CN103226634A - Computing method for unsteady flow field of rotary jet pump based on three-dimensional dynamic mesh - Google Patents

Abstract

The invention provides a calculation method for the unsteady flow field of a rotary jet pump based on a three-dimensional dynamic grid, which is applied in the technical field of a rotary jet pump. The method is based on a three-dimensional dynamic grid technology and solves the technical problem of grid reconstruction. The calculation of the unsteady flow field of the rotary jet pump is realized.

Description

Calculation method of unsteady flow field of rotary jet pump based on 3D dynamic mesh

technical field

The invention relates to the technical field of a rotary jet pump, in particular to a calculation method for an unsteady flow field of a rotary jet pump.

Background technique

Low-flow high-lift pumps are widely used in many industries such as petroleum, chemical industry, urban construction, metallurgy, and papermaking. The specific speed of this type of pump is very low (ns≤30), so the problem is that the operation efficiency of the pump is low and the energy loss is extremely serious. Rotary jet pump (also known as rotary jet pump, Pitot pump) is a low specific speed pump developed based on the principle of converting fluid kinetic energy into potential energy by Pitot tube. Figure 1 is the flow model of the rotary jet pump, which includes four main flow-passing components: the water inlet section, the impeller, the rotor chamber and the manifold, where the impeller and the rotor chamber are fixed together and rotate synchronously with the main shaft, while the manifold It is a stationary element inserted into the rotor cavity. When the rotary jet pump is running, the liquid enters the impeller from the water inlet section, gets energy from the blades of the rotating impeller, enters the rotor cavity, and then rotates with the rotor cavity to obtain higher kinetic energy, then enters the header, and converts the kinetic energy into potential energy through the diffusion section Then it flows out from the straight pipe section of the collector. Compared with conventional high-pressure pumps (such as multi-stage pumps, reciprocating pumps, etc.), rotary jet pumps have the characteristics of simple structure, small size, light weight, convenient use, and reliable operation, and have gradually become the main option for low-flow high-lift pumps. There is also a large potential for improvement in performance such as pump efficiency. From the perspective of application, as shown in Figure 1, the radial length of the collector is equivalent to the radius of the rotor cavity, and the cross-sectional diameter of the collector accounts for about 1/4 to 1/3 of the axial dimension of the rotor cavity. It is conceivable that when the impeller and the rotor cavity rotate, the fixed and stationary manifold will inevitably cause serious interference to the flow field of the impeller and rotor cavity rotating at high speed (some rotary jet pumps rotate at speeds above 4000rpm), and there will also be formation around the manifold itself. The pressure difference produces a large force on the collector, causing unfavorable factors such as vibration and noise.

From the perspective of scientific research, for the unsteady flow problem of ordinary vane pumps (such as centrifugal pumps, axial flow pumps, etc.), when the impeller rotates, the geometric shapes of the rotor (impeller) domain and the stator (pump body, pipeline) domain are always the same. remain unchanged; except for the connected interface, the rotor domain and the stator domain are independent and do not interfere with each other, so the "slip grid method" in which the rotor domain and the stator domain are both rigid grids do relative motion can be used for transient analysis. However, in the flow field of the jet pump, when the rotor (rotor chamber and impeller) domain rotates, the rotor domain conflicts with the stationary manifold entity; to avoid the interference of the manifold entity, the rotor calculation domain must change with time For large deformations, the calculation grid also needs to be adjusted accordingly, so the above-mentioned "slip grid method" for rigid grids cannot be used.

To solve the problem of large deformation of the flow domain and grid, it is necessary to adopt three-dimensional calculation of the rotating domain and its grid reconstruction technology, so that it is possible to analyze the unsteady flow field of the rotary jet pump. Due to the complexity of 3D mesh reconstruction, time-consuming calculation, and errors caused by multiple mapping interpolation of old and new meshes, so far the flow field analysis based on 3D mesh reconstruction technology is still in the exploratory stage, let alone Mature engineering application cases.

In 1923, Krogh applied the Pitot tube working principle to the design of the pump, which was later developed into a rotary jet pump for World War II. In the 1980s, Germany, Japan and other countries invested in the theoretical research and product development of rotary jet pumps. At present, many products have been put into the market, but most of the available rotary jet pump materials are product introductions, related research reports and technical reports. Information is still rare. In the mid-1990s, Sbom made a comprehensive introduction to the working principle, structural characteristics and applications of the rotary jet pump.

