Cavitation risk assessment and correction method and system for repaired contour of water pump impeller

Abstract

The invention discloses a cavitation risk assessment and correction method and system for a repaired contour of a water pump impeller, and belongs to computer aided design and manufacturing, and the method comprises the steps: generating a difference matrix through point cloud and standard data registration; calling operation data to calculate grid fluid shear and impact values; establishing a load and thickness mapping table according to the three-layer material parameters; constructing a simultaneous inequality group covering the fluid load; screening the thickness ratio solution with the lowest material consumption to generate a repairing instruction; and constructing and outputting a hierarchical repair visualization model. According to the method, the flow field can be actively optimized and cavitation can be inhibited in the repairing stage, and an optimized repairing scheme considering economical efficiency, reliability and hydraulic performance is generated.

Description

Technical Field

[0001] This invention relates to computer-aided design and manufacturing, and in particular to a method and system for assessing and correcting cavitation risk in the repair profile of a water pump impeller. Background Technology

[0002] As a core flow-through component of hydraulic machinery, the pump impeller is highly susceptible to cavitation damage during long-term operation due to the combined effects of high-speed fluid impact and abrasion from sediment particles. This damage alters the impeller profile, consequently affecting the pump's hydraulic performance and structural strength. Therefore, accurate cavitation risk assessment and effective profile repair of damaged impellers are crucial for ensuring the safe and stable operation of pump equipment.

[0003] In related technologies, Chinese invention patent application CN119538429A discloses a method for calculating the shape of the cavitation zone and material loss of a water pump blade under sediment-laden conditions. This method determines the cavitation zone by identifying characteristic parameters and conducting transient numerical simulations on the research object, and obtains relevant flow field parameters; it introduces velocity correction factors and sediment-laden correction factors to reflect the influence of water flow velocity and sediment particles on the material loss due to cavitation; it calculates the cavitation offset distance and determines the shape of the cavitation zone based on the cavitation offset distance and the geometric contour of the cavitation zone; and it calculates the material loss in the cavitation zone using the velocity correction factor and sediment-laden correction factor.

[0004] While the relevant technologies can quantitatively calculate the shape distribution and material loss of cavitation damage areas under specific working conditions, they mainly focus on post-damage assessment or risk prediction. They fail to generate targeted repair implementation parameters based on the specific distribution of local flow field loads, and cannot establish a quantitative matching relationship between fluid stress and the resistance of repair materials, making it difficult to directly guide subsequent physical remanufacturing and repair projects. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method and system for assessing and correcting cavitation risk in the repair profile of a water pump impeller. This method can proactively optimize the flow field and suppress cavitation during the repair stage, generating an optimized repair scheme that balances economy, reliability, and hydraulic performance.

[0006] The above objectives can be achieved through the following approach:

[0007] A method for cavitation risk assessment and correction of a water pump impeller repair profile includes: acquiring three-dimensional scanned point cloud data and standard geometric data of the impeller to be repaired; calculating the normal distance difference through coordinate registration to generate a difference data matrix; discretizing the difference data matrix into grid nodes; retrieving historical operating data of the water pump as boundary conditions; solving the pressure field and velocity field of the grid nodes to obtain fluid shear stress and fluid impact pressure values; accessing a preset material performance database to obtain material parameters, including base layer material, transition layer material, and surface layer material; establishing a load thickness mapping table based on the material parameters; calculating the maximum allowable impact force and maximum allowable shear force under different layer thickness combinations; establishing a system of simultaneous inequalities for the grid nodes using the material parameters; solving the system of simultaneous inequalities to obtain solutions with different thickness ratios; selecting the thickness ratio solution with the lowest material consumption as the execution parameter to generate a three-dimensional repair instruction; and, according to the three-dimensional repair instruction, identifying the base layer material, transition layer material, and surface layer material on the grid nodes using different layers to construct a layered repair visualization model and output it for display.

[0008] Optionally, generating the difference data matrix includes: performing a coarse alignment of the 3D scanned point cloud data and the standard geometric data based on principal component analysis, and performing a fine transformation based on the iterative nearest point to obtain aligned point cloud data and aligned geometric data; traversing the sampling points based on the aligned point cloud data, calculating the normal projection coordinates along the aligned geometric data, and solving the Euclidean distance between the sampling points and the normal projection coordinates to obtain the normal distance difference; mapping the normal distance difference to a two-dimensional plane coordinate index to construct the difference data matrix.

[0009] Optionally, obtaining the fluid shear stress and fluid impact pressure values ​​includes: mapping the difference data matrix to the aligned geometric data, constructing a three-dimensional spatial mesh, and defining the vertices of the three-dimensional spatial mesh as mesh nodes; parsing the flow-head curve data through the pump's historical operating data, extracting the flow rate value and converting it into a velocity value to be assigned to the inlet boundary node of the three-dimensional spatial mesh, and extracting the static pressure value to be assigned to the outlet boundary node of the three-dimensional spatial mesh; performing fluid dynamics iterative calculations on the mesh nodes, solving for the local velocity vector and static pressure scalar, calculating the fluid shear stress value based on the near-wall gradient of the local velocity vector, and synthesizing the fluid impact pressure value based on the static pressure scalar and the normal component of the local velocity vector.

[0010] Optionally, calculating the maximum allowable impact force and maximum allowable shear force under different layer thickness combinations includes: extracting the unit thickness impact resistance value and unit thickness shear resistance value from the base layer material, the transition layer material, and the surface layer material respectively; setting the thickness deviation length for the base layer material, the transition layer material, and the surface layer material to generate a thickness combination sequence, and constructing a load thickness mapping table based on the thickness combination sequence as row index and column index; for the thickness combination sequence, multiplying the thickness of each layer material by the unit thickness impact resistance value and performing a summation calculation to obtain the maximum allowable impact force, and multiplying the thickness of each layer material by the unit thickness shear resistance value to obtain the maximum allowable shear force, and filling the maximum allowable impact force and the maximum allowable shear force into the load thickness mapping table.

[0011] Optionally, establishing a system of simultaneous inequalities for the mesh nodes using the material parameters includes: setting a thickness variable to be solved based on the base layer material, the transition layer material, and the surface layer material; constructing a total strength constraint condition, limiting that the sum of the products of the thickness variable to be solved and the unit thickness impact resistance value is greater than or equal to the fluid impact pressure value, and the sum of the products of the thickness variable and the unit thickness shear resistance value is greater than or equal to the fluid shear stress value; constructing a surface brittleness constraint condition, limiting that the thickness variable to be solved is less than or equal to the inverse proportional function value of the fluid impact pressure value, and merging the total strength constraint condition and the surface brittleness constraint condition into a system of simultaneous inequalities.

