An aero-engine ice wind tunnel test section modeling simulation optimization integrated platform

Abstract

The application discloses an aero-engine ice wind tunnel test section modeling simulation optimization integrated platform, comprising a geometric parameterization modeling module, a mesh division and boundary condition setting module, an aero-engine performance calculation agent model module, a CFD simulation module, an iterative optimization module, a visualization and data analysis module, the application realizes deep coupling of geometric parameterization modeling, CFD simulation, aero-engine performance calculation and multidisciplinary optimization technology, and constructs an automatic closed-loop process of the aero-engine ice wind tunnel test section from geometric modeling to flow field simulation analysis and then to aerodynamic layout optimization, so that the high efficiency of calculation is ensured, the accuracy of simulation results is improved, a fast, accurate and scalable technical tool is provided for accurate flow field simulation and efficient aerodynamic layout optimization of the aero-engine ice wind tunnel test section, and the technical problems of long design period, low efficiency and high cost of the aero-engine ice wind tunnel test section are solved.

Description

Technical Field

[0001] This application relates to the field of wind tunnel technology, and in particular, to an integrated platform for modeling, simulation and optimization of wind tunnel test sections for aero-engines. Background Technology

[0002] An aero-engine icing wind tunnel is a special type of wind tunnel capable of artificially simulating icing meteorological conditions on the ground. It is primarily used for research on aero-engine icing mechanisms, verification of anti-icing and de-icing systems, and airworthiness compliance verification. It is a crucial ground-based testing facility for airworthiness certification and qualification during aero-engine development. The airflow and cloud fields within the aero-engine icing wind tunnel test section are influenced not only by the intake and exhaust systems and spray systems of the test section but also by the size and operating conditions of the aero-engine test object. Because different aero-engine test objects vary in size, performance, and icing test requirements, the aero-engine icing wind tunnel test section needs to be adapted to meet the icing test requirements of the aero-engine test object before conducting aero-engine icing wind tunnel tests. However, in the existing aero-engine icing wind tunnel test section modification design process, the various design stages (such as geometric modeling, flow field simulation, and aerodynamic layout optimization) are fragmented and cannot be efficiently iterated and optimized, resulting in long test section modification cycles, low icing test efficiency, and high test risks and costs. Summary of the Invention

[0003] This application provides an integrated platform for modeling, simulation and optimization of icing wind tunnel test sections for aero-engines. It is used to solve the technical problem that in the process of modifying and designing icing wind tunnel test sections for aero-engines to meet the icing test requirements of aero-engine test objects, the various design stages (such as geometric modeling, flow field simulation and aerodynamic layout optimization) are isolated from each other and cannot be efficiently iterated and optimized.

[0004] This application is achieved through the following solution:

[0005] An integrated platform for modeling, simulation, and optimization of an aero-engine ice wind tunnel test section includes:

[0006] The geometric parametric modeling module is used to obtain the initial geometric parameters required for the modification design of the ice wind tunnel test section. By setting control parameters and geometric relationships, it automatically generates the geometric model of the test section. The initial geometric parameters include the size and initial layout values ​​of each component of the test section.

[0007] The mesh generation and boundary condition setting module is used to discretize the computational domain of the test section, generate a three-dimensional mesh of the flow region inside the test section, set the boundary conditions of the computational domain of the test section, and select the flow field model and governing equations.

[0008] The aero-engine performance calculation proxy model module is used to construct an aero-engine performance calculation proxy model, calculate the inlet and outlet parameters of the aero-engine in real time, and bidirectionally couple with the test section flow field simulation system.

[0009] The CFD simulation module is used to start the CFD simulation solver and perform time-progressed numerical solutions of the flow field inside the test section. During the solution process, the aero-engine performance calculation proxy model and the CFD solver are tightly integrated in a bidirectional coupling manner. The two achieve real-time interaction through a mechanism of data exchange before the start of each iteration step.

[0010] The iterative optimization module is used to automatically adjust the aerodynamic layout parameters of the test section based on the genetic algorithm and simulation results. The goal of the iterative optimization is to improve the quality of the inlet flow field of the aero-engine test component, and the constraints are geometric size limitations, operating condition requirements, etc.