At the end of the 1980s, my country introduced rotary jet pump products one after another, and the country also started research on the working principle and hydraulic performance of rotary jet pumps. Yang Junhu, Qi Xueyi, Ma Xijin, etc. analyzed the flow conditions in the jet pump through prototype tests, proposed a composite impeller design method with alternate long and short blades, and gave the design formula and selection range of the header; they also used any quasi-ternary The quasi-three-dimensional flow field in the impeller of the rotary jet pump is solved by the orthogonal surface method, and the flow mechanism and design method of the rotary jet pump are studied preliminarily. Wang Chengmu et al. conducted different combination tests of three impellers and four headers of the rotary jet pump, and found that the efficiency of the composite conventional blade impeller was not as high as that of the straight blade impeller; it was also found that the cross-sectional shape, size and roughness of the header In addition to affecting the size of the flow resistance, the temperature also affects the working stability of the pump.

With the rapid development of computing technology, some scholars have carried out numerical simulation research on the viscous flow field of the jet pump in recent years. Wang Yunyun, Chen Cichang, Yang Changming, etc. used Fluent software, adopted the standard k-ε turbulence model and SIMPLEC algorithm, and analyzed the flow law in the impeller of the jet pump. Cheng Yunzhang, Zhu Bing, Chen Hongxun et al. carried out three-dimensional numerical simulation of the impeller and rotor cavity of the jet pump in the relative reference system, and the flow in the manifold in the absolute reference system. The simulation results show that the manifold has Significant backflow and swirl. Zou Xuelian and Chen Hongxun established the flow mathematical model of the straight blade impeller of the rotary jet pump, and carried out three-dimensional calculation and comparative analysis of the flow in the impeller. Xu Hongyuan, Tian Aimin, Wang Xiaodong, etc. have conducted in-depth research on the flow in the manifold, understood the distribution of pressure, flow velocity and turbulence in the manifold, and learned that the shape of the diffuser section of the manifold has a greater impact on pump efficiency. Yu Huiqiong, Ji Quankai, Fu Jie et al. carried out numerical simulation of the flow field in the rotary jet pump, and discussed the influence of geometric parameters such as impeller and manifold on the performance of the pump. Yi Tongxiang and Wang Chunlin used ANSYS-CFX software to carry out large eddy numerical simulation of the flow field in the rotary jet pump, analyzed the influence of the shape of the inlet of the manifold and the airfoil on the performance of the pump, and analyzed the impact of solid particles on the rotary jet pump under different working conditions. The wear condition of the wall surface of the jet pump was studied, and the pulsating pressure in the pump was studied under the condition that the collecting pipe interfered with the rotating flow field. Song Huaide and Liu Yi carried out numerical simulation of the internal flow field of the header and rotor cavity of the rotary jet pump, and studied the rotary jet pumps with different header shapes and inlet sizes. Huang Si, Su Lijuan and others used numerical simulation combined with performance experiments to discuss the impeller forms (one-side open, one-side closed and double-side open), blade shapes (straight blades, curved blades, and long and short blades) of vertical jet pumps. Combination), the number of blades, the shape of the manifold, the cross section of the manifold, the turning radius of the manifold and other parameters on the performance of the pump.

Due to the previously mentioned problem that due to the interference of the manifold entity, the calculation domain and the calculation grid of the rotor must be greatly deformed and adjusted over time, so the flow field calculation of the above-mentioned rotary jet pump can only be limited to the steady calculation, that is, for the rotor domain (Impeller and rotor cavity) adopted the "rotor freezing method": both the rotor domain and the stator domain (manifold) are rigid, there is no relative motion between them. Obviously, the steady state simulation cannot reflect the instantaneous flow conditions of the rotojet pump, especially the fixed header will seriously interfere with the high-speed rotating impeller and the flow field of the rotor chamber, and the pressure difference around the header itself will also generate pressure on the header. Large force and may cause vibration noise phenomenon.

With regard to the unsteady calculation method of hydraulic machinery, the unsteady flow problems of hydraulic machinery can be roughly divided into the following three categories: (1) the object is stationary but the flow is unsteady, such as the separation flow of stationary cascades, etc.; (2) a single object Unsteady flow problems with rigid motion, such as the rotation of the rotor around the shaft, etc.; (3) Unsteady problems with relative motion or deformation motion of multiple bodies, such as gear pump meshing, valve opening and closing, and piston reciprocating motion in the cylinder etc.