[0012] Optionally, the method further includes: before establishing the set of simultaneous inequalities, using the load thickness mapping table to pre-screen the mesh nodes, and eliminating thickness combination sequences where the maximum allowable impact force is lower than the fluid impact pressure value and the maximum allowable shear force is lower than the fluid shear stress value.

[0013] Optionally, setting the thickness variable to be solved based on the base layer material, the transition layer material, and the surface layer material includes: establishing non-negative continuous numerical variables for the base layer material, the transition layer material, and the surface layer material respectively; and constructing the non-negative continuous numerical variables into a one-dimensional vector as the thickness variable to be solved.

[0014] Optionally, the generation of the 3D repair instruction includes: traversing the feasible region for the system of simultaneous inequalities, extracting all thickness combination values ​​to form a feasible solution set; based on the feasible solution set, calculating the sum of the thicknesses of the base layer material, the transition layer material, and the surface layer material as a total consumption index, and sorting the total consumption index in ascending order; selecting the first value of the total consumption index as an execution parameter, and binding the execution parameter to the mesh node to generate the 3D repair instruction.

[0015] Optionally, the construction and output display of the layered repair visualization model includes: dividing the mesh nodes into a bottom layer of the base material, an intermediate layer of the transition material, and a top layer of the surface material according to the three-dimensional repair instructions; performing geometric stretching operations on the bottom layer, the intermediate layer, and the top layer according to the execution parameters to construct a three-dimensional sub-model; and merging and assembling the three-dimensional sub-models according to the spatial stacking order to construct and output the layered repair visualization model.

[0016] Based on the same inventive concept, this invention also provides a system for assessing and correcting cavitation risk in the repair profile of a water pump impeller. The system includes: a difference matrix generation module, used to acquire three-dimensional scanned point cloud data and standard geometric data of the impeller to be repaired, and to generate a difference data matrix by calculating the normal distance difference through coordinate registration; a load calculation module, used to discretize the difference data matrix into grid nodes, retrieve historical operating data of the water pump as boundary conditions, and solve the pressure field and velocity field on the grid nodes to obtain fluid shear stress values ​​and fluid impact pressure values; and a mapping table construction module, used to access a preset material property database to obtain material parameters, including the base layer material, overload... The system includes a base layer material and a surface layer material. Based on the material parameters, a load-thickness mapping table is established, and the maximum allowable impact force and maximum allowable shear force are calculated under different layer thickness combinations. A constraint establishment module is used to establish a system of simultaneous inequalities for the mesh nodes using the material parameters. A thickness ratio solution module is used to solve the system of simultaneous inequalities, obtain solutions with different thickness ratios, select the solution with the lowest material consumption as the execution parameter, and generate a three-dimensional repair instruction. A layered visualization output module is used to identify the base layer material, the transition layer material, and the surface layer material on the mesh nodes using different layers according to the three-dimensional repair instruction, construct a layered repair visualization model, and output and display it.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] 1. By constructing a dynamic coupling closed-loop optimization system between geometry, flow field, and material properties, impeller repair is transformed from traditional passive geometric restoration to proactive performance-driven redesign. The repair scheme no longer aims solely at restoring the shape, but rather actively reshapes the local profile to improve the flow field and suppress cavitation at its source. This paradigm shift from "anti-cavitation" to "cavitation suppression" can extend the service life of the repaired impeller and improve the long-term operational reliability of the equipment.

[0019] 2. Through an iterative mechanism of high-fidelity simulation verification and adaptive correction using a surrogate model, the self-consistency and high reliability of the repair scheme are ensured. Each round of initial optimization must undergo rigorous flow field simulation verification, and the verification results are fed back to the surrogate model, which serves as the core of continuous optimization decisions. This closed-loop "design-verification-feedback" process ensures that the final output repair scheme is the product of multiple rounds of virtual verification and self-correction, and its predictive performance is highly consistent with physical reality. This reduces the risk of rework due to substandard performance after actual repair work and guarantees the certainty of the repair effect.

[0020] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating a method for assessing and correcting cavitation risk in the repair profile of a water pump impeller according to an embodiment of the present invention.

[0023] Figure 2 This is a scatter plot of the fluid shear stress and impact pressure distribution of the impeller surface grid nodes in an embodiment of the present invention.

[0024] Figure 3 This is a parallel coordinate diagram of the adaptive optimization of mesh node repair parameters in an embodiment of the present invention.

[0025] Figure 4 This is a schematic diagram of a cavitation risk assessment and correction system for the repair profile of a water pump impeller according to an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Reference Figure 1 One embodiment of the present invention proposes a method for assessing and correcting cavitation risk in the repair profile of a water pump impeller, which can actively optimize the flow field and suppress cavitation during the repair stage, generating an optimized repair scheme that takes into account economy, reliability and hydraulic performance.

[0028] The method described in this embodiment specifically includes:

[0029] Acquire the 3D scanning point cloud data and standard geometric data of the impeller to be repaired, calculate the difference in normal distance through coordinate registration, and generate a difference data matrix;

[0030] The difference data matrix is ​​discretized into grid nodes. Historical operating data of the water pump is retrieved as boundary conditions. The pressure field and velocity field are solved on the grid nodes to obtain the fluid shear stress value and the fluid impact pressure value.

[0031] Access a preset material property database to obtain material parameters, which include base layer material, transition layer material and surface layer material. Establish a load thickness mapping table based on the material parameters, and calculate the maximum allowable impact force and maximum allowable shear force under different layer thickness combinations.

[0032] A system of simultaneous inequalities is established for the grid nodes using the material parameters;

[0033] Solve the system of simultaneous inequalities to obtain solutions with different thickness ratios, select the solution with the lowest material consumption as the execution parameter, and generate a three-dimensional repair instruction.

[0034] According to the three-dimensional repair instructions, the base layer material, the transition layer material and the surface layer material are identified by different layers on the grid nodes, a layered repair visualization model is constructed and output for display.

[0035] Optionally, the generation of the difference data matrix includes:

[0036] The three-dimensional scanned point cloud data and the standard geometric data are roughly aligned based on principal component analysis, and a fine transformation is performed based on the iterative nearest point to obtain aligned point cloud data and aligned geometric data.

