High-pressure diaphragm pump one-way valve structure multi-objective optimization method based on digital twinning

Abstract

The invention discloses a multi-objective optimization method for a high-pressure diaphragm pump one-way valve structure based on digital twinning, and belongs to the technical field of pump valve optimization design. The method comprises the following steps: establishing a simulation model by collecting geometric dimension parameters and material attributes of a high-pressure diaphragm pump one-way valve; carrying out experimental design by adopting an Isight and ANSYS joint simulation method to generate sample points and calculating response values; constructing an agent model according to the response values of each group of sample points, improving the calculation efficiency, and completing the construction of a digital twinborn model; taking the maximum equivalent stress of the valve core and the minimum maximum flow velocity of the valve gap flow field as objective functions, establishing a multi-objective optimization model of the one-way valve structure, solving by adopting a genetic algorithm to obtain a structure parameter combination with the best shock resistance of the one-way valve, and finally performing result verification. The method can achieve the high-precision simulation of the operation state of the one-way valve and the solving of the optimal structural parameters of the shock resistance, thereby guiding the structural design of the one-way valve, reducing the time cost of product research and development, and prolonging the service life.

Description

Technical Field

[0001] The present invention relates to the technical field of pump valve optimization design, and particularly to a multi-objective optimization method for the structure of a check valve of a high-pressure diaphragm pump based on digital twin. Background Technique

[0002] As one of the key mechanical equipment for long-distance transportation of slurry pipelines, the reciprocating high-pressure diaphragm pump has the advantages of high head, corrosion resistance, high pressure resistance, etc., and is suitable for transporting solid-liquid two-phase medium flows with complex operating conditions, and has a wide application prospect in industrial production fields such as coal, chemical industry, and metallurgy. As the mechanical component with the most frequent movement in the high-pressure diaphragm pump, the good one-way flow function of the check valve ensures the normal operation of the pump body for liquid suction and drainage. However, affected by the movement of the solid-liquid two-phase flow in the cavity, the stress on its structure is complex and variable, resulting in faults such as valve jamming and breakdown of the valve core. Since structural damage is the main factor leading to the formation of check valve faults, a good structural design can ensure better applicability in a harsh working environment and extend the service life. The traditional pump valve structure optimization design is completed by conducting multiple groups of experimental analyses and selecting the optimal group.

[0003] However, the traditional pump valve structure optimization design method not only has the disadvantages of long time consumption and high design cost, but also has certain safety hazards for this special industrial equipment. Most of the existing solutions use multiple groups of simulation experiments to explore the influence law of different structural parameter changes on performance to propose optimization schemes, ignoring the coupling influence between parameters and not finding the optimal combination of structural parameters. Therefore, by constructing a high-fidelity digital twin model that integrates simulation, mechanism, and surrogate models, considering the coupling influence between various structural parameters, and obtaining the optimal combination of structural parameters through a multi-objective optimization design method, a structural optimization scheme with guiding significance can be provided, which can greatly improve the design efficiency and reduce the experimental test cost. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a multi-objective optimization method for the structure of a check valve of a high-pressure diaphragm pump based on digital twin, which has the advantages of low cost and high efficiency, provides a guiding scheme for the optimization design of the product structure, and solves the above technical problems.

[0005] To achieve the above object, the present invention provides the following technical solution: A multi-objective optimization method for the structure of a check valve of a high-pressure diaphragm pump based on digital twin, comprising the following steps:

[0006] S1. Collect the geometric structure size parameters and material properties of the check valve of the high-pressure diaphragm pump, and perform geometric modeling and mechanism modeling on the check valve to obtain a simulation model and a mechanism model;

[0007] The simulation model includes: a geometric model, a numerical model;

[0008] The mechanism model includes: slurry turbulence model, fluid-structure interaction model and spool movement model;

[0009] Furthermore, the geometric model includes: valve body, spool, limiter, spring and valve seat; among them, the spool includes: gland, rubber gasket, valve cone;

[0010] Furthermore, the numerical model includes: mesh model, boundary conditions and material parameter attribute settings;

[0011] Furthermore, the slurry turbulence model is constructed using the RNG K-ε model, and the expression is as follows:

[0012]