For the unsteady problems of type (1) mentioned above, a static rigid grid can meet the requirements. For unsteady problems of type (2), static rigid grids can still be used, but non-inertial frames or multiple reference frames (inertial frames + non-inertial frames) need to be selected for steady or unsteady calculations. For example, the most common rotor flow problem for hydraulic machinery. The flow is unsteady for an inertial frame because the rotor is periodically grazing the solution domain. However, in the case of not considering the stationary components, the calculation domain that moves with the rotating components is in the rotating reference frame (non-inertial frame), and the flow can be regarded as steady, which simplifies the problem. If stationary components are also considered in addition to rotating components, such as the flow problem of both rotor and stator in the turbine, this situation must be analyzed using Multiple Reference Frame (MRF for short). Using a static rigid grid can save many troubles (such as geometric conservation rate, motion boundary, etc.) caused by using dynamic grid calculations. Therefore, most CFD software integrates multi-reference frame MRF technology. However, it must be pointed out that the positional relationship between the rotor and the stator under the MRF system is only the mutual position between them at a certain moment, which is called the "Frozen Rotor Approach". Obviously, the rotor freezing method cannot solve the unsteady flow problem caused by the relative motion of the rotor and the stator.

For the type (3) unsteady problems involving relative motion or deformation of objects, common solutions include sliding mesh and moving mesh. The basic idea of the slip mesh is to pre-draw a slip subdomain around the motion trajectory of the moving part. At the interface between the slip subdomain and other regions, the boundary conditions are used to connect with other regions, so as to realize the calculation of the overall flow field. Sliding mesh technology is strictly a kind of rigid moving mesh. During the whole movement process, the computational mesh performs rigid body motion together with the object in a known motion mode, and the computational mesh does not need to be regenerated, so the calculation amount is small, and the quality of the initial mesh can be maintained. Slip grid technology is widely used in unsteady flow problems such as conventional rotating impeller machinery and vehicle intersection, and this grid technology is also integrated in CFD software such as Ansys-CFX and Fluent. However, the sliding mesh method cannot solve complex problems such as deformable bodies or relative motions of multiple bodies, and the rotor domain of the rotary jet pump belongs to the category of deformable bodies.

Theoretically speaking, the moving mesh method is a general method to solve unsteady flow problems, and is generally used to solve the problem that the shape of the flow field changes with time due to boundary motion. The movement mode of the boundary can be known or unknown in advance, that is, the movement of the boundary is determined by the calculation result of the previous step, and the update of the grid is automatically completed according to the change of the boundary. The basic idea of the dynamic grid method is to update the grid of the fluid domain in each time step to realize the change of the solution domain due to boundary motion, and use the arbitrary Lagrangian-Eulerian (ALE) method to describe The governing equation of fluid motion, the physical quantities of the new grid are mapped from the old grid through interpolation operations, and the physical quantities on the grid of each time step are iteratively solved to obtain the dynamic evolution results of the flow.

There are three common dynamic grid methods: Spring-based smoothing, Dynamic layering and Remeshing based on spring theory.

(l) The spring smoothing method simplifies each grid edge into a spring with a certain stiffness and connected by nodes. Taking the boundary displacement as the boundary condition of the spring, the displacement increment of the node is obtained by solving the force balance equation of the spring system, and finally the node position of the new grid is obtained. In the spring smoothing method, the topological relationship of the old and new grid nodes remains unchanged, so the calculation accuracy can be guaranteed. However, the spring smoothing method stretches or compresses the grid by changing the position of the grid nodes, which may easily cause the grid to be too dense or too sparse; The grid quality deteriorates and affects the calculation accuracy; in severe cases, even negative volume grids appear, which makes the calculation error and terminated.

(2) The dynamic layering method is to dynamically increase or decrease the grid layer on the boundary according to the displacement of the boundary, that is, first assume an ideal grid layer height on the boundary, and when the boundary moves, if the grid layer next to the boundary When the height of the grid layer is stretched to a certain extent compared with the ideal height, it will be divided into two grid layers; if the adjacent grid is compressed to a certain extent, the two grid layers adjacent to the boundary will be merged into one layer to maintain a certain density on the mesh layer on the boundary. The dynamic layering method is faster in generating meshes, but it requires the meshes near the motion boundary to be hexahedral (three-dimensional), which is not suitable for flow domains with complex shapes.

(3) The mesh reconstruction method is a supplement to the elastic smoothing method, and the elastic smoothing method can generally only deal with small deformation flow field problems. For some problems of hydraulic machinery, such as gear pump meshing, valve switching process, etc., large deformation is inevitable. Therefore, it is necessary to use the mesh reconstruction method to re-divide the flow field mesh. The grid reconstruction method uses the grid size and distortion rate as the evaluation criteria. When the movement and deformation of the boundary are too large, and the local grid is severely distorted, the grid is re-divided for these areas. The physical quantities on the new grid are obtained from the old grid by volume conservation laws and interpolation mapping. It should be noted that the mesh reconstruction method needs to re-mesh every time the deformation moves, thus consuming more calculation time.

