi-working condition simulation and optimization with dynamic environment data, a runner design scheme that can adapt to dynamic changes of karst cave water quality is generated, solving the problem of poor environmental adaptability of traditional design.

[0008] In some possible embodiments, the design variables include a profile variable for controlling a blade skeleton line and a structure variable for defining a shape of a water outlet edge; The parameter model of the water turbine runner includes: Defining a three-dimensional curve path of the blade skeleton line, the three-dimensional curve path including a plurality of control points; Adjusting coordinates of the control points to generate the profile variable; Defining an asymmetric wedge angle of the water outlet edge and a thickness distribution coefficient along a blade height; Generating the structure variable based on the asymmetric wedge angle and the thickness distribution coefficient.

[0009] By precisely defining the blade skeleton line profile and the water outlet edge structure through the parameterized model, controllable design variables are provided for optimization.

[0010] In some possible embodiments, the simulation model includes a fluid dynamics model, a damage prediction model, and a life prediction model; The simulation model is established based on the dynamic environment data, including: A fluid dynamics model is constructed based on the hydro-meteorological data and the cavitation core data; A damage prediction model is constructed based on the water quality characteristic data and the fluid dynamics model; A life prediction model is constructed based on the fluid dynamics model.

[0011] By constructing a simulation model coupled with multiple physical fields, comprehensive quantitative evaluation of the runner's hydraulic performance, damage condition, and fatigue life is realized, replacing traditional single performance analysis.

[0012] In some possible embodiments, the simulation model includes a fluid dynamics model, a damage prediction model, and a life prediction model; The working conditions include a first working condition representing water quality in the dry season; In the first working condition, the simulation model is used to perform flow characteristic calculation on the initial runner to generate performance prediction results, including: The pressure distribution of a blade back surface area of the initial runner is calculated through the fluid dynamics model to obtain a lowest local pressure value of the back surface area, the blade back surface area including a water outlet edge back surface area; generate a pressure field and a pressure load based on the pressure distribution; calculate a cavitation void fraction of the back surface region of the exit edge based on the pressure field and by the damage prediction model; generate cavitation damage based on the damage prediction model; predict a crack risk of the back surface region of the exit edge based on the pressure load and the cavitation damage and by the life prediction model; generate a first performance prediction result based on the minimum local pressure value, the cavitation void fraction and the crack risk, the first performance prediction result being used to represent the anti-cavitation performance of the initial runner.

[0013] In the first working condition, the cavitation and crack risk of the low pressure area on the back surface of the blade are evaluated, and an anti-cavitation performance prediction result is generated to provide a basis for targeted improvement of the anti-cavitation ability.

[0014] In some possible embodiments, the simulation model includes a fluid dynamics model, a damage prediction model and a life prediction model; The working conditions include a second working condition representing water quality in the flat water period; In the second working condition, the simulation model is used to perform flow characteristic calculation on the initial runner to generate a performance prediction result, including: calculate energy conversion characteristics of the initial runner at different guide vane opening degrees by the fluid dynamics model to obtain an optimal efficiency point and operating parameters; generate a velocity field and a vorticity field of the whole flow passage based on the energy conversion characteristics; calculate flow stability of the initial runner in a non-design working condition based on the velocity field and the vorticity field and by the damage prediction model; generate an efficiency attenuation characteristic based on the damage prediction model; generate a second performance prediction result based on the optimal efficiency point, the operating parameters and the efficiency attenuation characteristic, the second performance prediction result being used to represent operating efficiency and stability of the initial runner.

[0015] In the second working condition, the efficiency and flow stability of the runner are evaluated to generate an efficiency performance prediction result, so that the optimization scheme has high efficient and stable operation ability in the flat water period.

[0016] In some possible embodiments, the simulation model includes a fluid dynamics model, a damage prediction model and a life prediction model; The working conditions include a third working condition representing water quality in the wet period; In the third working condition, the simulation model is used to perform flow characteristic calculation on the initial runner to generate a performance prediction result, including: calculating, by the fluid dynamics model, particle impact characteristics of the blade surface region of the initial runner, to obtain particle impact velocity and attack angle; generating particle trajectory distribution and local flow velocity field based on the particle impact characteristics; calculating, by the damage prediction model, material wear rate of the blade surface and the water inlet edge region based on the particle trajectory distribution and the flow velocity field; generating wear damage distribution based on the damage prediction model; calculating, by the life prediction model, material loss and strength degradation of the blade surface region under the sand-laden flow based on the particle impact velocity, attack angle and wear damage distribution; generating third performance prediction results based on the material wear rate, material loss and strength degradation, the third performance prediction results being used to characterize the anti-wear performance and structural strength of the initial runner.

[0017] In the third working condition, blade surface wear and strength degradation are evaluated to generate anti-wear performance prediction results, which provide a basis for enhancing the anti-erosion performance and structural reliability of the runner.

[0018] In some possible embodiments, the constructing a mapping relationship based on the design variables and the performance prediction results comprises: determining a plurality of sample points in the value space of the design variables by an experimental design method; for each of the sample points, performing flow characteristic calculation on the initial runner in multiple working conditions by the simulation model to generate performance prediction results corresponding to the sample point; based on all sample points and their corresponding performance prediction results, constructing a nonlinear mapping relationship from the design variables to the performance prediction results by a response surface model.

[0019] In some possible embodiments, the determining the target combination of the design variables based on the mapping relationship comprises: identifying a conflict relationship between the anti-cavitation performance characterized by the first performance prediction results and the anti-wear performance characterized by the third performance prediction results; constructing a multi-objective optimization function based on the conflict relationship, the optimization objectives of the function including maximizing the anti-cavitation performance and maximizing the anti-wear performance; solving the multi-objective optimization function in the space of the design variables to obtain a target solution set, each solution in the target solution set representing a design scheme achieving a different balance level between the anti-cavitation performance and the anti-wear performance; From the target solution set, a target solution is determined as a target combination of the design variables based on a preset rule, the preset rule being a decision rule balancing the conflict relationship.

[0020] By identifying and balancing the conflict between anti-cavitation and anti-wear performance, a best compromise solution is sought using multi-objective optimization to generate a design target combination of the runner that optimizes overall performance.

[0021] In some feasible embodiments, after the determination of the target solution as the target combination of the design variables from the target solution set based on the preset rule, the method further comprises: inputting the target combination of the design variables into the parametric model to generate three-dimensional geometric data of a target runner, the three-dimensional geometric data including spatial point, curve and surface data for defining the shape of the runner surface, and topological structure data defining the connection relationship between the upper crown, the lower crown and the blades; the three-dimensional geometric data of the target runner as the runner design scheme.

[0022] In a second aspect, the application provides a high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality, comprising an upper crown, a lower crown and a plurality of blades fixed to the upper crown and the lower crown, the blades being obtained based on the optimization method of the high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality of the first aspect.

[0023] The runner product obtained by the above optimization method has blade geometry with inherent high-performance characteristics suitable for adapting to the dynamic destruction environment of karst water quality.

[0024] In some feasible embodiments, the water outlet edge of the blade is a non-symmetrical structure defined by a structure variable in the design variables, and the structure variable includes a non-symmetrical wedge angle. The blade's spine curve is a three-dimensional curve defined by a profile variable in the design variables, and the profile variable includes the coordinates of a plurality of control points.

[0025] The runner blade has an optimized non-symmetrical water outlet edge and a three-dimensional curve spine, which directly strengthens the structure of the low-pressure area on the back of the water outlet edge and effectively suppresses cavitation and wear.

[0026] In some feasible embodiments, the thickness of the blade gradually changes along the height direction of the blade, and the thickness distribution is defined by a structure variable in the design variables, and the structure variable includes a thickness distribution coefficient.

[0027] The runner blade adopts an optimized thickness gradient distribution, so that the blade structure strength and hydraulic performance are optimally matched at different heights, improving the overall reliability.