[0013] In the formula, K represents turbulent kinetic energy; ρ represents liquid density; ε represents dissipation rate; G k represents the turbulent kinetic energy caused by the velocity gradient; G b represents the turbulent kinetic energy generated by buoyancy; Y M represents the contribution of the fluctuating expansion number to the total dissipation rate; α K and α ε respectively represent the turbulent reaction Prandtl numbers of the K equation (the first equation) and the ε equation (the second equation); represents the partial derivative operation with respect to x i ; represents the partial derivative with respect to x j ; x i and x j represent the components in the i and j directions of the coordinate; μ eff represents the viscosity coefficient; S K and S ε represent the source terms corresponding to K and ε; C 1ε 、C 2ε and C 3ε respectively represent the first empirical constant, the second empirical constant and the third empirical constant; R ε represents the ratio of the dissipation rate to the turbulent kinetic energy K;

[0014] Furthermore, the expression of the fluid-structure interaction model is as follows:

[0015]

[0016] In the formula, M S is the mass matrix; C S is the damping matrix; K S is the stiffness matrix; r s and r f respectively represent the spool displacement and the fluid displacement of the fluid-structure interaction surface; τ s and τ fThe force on the spool and the fluid force acting on the fluid-structure interaction surface, respectively; n f and n s respectively represent the cosines of the vertical direction of the contact surface between the fluid and the check valve;

[0017] Furthermore, the expression of the spool motion model is as follows:

[0018]

[0019] In the formula, m s is the mass of the spool; p s , p are the inlet pressure and outlet pressure of the valve respectively; g is the acceleration due to gravity; k and h 0 are the spring stiffness and preload respectively; h is the spool displacement; B is the fluid damping coefficient;

[0020] S2. Use the combined simulation method of Isight and ANSYS to simulate the simulation model and the mechanism model, generate experimental sample points and calculate the response values;

[0021] The combined simulation is to perform experimental design on the design variables (the half-cone angle α of the spool, the small diameter d of the spool, the cone surface width h of the spool) and the target values (the maximum equivalent stress S of the spool max , the maximum flow velocity V of the valve clearance flow field max ) set for the check valve structure through the DOE module of Isight, generate sample points using the optimal Latin hypercube method, and at the same time, the ANSYS software performs numerical solution on the model in step S1 to obtain the actual response values of each group of experimental sample points;

[0022] S3. According to the response values obtained in S2, construct a kriging surrogate model and perform feedback correction using the evaluation index of prediction accuracy and perform feedback correction using the evaluation index of prediction accuracy;

[0023] Specifically, the kriging surrogate model is to establish a mathematical model representing the relationship between input and output by training experimental sample points, that is, to obtain the regression equation between the design variables (the half-cone angle α of the spool, the small diameter d of the spool, the cone surface width h of the spool) and the target values (the maximum equivalent stress S of the spool max and the maximum flow velocity V of the valve clearance flow field max ), and perform feedback correction using the evaluation index of prediction accuracy. The mathematical model can be expressed in the form of the sum of a polynomial regression model and a stochastic process;

[0024] The expression of the kriging surrogate model is as follows:

[0025]

[0026] where num represents the number of polynomial functions; ω represents the polynomial coefficient vector; k(x) represents the basis function of the input parameter x; S(x) represents a statistical variable satisfying an expected value of 0 and a variance, and its covariance is:

[0027]

[0028] where is the variance of the Gaussian process; R(θ, x i , x j ) is the correlation function of the parameter θ, which is used to characterize the spatial correlation relationship between the sample points x i and x j , and its expression is:

[0029]

[0030] where m is the number of design variables; θ = {θ k : k = 1, 2,..., 3} are the model parameters to be determined, and this parameter can be taken as a fixed value or optimized as a model parameter; d k is the Euclidean distance of the sample point k;

[0031] The evaluation index of the prediction accuracy selects the coefficient of determination R 2 , and when R 2 > 0.9, the accuracy requirement is satisfied, and the calculation formula is as follows:

[0032]

[0033] where M is the number of test samples; is the predicted value of the test sample; is the mean value of the response values of the test samples; y p is the true response value of the test sample;

[0034] S4. Establish a multi-objective optimization mathematical model for the check valve structure according to the results of S3, and the expression is as follows:

[0035]

[0036] where α, d, and h represent the half-cone angle, minor diameter, and cone surface width of the valve core, which are set as optimization design variables; S max and V max respectively represent the maximum equivalent stress on the valve core and the maximum flow velocity in the valve gap flow field, which are set as optimization objectives; according to the structural dimension relationship of each design variable and combined with the single-factor analysis of the optimization objectives, the constraint range of the design variables is finally determined;

[0037] S5. Use the NSGA-II multi-objective genetic algorithm to iteratively solve the multi-objective optimization mathematical model of the check valve structure to obtain the optimal structural parameters;

[0038] S6. Use the digital twin model to simulate and verify the results of the optimal structural parameters;

[0039] Specifically, combine the obtained optimal parameter combinations of the spool half-cone angle α, small diameter d, and cone surface width h and the objective function values, construct the optimized structure of the check valve according to the parameters, and use the digital twin model for simulation verification.