At present, the application of dynamic grid technology in hydraulic machinery is mainly in three aspects: the rotation of eccentric pump or gear pump, the opening and closing process of valves, and the reciprocating motion of piston in cylinder. The general practice of these tasks is to use a dynamic boundary file (Profile) or a user-defined function (User-Defined Function, referred to as UDF) in the flow software to define the known movement of the dynamic boundary of the gear, spool or piston. Grid technology (spring smoothing method, dynamic layering method and grid reconstruction method) performs simulation calculation and visual analysis on the unsteady flow state in the calculation domain, and obtains the change of flow field and force with time. However, when the dynamic mesh technology is extended to 3D mesh deformation, not only the complexity of the algorithm increases and the efficiency of mesh deformation decreases, but also the quality of the deformed mesh is often unsatisfactory and the calculation is terminated. Therefore, the application of the above dynamic mesh technology is basically limited. In two-dimensional (such as in the gear pump model, it is approximately considered that the flow parameters do not change in the direction of the axis of rotation; axisymmetric assumptions are made in the opening and closing of the valve and the reciprocating motion of the piston in the cylinder) or quasi-three-dimensional (the two-dimensional domain is drawn in the axial direction extended into a three-dimensional domain) for simplified model calculations.

To sum up, because the static header of the rotary jet pump seriously interferes with the rotor domain, the flow field analysis of all current rotary jet pumps can only be based on the steady calculation of the rotor freezing method, and the unsteady calculation of the rotary jet pump Flow field calculation is still blank.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies of the prior art and provide a calculation method for the unsteady flow field of a rotary jet pump based on a three-dimensional dynamic grid. The method is based on a three-dimensional dynamic grid technology and solves the technical problem of grid reconstruction , the calculation of the unsteady flow field of the rotary jet pump is realized.

In order to achieve the above object, the present invention adopts the following technical scheme: the calculation method of the unsteady flow field of the rotary jet pump based on the three-dimensional dynamic grid, it is characterized in that, comprises the following steps:

In the first step, the rotor cavity, impeller, water inlet section and header in the jet pump are defined as the calculation domain, in which the rotor cavity and the impeller are defined as the rotation domain in the calculation domain, and the three-dimensional calculation domain is constructed using mechanical drawing software Entity, forming a three-dimensional entity file;

The second step is to read the 3D entity file by using the meshing software, mesh the computational domain, obtain the 3D initial mesh, and form a mesh file;

The third step is to use computational fluid dynamics software to read the grid file, and to set the physical parameter setting; the physical parameter setting includes the setting of the motion boundary; the setting of the motion boundary means that the interface of the motion boundary is set as fluid and The contact surface between the rotor cavity and the impeller, the motion rule is set to rotate around the axis, and the motion mode is set to Profile mode;

The fourth step is to use computational fluid dynamics software to calculate the values of each time step in the calculation cycle in turn; the numerical calculation of each time step is completed after the calculated value converges; when the numerical calculation of a time step is completed, By updating the grid of the rotation domain and the grid nodes adjacent to the rotation domain, the grid of the next time step is reconstructed. There are two schemes for reconstructing the grid of the next time step: (1) The grid size and distortion rate are within the tolerance In the area within the range, first use the spring smoothing method to obtain the grid node of the next time step, and then obtain the physical quantity of the grid of the next time step from the existing grid through interpolation operation, so as to reconstruct the grid of the next time step; (2) For the area where the grid size and distortion rate exceed the allowable range, the grid is re-divided as the grid of the next time step; after the grid reconstruction of the next time step, the value of the next time step is calculated until the last time step is completed step calculation.

Due to the particularity of the structure of the rotary jet pump, the current calculation of the flow field of the rotary jet pump at home and abroad can only use the steady method of "rotor freezing", while the general method for the calculation of the unsteady flow field of the rotary impeller machine is to use the "slip net Grid method", but this method still cannot solve the unsteady calculation of the flow field of the jet pump; the method of the present invention adopts the calculation strategy of combining the rotating domain and the rigid domain, based on the three-dimensional grid, using the spring smoothing method combined with the local grid weight The construction method realizes grid reconstruction, calculates the value of each time step, realizes the unsteady flow field calculation of the jet pump, and provides a powerful tool for the analysis of the transient flow characteristics of the jet pump.

A further solution is: in the fourth step, using the spring smoothing method to obtain the grid node of the next time step refers to adopting the spring smoothing method for the area where the grid size and distortion rate are within the allowable range, and the area The grid edge of the grid is simplified to a spring with a certain stiffness and connected by grid nodes, and the grid is stretched or compressed by changing the position of the grid nodes; the boundary movement is used as the boundary condition of the spring, and the force balance equation of the spring system is solved Get the displacement increment of the grid node, and finally get the grid node position in the next time step.