[0037] First, a coarse alignment based on Principal Component Analysis (PCA) is performed. The PCA algorithm extracts the three principal directions (eigenvectors) of the point cloud data by calculating the covariance matrix, and aligns the centroid and principal axis directions of the scanned data with those of the standard geometric data, thus eliminating large rotational and translational deviations from the initial position. Based on this coarse alignment, a finer transformation is then performed using the Iterative Closest Point (ICP) algorithm. The ICP algorithm iteratively finds the nearest point on the standard geometric surface for each point in the scanned point cloud and calculates the rigid transformation matrix (including the rotation matrix) that minimizes the sum of squared Euclidean distances between the two point sets. Translation vector To quantify registration accuracy, the following objective function is adopted. Perform iterative optimization:

[0038] ,

[0039] in, The objective function value representing the registration error is usually expressed in square millimeters. This indicates the total number of sampling points involved in the calculation; This indicates the first point in the point cloud data to be registered. The coordinate vector of each point ; Represents the relationship between standard geometric data and The coordinate vector of the nearest point; express rotation matrix; express Translation vector; This represents the Euclidean norm. The physical meaning of the formula lies in measuring the overall fit between the transformed point cloud and the standard model. It is continuously updated during the iteration process. and To minimize Until the change in error between two iterations Less than the convergence threshold, such as Or, it may reach the maximum number of iterations.

[0040] For example, suppose we obtain a point cloud of a rotor scan containing 5 million points. First, through PCA calculation, we find that the maximum principal axis direction of the scan data is a vector. The principal axis of standard geometry is An initial rotation matrix is ​​automatically generated to align the two elements. Then, the ICP algorithm is started, with a maximum of 100 iterations and a convergence threshold of [value missing]. The objective function value was calculated during the 15th iteration. From the initial Descending to The convergence condition is met. At this point, the output aligned point cloud data is the final convergence result. and The transformed coordinate set has a spatial overlap with standard geometry down to the micrometer level.

[0041] Based on the aligned point cloud data, the sampling points are traversed, the normal projection coordinates are calculated along the aligned geometric data, and the Euclidean distance between the sampling points and the normal projection coordinates is calculated to obtain the normal distance difference.

[0042] For each sampling point in the aligned point cloud data, the KD-Tree spatial indexing method is used to search for the nearest neighbor of that sampling point in the aligned geometry data, and the surface normal vector of that neighborhood is calculated. Then, the projected distance of that sampling point to the standard surface along the normal vector direction is calculated. This distance represents the normal deviation of the actual surface from the ideal surface, directly reflecting the depth of the cavitation pit or the wear thickness. Normal distance difference. The calculation formula is as follows:

[0043] ,

[0044] in, This represents the difference in normal distance, in millimeters. Negative values ​​usually indicate missing material, while positive values ​​indicate attachments or deformation. This indicates the spatial coordinates of the current sampling point in the aligned point cloud data; express Perpendicular coordinates on a standard geometric surface; Represents the standard geometric surface at point The unit normal vector at that location; "" indicates vector dot product operation. The formula eliminates the influence of tangential deviation through dot product operation, retaining only the deformation component perpendicular to the blade surface, thereby ensuring the accuracy of cavitation depth measurement.

[0045] For example, select a point on the suction surface of the blade. Its coordinates after alignment are Find the corresponding projection point on the standard geometric surface. Coordinates are The unit normal vector at that location for Substitute into the formula to calculate: vector ;distance This value This means that the location exists. The deviation. If inward concavity is defined as negative, it can be adapted by adjusting the direction of the normal vector or the formula sign.

[0046] The difference in normal distance is mapped to a two-dimensional plane coordinate index to construct a difference data matrix.

[0047] To facilitate structured numerical calculations, the unstructured three-dimensional discrete point deviations need to be converted into regular two-dimensional matrices. A flow surface parameterization mapping technique is employed to establish a matrix from the three-dimensional surface. To the two-dimensional parameter domain The mapping relationship. Usually set. The shaft corresponds to the streamline direction of the blade. The axis corresponds to the spanwise direction of the blade. The standard geometric surface is meshed, and the difference in normal distance for each mesh node is filled into the corresponding two-dimensional matrix element. For matrix nodes not directly covered by point cloud sampling points, bilinear interpolation is used for numerical filling.

[0048] For example, the resolution of the matrix is ​​set to... The blade inlet edge corresponds to column 1 of the matrix, and the outlet edge corresponds to column 1024; the blade root corresponds to row 1, and the blade tip corresponds to row 1024. If there exists a region in the middle of the blade, slightly off-center from the tip, with a diameter of approximately... For cavitation pits, the values ​​within the corresponding sub-matrix block in the difference data matrix will exhibit a series of consecutive negative values, for example... And so on, while the values ​​in other undamaged regions of the matrix are close to... .

[0049] Optionally, obtaining the fluid shear stress values ​​and fluid impact pressure values ​​includes:

[0050] The difference data matrix is ​​mapped to the aligned geometry data to construct a three-dimensional spatial mesh, and the vertices of the three-dimensional spatial mesh are defined as mesh nodes;

[0051] Since the standard design model cannot accurately represent actual cavitation pits and wear thinning, the difference data matrix is ​​first read. This matrix records the difference in normal distance between the impeller surface and the aligned geometry. All surface vertices of the aligned geometry are traversed, and the position of each vertex is moved along the normal direction based on its corresponding difference value in the difference data matrix, thus generating a reconstructed geometry. Subsequently, the fluid domain containing this reconstructed geometry is discretized, dividing the continuous fluid space into a finite number of tiny control volumes, such as tetrahedral or hexahedral elements. To capture near-wall flow details, such as boundary layer separation, mesh refinement is performed on the blade surface and the cavitation pit region, generating a three-dimensional spatial mesh. All corner points of this three-dimensional spatial mesh are defined as mesh nodes, which serve as the spatial carriers for storing physical quantities such as pressure and velocity in subsequent numerical calculations.

[0052] For example, suppose the difference data matrix shows that a certain region on the suction surface of the blade has a depth of... The cavitation pits. During the mapping process, the corresponding aligned geometric data surface vertices in this region will be recessed inwards. In the subsequently constructed 3D spatial mesh, the global mesh size is set to... In the cavitation pit area, the grid size was refined to... Furthermore, the boundary layer mesh generated in the wall normal direction has 15 layers, and the height of the first layer mesh is [missing information]. To ensure dimensionless wall distance The value meets the calculation requirements of the turbulence model.

[0053] The flow rate and head curve data are analyzed by the historical operation data of the water pump, the flow rate value is extracted and converted into the flow velocity value and assigned to the inlet boundary node of the three-dimensional spatial grid, and the static pressure value is extracted and assigned to the outlet boundary node of the three-dimensional spatial grid.

