Numerical simulation method of turbulent flow based on partial Reynolds stress model

By introducing ω-scale equations and algebraic relationships into the Reynolds stress model, the helical stress model is formed, and the problems of low computational efficiency and poor robustness of the existing Reynolds stress model in separated flow and three-dimensional turbulence simulation are solved. , achieving more accurate and efficient turbulent flow simulation.

CN118709605BActive Publication Date: 2024-11-19NAT UNIV OF DEFENSE TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411188415.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-19
Estimated Expiration
2044-08-28

AI Technical Summary

Technical Problem

The existing Reynolds stress model has problems of low computational efficiency and poor robustness when simulating separation flow and turbulent flow with obvious three-dimensional effects. It is particularly difficult to accurately capture the historical effects of Reynolds stress tensors, and the robustness of different nonlinear assumptions. The difference is large, resulting in not being widely used in the actual engineering field.

Method used

Using the turbulent flow numerical simulation method based on partial Reynolds stress model, the Reynolds normal stress transport equation is established and coupled with the ω scale equation to form a partial Reynolds stress model of four equations, obtain the Reynolds average Navier-Stokes equation system and couple it through algebraic relationships Simplify the shear stress three-component to form a new quadruple-equivalent RANS model.

Benefits of technology

The turbulent flow simulation is realized with few equations, low calculation cost and good numerical stability, which can predict the historical effect of Reynolds stress anisotropy and improve the physical accuracy and computational efficiency of the simulation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118709605B_ABST
    Figure CN118709605B_ABST
Patent Text Reader

Abstract

The present invention relates to a turbulent flow numerical simulation method based on a partial Reynolds stress model, comprising: S1. establishing a Reynolds normal stress transport equation, and introducing oh S1. Obtain the Reynolds-averaged Navier-Stokes equations and couple them with the four-equation partial Reynolds stress model to obtain a coupled equation group; S2. Construct the grid data of the turbulent flow field to be simulated, and set the boundary conditions and initial conditions of the turbulent flow field; S3. Numerically solve the coupled equation group based on the grid data to obtain the numerical simulation results of turbulent flow. The present invention has the advantages of fewer equations, low computational cost, and good numerical stability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of computational fluid dynamics, and in particular to a turbulent flow numerical simulation method based on a partial Reynolds stress model. Background Art

[0002] With the development of computing technology, computational fluid dynamics (CFD) has become one of the three major means (theory, experiment, and calculation) to study the generation and evolution mechanism of turbulent structures and solve practical engineering turbulence problems. In particular, the vigorous development of CFD software has brought great convenience to designers and researchers. The traditional numerical simulation methods of turbulence include three categories: Reynolds average simulation (RANS) method, large eddy simulation (LES) method, and direct numerical simulation (DNS) method. Due to the limitations of computer hardware conditions, the RANS method is still the main method for solving engineering problems.

[0003] As for the RANS method, it includes the algebraic eddy viscosity assumption (linear / nonlinear), the direct establishment of the Reynolds stress transport equation and others. Among them, the turbulence model (LEVM) based on the linear eddy viscosity assumption has been widely used in the industry in the past thirty years, and steady-state flow problems with weak three-dimensional effects have been well solved. Especially in the field of aerospace, the typical representatives are the Sparlart-Allmarsa (SA) one-equation model, the shear stress transport (SST) k-ω two-equation model and its derivative versions. However, some turbulent flows involving separation and obvious three-dimensional effects have challenged the above-mentioned eddy viscosity theory. For example, flow to corners is one of the key flows that is difficult to accurately simulate with LEVM at present, and it is commonly found in civil / military aircraft.