[0028] From the above technical solutions, the application provides a high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality and an optimization method. The method comprises the following steps: obtaining dynamic environment data of a target power station, wherein the dynamic environment data comprises hydro-meteorological data, water quality characteristic data and cavitation nucleus data under different seasonal typical working conditions; constructing a parameter model of a hydraulic turbine runner, to generate an initial runner scheme based on the parameter model, wherein the parameter model is used for design variables; establishing a simulation model based on the dynamic environment data; performing flow characteristic calculation on the initial runner through the simulation model under multiple working conditions, to generate performance prediction results; constructing a mapping relationship based on the design variables and the performance prediction results; determining a target combination of the design variables based on the mapping relationship, to generate a runner design scheme. The method drives multiple working condition simulation and optimization through dynamic environment data, generates a runner design scheme that can adapt to dynamic changes of karst water quality, and solves the problem of poor environmental adaptability in traditional design. BRIEF DESCRIPTION OF DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0030] Figure 1 The flowchart of the optimization method of the high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality provided by the embodiments of the present application is shown in the figure. Figure 2 The flowchart of establishing a simulation model provided by the embodiments of the present application is shown in the figure. Figure 3 The flowchart of generating a first performance prediction result provided by the embodiments of the present application is shown in the figure. Figure 4 The flowchart of generating a second performance prediction result provided by the embodiments of the present application is shown in the figure. Figure 5 The flowchart of generating a third performance prediction result provided by the embodiments of the present application is shown in the figure. Figure 6 The structural diagram of the high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under karst water quality provided by the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0031] The embodiments will be described in detail below, and examples are shown in the drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following embodiments do not represent all the embodiments consistent with the present application.

[0032] The runner of a hydraulic turbine is applied in a hydropower station. The Liulongdong hydropower station is located in Yunnan Province and is a hydropower station directly utilizing an underground river in a karst area. The underground water in the karst area flows through a limestone layer and dissolves a large amount of calcium carbonate (CaCO3), forming supersaturated hard water. The CaCO3 microcrystals and bubbles rich in the hard water are carriers of cavitation nuclei. Cavitation bubbles are more likely to be generated on the surfaces of the microcrystals and rapidly collapse, thereby intensifying the strength and frequency of cavitation.

[0033] On the surface of the runner blade, due to the sharp change of pressure and flow rate, the rapid precipitation of CaCO3, i.e., scaling and dissolution, is induced, thereby damaging the metal oxide protective layer on the surface of the blade, exposing the fresh metal directly, accelerating chemical corrosion, and forming a vicious cycle of corrosion, scaling and peeling.

[0034] Different from the round silt of a surface river, the solid particles carried by the underground river are mainly limestone debris with sharp edges and corners. The particles have high hardness and sharp edges and corners, and under the driving of high-speed water flow, can also cause severe microscopic cutting on the surface of the blade.

[0035] The micro-jet generated by the collapse of the cavitation bubble can first flush and soften the metal surface, and the rock debris removes the material. The damage caused by the cavitation is much greater than the simple addition of the two.

[0036] In the runner, the runner includes a front surface and a back surface. The water flow impacts the working surface, i.e., the front surface or the pressure surface, which is the side of the blade that pushes the water flow and does work on the water, and transmits energy to the runner. The pressure on the front surface is high. The back surface, i.e., the suction surface, is opposite to the working surface. In order to form a pressure difference to drive the runner to rotate, the pressure on the back surface is lower than that on the working surface. Therefore, the back surface of the water outlet edge is the part of the blade back surface close to the water outlet edge.

[0037] The water outlet edge of the blade is far away from the water inlet edge and the edge where the water flow finally leaves the blade, is located at the outer edge of the runner, and connects the upper crown and the lower crown. Therefore, the back surface of the water outlet edge is a strip-shaped area starting from the upper crown and the lower crown of the runner, across the entire height of the blade back surface, and is not a point, but a line or a narrow surface.

[0038] According to fluid mechanics (Bernoulli's principle), when the water flow flows at high speed through the back surface of the blade (especially the part with large curvature), the pressure will sharply decrease. Among all the areas on the back surface of the blade, the pressure near the water outlet edge is the lowest due to the separation effect of the fluid.

[0039] When the local pressure here is reduced to the saturation vapor pressure at the water temperature at that time, the water flow will boil, producing a large amount of steam cavitation (cavitation phenomenon), and when these cavitation bubbles move with the water flow to an area with higher pressure, they will instantaneously collapse. Cavitation collapse will produce extremely high micro-jet impact force, which will hit the metal surface, and this impact will cause metal fatigue, material spalling, i.e. cavitation damage.

[0040] At the same time, the blade itself also bears the rotating centrifugal force and water pressure, and in the stress concentration area of the water outlet edge, the material damage caused by cavitation will become the source of fatigue cracks, which will continuously expand under alternating loads and eventually lead to blade cracking.

[0041] In summary, the back of the runner blade water outlet edge is the area with the lowest pressure. Due to low pressure, a large number of cavitation bubbles are generated and collapse, and low pressure reduces the solubility of CaCO3, which may cause scale layer to fall off and expose the matrix. When the water flow passes around the sharp water outlet edge, solid particles will directly impact or scratch the back of the blade due to inertia.

[0042] Traditional straight or simple curved water outlet edges cannot effectively guide such multiphase turbulent flow, resulting in strong local vortex and extremely uneven pressure distribution. Ultimately, fatigue cracks start from the water outlet edge and rapidly expand under the combined action of stress concentration points and cavitation damage points.

[0043] Specifically, the existing design method ignores the dynamic characteristics of water quality, is usually based on static and single water quality component assumptions, and does not fully consider the dynamic characteristics of underground river water quality in karst areas that change with the seasons, for example, lacks monitoring of the seasonal changes in the saturation index of CaCO3 in water, which cannot reflect the problem of increased scaling tendency caused by high mineralization during dry season; for another example, lacks quantitative analysis of the seasonal fluctuations in the size distribution and concentration of cavitation nuclei, leading to deviations between cavitation inception prediction and actual working conditions; for another example, lacks description of the angularity, hardness, and concentration spectrum of suspended particulate matter, simplifying multi-angular rock debris into circular silt, and underestimating the actual wear intensity.

[0044] Moreover, the existing method does not analyze the flow field in key areas in detail, especially lacking specialized pressure distribution monitoring of the narrow band area on the back of the water outlet edge, which cannot capture the extremely low pressure state in this area. Lack of simulation of the interaction between solid, liquid and gas in multiphase turbulent flow simplifies complex multiphase flow into single-phase flow. Lack of dynamic tracking of blade surface boundary layer separation and vortex evolution leads to distorted pressure fluctuation prediction.

[0045] For different seasons, the adaptability to seasonal operation conditions is insufficient, the existing optimization design is carried out around a single design condition, there is a lack of multi-condition coordinated optimization framework, the performance conflict between anti-cavitation in dry season and anti-wear in wet season cannot be balanced, the introduction of dynamic environmental load spectrum cannot take seasonal water quality change as a design input condition, and the life cycle performance evaluation ignores the alternating dominant damage modes that the runner bears in the annual cycle.

[0046] The part embodiments of the application provide a high-efficiency anti-cavitation mixed-flow water turbine runner optimization method under karst water quality. The variation law of CaCO3 saturation, particle characteristics and cavitation nucleus distribution is described by obtaining water quality characteristic data under different seasonal typical conditions. The mutual enhancement effect of cavitation and wear is quantified by a coordinated factor in a damage prediction model, reflecting the combined damage mechanism under karst water quality. The performance evaluation of a special area is realized by special pressure monitoring and bubble volume fraction calculation of the outlet side back area. A multi-condition optimization framework including dry season, normal water period and wet season is constructed to ensure the optimal comprehensive performance of the runner design scheme in the annual operation cycle.

[0047] By establishing a parameter model and a simulation model, the performance of the runner is predicted and evaluated. Then, a mapping relationship between the design variables and the performance prediction results is constructed to convert the simulation process into mathematical calculation and improve the speed of design iteration. Through optimization based on the mapping relationship, the target combination of design variables is determined to output a runner design scheme that achieves an optimal balance among multiple performance indicators. This can effectively solve the problem that the traditional runner design is difficult to adapt to multiple modes and dynamically changing damage environments. The obtained runner design scheme can adapt to complex water quality environmental changes.

[0048] As shown in Figure 1 , the method comprises: S100: obtaining dynamic environmental data of a target power station, and constructing a parameter model of a water turbine runner, the parameter model being used for design variables.

[0049] The water turbine runner design method in the prior art is usually based on idealized stable hydrological conditions, and the hydro-meteorological data used is mostly annual average value or single design condition value, without fully considering complex hydrological characteristics such as the distinct dry and wet seasons and strong underground river flow mutation in karst areas in Yunnan. This simplification makes the designed runner difficult to adapt to the dramatic dynamic changes in the actual operating environment, which is an important reason for the performance degradation and damage aggravation of the runner.

[0050] To solve the above problems, a dynamic environmental data system is introduced as a compensation factor for the model. The dynamic environmental data is obtained from the long-term operation monitoring system of the target hydropower station and can accurately reflect the spatial and temporal variation of water source conditions under the special terrain and climate in Yunnan. The dynamic environmental data includes hydro-meteorological data, water quality characteristic data, and cavitation core data under different seasonal typical working conditions.