[0054] To ensure the realism of the simulation results, the boundary conditions must be set based on the actual operating conditions of the pump. Pre-stored historical operating data of the pump, containing time-series information on flow rate, inlet and outlet pressures, and rotational speed, is accessed. The pump's unique flow-head curve, the QH curve, is then extracted, reflecting the pump's hydraulic performance characteristics under different operating conditions. The flow rate value for the target operating condition is extracted and converted into a velocity value according to the fluid dynamics continuity equation, and assigned to the inlet boundary nodes of the 3D spatial mesh. Simultaneously, the corresponding static pressure value is extracted from the curve data and used as the pressure outlet boundary condition, assigned to the outlet boundary nodes.

[0055] For example, suppose historical operating data shows that the operating point of the water pump during cavitation is a flow rate of Inlet pipe diameter First, calculate the inlet cross-sectional area. Next, the average inlet velocity is calculated. .Should The numerical value is assigned to all grid nodes at the inlet. Simultaneously, the outlet static pressure value at this time is extracted. And assign it to the grid node at the exit.

[0056] Perform fluid dynamics iterative calculations on the mesh nodes to solve for the local velocity vector and static pressure scalar. Calculate the fluid shear stress value based on the near-wall gradient of the local velocity vector, and synthesize the fluid impact pressure value based on the static pressure scalar and the normal component of the local velocity vector.

[0057] After completing mesh construction and boundary assignment, the fluid dynamics solver is invoked to perform iterative calculations. The iteration process continues until the residual curves of all physical quantities decrease to the convergence criterion, such as... Below. At this point, each mesh node has a defined local velocity vector and a hydrostatic scalar. Based on the solution results, two core damage indices are further calculated: First, the fluid shear stress value is calculated. This value characterizes the tangential friction force caused by fluid viscosity on the blade surface, which is the main factor leading to material wear. Its calculation formula is as follows:

[0058] ,

[0059] in, This represents the numerical value of fluid shear stress, in Pascals. The dynamic viscosity of a fluid is expressed in Pascal-seconds (Pa·s), for example, in... The dynamic viscosity of water is approximately ; This represents the tangential velocity component parallel to the wall. Indicates the distance along the direction of the wall normal; The near-wall gradient represents the local velocity vector, i.e., the rate of change of velocity in the direction perpendicular to the wall. This value is calculated based on the ratio of the velocity difference to the distance difference between the first layer of mesh nodes adjacent to the wall. Second, the fluid impact pressure is calculated. This value combines the hydrostatic and hydrodynamic impact effects of the fluid and is directly related to the destructive force during cavitation collapse. Its calculation formula is as follows:

[0060] ,

[0061] in, This represents the numerical value of fluid impact pressure, in Pascals. Represents the static pressure scalar at the grid node; This represents the fluid density, expressed in kilograms per cubic meter. For water at room temperature, this value is typically taken as... ; This represents the normal component of the local velocity vector, i.e., the magnitude of the velocity of a fluid particle impacting the blade surface perpendicularly. This value is calculated by projecting the solved local velocity vector onto the surface normal vector where the mesh node is located.

[0062] Optionally, the calculation of the maximum permissible impact force and the maximum permissible shear force under different layer thickness combinations includes:

[0063] The impact resistance and shear resistance per unit thickness are extracted from the base layer material, the transition layer material, and the surface layer material, respectively.

[0064] The impact resistance and shear resistance per unit thickness are not simple material constants, but rather standardized characteristic parameters based on a material property database containing historical test data. In a laboratory setting, standard samples of the base layer, transition layer, and surface layer materials were prepared, and the failure loads at different thicknesses were measured using a drop hammer impact tester and a shear strength tester. Subsequently, linear regression was used to normalize the failure loads to the force values ​​that can be withstood per unit thickness. The physical significance of these two values ​​lies in their decoupling of material properties from geometric thickness, allowing for the prediction of composite material properties at arbitrary thicknesses through simple mathematical calculations. The extraction operation is performed through a database query interface, indexing the corresponding performance fields based on the selected material ID.

[0065] For example, suppose the repair scheme is selected as follows: the base layer material is "316L stainless steel powder", the transition layer material is "Ni60 nickel-based alloy", and the surface material is "WC-12Co tungsten carbide". Access the database and extract the impact resistance value per unit thickness of 316L stainless steel. That is, it can withstand an impact force of 1200 Newtons per millimeter of thickness, and the shear resistance value per unit thickness is... The impact resistance per unit thickness of the WC-12Co surface material was extracted as follows: The shear strength per unit thickness is Hard alloys have relatively weak shear resistance. This basic data is loaded into memory for subsequent calculations.

[0066] For the base layer material, the transition layer material, and the surface layer material, a thickness deviation length is set to generate a thickness combination sequence, and a load thickness mapping table is constructed based on the thickness combination sequence as row index and column index;

[0067] To transform the continuous thickness optimization problem into a computer-processable discrete search problem, the thickness of the repair layer needs to be discretized. First, the minimum process resolution of the laser cladding or additive manufacturing equipment is read and set as the thickness discrete step size, for example... Subsequently, according to the impeller repair process specifications, the allowable thickness range for each layer of material was set, such as the base layer. Using this step size as the increment, generate all thickness combination sequences that meet the range constraints. Each thickness combination sequence is a one-dimensional vector containing three elements. , representing the assumed thicknesses of the three material layers. Finally, a structured storage space is allocated in memory to construct a load-thickness mapping table. This table is essentially a multidimensional hash table, where the key is a unique code for the thickness combination sequence, and the value is the mechanical property data to be calculated.

[0068] For the thickness combination sequence, the thickness of each layer of material is multiplied by the unit thickness impact resistance value and summed to obtain the maximum allowable impact force. The thickness of each layer of material is multiplied by the unit thickness shear resistance value to obtain the maximum allowable shear force. The maximum allowable impact force and the maximum allowable shear force are then filled into the load thickness mapping table.

[0069] The strength contribution of each material layer to the overall structure is calculated by linear superposition. This process iterates through each thickness combination sequence and performs batch calculations using unit parameters to determine the maximum permissible impact force. The calculation formula is as follows:

[0070] ,

[0071] in, This indicates the maximum permissible impact force under the current thickness combination. Physically, it represents the ultimate load that the composite repair layer can withstand in the vertical direction without plastic deformation or breakage. The unit is usually Newton or Megapascal. The index representing the material level takes different values. Represents the basal layer, Represents the transition layer Represents the surface layer; The number of thickness combination sequences represents the first... The thickness of the layer material, in millimeters; Representing the The impact resistance value per unit thickness of a layered material, physically representing the impact resistance contribution provided by the material per unit thickness. Maximum allowable shear force. The calculation formula is as follows:

[0072] ,

[0073] in, This represents the maximum permissible shear force under the current thickness combination, which is the ultimate ability of the composite repair layer to resist interlayer slip or shear peeling under tangential fluid scouring. Representing the The shear strength per unit thickness of the layer material. The calculated value... and The data is packaged and written into the corresponding record in the load-thickness mapping table. The completed mapping table will serve as an offline "strength dictionary," accessible simply by indexing the thickness. The corresponding strength limit can be found within the time complexity, eliminating the need to repeatedly calculate the physical formula, thus improving the efficiency of the entire field optimization.