[0004] At present, there are two types of alternatives to solve the defects of LEVM: Reynolds stress model (RSM) and nonlinear eddy viscosity model (NLEVM). The former directly solves the transport equations of the six components of Reynolds stress without the eddy viscosity assumption, and is considered to be the most advanced practical RANS model. In particular, the influence of Reynolds stress anisotropy is taken into account, and the above-mentioned key flows can be simulated well. However, due to low computational efficiency (large number of transport equations, numerical rigidity problems of source terms, etc.), poor robustness (shear stress will provide negative physical dissipation, wall characteristics of scaling equations, etc.), etc., it has not been widely used in practical engineering fields; for example, Chinese patent application publication number CN115438598A discloses a numerical method for a Reynolds stress turbulence model based on a general time root square scale, which is based on the existing SSG / LRR- oh Reynolds stress model oh The scaling equation, according to The general time root square scale and oh The relationship between the scales is obtained, and the turbulence is about Scale Scaling equation; Scaling equations and predicted turbulence based on SSG / LRR- oh Model The equations are coupled to obtain SSG / LRR- Model, obtain RANS equations, and SSG / LRR- The model is coupled to obtain a coupled equation set, and this coupled equation set needs to follow the existing SSG / LRR- oh The model makes the entire coupled equation set more complex, which specifically includes seven partial differential equations. Generally speaking, the more equations there are, the more complex the solution and the greater the computational cost. For the simulation process of computational fluid dynamics (CFD), the greater the computational cost, the lower the computational efficiency and the higher the simulation cost.

[0005] The latter generally consists of one or two partial differential equations, but the historical effects of the Reynolds stress tensor cannot be considered, and the robustness of different nonlinear assumptions also varies greatly. In 2021, NASA's "US 2030 CFD Vision (Updated Version)" pointed out: Research from Europe, the United States, China, etc. has greatly improved the above problems, but it has not yet met expectations in the application of separated flows. The latter still maintains the algebraic form of the Reynolds stress and strain rate of LEVM, but believes that the relationship between them is nonlinear. The advantage is that it can inherit the characteristics of LEVM with a small number of equations and high computational efficiency; the disadvantage is that the historical effects of the Reynolds stress tensor cannot be considered, and the robustness of different nonlinear assumptions also varies greatly. Other methods, such as the algebraic Reynolds stress model (ARSM), can be proved to be classified as NLEVM type. Summary of the invention

[0006] The object of the present invention is to provide a turbulent flow numerical simulation method based on a partial Reynolds stress model.

[0007] To achieve the above-mentioned object of the invention, the present invention provides a turbulent flow numerical simulation method based on a partial Reynolds stress model, comprising:

[0008] S1. Establish the Reynolds normal stress transport equation and introduce oh The scale equation is coupled with the four-equation partial Reynolds stress model, wherein the four-equation partial Reynolds stress model is expressed as:

[0009]

[0010] in, is the time-averaged density, , , is the Reynolds normal stress without density in three directions,t For time, are the velocity components in three directions, x k , x j They are the coordinate components in three directions, and the subscripts k The value is 1, 2, 3, subscript j The value is 1, 2, 3, , , are the generated terms of the Reynolds normal stress transport equation in three directions, and P 11 , P 22 , P 33 are the generation of Reynolds normal stress in three directions, , , are the pressure-strain correlation terms of the Reynolds normal stress transport equation in three directions, and , , are the pressure-strain correlation terms without density, e is the turbulent kinetic energy dissipation rate and is expressed as , m is the dynamic viscosity coefficient, D is the empirical coefficient of the diffusion term, C μ is the empirical coefficient of turbulence and is expressed as , is the turbulent kinetic energy and is expressed as , oh represents Menter's specific dissipation rate scale, is the diffusion coefficient of the equation for the specific dissipation rate scale, is the turbulent kinetic energy generation term and is expressed as , is the coefficient of the dissipation term of the equation for the specific dissipation rate scale, s ω is the diffusion coefficient of the scaling equation, is the coefficient of the cross-derivative term of the equation for the specific dissipation rate scale;

[0011] S2. Obtain the Reynolds average Navier-Stokes equations and couple them with the four-equation partial Reynolds stress model to obtain a coupled equation group;

[0012] S3. constructing the grid data of the turbulent flow field to be simulated, and setting the boundary conditions and initial conditions of the turbulent flow field;

[0013] S4. Numerically solving the coupled equations based on the grid data to obtain numerical simulation results of turbulent flow.

[0014] According to one aspect of the present invention, in step S1, the generated term of the Reynolds normal stress transport equation is obtained based on the Reynolds stress model and is expressed as:

[0015]

[0016] in, P ij is the generation of Reynolds normal stress, , are the Reynolds stress without density, , They are the velocity components in three directions, k The value is 1, 2, 3, subscript j The value is 1, 2, 3, subscript i The value is 1, 2, or 3.