[0051] The hydro-meteorological data, as the core compensation factor, is used to correct the deviation between the ideal model and the actual working conditions. This data set includes the physical state and motion characteristics of the target power station under different seasons and weather conditions in Yunnan's special climate conditions. Specifically, it includes typical daily flow variation curves in dry and wet seasons, water level variation curves in corresponding periods, and transient flow pulse data caused by extreme weather events such as single-point rainstorms in the southwest region. The introduction of these parameters effectively compensates for the lack of consideration of climate mutations in traditional design.

[0052] The water quality characteristic data, as a compensation factor for the corrosion and wear model, solves the water quality problem specific to karst geological regions. This data includes key parameters such as the dynamic change of water CaCO3 saturation index with seasons, the mass concentration fluctuation of suspended particles, the particle size distribution characteristics from limestone strata, and the mineral hardness of particles, providing data support for accurately predicting the damage of the runner under special water quality.

[0053] The cavitation core data, as a compensation factor for the cavitation model, is optimized for high CaCO3 water quality characteristics. This data includes the size spectrum distribution and number density of cavitation cores obtained through optical or acoustic measurement methods in different seasons, which directly reflects the influence of the rich CaCO3 microcrystals in karst water bodies on cavitation initiation.

[0054] The parameter model is a model that describes the three-dimensional geometry of the water turbine runner through mathematical parameters. This model converts the physical structure of the runner into variables that can be processed by a computer. Design variables are defined in the parameter model and are used to control the adjustment parameters of the runner geometry. The variable set constitutes the search space of the optimization process, and the value change affects the shape of the runner. These variables can include profile variables for controlling the shape of the blade profile, and structure variables for defining the local geometric characteristics of the blade.

[0055] The parameter model describes the geometry of the runner through a set of pre-defined design variables. Any modification to the shape of the runner is converted into an adjustment of the values of these design variables. By integrating dynamic environmental data as a compensation factor into the optimization design, this application achieves a high degree of adaptation of the runner design to the special terrain and climate conditions in Yunnan, effectively improving the performance and service life of the runner in the actual operating environment.

[0056] S200: generating an initial runner scheme based on the parameter model.

[0057] Based on the parameter model, an initial runner scheme can be generated. The initial runner scheme is a three-dimensional geometry model of the water turbine runner set before the optimization starts, and its geometric shape is determined by the values of all design variables in the parameter model.

[0058] For example, an initial runner scheme can be a parameterized model constructed based on the HL runner geometry parameters. In this scheme, the basic skeleton of the Francis runner is first established in the three-dimensional modeling software, and the core profile variable, i.e. the three-dimensional spatial coordinate points that control the shape of the blade profile, is set to be consistent with the HL runner blade profile data. The initial coordinates of these control points define a spatial curve with a specific curvature distribution rule, which constitutes the working surface and back surface profile of the blade.

[0059] At the same time, the structural variables of the initial runner scheme are set to specific initial values. For example, the asymmetric wedge angle that defines the shape of the water exit edge is set to an empirical value, and the wedge angle on the working surface side of the blade and the wedge angle on the back surface side have a fixed difference. In addition, the thickness distribution coefficient along the height direction of the blade is set to a linear uniform change mode from the upper crown to the lower crown of the blade, so that the thickness of the blade is maximum at the root and minimum at the water exit edge.

[0060] Thus, through the initial values of the profile variables and the structural variables, a runner model is generated, which has the geometric information required for fluid dynamics calculation, but its anti-cavitation performance, anti-wear performance and efficiency under the karst water quality have not been optimized.

[0061] S300: establishing a simulation model based on dynamic environment data.

[0062] After obtaining the dynamic environment data, a simulation model needs to be established. The simulation model is established in a computer environment and is used to simulate the water turbine runner to predict and evaluate the performance of the runner scheme defined by the parameter model, and calculate the performance prediction results reflecting its efficiency, reliability and life.

[0063] The simulation model can simulate the running state of the runner under the actual working conditions described by the dynamic environment data. By running this model, the behavior of the runner under various conditions can be predicted.

[0064] S400: performing flow characteristic calculation on the initial runner through the simulation model under multiple working conditions to generate performance prediction results.

[0065] The multiple operating conditions are determined based on dynamic environmental data, representing different operating conditions faced by the hydraulic turbine in different periods of the year. These operating conditions are used to test the performance of the runner to ensure that it can operate well in different seasonal environments. For example, these operating conditions can include an operating condition representing low flow and high mineralization conditions in the dry season, an operating condition representing design flow conditions in the flat season, and an operating condition representing high flow and high sediment content conditions in the wet season.

[0066] It should be noted that the simulation model in the present application does not simulate the environment, but simulates the working state and performance of the specific runner geometry defined by the parameter model under given environmental conditions. The parameter model provides the tested runner geometry entity, the dynamic environmental data defines the boundary conditions of the simulation, and the simulation model is a virtual system established in the computer to simulate the internal physical process of the specific runner entity under the specified environmental conditions.

[0067] Under the defined multiple operating conditions, the initial runner scheme is executed by the simulation model to perform flow characteristic calculation, which is used to quantify the performance of the specific runner geometry defined by the parameter model under different operating conditions. The simulation model outputs performance prediction results that can reflect the efficiency, durability and stability of the runner scheme by solving the physical equations of control fluid flow, energy conversion and material response.

[0068] Specifically, the simulation model takes the runner geometry data generated by the parameter model as the calculation entity, and takes the operating condition parameters defined by the dynamic environmental data as the boundary conditions. The complex flow inside the runner is simulated by computational fluid dynamics method, the evolution process of the material is predicted by damage model, and the mechanical response of the component is evaluated by structure model. The final output performance prediction result is a quantitative representation of the comprehensive performance of the specific runner geometry under specific environmental conditions.

[0069] This analysis method of coupling the parameter model (defining geometry) and the simulation model (evaluating performance) under multiple environmental operating conditions enables the designer to evaluate the adaptability of the same runner design scheme under different seasonal environments throughout the year, thereby overcoming the problem of performance degradation in actual complex environments caused by optimization only for a single ideal environment.

[0070] S500: Based on the design variables and the performance prediction results, a mapping relationship is constructed.

[0071] The mapping relationship represents the nonlinear correspondence between the design variables and the performance prediction results, and is used to simplify the simulation calculation process into a mathematical mapping, thereby screening different variable combinations in the design space.

[0072] S600: Based on the mapping relationship, a target combination of design variables is determined to generate a runner design scheme.

[0073] The target combination is determined from the design variable space to filter out a specific variable value set that can optimize the overall performance of the runner. By using the performance prediction provided by the mapping relationship, the design scheme is searched and determined through an optimization algorithm, thereby generating a runner design scheme that not only meets the efficiency requirement but also has high damage resistance.

[0074] For example, the seasonal flow, water quality composition and cavitation core concentration of the underground river hydropower station within three years are collected as dynamic environmental data, a parameter model containing 15 design variables is created to define the basic shape of the runner blade, then a simulation model is constructed using CFD and FEA software, which can simulate the operating state of the runner under three typical working conditions of dry season, normal water period and flood season. Through simulation calculation, a large number of design variable-performance prediction result data pairs are obtained, and the response surface model is trained as the mapping relationship. Finally, the genetic algorithm is used to perform multi-objective optimization on the response surface model to determine the target combination that achieves the best balance of efficiency, anti-cavitation and anti-wear performance, thereby outputting the three-dimensional design drawing of the runner.

[0075] For the design variables of the water turbine runner, in the embodiment, the profile variables for controlling the blade profile and the structure variables for defining the shape of the outflow edge are included.

[0076] The profile variables are used to control the three-dimensional space shape of the water turbine runner blade profile, and by adjusting the values, the curvature, torsion angle and placement position of the blade profile can be changed to determine the hydraulic performance of the runner. For example, the profile variables can be a set of coordinates of three-dimensional space control points, by changing the spatial position of these points, various different blade profile shapes from thick to thin, from flat to twisted can be generated.

[0077] The structure variables are used to define the local key geometric features of the runner blade. The variable describes the local structure that affects the anti-cavitation and anti-wear performance of the runner, and the value affects the geometric shape of these areas. The structure variables include the asymmetric wedge angle for describing the asymmetric geometric shape of the outflow edge, and the thickness distribution coefficient for describing the thickness variation rule of the blade from the root to the edge.

[0078] In some embodiments, the parameter model of the water turbine runner is constructed, including: defining a three-dimensional curve path of the blade profile; adjusting the coordinates of the control points to generate the profile variables; defining the asymmetric wedge angle of the outflow edge and the thickness distribution coefficient along the blade height; generating the structure variables based on the asymmetric wedge angle and the thickness distribution coefficient.