[0074] For example, suppose the current traversal of a thickness combination sequence is: basal layer transition layer ,surface layer Known impact resistance per unit thickness: base layer transition layer ,surface layer Known shear strength per unit thickness: base layer transition layer ,surface layer Then perform the following calculation: maximum permissible impact force. Unit force value. Maximum permissible shear force. Unit force value. Finally, fill the numerical pair {impact force: 500, shear force: 310} into the cell with index [2.0, 1.0, 0.5] in the load thickness mapping table.

[0075] Optionally, establishing a system of simultaneous inequalities for the mesh nodes using the material parameters includes:

[0076] The thickness variable to be solved is set based on the base layer material, the transition layer material, and the surface layer material;

[0077] For each discrete grid node, the thickness of each of the three material layers to be deposited during the repair process needs to be determined. Therefore, a set of non-negative continuous variable vectors is defined. As the thickness variable to be solved. Represents the thickness of the base layer material. This represents the thickness of the transition layer material. These three variables represent the thickness of the surface material. They are the target solutions we are looking for. To conform to physical reality, basic constraints are also set. ( This means that the material thickness cannot be negative.

[0078] Construct a total strength constraint condition, which stipulates that the sum of the product of the thickness variable to be solved and the impact resistance value per unit thickness is greater than or equal to the fluid impact pressure value, and the sum of the product of the thickness variable and the shear resistance value per unit thickness is greater than or equal to the fluid shear stress value.

[0079] The overall strength constraint consists of two independent inequalities, corresponding to the compressive strength requirement in the vertical direction and the shear strength requirement in the tangential direction, respectively. The first inequality is the impact strength constraint, and its mathematical expression is:

[0080] ,

[0081] in, Indicates the first The unsolved thickness variable of the layer material; Indicates the first Impact resistance per unit thickness of the layer material; This represents the fluid impact pressure at that grid node. The total compressive strength provided by the three layers of material must be sufficient to cover the maximum impact load applied by the fluid at that point. The second inequality is the shear strength constraint, and its mathematical expression is:

[0082] ,

[0083] in, Indicates the first Shear strength per unit thickness of the layer material; This represents the fluid shear stress value at this grid node. The overall resistance of the repair layer to tangential slip must be greater than the shear force generated by fluid erosion to prevent coating peeling.

[0084] For example, suppose a certain grid node The fluid load at the point is: impact pressure Shear stress The selected material parameters are: base layer Impact resistance Shear resistance Transition layer Impact resistance Shear resistance ;surface layer Impact resistance Shear resistance The system of total strength constraint inequalities established for this node is as follows:

[0085] ,

[0086] .

[0087] Construct surface brittleness constraints, limiting the thickness variable to be solved to be less than or equal to the inverse proportional function value of the fluid impact pressure, and combine the total strength constraints and the surface brittleness constraints into a system of simultaneous inequalities.

[0088] Surface materials, such as tungsten carbide and ceramic composites, have extremely high hardness but poor toughness. If the coating is too thick, the residual stress accumulated inside cannot be released under strong impact, easily leading to cracking. Therefore, it is necessary to construct brittle constraint conditions for the surface layer and limit its thickness. The upper limit is determined by an inverse proportional function, meaning that the greater the fluid impact pressure, the smaller the allowable surface thickness. Its mathematical expression is:

[0089] ,

[0090] in, This specifically refers to the unsolved thickness variable of the surface material; the base layer and transition layer are not subject to this restriction. These represent the material's toughness characteristic constants and are stored in a database. This represents the fluid impact pressure value of that grid node; For a very small positive number, such as This is used to prevent the denominator from being zero. Finally, all inequalities are packaged and merged into a system of simultaneous inequalities for that grid node. This system of inequalities encloses a multidimensional geometric space, which is the feasible solution domain for that node. For example... Figure 2 As shown, the scatter points represent the discretized grid node loads, the dashed steps represent the maximum load envelope that different thickness combinations can withstand in the load thickness mapping table, and the nodes located in the lower left of the dashed lines indicate that the thickness combination can meet its repair strength requirements.

[0091] For example, the node The toughness characteristic constant of the surface material was obtained from the database. The surface brittle constraint is then: This means that no matter how thick the surface material needs to be to meet strength requirements, At most, you can lay Insufficient strength must be addressed by increasing the base layer. or transition layer The thickness is used to supplement it. The final system of simultaneous inequalities is: We will look for the lowest-cost combination within this range.

[0092] Optionally, the method further includes:

[0093] Before establishing the set of simultaneous inequalities, the mesh nodes are pre-screened using the load thickness mapping table to remove thickness combination sequences where the maximum allowable impact force is lower than the fluid impact pressure value and the maximum allowable shear force is lower than the fluid shear stress value.

[0094] By quickly eliminating thickness combinations that clearly do not meet basic strength requirements, the feasible region for solving simultaneous inequalities is reduced, thereby improving the algorithm's computational efficiency and reducing overall computation time. It avoids redundant calculations of numerous invalid solutions during complex constraint solving processes. Before establishing a system of simultaneous inequalities for a specific mesh node, it first triggers a pre-screening procedure. It retrieves two key input data: one is a load-thickness mapping table containing all discrete thickness combinations and their corresponding mechanical properties; the other is the precisely calculated fluid impact pressure and fluid shear stress values ​​for that mesh node. Then, it iterates through each record in the load-thickness mapping table, executing a double-condition judgment logic with the following conditions:

[0095] ,

[0096] in, This indicates the maximum permissible impact force currently recorded in the load thickness mapping table; This represents the fluid impact pressure value of the current grid node; This indicates the maximum permissible shear force currently recorded in the load-thickness mapping table; This represents the fluid shear stress value of the current mesh node; This represents a logical "OR" operation. If the maximum allowable impact force of the thickness combination sequence is lower than the fluid impact pressure value, or its maximum allowable shear force is lower than the fluid shear stress value, it indicates that the repair scheme is fundamentally unable to withstand the fluid load at that point. Thickness combination sequences that satisfy either of the above conditions will be marked as "invalid schemes" and removed from the candidate scheme list of that grid node. Only those thickness combination sequences that simultaneously satisfy both strength conditions, i.e., their maximum allowable impact force and maximum allowable shear force are both greater than or equal to the corresponding fluid load values, will be retained to form a smaller subset of feasible solutions after preliminary screening, and will serve as valid input for the system of simultaneous inequalities.