[0040] Advantages of the present invention:

[0041] By constructing the digital twin model of the check valve, performing multi-objective optimization design on this model to obtain the optimized model, and using the genetic algorithm to iteratively solve the optimized model, the optimal structural parameter combination is obtained to achieve the multi-objective optimization design of the check valve structure, reduce the product maintenance cost and improve the design efficiency, providing important guiding opinions for the optimization design of the pump valve structure. Description of the drawings

[0042] Figure 1 is the step flow chart of the present invention;

[0043] Figure 2 is the flow structure schematic diagram of the present invention;

[0044] Figure 3 is the schematic diagram of the parametric modeling of the check valve structure of the present invention;

[0045] Figure 4 is the flow chart of the NSGA-II genetic algorithm of the present invention;

[0046] Figure 5 is the convergence curve diagram of the present invention; among them, part (a) is the convergence curve diagram of V max ; part (b) is the convergence curve diagram of S max ;

[0047] Figure 6 is the velocity contour map of the flow field of the check valve before and after optimization of the present invention; among them, part (a) is the velocity contour map of the flow field of the check valve before optimization; part (b) is the velocity contour map of the flow field of the check valve after optimization;

[0048] Figure 7 is the equivalent stress contour map of the check valve before and after optimization of the present invention; among them, part (a) is the equivalent stress contour map of the check valve before optimization; part (b) is the equivalent stress contour map of the check valve after optimization. Detailed implementation manners

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0050] As Figure 1 and Figure 2 shown, a multi-objective optimization method for the one-way valve structure of a high-pressure diaphragm pump based on digital twin includes the following steps:

[0051] S1. Collect the geometric structure size parameters and material properties of the one-way valve of the high-pressure diaphragm pump, and perform geometric modeling and mechanism modeling on the one-way valve to obtain a simulation model and a mechanism model;

[0052] The simulation model includes: a geometric model and a numerical model;

[0053] The mechanism model includes: a slurry turbulence model, a fluid-structure interaction model, and a spool movement model;

[0054] Furthermore, the geometric model includes: a valve body, a spool, a limiter, a spring, and a valve seat; as Figure 3 shown; wherein, the spool includes: a gland, a rubber gasket, and a valve cone;

[0055] Furthermore, the numerical model includes: a mesh model, boundary conditions, and material parameter attribute settings;

[0056] In this embodiment, by performing mesh division on the geometric model, tetrahedral elements are used for discrete processing and the surface area of the spool is subjected to mesh encryption processing. The boundary conditions are set as a velocity-type inlet and a pressure-type outlet, and the specific values are v = 2.6sin(4.54t) m / s and P = 0.78 MPa. The material of the rubber gasket is polyurethane, and the materials of the gland and the valve cone are 20CrNiMo alloy structural steel;

[0057] Furthermore, the slurry turbulence model is constructed using the RNG K-ε model, and the expression is as follows:

[0058]

[0059] In the formula, K represents the turbulent kinetic energy; ρ represents the liquid density; ε represents the dissipation rate; G k represents the turbulent kinetic energy caused by the velocity gradient; G b represents the turbulent kinetic energy generated by buoyancy; Y M represents the contribution of the fluctuating expansion number to the total dissipation rate; α K and α εrespectively represent the turbulent reaction Prandtl numbers of the K equation (the first equation) and the ε equation (the second equation); represents the partial derivative operation with respect to x i ; represents the partial derivative with respect to x j ; x i and x j represent the components in the i and j coordinate directions; μ eff represents the viscosity coefficient; S K and S ε represent the source terms corresponding to K and ε; C 1ε , C 2ε and C 3ε respectively represent the first empirical constant, the second empirical constant, and the third empirical constant; R ε represents the ratio of the dissipation rate to the turbulent kinetic energy K;

[0060] Furthermore, the expression of the fluid-structure interaction model is as follows:

[0061]