[0079] The three-dimensional curve path is a continuous spatial curve describing the center profile of the blade profile, which is used to generate the blade surface and determine the flow direction and energy conversion efficiency when the water flows through the blade. The three-dimensional curve path is defined by control points, and by adjusting the coordinates of these control points, the curve path can be smoothed or locally modified in shape.

[0080] For the profile variable, after the three-dimensional curve path is defined, the coordinates of the control points are adjusted, for example, moving a control point in the radial, circumferential or axial direction to control the blade setting angle, wrap angle and other hydraulic geometric parameters, to generate the profile variable.

[0081] The asymmetric wedge angle is used to set the blade outlet edge, which is the angle parameter of the geometric difference between the working surface and the back surface. This angle describes the degree of asymmetry of the cross-sectional shape of the outlet edge, and its value affects the pressure distribution in the low pressure area of the back surface of the outlet edge and the stability of the flow state, and is a geometric feature for resisting cavitation damage.

[0082] The thickness distribution coefficient is used to control the variation of the blade thickness along its height direction. This coefficient defines how the metal thickness of the blade changes from the upper crown connection to the lower crown connection, and from the inlet edge to the outlet edge, in order to balance the structural strength and hydraulic performance of the blade. Based on the asymmetric wedge angle and the thickness distribution coefficient, the structural variable is generated, which integrates the above local geometry definition into a data form that can be processed by the optimization algorithm.

[0083] Based on these two design variables, an initial runner scheme is generated. Based on the dynamic environmental data, a simulation model is established, which includes a fluid dynamics model, a damage prediction model and a life prediction model.

[0084] The fluid dynamics model solves the control of fluid motion by numerical methods to simulate the calculation model of the internal flow field of the hydraulic turbine, which is used to calculate the velocity distribution, pressure distribution, vorticity field and cavitation occurrence area in the runner flow passage, to predict the hydraulic efficiency, cavitation characteristics and pressure pulsation characteristics of the runner.

[0085] In this embodiment, the construction of the fluid dynamics model emphasizes the influence of weather and seasonal climate factors on the boundary conditions. Unlike the conventional design method of using annual average flow or fixed design value, the seasonal climate driving factor is used as the basis for setting the boundary conditions.

[0086] For example, instead of using static theoretical values for the inlet velocity boundary and outlet pressure boundary, hydrodynamic data determined by seasonal weather patterns are coupled. For example, when simulating dry season conditions, the inlet velocity boundary is derived from typical low flow data (e.g., 0.5 cubic meters per second) of the underground river system during the dry winter and spring seasons in Yunnan. When simulating wet season conditions, high flow data (e.g., 2.5 cubic meters per second) at the outlet of the underground river after being replenished by strong summer monsoon rains is used. The outlet pressure boundary is also set according to the water level observation curve for the corresponding season.

[0087] The core of this setup is to map the chain reaction process of weather rainfall as input conditions for the simulation. The distinct dry and wet season climate characteristics in Yunnan result in a dramatic cyclical fluctuation in water replenishment throughout the year. This climate-driven dynamic change in flow and water level is a key environmental variable that has not been fully considered in traditional design methods.

[0088] By using dynamic sequences instead of static values as boundary conditions, the fluid dynamics model can more realistically simulate the non-steady state operating conditions of the runner under actual climate-driven conditions, thereby reproducing extreme flow phenomena and potential damage patterns that only occur under specific seasonal heavy rainfall or prolonged drought conditions, providing a simulation environment for subsequent performance evaluation and optimization. For example, the model can capture the impact of a transient flow pulse caused by a single summer rainstorm on the stability of the runner's output and the structural load.

[0089] At the same time, cavitation core data, such as the size distribution and concentration of cavitation cores, are input into the parameter settings of the cavitation model (e.g., the Schnerr-Sauer model or the Zwart-Gerber-Belamri model) to simulate the initiation and development of cavitation. After the settings are complete, the fluid dynamics model can be used to solve the Navier-Stokes equation, outputting flow information including the full passage pressure field, velocity field, etc.

[0090] When establishing the damage prediction model, its construction is not only based on water quality characteristic data and fluid dynamics models, but more importantly, it is configured as a coupled damage model that can handle water-sand-air three-phase media. This model calculates cavitation, wear, and their synergistic effects in a unified framework by introducing local gas volume fraction (from the simulation of free gas in the fluid dynamics model) and solid particle concentration and properties, ensuring that the performance prediction results generated under all operating conditions fully consider the comprehensive impact of three-phase fluids in the karst water.

[0091] For example, Figure 2As shown, in some embodiments, a fluid dynamics model is constructed based on hydro-meteorological data and cavitation kernel data. The hydro-meteorological data is used to set the boundary conditions for the calculation, such as inlet flow rate and outlet pressure. At the same time, the cavitation kernel data is used to ensure the physical accuracy of the cavitation prediction.

[0092] The damage prediction model couples fluid dynamics calculation results with material response mechanisms to quantitatively predict the material loss rate of the impeller flow surface due to cavitation and wear in complex water environments.

[0093] In some embodiments, a damage prediction model is constructed based on water quality characteristic data and a fluid dynamics model. The water quality characteristic data provides the concentration, hardness, particle size distribution, and water chemical properties of solid particles in the water. The model combines these data with the flow field information calculated by the fluid dynamics model and estimates the damage rate of the material through semi-empirical or physical damage formulas.

[0094] For example, water quality characteristic data are read, such as the saturation index of CaCO3 in the water, the concentration, hardness and particle size distribution of suspended particles. The damage prediction model includes a cavitation damage sub-model and an abrasion damage sub-model. The cavitation damage sub-model calculates material loss based on local pressure changes and cavitation dynamics in the flow field, while the abrasion damage sub-model calculates the erosion wear rate by tracking the trajectory of solid particles and applying the Arcard abrasion formula or other abrasion models. The outputs of the two sub-models are combined to generate an overall damage distribution cloud map of each region on the surface of the impeller blade.

[0095] The life prediction model is based on structural mechanics and fatigue analysis theory. It is used to assess the structural integrity and service life of the runner under long-term alternating loads, and to predict the risk of fatigue crack initiation and life extension of blades and other key components under cyclic stress.

[0096] In some embodiments, a life prediction model is constructed based on a fluid dynamics model. Specifically, the unsteady pressure load acting on the surface of the runner blades, calculated by the fluid dynamics model, is used as the input load boundary condition.

[0097] The simulation model, constructed in a hierarchical manner, enables the evaluation of the turbine runner's performance. By constructing damage prediction and life prediction models, which rely on the calculation results of the fluid dynamics model respectively, the data source for performance prediction is ensured to be consistent, avoiding the problem of disconnect between the flow field, damage field, and stress field in traditional analysis.

[0098] After constructing the fluid dynamics model, damage prediction model, and life prediction model, the initial impeller is subjected to flow characteristic calculations under various operating conditions through simulation models to generate performance prediction results.

[0099] Because the runner is applied to the underground river in karst area, the water level of the underground river is different in the wet season and the dry season. In the wet season, the generator set often operates under super high load, the normal pressure and shear force of water flow on the surface of the blade increase, the mechanical stress level is high, and fatigue is accelerated. Rainwater erodes the surface and rock stratum, carries a large amount of silt and rock debris into the underground river, so that abrasive wear becomes the dominant failure mode, cooperates with cavitation, and the material loss rate reaches the annual peak.

[0100] In the dry season, the mechanical stress decreases, but due to the decrease of groundwater recharge, the water flow stays in the rock stratum for a longer time, the concentration of dissolved CaCO3 increases relatively, leading to the tendency of chemical scaling and the increase of cavitation nucleus concentration, so that cavitation and corrosion become the dominant failure mode.

[0101] Under different working conditions, different simulation models can be used to perform flow characteristic calculation to generate performance prediction results.

[0102] In some embodiments, the simulation model includes a model scheduler with dynamic environmental data, which can automatically select and combine the most suitable physical model and evaluation criteria based on the current input working condition attributes, such as seasonal type, to perform the most targeted flow characteristic calculation.

[0103] In some embodiments, in the first working condition, i.e. the dry season, the model scheduler identifies the characteristic parameters of low flow and high salinity, and automatically configures the simulation process with high-precision cavitation model and structural fatigue model as the core. Under this process, the minimum local pressure value, the cavitation volume fraction from the damage prediction model, and the crack risk are combined into the first performance prediction result to represent the anti-cavitation performance.

[0104] In the first working condition, in addition to the characteristics of high salinity water quality prone to scaling, the water is rich in free gas, such as CO2 precipitated from supersaturated water body, air, etc. These free gas microbubbles play the role of cavitation nucleus in the flow field, significantly reducing the energy threshold of cavitation.