[0017] According to one aspect of the present invention, in step S1, the pressure-strain correlation term of the Reynolds normal stress transport equation is obtained using the SSG / LRR hybrid model and is expressed as:

[0018]

[0019] in, is the pressure-strain correlation term of the Reynolds normal stress transport equation, C 1. C 2. C 3. C 4. C 5. , are the pressure-strain correlation coefficients, is the sum of the Reynolds normal stresses, , , , , is the anisotropy tensor, is the Kroneck operator, , , , is the average strain rate tensor, , is the mean vorticity tensor, subscript k The value is 1, 2, 3, subscript j The value is 1, 2, 3, subscript i The value is 1, 2, or 3.

[0020] According to one aspect of the present invention, in step S1, the pressure-strain correlation term coefficient in the pressure-strain correlation term of the Reynolds normal stress transport equation is obtained by using Menter's transition function F 1 is obtained, which is expressed as:

[0021]

[0022]

[0023] in, f Respectively refer to the pressure-strain correlation coefficient C 1. C 2. C 3. C 4. C 5. , , F 1 is the transition function, for f and LRR-ω The mode-related part, for f and SSG-e The part related to the mode, ζ is the transition function judgment function, d w is the wall distance, is the kinematic viscosity coefficient, are the parameters related to the zeta function, are cross-derivative related parameters.

[0024] According to one aspect of the present invention, in step S2, the Reynolds averaged Navier-Stokes equations are obtained and coupled with the four-equation partial Reynolds stress model to obtain the coupled equations, wherein the Reynolds averaged Navier-Stokes equations are expressed as:

[0025]

[0026] in, is the time average pressure, is the viscous stress tensor, which is expressed as , is the average strain rate tensor, is the Reynolds shear stress tensor, which is expressed as: , is the average temperature in Favre, Average total energy for Favre, is the specific heat ratio at constant pressure, Pr is the laminar Prandtl constant, Pr t is the turbulent Prandtl constant, mt is the turbulent viscosity coefficient, R is the ideal gas molecular constant, is the Reynolds stress without density.

[0027] According to one aspect of the present invention, the Reynolds shear stress tensor is expanded using an algebraic relationship and is expressed as:

[0028]

[0029] in, , , is the Reynolds shear stress without density in three directions, for x Direction speed, for y Direction speed, for z Direction speed, x , y , z are the three directions of the rectangular coordinate system, , , is the anisotropic eddy viscosity coefficient and is expressed as:

[0030]

[0031] in, n The value is 1, 2, or 3.

[0032] According to one aspect of the present invention, in step S4, in the step of numerically solving the coupled equations based on the grid data to obtain the numerical simulation results of the turbulent flow, the 12 independent variables included in the coupled equations are solved: , , , , , , , , and oh The numerical solutions of other variables are then derived through the relationship.

[0033] According to a scheme of the present invention, the present invention only retains the transport equation of the three components of Reynolds normal stress on the traditional Reynolds stress model, and the three components of shear stress are obtained through algebraic relations. At the same time, a mature scale providing equation (such as the ω scale model) is introduced to form a new four-equation RANS model. Before the present invention, the RANS model is mainly divided into a (full) Reynolds stress model and an eddy viscosity model. Compared with the (full) Reynolds stress model, the present invention has the advantages of fewer equations, low computational cost, and good numerical stability; compared with the eddy viscosity model, it can predict the historical effect of Reynolds stress anisotropy, that is, it is more accurate physically. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flowchart schematically showing a method for numerically simulating turbulent flow according to an embodiment of the present invention;

[0035] Figure 2 1 is a diagram schematically showing a comparison of calculation results of a flat plate turbulent boundary layer using a turbulent flow numerical simulation method according to an embodiment of the present invention and a traditional method, wherein (a) shows a comparison of calculation results of a streamwise velocity type, and (b) shows a comparison of calculation results of a three-component Reynolds normal stress;

[0036] Figure 3 Schematically shows the comparison of the calculation results of the square cavity flow by the turbulent flow numerical simulation method according to one embodiment of the present invention and the traditional method (cloud map is colored by Mach number; streamlines are colored by the transverse velocity vector v and w ), where (a) represents the square cavity flow calculation result diagram of the traditional k-ω SST method, (b) represents the square cavity flow calculation result diagram of the traditional SSG / LRR-ωRSM method, and (c) represents the square cavity flow calculation result diagram of the Semi-RSM method of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments cannot be described one by one here, but the embodiments of the present invention are not therefore limited to the following embodiments.