[0062] In the formula, M S is the mass matrix; C S is the damping matrix; K S is the stiffness matrix; r s and r f are respectively the spool displacement and the fluid displacement of the fluid-structure interaction surface; τ s and τ f are respectively the spool force and the fluid force on the fluid-structure interaction surface; n f and n s respectively represent the cosines of the vertical directions of the fluid and the check valve contact surface;

[0063] Furthermore, the expression of the spool motion model is as follows:

[0064]

[0065] In the formula, m s is the spool mass; p s , p are respectively the inlet pressure and the outlet pressure of the valve; g is the acceleration due to gravity; k and h 0 are respectively the spring stiffness and the pre-tightening amount; h is the spool displacement; B is the fluid damping coefficient;

[0066] S2. Use the combined simulation method of Isight and ANSYS to simulate the simulation model and the mechanism model, generate experimental sample points and calculate the response values;

[0067] Co-simulation means that the design variables (semi-cone angle α of the spool, minor diameter d of the spool, cone surface width h of the spool) and target values (maximum equivalent stress S of the spool max and maximum flow velocity V of the valve clearance flow field max ) set for the check valve structure are designed through the DOE module of Isight. The optimal Latin hypercube method is used to generate sample points. At the same time, the ANSYS software numerically solves the model in step S1 to obtain the actual response values of each group of experimental sample points; it is automatically completed by computer software;

[0068] S3. According to the response values obtained in S2, construct a kriging surrogate model and perform feedback correction using the evaluation index of prediction accuracy;

[0069] Specifically, the kriging surrogate model is to establish a mathematical model representing the relationship between input and output by training experimental sample points, that is, to obtain the regression equation between the design variables (semi-cone angle α of the spool, minor diameter d of the spool, cone surface width h of the spool) and the target values (maximum equivalent stress S of the spool max and maximum flow velocity V of the valve clearance flow field max ), and perform feedback correction using the evaluation index of prediction accuracy. The mathematical model can be expressed in the form of the sum of a polynomial regression model and a stochastic process;

[0070] The expression of the kriging surrogate model is as follows:

[0071]

[0072] In the formula, num represents the number of polynomial functions; ω represents the polynomial coefficient vector; k(x) represents the basis function of the input parameter x; S(x) represents the statistical variable satisfying the 0 expectation and variance, and its covariance is:

[0073]

[0074] In the formula, is the variance of the Gaussian process; R(θ, x i , x j ) is the correlation function of the parameter θ, which is used to characterize the spatial correlation relationship between the sample points x i and x j , and its expression is:

[0075]

[0076] In the formula, m is the number of design variables; θ = {θ k : k = 1, 2,..., 3} are the model parameters to be determined, and these parameters can be taken as fixed values or optimized as model parameters; d kIs the Euclidean distance of sample point k;

[0077] The evaluation index for prediction accuracy selects the coefficient of determination R 2 , when R 2 > 0.9, it meets the accuracy requirements. The calculation formula is as follows:

[0078]

[0079] In the formula, M is the number of test samples; is the predicted value of the test sample; is the mean value of the response values of the test samples; y p is the true response value of the test sample;

[0080] S4. Establish a multi-objective optimization mathematical model for the check valve structure according to the results of S3. The expression is as follows:

[0081]

[0082] In the formula, α, d, and h represent the half-cone angle, minor diameter, and cone surface width of the valve core, which are set as optimization design variables; S max and V max represent the maximum equivalent stress on the valve core and the maximum flow velocity in the valve gap flow field respectively, which are set as optimization objectives; According to the structural dimension relationship of each design variable and combined with the single-factor analysis of the optimization objectives, the constraint range of the design variables is finally determined;

[0083] S5. Use the NSGA-II multi-objective genetic algorithm to iteratively solve the multi-objective optimization mathematical model of the check valve structure to obtain the optimal structural parameters;

[0084] Specifically, the NSGA-II multi-objective genetic algorithm iteratively solves the optimization model constructed in S4, and after iteration, an optimal combination of structural parameters is obtained; The algorithm flow is as Figure 4 shown. The population size is taken as 20, the maximum number of iterations is 1000, the crossover probability is taken as 0.9, the mutation probability is taken as 0.05, and the optimal solution is obtained after 800 iterations. The results of the 800th iteration are selected as the optimization scheme, and the convergence curve of the objective function is as Figure 5 shown in parts (a) and (b) of

[0085] S6. Use the digital twin model to simulate and verify the results of the optimal structural parameters;