[0011] The visualization and data analysis module is used for monitoring the simulation process and visualizing the results, including real-time display of the geometric model of the test section, the mesh, and the intermediate and final results of the flow field simulation.

[0012] Furthermore, the initial values ​​for the dimensions and layout of each component in the test section include the rear compartment of the test cabin, corner sections, engine air intake ducts, free jet nozzles, test benches, test platforms, exhaust diffusers, the outer contour of the aero-engine, and the dimensions and relative positions of the drive shaft. When generating the geometric model of the test section, spline curves or NURBS surfaces (non-uniform rational B-spline surfaces) are used to describe the surface profiles of components such as engine air intake ducts and free jet nozzles. After automatically generating the geometric model of the test section, based on the preset parameters, the automated tools are called to pass the geometric model of the test section to the mesh generation and boundary condition setting module.

[0013] Furthermore, when dividing the three-dimensional mesh of the internal flow region of the test section, the mesh type is selected as structured, unstructured, or hybrid mesh. The rear compartment of the test chamber uses a structured mesh to ensure the orderly distribution of the mesh, while complex components use unstructured meshes. For key areas such as the free jet nozzle outlet, the aero-engine inlet and outlet, and the exhaust diffuser inlet, local densification or unstructured meshes are used to capture the flow details of the airflow field and cloud field. The mesh quality is ensured by optimizing indicators such as element distortion, orthogonality, and volume distribution.

[0014] When setting the boundary conditions of the computational domain of the test section, the inlet boundary of the test section is set as a pressure inlet or a mass flow inlet, and the outlet boundary is set as a pressure outlet; the solid surface inside the test section is set as a wall boundary condition; the initial parameters of the atomized water droplets need to be defined at the outlet of the spray device, and a large number of simulated particles are released at the nozzle position through the discrete phase model (DPM) to represent the atomized water droplets.

[0015] Furthermore, when selecting the flow field model and governing equations, a numerical model is constructed within the Eulerian-Lagrange framework by coupling the discrete phase model and the component transport model. This model is used to numerically simulate the motion, evaporation, and heat and mass transfer processes of atomized water droplets in cold air. The governing equations for the experimental flow field of the numerical model include:

[0016] Continuity equation: ;

[0017] in, Density (kg / m³) 3 ), For time (s), Flow velocity (m / s);

[0018] Momentum equation: ;

[0019] in, Spatial coordinate vector The i-th (i=1,2,3) component (m). For the first Velocity vector components in each direction (m / s) Pressure (Pa). The value is the dynamic viscosity (Pa·s). For the first Volume force vector components (N) in each direction;

[0020] Energy equation: ;

[0021] in, Temperature (K) Thermal conductivity (W / m·K) Specific heat capacity (J / kg·K) Internal heat source (W / m 3 );

[0022] Component equation: ;

[0023] in, For matter Concentration (kg / m³) 3 ), For matter diffusion coefficient (m) 2 / s), For matter The source item;

[0024] Equations of state for each component: , ;

[0025] in, , The densities (kg / m³) of components a and b in the mixture are respectively. 3 ), These are the partial pressures (Pa) of components a and b, respectively. It is the ideal gas constant (8.314 J / (mol·k));

[0026] Pressure division equation: ;

[0027] The turbulence model uses the RNG k-ε model, and its governing equations are:

[0028] ;

[0029] ;

[0030] in, Turbulent kinetic energy (m 2 / s 2 ), Turbulent dissipation rate (m 2 / s 3 ), Let be the model constants of equation k. The turbulent viscosity is (Pa·s). Effective viscosity (Pa·s). The strain rate is 1 / s. For the model constants of the generated terms, for The model constants of the equation, For the dissipation term, the model constants are... Additional terms for analyzing flow problems under rapid strain;

[0031] ;

[0032] in, , airflow velocity vector The i-th and j-th components (m / s). Spatial coordinate vector The i-th and j-th components (m);

[0033] Considering the effects of aerodynamic drag, gravity and buoyancy, additional mass force, pressure gradient force, and thermophoretic force on the motion of the water droplet, the governing equation for the motion of the water droplet is:

[0034] ;