The calculation cycle in the fourth step refers to the time required for the rotation domain to rotate 360°.

There are two methods for judging the numerical convergence in the fourth step: (1) judging by the change of the residual value; (2) writing a program to monitor the numerical value.

A further solution is: the physical parameter setting in the third step also includes: setting the inlet of the calculation domain as a pressure boundary condition; setting the outlet of the calculation domain as a mass flow boundary condition; setting the standard k-ε turbulence model for unsteady flow calculation ; Set the time step Δt; Set the initial conditions for the unsteady flow calculation, the initial conditions adopt the convergent solution of the steady flow in the jet pump.

其中，n为旋喷泵的转速值，Z为叶片数，k为＞1的整数。The value range of the time step Δt in the third step is:

Wherein, n is the rotational speed value of the rotary jet pump, Z is the number of blades, and k is an integer >1.

The convergent solution of the steady flow adopted by the initial conditions in the third step is calculated by the rotor freezing method.

The grid division software in the second step adopts ICEM software.

The computational fluid dynamics software in the third step and the fourth step adopts Ansys-Fluent software.

Compared with the prior art, the present invention has the following outstanding advantages and effects:

1. The method of the present invention adopts the calculation strategy of combining the rotating domain and the rigid domain, solves the technical problem of grid reconstruction in the three-dimensional calculation domain, and overcomes the technical difficulty that the existing technology cannot calculate the unsteady flow field of the jet pump, The calculation of the unsteady flow field of the rotary jet pump is realized, which provides a powerful tool for the analysis of the transient flow characteristics of the rotary jet pump;

2. The method of the present invention effectively uses the spring smoothing method combined with the local mesh reconstruction method to realize mesh reconstruction without consuming too much calculation time.

Description of drawings

Figure 1 is a sectional view of a rotary jet pump;

Fig. 2 is a flow chart of the inventive method;

Figure 3(a) is a schematic diagram of the computational domain surface grid at t=0;

Figure 3(b) is a schematic diagram of the computational domain surface grid at t=5Δt;

Figure 3(c) is a schematic diagram of the computational domain surface grid at t=10Δt;

Figure 3(d) is a schematic diagram of the computational domain surface grid at t=15Δt;

Fig. 4(a) is a schematic diagram of the computational domain near the inlet of the collector on the central section of the computational domain at t=0;

Figure 4(b) is a schematic diagram of the computational domain near the inlet of the collector on the central section of the computational domain at t=5Δt;

Fig. 5 is the outlet flow figure of the rotary jet pump that utilizes the inventive method to draw;

图；Fig. 6 is the dimensionless radial force that utilizes the method of the present invention to draw

picture;

图；Fig. 7 is the dimensionless radial force that utilizes the method of the present invention to draw

picture;

Among them, 1 is the rotor cavity, 2 is the impeller, 3 is the water inlet section, 4 is the header, and 4.1 is the outlet of the header.

Detailed ways

The present invention will be further described in detail below in conjunction with examples, but the embodiments of the present invention are not limited thereto.

Example

The method for calculating the unsteady flow field of the present invention is applied to a rotary jet pump, and its flow chart is shown in Figure 2, including the following steps:

In the first step, the rotor cavity, impeller, water inlet section and header in the jet pump are defined as the calculation domain of the fluid, in which the rotor cavity and impeller are rotating domains, and the water inlet section and header are rigid domains; using Mechanical drawing software (such as Pro/E software) constructs the 3D entity of the calculation domain to form a 3D entity file;

The second step is to use meshing software (such as ICEM software) to read the 3D solid file, mesh the computational domain, obtain the initial unit of the structured/unstructured hybrid grid, and form a grid file;

The third step is to use computational fluid dynamics software (such as Ansys-Fluent) to read the grid file and set the physical property parameters;

The physical property parameter setting includes setting the motion boundary. The setting of the motion boundary means that the interface of the motion boundary is set as the surface where the fluid contacts the rotor cavity and the impeller, the motion law is set to rotate around the axis, and the motion mode is set to Profile mode; Setting and control are the focus of moving mesh calculation; for the moving boundary with known motion rules, it is necessary to define the movement mode of the moving boundary, which can generally be controlled by two methods: Profile (moving boundary file) and UDF (user-defined function) ;The Profile method is suitable for relatively simple motion laws, such as translation and rotation; for more complex dynamic boundary motion, UDF method is required for control; although the motion boundary of this method is the surface in contact with the rotor cavity and impeller, the curved surface The shape is relatively complex, but the law of motion is simple rotation around the axis, so this method uses the Profile method to define the motion mode of the motion boundary;