[0097] For example, suppose a grid node is located at the inlet edge of the impeller blade to be repaired. Based on fluid dynamics analysis and calculation, the fluid impact pressure it withstands is... The fluid shear stress value is Traverse the load-thickness mapping table and read the first thickness combination sequence, for example: base layer. transition layer ,surface layer According to the table, the maximum permissible impact force for this combination is... The maximum permissible shear force is Execution comparison: Due to Less than The rejection criteria are met. Therefore, this thickness combination sequence cannot satisfy the node rejection criteria. The pressure resistance requirement is directly eliminated, and it is no longer substituted into the system of simultaneous inequalities for calculation.

[0098] For example, suppose a grid node is located in the middle of the impeller blade to be repaired. The flow velocity at this location is extremely high, and the calculated fluid impact pressure is... However, the fluid shear stress value is as high as Continuing to traverse the load-thickness mapping table, another thickness combination sequence is read. Looking up the table, the maximum permissible impact force corresponding to this combination is... However, due to the thin surface layer and soft substrate, its maximum allowable shear force is only... Execution comparison: Although ,but Less than The rejection criteria are met. Therefore, it is determined that this thickness combination sequence cannot resist the nodes. Strong shear flows at certain locations are also eliminated. Only when a combination is found that results in sufficient shear resistance... Only when this condition is met will the combination be retained for the next round of optimization.

[0099] Optionally, setting the thickness variable to be solved based on the base layer material, the transition layer material, and the surface layer material includes:

[0100] Non-negative continuous numerical variables are established for the base layer material, the transition layer material, and the surface layer material, respectively;

[0101] For each mesh node requiring a solution, a non-negative continuous numerical variable is established for the base layer material, transition layer material, and surface layer material. This definition clarifies the variable's properties: the "non-negative" property ensures that the calculated thickness value is physically feasible, meaning the material deposition thickness cannot be negative, which conforms to the physical laws of additive manufacturing; the "continuous" property means that the solver can find the optimal solution within a continuous real number range, rather than being forced to adhere to an discrete step size. This definition method eliminates the quantization error introduced by discretization, allowing the final calculated repair scheme to theoretically achieve the ultimate mathematical precision, thereby enabling precise control of material consumption.

[0102] For example, suppose the suction surface of the blade is numbered as For the grid nodes, define three variables: This represents the thickness of the 316L substrate. This represents the thickness of the Ni60 transition layer. This represents the thickness of the tungsten carbide surface layer. The range of values ​​for these three variables is set to be [missing information]. The continuous real number field. If the subsequent calculation result is... If the physical meaning is that a base layer does not need to be laid at that point; if the calculation result is This means that the process instruction will require the cladding head to be precisely deposited at that point. Thick transition layer material.

[0103] The non-negative continuous numerical variables are constructed as a one-dimensional vector to form the thickness variable to be solved.

[0104] These three independent, non-negative, continuous numerical variables are organized in logical order and constructed into a one-dimensional vector representing the thickness variable to be solved. This vector is the core data structure passed to the optimization solver, and its mathematical representation is as follows:

[0105] ,

[0106] in, This represents the vector of thickness variables to be solved; , , These are non-negative continuous numerical variables corresponding to the thicknesses of the base layer, transition layer, and surface layer materials, respectively. This represents the matrix transpose operation. This one-dimensional vector form is the standard input format for numerical optimization algorithms. It packages multiple unknowns into a single entity, allowing constraints to be represented as a product of a matrix and a vector. This facilitates efficient matrix operations by the computer, enabling rapid searching for the optimal solution that minimizes the objective function within a multidimensional feasible region.

[0107] For example, for a node The thickness variable to be solved is constructed as a vector. When performing impact strength verification, the impact strength numerical vector per unit thickness corresponding to that node is retrieved. At this point, the complex intensity calculation is transformed into a concise vector dot product operation within the solver: The optimizer will continuously adjust the vectors. The values ​​of the three components are calculated until a vector is found. This ensures that the dot product is greater than the fluid load and that the sum of the vector elements is minimized.

[0108] Optionally, the three-dimensional repair generation instructions include:

[0109] For the system of simultaneous inequalities, traverse the feasible region and extract all thickness combinations to form a feasible solution set;

[0110] For each grid node, a system of simultaneous inequalities is established, traversing the feasible region comprised of the overall strength constraint and the surface brittleness constraint. Within this region, all thickness combinations satisfying the conditions are extracted, forming a set of feasible solutions containing multiple feasible solutions. Each feasible solution is a one-dimensional vector containing the thickness values ​​of the base layer, transition layer, and surface layer.

[0111] For example, suppose for a grid node A system of simultaneous inequalities constrains the three thickness variables. The solver performs a discrete scan within the feasible region, assuming a scan accuracy of . Three efficient combinations satisfying all compressive, shear, and brittle constraints were found: Combination A: Combination B: Combination C: These three vectors constitute the set of feasible solutions for this node.

[0112] Based on the feasible solution set, the sum of the thicknesses of the base layer material, the transition layer material, and the surface layer material is calculated as the total consumption index, and the total consumption index is sorted in ascending order;

[0113] The total consumption index is used as the evaluation standard. This index is defined here as the sum of the thicknesses of the three layers of material. Physically, it represents the total volume or deposition height of material required for repair at a single point, directly related to repair cost and processing time. The calculation formula is as follows:

[0114] ,

[0115] in, This represents the total consumption, expressed in millimeters. , , These represent the thickness values ​​of the base layer material, transition layer material, and surface layer material in the feasible solution, respectively. After calculating the total consumption index for each feasible solution, all feasible solutions are arranged in ascending order according to their total consumption index to form an ordered list of solutions.

[0116] For example, the total consumption index is calculated for three combinations in the feasible solution set: the total consumption of combination A. Total consumption of combination B Total consumption of combination C After the calculation is complete, an ascending sort operation is performed. The sorted list is as follows: first position: combination C; second position: combination B; third position: combination A. It is evident that although all three schemes meet the strength requirements, combination C saves the most materials.

[0117] The first value of the total consumption index is selected as the execution parameter, and the execution parameter is bound to the mesh node to generate a three-dimensional repair instruction.