[0038] like Figure 1 As shown, according to one embodiment of the present invention, a turbulent flow numerical simulation method based on a partial Reynolds stress model of the present invention comprises:

[0039] S1. Establish the Reynolds normal stress transport equation, introduce the ω-scale equation into the Reynolds normal stress transport equation and couple it with it to form a four-equation partial Reynolds stress model, where the four-equation partial Reynolds stress model is expressed as:

[0040]

[0041] in, is the time-averaged density, , , is the Reynolds normal stress without density in three directions, t For time, are the velocity components in three directions, x k , x j They are the coordinate components in three directions, and the subscripts k The value is 1, 2, 3, subscript j The value is 1, 2, 3, , , are the generated terms of the Reynolds normal stress transport equation in three directions, and P 11 , P 22 , P 33 are the generation of Reynolds normal stress in three directions, , , are the pressure-strain correlation terms of the Reynolds normal stress transport equation in three directions, and , , are the pressure-strain correlation terms without density, e is the turbulent kinetic energy dissipation rate and is expressed as , m is the dynamic viscosity coefficient, D is the empirical coefficient of the diffusion term, C μ is the empirical coefficient of turbulence and is expressed as , is the turbulent kinetic energy and is expressed as , oh represents Menter's specific dissipation rate scale, is the diffusion coefficient of the equation for the specific dissipation rate scale, is the turbulent kinetic energy generation term and is expressed as , is the coefficient of the dissipation term of the equation for the specific dissipation rate scale, s ω is the diffusion coefficient of the scaling equation, is the coefficient of the cross-derivative term of the equation for the specific dissipation rate scale;

[0042] S2. Obtain the Reynolds-averaged Navier-Stokes equations and couple them with the four-equation partial Reynolds stress model to obtain a coupled equation system;

[0043] S3. constructing the grid data of the turbulent flow field to be simulated, and setting the boundary conditions and initial conditions of the turbulent flow field;

[0044] S4. Numerically solve the coupled equations based on the grid data to obtain the numerical simulation results of turbulent flow.

[0045] According to one embodiment of the present invention, in step S1, the generated term of the Reynolds normal stress transport equation is obtained based on the Reynolds stress model and is expressed as:

[0046]

[0047] in, P ij is the generation of Reynolds normal stress, is the Reynolds stress without density, is the Reynolds stress without density, are the velocity components in three directions, are the velocity components in three directions, subscript k The value is 1, 2, 3, subscript j The value is 1, 2, 3, subscript i The value is 1, 2, or 3.

[0048] According to one embodiment of the present invention, in step S1, the pressure-strain correlation term of the Reynolds normal stress transport equation is obtained using the SSG / LRR hybrid model, and is expressed as:

[0049]

[0050] in, is the pressure-strain correlation term of the Reynolds normal stress transport equation, C 1. C 2. C 3. C 4. C 5. , are the pressure-strain correlation coefficients, is the sum of the Reynolds normal stresses, , , , , is the anisotropy tensor, is the Kroneck operator, , , , is the average strain rate tensor, , is the mean vorticity tensor, subscript k The value is 1, 2, 3, subscript jThe value is 1, 2, 3, subscript i The value is 1, 2, or 3.

[0051] According to one embodiment of the present invention, in step S1, the pressure-strain correlation term coefficient in the pressure-strain correlation term of the Reynolds normal stress transport equation is obtained by using Menter's transition function F 1 is obtained, which is expressed as:

[0052]

[0053]

[0054] in, f Respectively refer to the pressure-strain correlation coefficient C 1. C 2. C 3. C 4. C 5. , , F 1 is the transition function, for f and LRR-ω The mode-related part, for f and SSG-e The part related to the mode, ζ is the transition function judgment function, d w is the wall distance, is the kinematic viscosity coefficient, are the parameters related to the zeta function, are cross-derivative related parameters.