[0035] in: Let be the acceleration vector of the water droplet. The mass of the water droplet is (kg). Let be the position vector of the water droplet. The dynamic viscosity of the gas is (Pa·s). The diameter of the water droplet is in meters (m). The dynamic drag coefficient of the water droplet. For relative Reynolds number, The velocity vector of the airflow. Let be the velocity vector of the water droplet. The velocity of the water droplet is denoted as m / s. The volume of the water droplet (m 3 ), Water droplet density (kg / m³) 3 ), Gas density (kg / m³) 3 ), It is the gravitational acceleration vector. This is the virtual mass force coefficient (with a value of 0.5). For airflow pressure gradient, For airflow temperature gradient, Thermophoretic force coefficient, Let the gas temperature be K. Rewrite the above equation in a form that does not show the diameter of the water-containing droplet:

[0036] ;

[0037] in, The drag coefficient is calculated using the following formula:

[0038] ;

[0039] Relative Reynolds Number The calculation formula is:

[0040] ;

[0041] in, , The speeds of the airflow and water droplets are respectively (m / s);

[0042] By combining the drag coefficients of spherical and disc-shaped water droplets and performing linear interpolation with the relative deformation of the water droplets as the variable, the dynamic drag coefficient of the water droplets is obtained. The calculation method is as follows:

[0043] ;

[0044] in, The drag coefficient of a spherical water droplet. This is the linear interpolation result corresponding to the relative deformation of the water droplet;

[0045] in,

[0046] ;

[0047] in, Let the Reynolds number be the number of the water droplet. The Reynolds number is used as a reference to determine the state of the flow.

[0048] Near the wall, consider the Saffman force of the shear layer. Its expression is:

[0049] ;

[0050] in, The lift coefficient of a spherical water droplet in an ideal fluid shear flow is taken as 0.5;

[0051] The turbulent effect in the flow field is described using a Discrete Random Walk (DRW) model, and the turbulent diffusivity of the water droplets is... ,in, , These are the airflow pulsation velocity vectors. The i-th and j-th components, The integration time scale;

[0052] In the calculation of turbulent diffusion of water droplets, Taken as the vortex survival time T L Time of water droplets passing through vortices T cross Minimum value:

[0053] ;

[0054] Where τ is the characteristic relaxation time of the water droplet (s). Scale for vortex length (m);

[0055] airflow pulsation velocity in turbulent vortex Given by the turbulent kinetic energy k:

[0056] ;

[0057] in: This represents the mean square value of the airflow pulsation velocity. These are random numbers that follow a normal distribution.

[0058] Furthermore, the aero-engine performance calculation proxy model module is specifically used for:

[0059] Based on mathematical models or experimental data of aero-engines, a proxy model for calculating aero-engine performance is constructed using proxy modeling technology. This model is used to calculate the inlet and outlet parameters of aero-engines in real time. The inputs of the proxy model are parameters such as total intake pressure, total intake temperature, static pressure of exhaust environment, and speed or fuel flow rate of aero-engines. The outputs are the inlet and outlet flow rates, temperature, and pressure of the engine.

[0060] For the selected aero-engine test object, firstly, sample data of surrogate models are collected through aero-engine whole-machine test or engine mathematical model, and then multiple candidate surrogate models are established using surrogate model modeling method;

[0061] The optimal surrogate model is selected as the surrogate model for engine performance calculation by using surrogate model evaluation indicators;

[0062] When the test section flow field simulation system specifies the engine intake and exhaust environment parameters (total intake temperature, total pressure, and static exhaust pressure) and engine speed or fuel flow rate for the aero-engine test object, the aero-engine performance calculation proxy model predicts the engine's inlet and outlet flow rates, temperature, and pressure within milliseconds. This is used to update the aerodynamic interface parameters of the aero-engine test object's intake and exhaust devices and the rear compartment of the test cabin in real time during the test section CFD simulation.

[0063] Furthermore, the proxy modeling methods include polynomial response surface method, radial basis function method, Kriging model method, and orthogonal polynomial method.

[0064] Furthermore, the evaluation metrics for the surrogate model include the coefficient of determination, root mean square error, and maximum absolute error.