其中，n为旋喷泵的转速值，Z为叶片数，k为＞1的整数；设置非定常流动计算的初始条件，初始条件的值采用旋喷泵内定常流动的收敛解，定常流动的收敛解可通过转子冻结法计算得出；The physical property parameter setting also includes setting the inlet of the calculation domain as the pressure boundary condition; setting the outlet of the calculation domain as the mass flow boundary condition, and setting the mass flow boundary condition according to the flow rate of the operating point; setting the standard k-ε turbulence model for unsteady flow calculation; Set the time step Δt, Δt is not greater than the time difference between adjacent impeller blades passing over the same position, the time step Δt is determined according to the speed value n and the number of blades Z of the jet pump,

Among them, n is the speed value of the rotary jet pump, Z is the number of blades, and k is an integer >1; the initial conditions for unsteady flow calculation are set, and the value of the initial conditions adopts the convergence solution of the steady flow in the rotary jet pump, and the steady flow The converged solution can be calculated by the rotor freezing method;

The fourth step is to use computational fluid dynamics software (such as Ansys-Fluent software), take the rotation domain to rotate 360° as a calculation cycle, and calculate the values of each time step in the calculation cycle in turn. The value of each time step is calculated in The calculated value is completed after the convergence; when the numerical calculation of a time step is completed, the 3D entity of the next time step is reconstructed, and the 3D entity of the next time step is reconstructed by updating the rotation domain grid and the grid nodes adjacent to the rotation domain. Grid, calculate the value of the next time step until the calculation of the last time step is completed; there are two methods for judging numerical convergence: (1) judging by the change of residual value; (2) writing a program to monitor the value; In different time steps, due to the movement of the motion boundary, the calculation domain is deformed, so the grid of the next time step must be obtained before calculation, and two schemes are used to reconstruct the grid of the next time step: (1) the grid In the area where the size and distortion rate are within the allowable range, the spring smoothing method is used to obtain the grid node of the next time step, and then the physical quantity of the next time step grid is obtained from the existing grid through interpolation operation, so as to reconstruct the next time step grid node. The grid of the time step; (2) For the area where the grid size and distortion rate exceed the allowable range, the grid is re-divided as the grid of the next time step; after the grid reconstruction of the next time step, the grid of the next time step is calculated value until the last time step is completed;

Using the spring smoothing method to obtain the grid nodes of the next time step means that for the area where the grid size and distortion rate are within the allowable range, the spring smoothing method is used to simplify the grid edges of the area to have a certain stiffness and pass through the mesh. The spring connected to the grid nodes stretches or compresses the grid by changing the position of the grid nodes; taking the boundary movement as the boundary condition of the spring, the displacement increment of the grid nodes is obtained by solving the force balance equation of the spring system, and finally the following Grid node positions at a time step;

When the rotor chamber and the impeller rotate around the shaft through a certain angle, due to the existence of the fixed header, the rotating domain of the jet pump conflicts with the collecting pipe, forcing the rotating domain to deform; the rotating domain deforms in two parts, that is, the original part The domain of rotation becomes the header, and the other part is that the space left by the header after turning this angle becomes a part of the new domain of rotation; therefore, the new domain of rotation is the original domain of rotation minus the new collector The overlapping part of the pipe entity is added to the part left by the original header entity; at any time, the shape and movement mode of the boundary surface of the rotation domain are known, and the boundary surface of the rotation domain includes two parts: one part is with The static interface adjacent to the collector, and the rest is the rotating interface adjacent to the impeller and rotor cavity; with the geometry and motion laws of the boundary, the calculation domain at any time can be updated and the mesh can be reconstructed;

After the fourth step, the fifth step can also be included, which is post-calculation processing; the post-calculation processing includes displaying the calculation results of the unsteady flow field of the roto-jet pump, predicting the external characteristic curve of the roto-jet pump in a steady state, and obtaining the pulsation value of the pump internal pressure (including pulsation frequency and pulsation amplitude) and other unsteady results, the magnitude and direction of the force on the header due to the pressure difference around the header can be obtained.