[0118] The first value in this sorted list, representing the feasible solution with the lowest total consumption index, is automatically selected and designated as the execution parameter for that mesh node. This execution parameter is then bound to the spatial coordinates and normal vector information of the currently processed mesh node, forming an instruction record containing location and repair layer thickness information. By repeating the above process for all mesh nodes to be repaired, a set containing optimized repair parameters for all nodes is finally generated. This set constitutes the complete 3D repair instruction, providing precise data input for layered visualization and automated repair. Figure 3As shown in the figure, each broken line represents the solution result of a grid node. The horizontal axis corresponds to the fluid load input and the optimization output, respectively. The black solid line represents the dangerous working condition node subjected to high fluid impact pressure, and the gray dashed line represents the normal working condition node subjected to low fluid impact pressure. When the fluid impact pressure is high, the corresponding black solid line shows a significant decreasing trend at the surface thickness. This indicates that the algorithm automatically activates the brittle surface constraint condition under high impact conditions, preventing cracking by thinning the hard surface layer. At the same time, the black solid line shows an increasing trend at the base layer thickness, indicating that the algorithm automatically increases the thickness of the tough base layer to compensate for the total strength requirement.

[0119] For example, the combination C that locks the first position of the sorted list. As the optimal solution. Obtain the current grid node. spatial coordinates and normal vector This data is encapsulated into machine-readable instructions: {Node:1024,Pos:[120.5,-45.0,88.2],Norm:[0.8,0.6,0],Layer_Thick:[2.5,0.5,0.3]}. These instructions are then transmitted to the control system of the additive manufacturing equipment, guiding the laser cladding head to move to the specified coordinates and precisely follow the substrate layer. Transition layer ,surface layer The process parameters are used to repair the layer by layer.

[0120] Optionally, the construction and output display of the hierarchical repair visualization model includes:

[0121] Based on the three-dimensional repair instructions, the mesh nodes are divided into the bottom layer of the base layer material, the middle layer of the transition layer material, and the top layer of the surface material;

[0122] The execution parameters bound to each mesh node are read, namely the optimal thickness values ​​of the base layer, transition layer, and surface layer materials. Based on this data, all mesh nodes to be repaired are logically divided into three independent geometric layers: a bottom layer representing the base layer material repair area, an intermediate layer representing the transition layer material, and a top layer representing the surface layer material. These three layers are topologically based on the same set of mesh node indices, but each is associated with different material property labels and corresponding thickness variables, thus preparing the data for layered modeling.

[0123] Based on the execution parameters, perform geometric stretching operations on the bottom layer, the middle layer, and the top layer to construct a three-dimensional sub-model;

[0124] For the bottom layer, based on the original worn surface of the impeller to be repaired, a geometric extrusion operation is performed along the normal direction of the surface of each mesh node according to the base layer material thickness value corresponding to that node in the execution parameters, generating a 3D sub-model that accurately describes the repair volume of the base layer. To ensure sufficient disclosure, the coordinate calculation formula for the new vertices in the geometric extrusion operation is as follows:

[0125] ,

[0126] in, Represents the newly generated vertex space coordinates after stretching. ; Represents the vertex space coordinates of the reference plane. For the bottom layer, the reference surface is the original worn surface; for the middle layer, the reference surface is the top surface of the bottom layer; and for the top layer, the reference surface is the top surface of the middle layer. This indicates the execution parameters of the current layer at the corresponding grid node; This represents the unit normal vector at the mesh node. The bottom sub-model is constructed using the original surface as the reference; the middle sub-model is constructed by extruded using the outer surface of the bottom sub-model as the new reference; finally, the top sub-model is constructed by extruded using the outer surface of the middle sub-model as the reference.

[0127] For example, suppose we are processing mesh nodes. Its original coordinates Unit normal vector The execution parameters are: base layer. transition layer ,surface layer Building the underlying structure: Based on the reference, stretch New Peak The underlying sub-model consists of and Composed of prisms. Constructing the intermediate layer: using... Based on the reference, stretch New Peak Building the top layer: with Based on the reference, stretch New Peak This creates a three-layered, tightly fitted solid mesh.

[0128] The three-dimensional sub-models are merged and assembled according to the spatial stacking order to construct a layered repair visualization model and output it for display.

[0129] Strictly following the physical spatial stacking order—bottom layer, middle layer, and top layer—three independent 3D sub-models are spatially positioned and subjected to Boolean union operations. This operation merges them into a single, seamless, unified repair geometry, eliminating internal overlapping surfaces and ensuring the model's topological correctness. This final model is the layered repair visualization model. For ease of observation, different rendering attributes are automatically assigned to geometric parts representing different materials, and the results are rendered and output in the graphical user interface. Users can perform interactive operations such as rotation, scaling, and sectioning on the model to intuitively review the thickness distribution and interlayer structure of the repair layers, ultimately confirming the final solution.

[0130] Based on the same inventive concept, this invention also provides a system for assessing and correcting cavitation risk in the repair profile of a water pump impeller, such as... Figure 4 As shown, the system includes:

[0131] The difference matrix generation module is used to acquire the three-dimensional scan point cloud data and standard geometric data of the impeller to be repaired, and to generate a difference data matrix by calculating the difference in normal distance through coordinate registration.

[0132] The load calculation module is used to discretize the difference data matrix into grid nodes, retrieve historical operating data of the water pump as boundary conditions, and solve the pressure field and velocity field for the grid nodes to obtain the fluid shear stress value and the fluid impact pressure value.

[0133] The mapping table construction module is used to access a preset material property database to obtain material parameters, including base layer material, transition layer material and surface layer material. Based on the material parameters, a load thickness mapping table is established, and the maximum allowable impact force and maximum allowable shear force are calculated under different layer thickness combinations.

[0134] A constraint establishment module is used to establish a set of simultaneous inequalities for the mesh nodes based on the material parameters.

[0135] The thickness ratio solution module is used to solve the system of simultaneous inequalities, obtain solutions with different thickness ratios, select the thickness ratio solution with the lowest material consumption as the execution parameter, and generate a three-dimensional repair instruction.

[0136] The layered visualization output module is used to identify the base layer material, the transition layer material and the surface layer material on the grid nodes according to the three-dimensional repair instructions, construct a layered repair visualization model and output it for display.

[0137] It should be noted that the functional division and information interaction between the various modules described above are logical, but in terms of physical implementation, they can be integrated on the same software platform or deployed in a distributed manner. The connections between them represent data flow and control flow, aiming to collaboratively achieve the objectives of this invention. The above descriptions are merely exemplary embodiments of this invention and should not be construed as limiting the scope of protection of this invention.