[0065] Furthermore, the coupling process between the aero-engine performance calculation proxy model and the CFD solver includes:

[0066] The program initializes the data and starts the coupled module;

[0067] During the iteration process, the coupling module establishes connections and neighborhood search links between code, each code handles its own computational tasks, and exchanges data at predetermined times;

[0068] Finally, disconnect the code and stop all calculations;

[0069] In this process, while the CFD simulation solver solves the airflow field, the simulation module considers multiphase flow and phase change processes such as the movement of water droplets, deposition, and ice formation. It uses the Lagrange discrete phase model to solve the force and motion equations of atomized water droplets in the airflow field and simulates their transport process.

[0070] Furthermore, the iterative optimization module is specifically used for:

[0071] The aerodynamic layout parameters of the test section are automatically adjusted based on the simulation results. The iterative optimization objective is to improve the quality of the inlet flow field of the aero-engine. The constraints include geometric size restrictions and operating condition requirements. The geometric size restrictions include the size of the free jet nozzle, the size of the exhaust diffuser, and the axial distance between the spray device and the aero-engine. The operating condition requirements include the area and location of the core area of ​​the cloud and fog field. Specifically, improving the quality of the inlet flow field of the aero-engine involves reducing the non-uniformity of the pressure field, temperature field, velocity field, and cloud and fog field of the aero-engine inlet flow field, ensuring that they are less than or equal to the set values.

[0072] When there are multiple optimization objectives, the weighted aggregation method or Pareto front method is used to solve them. For weighted objectives, including simultaneously reducing the non-uniformity of the engine inlet pressure field and the non-uniformity of the temperature field, a comprehensive objective is constructed, the weights are adjusted to reflect the emphasis, and then a single-objective optimization method is used to solve it.

[0073] The system can plot the optimal value and population diversity index curves during the iteration process of the genetic algorithm in real time to monitor the convergence trend. After optimization, users can view the sensitivity analysis results of the optimization objective to each design parameter. The system can analyze the impact of the design parameters on the optimization objective through surrogate models or response surface methodology, thereby obtaining engineering design guidance.

[0074] Furthermore, the visualization and data analysis module is developed based on a B / S architecture, which displays the geometric model, mesh, intermediate state and final results of the flow field simulation of the test section in real time on the browser, including cloud maps of parameters such as pressure, temperature, velocity, and liquid water content, as well as droplet size distribution, so that users can analyze the simulation results and evaluate the optimized design scheme of the test section.

[0075] Compared with the prior art, this application has the following beneficial effects:

[0076] This application proposes an integrated platform for modeling, simulation, and optimization of an aero-engine icing wind tunnel test section, including a geometric parametric modeling module, a mesh generation and boundary condition setting module, an aero-engine performance calculation proxy model module, a CFD simulation module, an iterative optimization module, and a visualization and data analysis module. Compared with existing technologies, this application demonstrates significant advantages in the following aspects:

[0077] On the one hand, this application realizes the integration and automation of the entire process of experimental section renovation design from data input to scheme output, solving the problems of independent geometric modeling, mesh generation, CFD simulation and optimization design, cumbersome data transmission and large workload of repetitive operations in the traditional process. By integrating multiple links into the same platform and automatically connecting them, the work efficiency of experimental section renovation design is significantly improved.

[0078] On the other hand, this application introduces a proxy model for aero-engine performance calculation to participate in the flow field simulation calculation of the test section, thus overcoming the deficiency of existing methods in failing to analyze the aerodynamic coupling characteristics of the test section and the aero-engine test object at a lower computational cost. This application embeds the aero-engine performance calculation proxy model into the flow field simulation of the test section, achieving real-time bidirectional coupling between the operating conditions of the aero-engine test object and the flow field of the test section. This can significantly improve the accuracy and reliability of the flow field simulation results of the test section without increasing computational costs.

[0079] Furthermore, by incorporating an iterative optimization module, this application constructs a simulation-driven automatic optimization closed loop, reducing reliance on human experience and enabling intelligent adjustment of design parameters. Compared to the traditional approach of repeated manual calculations, this automatic optimization mechanism can explore a wider design space in a shorter time and quickly converge to an optimized solution that meets the test requirements. Therefore, this application can significantly shorten the design cycle for modifying the wind tunnel test section of an aero-engine and reduce the risks and costs of aero-engine icing tests.