The outstanding advantages of the present invention are: due to the particularity of the structure of the rotary jet pump, the current calculation of the flow field of the rotary jet pump at home and abroad can only use the steady method of "rotor freezing", while the general method for the calculation of the unsteady flow field of the rotary impeller machinery The method is to use the "slip grid method", which allows relative motion between two rigid domains (rotor and stator parts), but this method still cannot solve the unsteady calculation of the flow field of the jet pump; The method of the present invention adopts a calculation strategy that combines the rotating domain (rotating parts such as rotor cavity and impeller) with the rigid domain (fixed parts such as water inlet section and header), wherein the rotating domain will deform with time, and the grid nodes The position also changes; while the position of the grid nodes adjacent to the rotating domain may change, the rest of the calculation domain does not deform with time; the calculation of the unsteady flow field of the jet pump is realized, which is the analysis of the transient flow characteristics of the jet pump Provide powerful tools. At the same time, the spring smoothing method combined with the local grid reconstruction method is effectively used to obtain the new grid node positions in the 3D calculation domain without consuming too much calculation time.

The calculation of the unsteady flow field of the rotary jet pump in the present invention can grasp the dynamic and static interference phenomenon of the flow field of the rotary jet pump, clarify the dynamic flow characteristics of the rotary jet pump and its unsteady flow mechanism, and improve the hydraulic performance and working reliability of the rotary jet pump , reducing vibration and noise and other unfavorable factors will have important scientific research value and engineering application value.

A small vertical jet pump is selected as the research object. The design parameters of the jet pump are: speed n=2900r/min, flow rate Q=1.8m3/h, head H=100m. The working medium is water, density ρ=998.2kg/m3, dynamic viscosity μ=1.003×10-3Pa·s, number of impeller blades Z=6. The calculation domain of the jet pump is established by using Pro-E, and Gambit is used to divide the grid of the calculation domain to obtain the grid units. Among them, there are 23,355 units in the water inlet section, 220,239 units in the rotating domain, and 159,237 units in the header. The total number of grid units is 402,831, and the total number of grid nodes is 112,104.

Fluent flow software was used for the calculation, and the rotation period of the impeller was calculated from the pump speed and the number of impeller blades to be 2.069×10 ^-2 s, and the blade sweeping period was 3.448×10 ^-3 s, and the calculation time step Δt=2.0×10 -3 s was selected ^{. 4} s. Use the Profile method to define the steering and rotation speed of the boundary surface of the computational domain. To simplify calculations, the deformed mesh is limited to the rotation domain. Mesh deformation is realized by spring smoothing method and local mesh reconstruction method. Figure 3(a)-(d) are schematic diagrams of the surface mesh of the computational domain at t=0, 5Δt, 10Δt, and 15Δt, respectively. It can be seen from Figure 3(a)-(d) that in addition to the outer surface of the rotating domain rotating with the solid, the inner surface in contact with the header is also adjusted over time, thus realizing the deformation and mesh reconstruction of the computational domain . Figure 4(a) and Figure 4(b) are schematic diagrams of the grids of the calculation domain near the inlet of the collector on the central section of the calculation domain at t=0 and t=5Δt, respectively, and the local grid near the inlet of the collector can be compared changes over time. It can be seen from Figure 4(a) and Figure 4(b) that the calculation grid in the fixed header remains unchanged, but the rotating domain needs to continuously adjust the boundary adjacent to the header over time, so the rotating domain grid appears different degrees of deformation and reconstruction (see the position indicated by the arrow).

Figure 5 shows the calculated change curve of the outlet flow of the jet pump with time. It can be seen from Figure 5 that after starting for a period of time (t≈0.010s, about half a rotation period), the outlet flow value of the jet pump tends to It is stable and periodically pulsates with time. In one rotation cycle, the outlet flow of the jet pump pulsates 6 times, that is, the flow pulsation frequency corresponds to the number Z of the impeller blades.

和

随时间的变化曲线。无量纲径向力的定义为：Figure 6 and Figure 7 respectively show the calculated dimensionless radial force

and

Variation curve over time. The dimensionless radial force is defined as:

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; /</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; \&rho;A</span>}{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; in</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}$

趋近于负值，纵向的

趋近于正值，即径向力方向是由轴心指向集流管的背流一侧。Where A is the total force area, F _x and F _y are the radial forces in the x and y directions, respectively, and U _in is the average flow velocity at the pump inlet. It can be seen from Figure 6 and Figure 7 that the pulsation amplitude of the radial force is relatively large, indicating that the interference phenomenon between the liquid flow in the rotor cavity and the collector cannot be ignored. When the rotary jet pump operates normally, the radial force in the horizontal direction

tends to be negative, the longitudinal

It tends to a positive value, that is, the radial force direction is from the axis to the back flow side of the collector.