[0105] Therefore, in this embodiment, when the cavitation volume fraction of the water edge back surface area is calculated by the damage prediction model, the influence of free gas is coupled in the calculation process. Specifically, the cavitation model not only considers the vaporization of water body caused by pressure reduction, but also takes into account the expansion and collapse of free gas microbubbles in the low pressure area. The extracted cavitation volume fraction result is essentially the gas phase volume fraction composed of water vapor and free gas, which can more truly reflect the initial size and intensity of the cavitation cloud under the water quality of the solution cave. Based on this more accurate cavitation damage prediction, the crack risk evaluated by the life prediction model is also more accurate.

[0106] By introducing a model scheduling mechanism based on working condition characteristics, the limitations of traditional single simulation mode in dealing with complex and variable environments are overcome, ensuring that the focus and accuracy of performance prediction match the current damage risk in different seasonal working conditions, thereby improving the pertinence and reliability of optimization design.

[0107] As shown in Figure 3 In some embodiments, in the first working condition, the flow characteristic calculation is performed on the initial runner by the simulation model to generate performance prediction results, including: The pressure distribution of the blade back area of the initial runner is calculated by the fluid dynamics model to obtain the minimum local pressure value of the back area, and the blade back area includes the back area of the water outlet edge; Based on the pressure distribution, a pressure field and a pressure load are generated; Based on the pressure field, the cavitation volume fraction of the back area of the water outlet edge is calculated by the damage prediction model; Based on the damage prediction model, cavitation damage is generated; Based on the pressure load and the cavitation damage, the crack risk of the back area of the water outlet edge is predicted by the life prediction model; Based on the minimum local pressure value, the cavitation volume fraction and the crack risk, a first performance prediction result is generated, which is used to characterize the anti-cavitation performance of the initial runner.

[0108] Under this working condition, the fluid dynamics model is numerically solved according to the set boundary conditions to obtain pressure field data including the blade back area.

[0109] After obtaining the pressure field data, the pressure load is generated to realize data conversion. Through the data mapping algorithm, the blade surface pressure distribution is transmitted to the finite element grid nodes of the runner structure to form a set of load boundary conditions.

[0110] Based on the pressure field data, the cavitation volume fraction of the back area of the water outlet edge is calculated by the damage prediction model, and the calculation of the cavitation volume fraction aims to quantify the cavitation intensity. Based on the comprehensive calculation of the damage prediction model, the quantitative results of cavitation damage are generated, i.e. the cavitation intensity is converted into material degradation prediction. Using the calculated cavitation volume fraction, combined with the water flow parameters and material characteristics, the mass loss rate caused by the impact of bubble collapse on the material surface is estimated by the cavitation prediction model, and a distribution map of cavitation damage is formed on the blade surface.

[0111] By combining pressure load and cavitation damage data, a life prediction model is used to predict the crack risk in the back region of the water edge. The life prediction model applies pressure load to the finite element model for stress analysis, while also considering the material property degradation and stress concentration effects caused by cavitation damage. A fatigue analysis algorithm is used to comprehensively evaluate the fatigue life of the back region of the water edge under the combined effects of alternating load and cavitation, thereby quantifying the crack risk.

[0112] Finally, by integrating the minimum local pressure value, cavitation volume fraction, and crack risk, the first performance prediction result is generated. This result characterizes the anti-cavitation performance of the initial runner scheme under the first operating condition from three dimensions: pressure level, cavitation intensity, and structural failure probability.

[0113] By calculating the cavitation volume fraction in the backwater region, the local details and intensity of cavitation can be represented, overcoming the limitations of relying solely on pressure thresholds for judgment. By coupling cavitation damage with mechanical pressure loads in a lifetime prediction model, the weakening effect of cavitation on material fatigue strength can be reflected, thus more accurately predicting structural failure risks.

[0114] like Figure 4 As shown, in the second operating condition, namely the level water period, the main goal is to improve efficiency. Therefore, the optimal efficiency point and operating parameters from the fluid dynamics model, as well as the efficiency decay characteristics from the damage prediction model, are extracted and combined into the second performance prediction results to characterize the operating efficiency and stability.

[0115] In some embodiments, under the second operating condition, flow characteristic calculations are performed on the initial runner using a simulation model to generate performance prediction results, including: The energy conversion characteristics of the initial runner under different guide vane openings were calculated using a fluid dynamics model to obtain the optimal efficiency point and operating parameters. Based on energy conversion characteristics, the velocity field and vorticity field of the entire flow channel are generated; Based on the velocity field and vorticity field, the flow stability of the initial impeller under non-design conditions is calculated using a damage prediction model. Based on the damage prediction model, efficiency decay characteristics are generated; Based on the optimal efficiency point, operating parameters, and efficiency decay characteristics, a second performance prediction result is generated, which is used to characterize the initial runner operating efficiency and stability.

[0116] The energy conversion characteristics of the initial runner under different guide vane openings are calculated using a fluid dynamics model. The calculated energy conversion characteristics are used to plot the runner's curves and define the working area. The fluid dynamics model performs calculations by continuously changing the boundary conditions of the guide vane opening while keeping the water head constant. Each calculation corresponds to one guide vane opening.

[0117] By recording the corresponding runner output power and hydraulic efficiency of each opening, the efficiency characteristic curve with guide vane opening or unit flow is obtained, and the peak efficiency point, i.e. the optimal efficiency point, and all the operating parameters corresponding to the point are identified from the curve.

[0118] Secondly, the velocity field and vorticity field of the full flow passage are further generated, and the flow field data calculated by the fluid dynamics model under multiple guide vane openings are extracted. The velocity field provides global information of flow line distribution and flow velocity size, and the vorticity field is obtained by mathematical operation on the velocity field, indicating the generation position and intensity of vortex structures such as blade back flow separation zone and draft tube vortex band.

[0119] Subsequently, based on the velocity field and vorticity field data, the flow stability of the initial runner under non-design conditions is calculated. The flow analysis module reads the velocity field and vorticity field data, and quantifies the instability degree of flow by analyzing the fluctuation amplitude of pressure and velocity at specific monitoring points and the non-stationary evolution characteristics of the vorticity field. For example, by evaluating the shape and swing amplitude of the vortex band in the draft tube or the shedding frequency of the vortex in the blade passage, it is determined whether there is a risk of inducing pressure pulsation and vibration.

[0120] Based on the comprehensive analysis of the damage prediction model, the efficiency decay characteristic is generated. The efficiency decay characteristic is the correlation between flow stability and macroscopic performance, and the correlation between the efficiency values calculated under different guide vane openings and the corresponding flow stability evaluation results. When the flow stability is detected to deteriorate at a certain opening, the efficiency value near the opening will drop sharply. By fitting the curve of efficiency change with opening, the trend and rate of efficiency decay from the peak value can be obtained, i.e. the efficiency decay characteristic.

[0121] Integrating the optimal efficiency point, the corresponding operating parameters and the efficiency decay characteristic, the second performance prediction result is generated. This result not only represents the highest efficiency level that the runner can achieve under normal water period conditions and the ideal operating state corresponding to it, but also predicts the maintenance ability of runner efficiency and the stability of flow when the actual operating conditions deviate from the ideal state.

[0122] By combining the energy conversion characteristic with the flow field stability, the evaluation of the operating performance of the runner under normal water period conditions is realized. By calculating the energy conversion characteristic under different guide vane openings, the highest efficiency point of the runner and its best operating parameters can be determined, providing targets for the optimization design and control of the water turbine. By analyzing the velocity field and vorticity field and evaluating the flow stability, the internal flow mechanism leading to efficiency decay is revealed, and the operating risk of the runner under non-design conditions is predicted.

[0123] In the third operating condition, namely the high-water season, wear is the primary concern. Therefore, particle impact velocity and angle of attack from the hydrodynamic model, material wear rate from the damage prediction model, and material loss and strength degradation from the life prediction model are extracted and combined into the third performance prediction results to characterize wear resistance and structural strength.

[0124] In the high-speed turbulent flow of the third operating condition, a large amount of free gas is carried, forming a three-phase mixed flow of water, sand and gas. The damage mechanism of this three-phase flow to the working surface of the blade is more complex. There is a significant synergistic enhancement effect between the micro-jet generated by cavitation collapse and the micro-cutting of solid particles.

[0125] Therefore, in this embodiment, when calculating the material wear rate of the blade working surface using the damage prediction model, a gas-solid coupled damage mechanism is introduced. This model not only calculates the direct impact wear of solid particles through particle trajectory tracking, but also analyzes the local high-pressure microjets generated by the collapse of free gas cavitation near the particle impact point. These microjets pre-soften or micro-damage the material surface, enabling the solid particle impact to remove more material. The material wear rate calculated based on this coupled model can more accurately predict the actual material loss under the combined action of three-phase flow.