[0055] In this embodiment, the value of the pressure-strain correlation term coefficient in the pressure-strain correlation term of the Reynolds normal stress transport equation is shown in Table 1.

[0056] Table 1

[0057]

[0058] in, .

[0059] In this embodiment, oh The values ​​of the correlation coefficients in the scaling equation are shown in Table 2.

[0060] Table 2

[0061]

[0062] According to an embodiment of the present invention, in step S2, the Reynolds averaged Navier-Stokes equations are obtained and coupled with the four-equation partial Reynolds stress model to obtain the coupled equations. The Reynolds averaged Navier-Stokes equations are expressed as:

[0063]

[0064] in, is the time average pressure, is the viscous stress tensor, which is expressed as , is the average strain rate tensor, is the Reynolds shear stress tensor, which is expressed as: , is the average temperature in Favre, Average total energy for Favre, is the specific heat ratio at constant pressure, Pr is the laminar Prandtl constant, Pr t is the turbulent Prandtl constant, m t is the turbulent viscosity coefficient, R is the ideal gas molecular constant, is the Reynolds stress without density.

[0065] In this embodiment, the Reynolds shear stress tensor is expanded using an algebraic relationship and is expressed as:

[0066]

[0067] in, , , is the Reynolds shear stress without density in three directions, for x Direction speed, for y Direction speed, for z Direction speed, x , y , z are the three directions of the rectangular coordinate system, , , is the anisotropic eddy viscosity coefficient and is expressed as:

[0068]

[0069] in, n The value is 1, 2, or 3.

[0070] According to one embodiment of the present invention, in step S3, grid data of the turbulent flow field to be simulated is constructed, and in the step of setting the boundary conditions and initial conditions of the turbulent flow field, the turbulent flow field includes a flat boundary layer, an airfoil wake, a curved boundary layer, a hump separation flow, a turbine blade flow field, a wingtip vortex, a wing-body combination flow field, a wing flow field, a square cavity transition configuration, a wing critical stall flow, an airfoil stall pitch motion, a delta wing high angle of attack flow, etc.

[0071] According to one embodiment of the present invention, in step S4, in the step of numerically solving the coupled equations based on the grid data to obtain the numerical simulation results of the turbulent flow, the 12 independent variables included in the coupled equations are solved: , , , , , , , , and oh The numerical solutions of other variables are then derived through the relationship.

[0072] It should be noted that this scheme retains part of the Reynolds stress transport equation, so it is called the partial Reynolds stress model (Semi-RSM).

[0073] According to one embodiment of the present invention, a device of the present invention using the aforementioned turbulent flow numerical simulation method comprises:

[0074] The first model building module is used to obtain a four-equation partial Reynolds stress model, where the four-equation partial Reynolds stress model is expressed as:

[0075]

[0076] in, is the time-averaged density, , , is the Reynolds normal stress without density in three directions, t For time, are the velocity components in three directions, x k , x j They are the coordinate components in three directions, and the subscripts k The value is 1, 2, 3, subscript j The value is 1, 2, 3, , , are the generated terms of the Reynolds normal stress transport equation in three directions, and P 11, P 22 , P 33 are the generation of Reynolds normal stress in three directions, , , are the pressure-strain correlation terms of the Reynolds normal stress transport equation in three directions, and , , are the pressure-strain correlation terms without density, e is the turbulent kinetic energy dissipation rate and is expressed as , m is the dynamic viscosity coefficient, D is the empirical coefficient of the diffusion term, C μ is the empirical coefficient of turbulence and is expressed as , is the turbulent kinetic energy and is expressed as , oh represents Menter's specific dissipation rate scale, is the diffusion coefficient of the equation for the specific dissipation rate scale, is the turbulent kinetic energy generation term and is expressed as , is the coefficient of the dissipation term of the equation for the specific dissipation rate scale, s ω is the diffusion coefficient of the scaling equation, is the coefficient of the cross-derivative term of the equation for the specific dissipation rate scale;