Claims

1. A method for cavitation risk assessment and correction of a water pump impeller repair profile, characterized in that, The method includes: Acquire the 3D scanning point cloud data and standard geometric data of the impeller to be repaired, calculate the difference in normal distance through coordinate registration, and generate a difference data matrix; The difference data matrix is ​​discretized into grid nodes. Historical operating data of the water pump is retrieved as boundary conditions. The pressure field and velocity field are solved on the grid nodes to obtain the fluid shear stress value and the fluid impact pressure value. Access a preset material property database to obtain material parameters, which include base layer material, transition layer material and surface layer material. Establish a load thickness mapping table based on the material parameters, and calculate the maximum allowable impact force and maximum allowable shear force under different layer thickness combinations. A system of simultaneous inequalities is established for the grid nodes using the material parameters; Solve the system of simultaneous inequalities to obtain solutions with different thickness ratios, select the solution with the lowest material consumption as the execution parameter, and generate a three-dimensional repair instruction. According to the three-dimensional repair instructions, the base layer material, the transition layer material and the surface layer material are identified by different layers on the grid nodes, a layered repair visualization model is constructed and output for display.

2. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 1, characterized in that, The generated difference data matrix includes: The three-dimensional scanned point cloud data and the standard geometric data are roughly aligned based on principal component analysis, and a fine transformation is performed based on the iterative nearest point to obtain aligned point cloud data and aligned geometric data. Based on the aligned point cloud data, the sampling points are traversed, the normal projection coordinates are calculated along the aligned geometric data, and the Euclidean distance between the sampling points and the normal projection coordinates is calculated to obtain the normal distance difference. The difference in normal distance is mapped to a two-dimensional plane coordinate index to construct a difference data matrix.

3. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 2, characterized in that, The obtained fluid shear stress values ​​and fluid impact pressure values ​​include: The difference data matrix is ​​mapped to the aligned geometry data to construct a three-dimensional spatial mesh, and the vertices of the three-dimensional spatial mesh are defined as mesh nodes; The flow rate and head curve data are analyzed by the historical operation data of the water pump, the flow rate value is extracted and converted into the flow velocity value and assigned to the inlet boundary node of the three-dimensional spatial grid, and the static pressure value is extracted and assigned to the outlet boundary node of the three-dimensional spatial grid. Perform fluid dynamics iterative calculations on the mesh nodes to solve for the local velocity vector and static pressure scalar. Calculate the fluid shear stress value based on the near-wall gradient of the local velocity vector, and synthesize the fluid impact pressure value based on the static pressure scalar and the normal component of the local velocity vector.

4. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 1, characterized in that, The calculation of the maximum permissible impact force and the maximum permissible shear force under different layer thickness combinations includes: The impact resistance and shear resistance per unit thickness are extracted from the base layer material, the transition layer material, and the surface layer material, respectively. For the base layer material, the transition layer material, and the surface layer material, a thickness deviation length is set to generate a thickness combination sequence, and a load thickness mapping table is constructed based on the thickness combination sequence as row index and column index; For the thickness combination sequence, the thickness of each layer of material is multiplied by the unit thickness impact resistance value and summed to obtain the maximum allowable impact force. The thickness of each layer of material is multiplied by the unit thickness shear resistance value to obtain the maximum allowable shear force. The maximum allowable impact force and the maximum allowable shear force are then filled into the load thickness mapping table.

5. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 4, characterized in that, The step of establishing a set of simultaneous inequalities for the grid nodes using the material parameters includes: The thickness variable to be solved is set based on the base layer material, the transition layer material, and the surface layer material; Construct a total strength constraint condition, which stipulates that the sum of the product of the thickness variable to be solved and the impact resistance value per unit thickness is greater than or equal to the fluid impact pressure value, and the sum of the product of the thickness variable and the shear resistance value per unit thickness is greater than or equal to the fluid shear stress value. Construct surface brittleness constraints, limiting the thickness variable to be solved to be less than or equal to the inverse proportional function value of the fluid impact pressure, and combine the total strength constraints and the surface brittleness constraints into a system of simultaneous inequalities.

6. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 5, characterized in that, The method further includes: Before establishing the set of simultaneous inequalities, the mesh nodes are pre-screened using the load thickness mapping table to remove thickness combination sequences where the maximum allowable impact force is lower than the fluid impact pressure value and the maximum allowable shear force is lower than the fluid shear stress value.

7. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 5, characterized in that, The process of setting the thickness variable to be solved based on the base layer material, the transition layer material, and the surface layer material includes: Non-negative continuous numerical variables are established for the base layer material, the transition layer material, and the surface layer material, respectively; The non-negative continuous numerical variables are constructed as a one-dimensional vector to form the thickness variable to be solved.

8. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 1, characterized in that, The generated 3D repair instructions include: For the system of simultaneous inequalities, traverse the feasible region and extract all thickness combinations to form a feasible solution set; Based on the feasible solution set, the sum of the thicknesses of the base layer material, the transition layer material, and the surface layer material is calculated as the total consumption index, and the total consumption index is sorted in ascending order; The first value of the total consumption index is selected as the execution parameter, and the execution parameter is bound to the mesh node to generate a three-dimensional repair instruction.

9. The method for cavitation risk assessment and correction of the repair profile of a water pump impeller according to claim 1, characterized in that, The construction and output display of the layered repair visualization model includes: Based on the three-dimensional repair instructions, the mesh nodes are divided into the bottom layer of the base layer material, the middle layer of the transition layer material, and the top layer of the surface material; Based on the execution parameters, perform geometric stretching operations on the bottom layer, the middle layer, and the top layer to construct a three-dimensional sub-model; The three-dimensional sub-models are merged and assembled according to the spatial stacking order to construct a layered repair visualization model and output it for display.

10. A cavitation risk assessment and correction system for a water pump impeller repair profile, applied to the cavitation risk assessment and correction method for a water pump impeller repair profile as described in any one of claims 1-9, characterized in that, The system includes: The difference matrix generation module is used to acquire the three-dimensional scan point cloud data and standard geometric data of the impeller to be repaired, and to generate a difference data matrix by calculating the difference in normal distance through coordinate registration. The load calculation module is used to discretize the difference data matrix into grid nodes, retrieve historical operating data of the water pump as boundary conditions, and solve the pressure field and velocity field for the grid nodes to obtain the fluid shear stress value and the fluid impact pressure value. The mapping table construction module is used to access a preset material property database to obtain material parameters, including base layer material, transition layer material and surface layer material. Based on the material parameters, a load thickness mapping table is established, and the maximum allowable impact force and maximum allowable shear force are calculated under different layer thickness combinations. A constraint establishment module is used to establish a set of simultaneous inequalities for the mesh nodes based on the material parameters. The thickness ratio solution module is used to solve the system of simultaneous inequalities, obtain solutions with different thickness ratios, select the thickness ratio solution with the lowest material consumption as the execution parameter, and generate a three-dimensional repair instruction. The layered visualization output module is used to identify the base layer material, the transition layer material and the surface layer material on the grid nodes according to the three-dimensional repair instructions, construct a layered repair visualization model and output it for display.

Images