[0126] By introducing and quantifying the influence of free gas in the performance prediction of operating conditions, the simulation model can improve the completeness and accuracy of predicting the service behavior of the turbine runner in real karst cave water quality environments by changing from two-phase flow (water-sand) to three-phase flow (water-sand-gas).

[0127] like Figure 5 As shown, in some embodiments, under the third operating condition, flow characteristic calculations are performed on the initial impeller using a simulation model to generate performance prediction results, including: The particle impact characteristics of the blade working surface region of the initial impeller are calculated using a fluid dynamics model to obtain the particle impact velocity and angle of attack. Based on the particle impact characteristics, particle trajectory distribution and local velocity field are generated; Based on particle trajectory distribution and flow velocity field, the material wear rate of the blade working surface and the inlet edge area is calculated using a damage prediction model. Based on the damage prediction model, the wear damage distribution is generated; Based on particle impact velocity, angle of attack, and wear damage distribution, a life prediction model is used to calculate the material loss and strength degradation of the blade working surface area under sand-laden water flow. Based on material wear rate, material loss, and strength degradation, a third performance prediction result is generated, which is used to characterize the wear resistance and structural strength of the initial impeller.

[0128] Under the third operating condition of high-water-content water quality during the high-water season, the impeller operates under high-sediment-content flow conditions to evaluate its wear resistance and structural integrity. Under high-sediment-content conditions, the fluid dynamics model is set based on a discrete phase model. By tracking the motion trajectory of solid particles, the average impact velocity and typical angle of attack distribution of particles when impacting the blade working surface are obtained. These parameters together describe the impact intensity and action mode of particles on the blade surface.

[0129] Among them, particle impact velocity is the relative velocity of solid particles at the instant of impact with the blade surface, which determines the magnitude of particle impact kinetic energy. Angle of attack is the angle between the trajectory of the solid particle and the tangent direction of the blade surface, which determines whether the particle impact is a normal impact or oblique scraping. The wear mechanism and wear rate of the material differ under different angles of attack.

[0130] Then, particle motion path data are extracted to form a trajectory distribution map of particles in the flow channel. At the same time, detailed velocity field information near the blade working surface is obtained. These data together reveal the formation mechanism of particle transport mode and impact location. Among them, particle trajectory distribution is the spatial statistical result of the motion path of solid particles in the impeller flow channel, and local velocity field is the spatial distribution of water flow velocity vector in the impeller flow channel, especially in the region near the blade working surface.

[0131] The purpose of generating wear damage distribution is to visualize the spatial distribution characteristics of wear. By mapping the material wear rate data onto the blade geometry model, a wear damage distribution cloud map is formed, which marks the locations of severe wear on the blade working surface and the inlet edge region.

[0132] Based on data on particle impact velocity, angle of attack, and wear damage distribution, a life prediction model is used to calculate the material loss and strength degradation of the blade's working surface under sediment-laden water flow. The calculation of material loss and strength degradation aims to assess the impact of wear on structural integrity. The life prediction model first obtains the material loss after a specific operating period by integrating the wear rate. Then, based on the post-wear geometric model, a structural mechanics analysis is performed to assess the stress level changes and load-bearing capacity reduction in key parts of the blade, i.e., the strength degradation.

[0133] Finally, by integrating material wear rate, material loss, and strength degradation, a third performance prediction result is generated. This result characterizes the wear resistance and structural strength of the initial impeller design under the third operating condition from three aspects: instantaneous wear rate, cumulative material loss, and structural performance degradation.

[0134] The anti-wear performance and structural strength evaluation method for the third working condition can provide data on the durability of the runner scheme in the high-sediment water flow environment in the wet season, ensure the judgment of the runner wear resistance, and not only focus on the instantaneous wear rate, but also consider the long-term effects such as cumulative material loss and structural strength degradation, thereby guiding the optimization process and improving the service life and operation reliability of the runner in harsh water quality conditions.

[0135] For the wet season, the inherent anti-wear characteristics of the runner under the erosion of high-concentration, high-hardness, and sharp-edged solid particles are quantified. Specifically, it includes particle impact kinetic energy on the blade working surface, material wear rate, and structural strength degradation.

[0136] For different working conditions and design variables, multiple mapping relationships can be generated, which represent the internal relationship between design variables and particle trajectories, impact angles, and energy, and describe how to actively guide the solid phase flow through geometric design to change the positive cutting to sliding guidance, thereby cutting off the cavitation erosion synergy effect.

[0137] However, the final determined target combination is not only suitable for a single working condition, but also deeply understands and adapts to the complexity of karst water quality.

[0138] Therefore, in the process of generating the mapping relationship, in some embodiments, a plurality of sample points are determined in the value space of the design variables by an experimental design method; For each sample point, flow characteristic calculation is performed on the initial runner under multiple working conditions by a simulation model to generate performance prediction results corresponding to the sample point; Based on all sample points and their corresponding performance prediction results, a response surface model is used to construct a nonlinear mapping relationship from design variables to performance prediction results.

[0139] The experimental design method is used to select a sample point set in a given design variable space, and the characteristics of the design space are reflected through limited sample points, which can minimize the simulation calculation cost while ensuring the construction accuracy.

[0140] The sample point is a specific parameter combination selected in the value space of the design variables by the experimental design method. Each sample point represents a possible hydraulic turbine runner design scheme. After determining the sample point set, flow characteristic calculation is performed on the initial runner under multiple working conditions by a simulation model for each sample point. The sample point simulation is performed to provide training data for constructing the mapping relationship.

[0141] The design variable values of each sample point are assigned to the parameter model to generate the corresponding runner geometry. The simulation model is run under the first working condition, the second working condition, and the third working condition, respectively, to calculate the complete performance prediction results corresponding to the sample point, including anti-cavitation performance, operating efficiency, and anti-wear performance.

[0142] Based on all sample points and their corresponding performance prediction results, a response surface model is used to construct a nonlinear mapping relationship between design variables and performance prediction results. The design variable data of the sample points are used as input, and the corresponding performance prediction results are used as output to train the Kriging response surface model. This model learns the complex relationship between variables and performance using Gaussian process regression. After cross-validation confirms the model's accuracy, a nonlinear mapping relationship capable of predicting performance under various operating conditions for any combination of design variables is established.

[0143] Based on the mapping relationship, determine the target combination of design variables, including: Identify the conflicting relationship between the cavitation resistance performance characterized by the first performance prediction result and the wear resistance performance characterized by the third performance prediction result; Based on the conflict relationship, a multi-objective optimization function is constructed. The optimization objectives of the function include maximizing anti-cavitation performance and maximizing anti-wear performance. Within the space of design variables, a multi-objective optimization function is solved to obtain an objective solution set. Each solution in the objective solution set represents a design scheme that achieves different balance levels between anti-cavitation performance and anti-wear performance. From the set of target solutions, based on preset rules, the target solutions are determined as the target combination of design variables. The preset rules are decision rules for balancing conflicting relationships.

[0144] For the dry season, the quantification focuses on the inherent cavitation resistance characteristics of the impeller under conditions of high cavitation nucleus concentration and low flow velocity. Specifically, this includes the minimum local pressure value on the back of the blades, the volume fraction of cavitation bubbles on the back of the outlet edge, and the risk of microcrack initiation.

[0145] The mapping relationship characterizes the intrinsic link between design variables (such as blade curvature and outlet edge shape) and cavitation initiation tendency and cavitation fatigue sensitivity, describing how to directly intervene in and improve the physicochemical environment of the low-pressure zone by adjusting geometry, thereby counteracting the initiation of phase change catalysis and dynamic scaling.

[0146] Conflict relationships refer to the mutual constraints between different performance indicators, where improving one indicator may lead to the deterioration of another. In this embodiment, the conflict relationship is between anti-cavitation performance and anti-wear performance.

[0147] By analyzing the data provided by the mapping relationship, we can observe the changing trend of anti-cavitation performance as anti-cavitation performance improves. When the two performance indicators show a significant negative correlation, we can confirm that there is a conflict between them that needs to be resolved.

[0148] After confirming the existence of conflict relations, a multi-objective optimization function is constructed. The multi-objective optimization function is a mathematical function containing two or more conflicting optimization objectives, used to find the best solution when all objectives cannot be optimized simultaneously. It is constructed directly based on the identified conflict relations, with the conflicting performance indicators as optimization objectives.

[0149] The optimization objectives of the multi-objective optimization function include maximizing anti-cavitation performance and maximizing anti-wear performance. These two objectives represent the performance requirements of the runner in different operating conditions, such as dry season and wet season.