In order to verify the validity and applicability of the method of the present invention, the performance curve obtained by the method of the present invention is compared with the measured performance results. The microcomputer test system adopts multi-function parameter measuring instrument, through which the microcomputer controls the start, stop and adjustment of external equipment, and realizes data collection, analysis and processing, etc. Measurements include motor three-phase AC voltage, current, power, grid frequency, speed, inlet and outlet pressure and flow rate of the rotary jet pump at the design speed, and finally obtain the performance curve of the rotary jet pump. The unsteady pressure is measured by a 10ms high frequency dynamic pressure sensor. The performance curve obtained by the test is the same as the performance curve obtained by the method of the present invention.

The above-mentioned embodiment is a preferred embodiment of the present invention, but the embodiment of the present invention is not limited by the above-mentioned embodiment, and any other changes, modifications, substitutions, combinations, Simplifications should be equivalent replacement methods, and all are included in the protection scope of the present invention.

Claims (9)

1. based on the computing method of the rotary-jet pump nonstationary flow field of three-dimensional dynamic mesh, it is characterized in that, comprise the steps:

The first step is defined as computational fields with rotary-jet pump rotor chamber, impeller, inlet segment and header, and wherein, rotor chamber and impeller are defined as the rotation territory in the computational fields, uses the 3D solid of mechanical drawing software construction computational fields, forms the 3D solid file;

Second step, use grid dividing software to read the 3D solid file, computational fields is carried out grid dividing, obtain three-dimensional initial mesh, form grid file;

In the 3rd step, use computational fluid dynamics software to read grid file, and carry out the physical parameter setting; Described physical parameter setting comprises the setting of moving boundaries; The setting of described moving boundaries is meant, the layout setting of the moving boundaries surface that to be fluid contacted with impeller with rotor chamber, and the characteristics of motion is set to sway, and mode of motion is set to the Profile mode;

The 4th step, use computational fluid dynamics software, the numerical value to each time step in the computation period calculates successively; The numerical evaluation of each time step is finished after the numerical value convergence of calculating gained; After the numerical evaluation of a time step is finished, by upgrading rotation territory grid and closing on the grid node of rotation territory part, next time step grid of reconstruct, next time step grid of reconstruct adopts two kinds of schemes: (1) is the zone in permissible range to size of mesh opening and aberration rate, adopt the spring theory of adjustment to obtain the grid node of next time step earlier, obtain the physical quantity of next time step grid by interpolation arithmetic from existing grid then, thus the grid of next time step of reconstruct; (2) size of mesh opening and aberration rate are exceeded the zone of permissible range, repartition grid, as next time step grid; Behind next time step grid reconstruction, calculate the numerical value of next time step, up to finishing the final time calculating in step.

2. the computing method of the rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh according to claim 1, it is characterized in that, the grid node that employing spring theory of adjustment in described the 4th step obtains next time step is meant, the zone in permissible range to size of mesh opening and aberration rate, take the spring theory of adjustment, the grid limit that this is regional is reduced to the spring that has certain rigidity and connect by grid node, stretches or compresses grid by changing the grid node position; With the border amount of movement is the boundary condition of spring, obtains the displacement increment of grid node by the equilibrium equation of finding the solution spring system, finally obtains the grid node position of next time step.

3. the computing method of the rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh according to claim 2 is characterized in that, the computation period in described the 4th step is meant, rotation territory 360 ° of required times of rotation.

4. the computing method of the rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh according to claim 3 is characterized in that, the numerical value convergent determination methods in described the 4th step has two kinds: (one) judges with the variation of residual values; (2) the coding logarithm value is monitored.

5. the computing method of the rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh according to claim 1, the physical parameter setting in described the 3rd step also comprises: the computational fields inlet is set is the pressure boundary condition; The computational fields outlet is set is the mass rate boundary condition; UNSTEADY FLOW is set calculates employing standard k-ε turbulence model; Time step Δ t is set; The starting condition that UNSTEADY FLOW is calculated is set, and described starting condition adopts the moving convergence solution of steady flow in the rotary-jet pump.

6. the computing method of the rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh according to claim 5 is characterized in that, the span of the time step Δ t in described the 3rd step is: Wherein, n is the tachometer value of rotary-jet pump, and Z is the number of blade, and k is＞1 integer.

7. the computing method of the rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh according to claim 5 is characterized in that, the permanent mobile convergence solution that the starting condition in described the 3rd step adopts is to calculate by the rotor freezing process.

8. according to the computing method of each described rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh among the claim 1-7, it is characterized in that the grid dividing software in described second step adopts ICEM software.

9. according to the computing method of each described rotary-jet pump nonstationary flow field based on three-dimensional dynamic mesh among the claim 1-7, it is characterized in that the computational fluid dynamics software employing Ansys-Fluent software in described the 3rd step and the 4th step.

Images