[0086] Specifically, the optimal parameter combinations of the valve core half-cone angle α, minor diameter d, and cone surface width h and the objective function values obtained are shown in Table 1. According to the parameters, an optimized structure of the check valve is constructed and simulated and verified using the digital twin model;

[0087] Table 1 Comparison of valve core structure parameters and target optimization before and after

[0088]

[0089] After completing all the above steps, the velocity contour and equivalent stress contour of the check valve flow field before and after optimization can be obtained, as Figures 6 - 7 shown;

[0090] From Figure 6 parts (a) and (b) therein, it can be seen that the maximum velocities of the internal flow field before and after optimization are 1.5 m / s and 1.3 m / s respectively, and the maximum flow velocity is distributed in the valve gap channel formed by the valve core and the valve seat. The optimized structure reduces the maximum flow velocity of the valve gap flow field. The fluid flow velocity is positively correlated with the erosion wear rate of the valve core. The erosion wear rate of the valve core structure can be reduced by reducing the maximum flow velocity;

[0091] From Figure 7 parts (a) and (b) therein, it can be seen that the maximum stresses of the valve core structure before and after optimization are 5.5 MPa and 4.6 MPa respectively, reducing the maximum equivalent stress borne by the valve core structure, thereby weakening the influence of stress concentration on the structural strength to achieve the extension of the service life of the check valve.

[0092] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-objective optimization method for the one-way valve structure of a high-pressure diaphragm pump based on digital twins, characterized in that: The following steps are involved: S1. Collecting geometric structure size parameters and material properties of the one-way valve of the high-pressure diaphragm pump, and performing geometric modeling and mechanism modeling on the one-way valve to obtain a simulation model and a mechanism model; The simulation model includes: geometric model, numerical model; The mechanism models include: slurry turbulence model, fluid-solid coupling model and valve core motion model; S2, using Isight and ANSYS joint simulation method to simulate the simulation model and mechanism model, generate experimental sample points and calculate the response value; S3, based on the response value obtained in S2, construct the kriging proxy model and use the evaluation index of prediction accuracy for feedback correction; S4. Establishing a multi-objective optimization mathematical model for the one-way valve structure according to the result of S3; S5. Use NSGA-II multi-objective genetic algorithm to iteratively solve the multi-objective optimization mathematical model of the check valve structure to obtain the optimal structural parameters; S6. The results of the optimal structural parameters are simulated and verified using a digital twin model.

2. According to claim 1, a multi-objective optimization method for a high-pressure diaphragm pump check valve structure based on digital twinning is characterized in that: In the method of using Isight and ANSYS to jointly simulate the simulation model and the mechanism model, generate experimental sample points and calculate the response value, the joint simulation is to perform experimental design on the design variables and target values ​​set for the one-way valve structure through the DOE module of Isight, and generate sample points by using the optimal Latin hypercube method. At the same time, ANSYS software numerically solves the model in step S1 to obtain the actual response value of each group of experimental sample points, wherein the design variables include the semi-cone angle α of the valve core, the minor diameter d of the valve core, and the cone width h of the valve core, and the target value includes the maximum equivalent stress S of the valve core. max and the maximum flow velocity V of the valve gap flow field max .

3. According to claim 1, a multi-objective optimization method for a high-pressure diaphragm pump check valve structure based on digital twinning is characterized in that: According to the response value obtained by S2, a kriging proxy model is constructed and feedback correction is performed using the evaluation index of prediction accuracy. The expression of the kriging proxy model is as follows: In the formula, num represents the number of polynomial functions; ω represents the polynomial coefficient vector; k(x) represents the basis function of the input parameter x; S(x) represents the function that satisfies the zero expectation and statistical variables of variance; The evaluation index of prediction accuracy is the coefficient of certainty R. 2 , when R 2 When it is greater than 0.9, the accuracy requirement is met, and the calculation formula is as follows: Where M is the number of test samples; is the predicted value of the test sample; is the mean of the test sample response values; y p is the true response value of the test sample.

4. According to claim 1, a multi-objective optimization method for a high-pressure diaphragm pump check valve structure based on digital twinning is characterized in that: The mathematical model of multi-objective optimization of the one-way valve structure is established according to the result of S3, and the expression is as follows: Where α, d and h represent the semi-cone angle, minor diameter and cone width of the valve core, which are set as the optimization design variables; S max and V max They respectively represent the maximum equivalent stress on the valve core and the maximum flow velocity in the valve gap flow field.

Images