[0150] Solving the multi-objective optimization function in the space of design variables is a process of finding all possible balanced solutions. Using multi-objective optimization algorithms, we search in the high-dimensional space composed of all design variables to find solutions that cannot be simultaneously surpassed by other solutions in terms of anti-cavitation performance and anti-wear performance. These solutions constitute the target solution set, each solution representing a design scheme with different balance levels between the two performances.

[0151] From the target solution set, the target solution is determined as the target combination of design variables based on pre-set rules. The final scheme is determined based on engineering judgment. The pre-set rules balance the identified conflict relations and can use weight allocation method, ideal point method, or preference selection method based on designer experience to determine the best solution between anti-cavitation and anti-wear performance as the final target combination of design variables.

[0152] In some embodiments, the target combination of design variables is input into the parametric model to generate three-dimensional geometric data of the target runner; The three-dimensional geometric data of the target runner is used as the runner design scheme.

[0153] The three-dimensional geometric data includes spatial points, curves, and surface data for defining the shape of the runner surface, as well as topological structure data for defining the connection relationship between the upper crown, lower crown, and blades. The spatial point, curve, and surface data are used to describe the geometric shape of the runner surface. These data include control vertex coordinates for defining blade profiles, spline curve parameters for constructing surfaces, and mesh or parameterized surface information for describing complex surfaces, which together determine the fluid contact surface morphology of the runner.

[0154] The topological structure data is a collection of information describing the connection relationship and relative position between the components of the runner. This data defines the assembly constraints and spatial fitting relationship between the upper crown, lower crown, and blades, ensuring that the components can be correctly assembled into a functional whole runner.

[0155] By converting the optimized design variables into three-dimensional geometric data, it is ensured that the best parameter combination obtained by optimization can be reflected in the specific product geometric features, avoiding the problem of disconnection between optimization and modeling in traditional design.

[0156] The method drives multi-working condition simulation and optimization by dynamic environment data, generates a runner design scheme that can adapt to dynamic changes of karst cave water quality, and solves the poor environmental adaptability problem of traditional design.

[0157] Based on the above-mentioned optimization method of the high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under the karst cave water quality, as shown in Figure 6 the embodiments of the present application provide a high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under the karst cave water quality, which comprises an upper crown, a lower crown and a plurality of blades fixed on the upper crown and the lower crown, and the blades are obtained based on the optimization method of the high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under the karst cave water quality.

[0158] The upper crown is a rotating structural component in the mixed-flow hydraulic turbine runner, which is located at the top of the blade and connected thereto, constitutes the upper boundary of the runner, is connected with the main shaft to transmit torque, and its geometric shape affects the overall stiffness of the runner and the flow characteristics of the upper crown gap of the water flow.

[0159] The lower crown is a rotating structural component in the mixed-flow hydraulic turbine runner, which is located at the bottom of the blade and connected thereto, constitutes the lower boundary of the runner, and forms an installation skeleton of the blade together with the upper crown, and its profile affects the hydraulic efficiency and structural stability of the runner.

[0160] The blade is fixed between the upper crown and the lower crown, and is used for converting water flow energy into mechanical energy. The geometric shape of the component determines the energy conversion efficiency, cavitation characteristics and anti-wear ability of the runner.

[0161] The upper crown of the runner can be casted by high-strength stainless steel, including a central flange plate and a peripheral annular bearing body. The central flange plate is provided with a tapered hole and a key groove matched with the main shaft for transmitting torque. The peripheral annular bearing body has an installation curved surface connected with the upper end of the blade. The profile of the curved surface is optimized and designed to realize smooth transition with the blade and reduce vortex loss of water flow in the upper crown gap.

[0162] The lower crown is made of the same material as the upper crown, including an inner ring body and an outer ring body. The inner ring body is provided with an installation curved surface connected with the lower end of the blade. The profile of the curved surface is matched with the shape of the root of the blade. The outer ring body has a guide curved surface for ensuring smooth transition of water flow. The curvature radius of the curved surface is optimized and calculated to effectively guide water flow into the blade flow passage and reduce the inlet impact loss.

[0163] A plurality of blades are fixed between the upper crown and the lower crown in a circumferential uniform distribution, the spatial position and geometric shape of each blade are embodied by the target combination obtained after the above optimization method. The blade includes a water inlet edge, a water outlet edge, a working surface and a back surface. The water inlet edge adopts an optimized curved shape, which can reduce the impact loss of the incoming flow; the water outlet edge adopts a specific profile design, which can control the generation of wake vortex; the working surface as the main working surface, its profile ensures efficient energy conversion; the back surface profile is optimized to improve the pressure distribution and suppress cavitation.

[0164] In some embodiments, the blade's bone line is a three-dimensional curve defined by a profile variable in the design variable, and the profile variable includes the coordinates of a plurality of control points.

[0165] The profile variable defines the three-dimensional curve path of the blade's bone line, and the water inlet edge is the starting contour line of this three-dimensional curve. Adjusting the control point coordinates of the profile variable in the optimization process will directly change the spatial curvature, inclination and shape of the water inlet edge to optimize the water flow angle of attack.

[0166] During the wet season, the particle impact characteristics of the blade working surface area are simulated to set the optimization direction of the water inlet edge. To avoid sharp rock particles causing severe cutting to the water inlet edge, the optimization algorithm will tend to generate a smoother, smoother water inlet edge curve to guide rather than impact the sediment-laden water flow into the blade flow passage.

[0167] After optimization, the water inlet edge can be optimized from the traditional relatively sharp wedge shape to a circular arc shape or an elliptical curve shape with a specific radius, thereby effectively reducing particle impact and inlet vortex.

[0168] In some embodiments, the water outlet edge of the blade is a non-symmetrical structure defined by a structure variable in the design variable, and the structure variable includes a non-symmetrical wedge angle.

[0169] The structure variable defines the non-symmetrical wedge angle of the water outlet edge. During the dry season, the void fraction and crack risk of the back surface area of the water outlet edge are evaluated, and the evaluation results drive the optimization of the shape of the water outlet edge. To improve the minimum pressure of the back surface and suppress cavitation, the optimization algorithm will adjust the non-symmetrical angle to form a back surface side with a smoother, thicker profile.

[0170] For example, the working surface side of the water outlet edge is relatively steep to maintain the water outlet efficiency, and the back surface side adopts a large rounding radius (e.g. R=2mm) to form a pressure relief area, thereby effectively suppressing the generation and collapse of cavitation.

[0171] After the water inlet edge is determined, the profile variable also defines the curved surface of the working face. In the high water period, the energy conversion characteristics are optimized to maximize efficiency, and the overall profile of the working face must meet the efficient energy conversion law, which can be a smooth, continuous and aerodynamic design surface.

[0172] In the wet season, the particle impact characteristics and material wear rate of the working surface are evaluated, which will affect the optimization of the local curvature of the working surface, avoid forming pits or bosses that are easy to be subjected to frontal impact, and guide the smooth sliding of particles.

[0173] For example, the working surface can be a composite curved surface with a specific pressure distribution, which can maintain laminar flow under designed conditions to achieve high efficiency, and make the particle trajectory smooth under sediment-laden flow to reduce local impact wear.

[0174] The profile variable also defines the curved surface of the back surface. In the dry season, the pressure distribution and the lowest local pressure value of the back surface area are analyzed to guide the optimization of the back surface profile. The goal is to eliminate local low pressure peaks and make the pressure distribution as smooth and uniform as possible.

[0175] For example, compared with the working surface, the back surface curve can be set to have different curvature changes, such as forming a pressure transition zone near the water outlet edge to avoid a sharp drop in pressure below the vaporization pressure.

[0176] The blade is fixed by welding connection, the installation curve of the upper and lower crowns is beveled, the end of the blade is processed into a corresponding butt joint shape, and the blade is firmly connected by full penetration welding process. After welding, the whole heat treatment is carried out to eliminate the welding residual stress and ensure the structural integrity of the runner under high speed rotating working condition.

[0177] In some embodiments, along the height direction of the blade, the thickness of the blade gradually changes, and the thickness distribution is defined by the structure variable in the design variable, and the structure variable includes a thickness distribution coefficient.

[0178] The thickness distribution coefficient as a component of the structure variable is in the form of a function parameter or a discrete numerical sequence, which defines the thickness change law from the upper crown connection of the blade to the lower crown connection.

[0179] In the manufacturing process of the runner, the manufacturing system reads the thickness distribution coefficient data output by the optimization method, and generates the machining path of the blade blank according to the coefficient. The machining equipment continuously adjusts the cutting depth in the height direction of the blade according to the change law specified by the thickness distribution coefficient, so that the blade at different height positions presents the thickness value required by the design, forming a smooth transition of the thickness gradual distribution.