[0077] The second model building module is used to obtain the Reynolds average Navier-Stokes equations and couple them with the four-equation partial Reynolds stress model to obtain a coupled equation group; in this embodiment, the obtained Reynolds average Navier-Stokes equation group is:

[0078]

[0079] in, is the time average pressure, is the viscous stress tensor, which is expressed as , is the average strain rate tensor, is the Reynolds shear stress tensor, which is expressed as: , is the average temperature in Favre, Average total energy for Favre, is the specific heat ratio at constant pressure, Pr is the laminar Prandtl constant, Pr t is the turbulent Prandtl constant, m t is the turbulent viscosity coefficient, R is the ideal gas molecular constant, is the Reynolds stress without density.

[0080] In this embodiment, the Reynolds shear stress tensor is expanded using an algebraic relationship and is expressed as:

[0081]

[0082] in, , , is the Reynolds shear stress without density in three directions, for x Direction speed, for y Direction speed, for z Direction speed, x , y , z are the three directions of the rectangular coordinate system, , , is the anisotropic eddy viscosity coefficient and is expressed as:

[0083]

[0084] in, n The value is 1, 2, or 3.

[0085] An input module is used to construct the grid data of the turbulent flow field to be simulated and set the boundary conditions and initial conditions of the turbulent flow field;

[0086] The numerical simulation module numerically solves the coupled equations based on the grid data to obtain the numerical simulation results of the turbulent flow. In this embodiment, in the step of numerically solving the coupled equations based on the grid data to obtain the numerical simulation results of the turbulent flow, 12 independent variables included in the coupled equations are solved: , , , , , , , , and oh The numerical solutions of other variables are then derived through the relationship.

[0087] To further illustrate this scheme, an example is given to illustrate it.

[0088] The method of the present invention is used to simulate the classic zero pressure gradient turbulent boundary layer example. In this embodiment, the calculation grid used in this embodiment is 137×97. Further, in this embodiment, the incoming flow Mach number is 0.2, and the Reynolds number Re=5×10 6 Therefore, based on the above conditions, the coupled equations (nine equations) consisting of the four-equation partial Reynolds stress model and the Reynolds averaged Navier-Stokes equations of the present invention are numerically solved on the computational grid, and the results are shown in Figure 2 It can be seen that the method of the present invention can simulate the Reynolds normal stress anisotropy effect, which is consistent with the traditional Reynolds stress model (SSG / LRR-ω) but has fewer equations (less cost), while the traditional eddy viscosity model (SST k-ω) cannot give the Reynolds normal stress anisotropy.

[0089] Furthermore, the method of the present invention is used to simulate the supersonic cavity flow example. In this embodiment, the calculation grid used in this embodiment is 241×41×41. Further, in this embodiment, the incoming flow Mach number is 0.5, and the Reynolds number Re=5.08×10 5 Therefore, based on the above conditions, the coupled equations (nine equations) consisting of the four-equation partial Reynolds stress model and the Reynolds averaged Navier-Stokes equations of the present invention are numerically solved on the computational grid, and the results are shown in Figure 3 It can be seen that the method of the present invention can capture the corner vortex of the flow direction corner, which is consistent with the traditional Reynolds stress model (SSG / LRR-ω) but has fewer equations (less cost), while the traditional eddy viscosity model (SST k-ω) cannot obtain the corner vortex.

[0090] The above contents are merely examples of specific solutions of the present invention. For devices and structures not described in detail therein, it should be understood that they can be implemented by adopting general devices and general methods available in the art.

[0091] The above is only one solution of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A numerical simulation method for turbulent flow based on a partial Reynolds stress model, characterized in that: include: S1. Establish the Reynolds normal stress transport equation and introduce ω The scale equation is coupled with the four-equation partial Reynolds stress model, wherein the four-equation partial Reynolds stress model is expressed as: in, is the time-averaged density, , , is the Reynolds normal stress without density in three directions, t For time, are the velocity components in three directions, x k , x j They are the coordinate components in three directions, and the subscripts k The value is 1, 2, 3, subscript j The value is 1, 2, 3, , , are the generated terms of the Reynolds normal stress transport equation in three directions, and P 11 , P 22 , P 33 are the generation of Reynolds normal stress in three directions, , , are the pressure-strain correlation terms of the Reynolds normal stress transport equation in three directions, and , , are the pressure-strain correlation terms without density, ε is the turbulent kinetic energy dissipation rate and is expressed as , μ is the dynamic viscosity coefficient, D is the empirical coefficient of the diffusion term, C μ is the empirical coefficient of turbulence and is expressed as , is the turbulent kinetic energy and is expressed as , ω represents Menter's specific dissipation rate scale, is the diffusion coefficient of the equation for the specific dissipation rate scale, is the turbulent kinetic energy generation term and is expressed as , is the coefficient of the dissipation term of the equation for the specific dissipation rate scale, σ ω is the diffusion coefficient of the scaling equation, is the coefficient of the cross-derivative term of the equation for the specific dissipation rate scale; S2. Obtain the Reynolds averaged Navier-Stokes equations and couple them with the four-equation partial Reynolds stress model to obtain a coupled equation group; wherein the Reynolds averaged Navier-Stokes equation group is expressed as: in, is the time average pressure, is the viscous stress tensor, which is expressed as , is the average strain rate tensor, is the Reynolds shear stress tensor, which is expressed as: , is the average temperature in Favre, Average total energy for Favre, is the specific heat ratio at constant pressure, Pr is the laminar Prandtl constant, Pr t is the turbulent Prandtl constant, μ t is the turbulent viscosity coefficient, R is the ideal gas molecular constant, is the Reynolds stress without density; The Reynolds shear stress tensor is expanded using an algebraic relationship and is expressed as: in, , , is the Reynolds shear stress without density in three directions, for x Direction speed, for y Direction speed, for z Direction speed, x , y , z are the three directions of the rectangular coordinate system, , , is the anisotropic eddy viscosity coefficient and is expressed as: in, n The value is 1, 2, 3; S3. constructing the grid data of the turbulent flow field to be simulated, and setting the boundary conditions and initial conditions of the turbulent flow field; S4. Numerically solving the coupled equations based on the grid data to obtain numerical simulation results of turbulent flow.

2. The turbulent flow numerical simulation method according to claim 1, characterized in that: In step S1, the generation term of the Reynolds normal stress transport equation is obtained based on the Reynolds stress model and is expressed as: in, P ij is the generation of Reynolds normal stress, , are the Reynolds stress without density, , They are the velocity components in three directions, k The value is 1, 2, 3, subscript j The value is 1, 2, 3, subscript i The value is 1, 2, or 3.

3. The turbulent flow numerical simulation method according to claim 2, characterized in that: In step S1, the pressure-strain correlation term of the Reynolds normal stress transport equation is obtained using the SSG / LRR hybrid model and is expressed as: in, is the pressure-strain correlation term of the Reynolds normal stress transport equation, C 1. C 2. C 3. C 4. C 5. , are the pressure-strain correlation coefficients, is the sum of the Reynolds normal stresses, , , , , is the anisotropy tensor, is the Kroneck operator, , , , is the average strain rate tensor, , is the mean vorticity tensor, subscript k The value is 1, 2, 3, subscript j The value is 1, 2, 3, subscript i The value is 1, 2, or 3.

4. The turbulent flow numerical simulation method according to claim 3, characterized in that: In step S1, the pressure-strain correlation coefficient in the pressure-strain correlation term of the Reynolds normal stress transport equation is calculated by Menter's transition function F 1 is obtained, which is expressed as: in, φ Respectively refer to the pressure-strain correlation coefficient C 1. C 2. C 3. C 4. C 5. , , F 1 is the transition function, for φ and LRR-ω The mode-related part, for φ and SSG-ε The part related to the mode, ζ is the transition function judgment function, d w is the wall distance, is the kinematic viscosity coefficient, are the parameters related to the zeta function, are cross-derivative related parameters.

5. The turbulent flow numerical simulation method according to claim 4, characterized in that: In step S4, in the step of numerically solving the coupled equations based on the grid data to obtain the numerical simulation results of the turbulent flow, the 12 independent variables included in the coupled equations are solved: , , , , , , , , and ω The numerical solutions of other variables are then derived through the relationship.