[0180] According to the thickness distribution coefficient, the thickness values of each section of the blade are calculated, the sections are sequentially arranged in the height direction, and a three-dimensional blade model with continuous thickness change is generated through surface fitting technology. The model serves as the geometric basis for the subsequent manufacturing process, ensuring that the finished blade is consistent with the optimized design scheme.

[0181] The optimized design of the thickness gradient distribution considers the structure of the runner in the karst water quality environment, adopts a larger thickness value at the upper crown connection area of the blade to ensure structural strength, adopts a moderate thickness at the middle area of the blade to balance the hydraulic performance and structural weight, and adopts a smaller thickness at the water outlet edge area of the blade to improve the cavitation performance.

[0182] The runner provided in the embodiment realizes the collaborative optimization of the structural performance and the hydraulic performance of the blade by adopting the thickness gradient distribution based on the thickness distribution coefficient.

[0183] The similar parts among the embodiments provided in the application can be referred to each other, the specific implementation provided above is only a few examples under the general concept of the application, and does not limit the protection scope of the application. For those skilled in the art, any other implementation extended according to the application scheme without creative labor belongs to the protection scope of the application.

Claims

1. A method for optimizing a high-efficiency anti-cavitation mixed-flow water turbine runner under the quality of cave water, characterized in that, The method comprises: acquiring dynamic environment data of a target power station, the dynamic environment data comprising hydro-meteorological data, water quality characteristic data, and cavitation core data under typical working conditions in different seasons; constructing a parameter model of a water turbine runner to generate an initial runner scheme based on the parameter model, the parameter model being used for design variables; establishing a simulation model based on the dynamic environment data; performing flow characteristic calculation on the initial runner under multiple working conditions by using the simulation model to generate performance prediction results; constructing a mapping relationship based on the design variables and the performance prediction results; determining a target combination of the design variables based on the mapping relationship to generate a runner design scheme.

2. The method of claim 1, wherein the method is characterized by: The design variables comprise profile variables for controlling a blade skeleton line and structure variables for defining a shape of a runner outlet edge; The method of constructing a parameter model of a water turbine runner comprises: defining a three-dimensional curve path of a blade skeleton line, the three-dimensional curve path comprising multiple control points; adjusting coordinates of the control points to generate the profile variables; defining an asymmetric wedge angle of a runner outlet edge and a thickness distribution coefficient along a blade height; generating the structure variables based on the asymmetric wedge angle and the thickness distribution coefficient.

3. The method of claim 1, wherein the method is characterized by: The simulation model comprises a fluid dynamics model, a damage prediction model, and a life prediction model; The method of establishing a simulation model based on the dynamic environment data comprises: constructing a fluid dynamics model based on the hydro-meteorological data and the cavitation core data; constructing a damage prediction model based on the water quality characteristic data and the fluid dynamics model; constructing a life prediction model based on the fluid dynamics model.

4. The method of claim 1, wherein the method is characterized by: The simulation model comprises a fluid dynamics model, a damage prediction model, and a life prediction model; The working conditions comprise a first working condition representing water quality in a dry season; In the first working condition, the method of performing flow characteristic calculation on the initial runner by using the simulation model to generate performance prediction results comprises: calculating, by using the fluid dynamics model, a pressure distribution of a blade back surface area of the initial runner to obtain a minimum local pressure value of the back surface area, the blade back surface area comprising a runner outlet edge back surface area; generating a pressure field and a pressure load based on the pressure distribution; calculating, by using the damage prediction model, a cavitation void fraction of the runner outlet edge back surface area based on the pressure field; generating cavitation erosion damage based on the damage prediction model; predicting, by using the life prediction model, a crack risk of the runner outlet edge back surface area based on the pressure load and the cavitation erosion damage; generating a first performance prediction result based on the minimum local pressure value, the cavitation void fraction, and the crack risk, the first performance prediction result being used to represent anti-cavitation performance of the initial runner.

5. The method of claim 1, wherein the method is characterized by: The simulation model comprises a fluid dynamics model, a damage prediction model, and a life prediction model; The working conditions comprise a second working condition representing water quality in a normal season; In the second working condition, the method of performing flow characteristic calculation on the initial runner by using the simulation model to generate performance prediction results comprises: The fluid dynamics model is used to calculate energy conversion characteristics of the initial runner at different guide vane opening degrees, so as to obtain an optimal efficiency point and operating parameters; Based on the energy conversion characteristics, a velocity field and a vorticity field of the whole flow passage are generated; Based on the velocity field and the vorticity field, the damage prediction model is used to calculate flow stability of the initial runner at off-design conditions; Based on the damage prediction model, an efficiency attenuation characteristic is generated; Based on the optimal efficiency point, the operating parameters and the efficiency attenuation characteristic, a second performance prediction result is generated, which is used to represent operating efficiency and stability of the initial runner.

6. The method of claim 1, wherein the method is characterized by: The simulation model comprises a fluid dynamics model, a damage prediction model and a life prediction model; The working conditions comprise a third working condition representing water quality in the wet season; In the third working condition, the simulation model is used to perform flow characteristic calculation on the initial runner, so as to generate a performance prediction result, comprising: The fluid dynamics model is used to calculate particle impact characteristics of a blade working surface area of the initial runner, so as to obtain a particle impact velocity and an attack angle; Based on the particle impact characteristics, a particle trajectory distribution and a local flow velocity field are generated; Based on the particle trajectory distribution and the flow velocity field, the damage prediction model is used to calculate a material wear rate of the blade working surface and a water inlet edge area; Based on the damage prediction model, a wear damage distribution is generated; Based on the particle impact velocity, the attack angle and the wear damage distribution, the life prediction model is used to calculate material loss and strength degradation of the blade working surface area under the sand-laden water flow; Based on the material wear rate, the material loss and the strength degradation, a third performance prediction result is generated, which is used to represent anti-wear performance and structural strength of the initial runner.

7. The method of claim 1, wherein the method is characterized by: The mapping relationship is constructed based on the design variables and the performance prediction result, comprising: A plurality of sample points are determined in a value space of the design variables by an experimental design method; For each sample point, flow characteristic calculation is performed on the initial runner under a plurality of working conditions by the simulation model, so as to generate a performance prediction result corresponding to the sample point; Based on all sample points and their corresponding performance prediction results, a response surface model is used to construct a nonlinear mapping relationship from the design variables to the performance prediction result.

8. The method of claim 1, wherein the method is characterized by: The target combination of the design variables is determined based on the mapping relationship, comprising: A conflict relationship between anti-cavitation performance represented by the first performance prediction result and anti-wear performance represented by the third performance prediction result is identified; Based on the conflict relationship, a multi-objective optimization function is constructed, and optimization objectives of the function include maximizing the anti-cavitation performance and maximizing the anti-wear performance; The multi-objective optimization function is solved in the space of the design variables, so as to obtain a target solution set, and each solution in the target solution set represents a design scheme achieving different balance levels between the anti-cavitation performance and the anti-wear performance. From the target solution set, a target solution is determined as the target combination of the design variables based on a preset rule, the preset rule being a decision rule balancing the conflict relations.

9. The method of claim 8, wherein the method is characterized by: After the step of determining the target solution as the target combination of the design variables based on the preset rule, the method further comprises: inputting the target combination of the design variables into the parameter model to generate three-dimensional geometric data of a target runner, the three-dimensional geometric data including spatial points, curve and surface data for defining the shape of the runner surface, and topological structure data for defining the connection relationship between the upper crown, the lower crown and the blades; the three-dimensional geometric data of the target runner is the runner design scheme.

10. A Francis turbine runner with high anti-cavitation efficiency under the quality of water in a karst cave, characterized in that, comprises: an upper crown, a lower crown and a plurality of blades fixed on the upper crown and the lower crown, the blades being obtained based on the optimization method of the high-efficiency anti-cavitation mixed-flow hydraulic turbine runner under the karst cave water quality according to any one of claims 1-9.

11. The high-efficiency anti-cavitation mixed-flow water turbine runner under the quality of the cave water according to claim 10, characterized in that, The water outlet edge of the blade is an asymmetric structure, and the asymmetric structure is defined by a structure variable in the design variables, the structure variable including an asymmetric wedge angle. The blade's backbone curve is a three-dimensional curve defined by a profile variable in the design variables, the profile variable including coordinates of a plurality of control points.

12. The high-efficiency anti-cavitation mixed-flow water turbine runner under the quality of the cave water according to claim 10, characterized in that, Along the height direction of the blade, the thickness of the blade is gradually distributed, and the thickness distribution is defined by a structure variable in the design variables, the structure variable including a thickness distribution coefficient.

Images