[0080] In addition to the purposes, features, and advantages described above, this application has other purposes, features, and advantages. A further detailed description of this application will be provided below with reference to the figures. Attached Figure Description

[0081] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0082] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0083] Figure 1 This is a schematic diagram of the modules of the integrated modeling, simulation and optimization platform for the air-to-air engine ice wind tunnel test section according to a preferred embodiment of this application;

[0084] Figure 2 This is a schematic diagram of a parametric geometric model;

[0085] Figure 3 This is a schematic diagram of a tetrahedral unstructured mesh;

[0086] Figure 4 This is a schematic diagram of the process for constructing a proxy model for calculating the performance of an aero-engine.

[0087] Figure 5 This is a schematic diagram of the workflow of the integrated platform for modeling, simulation and optimization of the ice tunnel test section of an aero-engine. Detailed Implementation

[0088] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0089] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0090] Terminology Explanation:

[0091] Aircraft engine icing wind tunnel: An aircraft engine icing wind tunnel is a test facility specifically designed to simulate the operation of aircraft engines in icing environments. It generates low-temperature, high-humidity airflow within a sealed compartment, and injects liquid water droplets into the airflow through atomizing nozzles to create a cloud-like field. During testing, parameters such as incoming flow velocity, temperature, pressure, liquid water content, and median droplet diameter are typically controlled to approximate icing meteorological conditions.

[0092] Non-uniformity: refers to the degree of uniformity of the distribution of a physical quantity on a cross section, commonly expressed by the formula δ=(X max -X min ) / X avg This indicates that one of the design goals of the modified wind tunnel test section for aero-engines is to ensure that the non-uniformity of the flow field at the aero-engine inlet is below a certain threshold, in order to reproduce specific icing conditions in nature.

[0093] Proxy models are approximate models trained using machine learning, response surface methodology, and other methods based on high-fidelity simulation results or experimental data. Proxy models can quickly predict the output of complex systems with extremely low computational cost. In aero-engine performance calculations, proxy models predict engine inlet and outlet flow rates, temperatures, and pressures based on input parameters such as total intake pressure, total intake temperature, exhaust static pressure, and engine speed or fuel flow rate. This can replace complex engine performance calculations, saving simulation time.

[0094] Zero-dimensional model: A physical model that ignores spatial distribution and expresses changes in matter or energy only at a macroscopic scale. Zero-dimensional models cannot provide information on local flow fields and microscopic processes, but they are extremely fast to compute and are used in preliminary design.

[0095] CFD, or Computational Fluid Dynamics, is a computer-aided engineering technique that uses numerical calculation methods and computer simulations to analyze and predict systems involving fluid flow, heat conduction, and related physical phenomena.

[0096] like Figure 1 As shown, a preferred embodiment of this application provides an integrated platform for modeling, simulation, and optimization of an aero-engine ice wind tunnel test section, comprising:

[0097] The geometric parametric modeling module is used to obtain the initial geometric parameters required for the modification design of the ice wind tunnel test section. By setting control parameters and geometric relationships, it automatically generates the geometric model of the test section (see...). Figure 2 The initial geometric parameters include the dimensions and initial layout values ​​of each component in the test section;

[0098] The mesh generation and boundary condition setting module is used to discretize the computational domain of the test section, generating a three-dimensional mesh of the flow region within the test section (see...). Figure 3 Set the boundary conditions of the computational domain for the test section, and select the flow field model and governing equations;

[0099] The aero-engine performance calculation proxy model module is used to construct an aero-engine performance calculation proxy model, calculate the inlet and outlet parameters of the aero-engine in real time, and bidirectionally couple with the test section flow field simulation system.

[0100] The CFD simulation module is used to start the CFD simulation solver and perform time-progressed numerical solutions of the flow field inside the test section. During the solution process, the aero-engine performance calculation proxy model and the CFD solver are tightly integrated in a bidirectional coupling manner. The two achieve real-time interaction through a mechanism of data exchange before the start of each iteration step.

[0101] The iterative optimization module is used to automatically adjust the aerodynamic layout parameters of the test section based on the genetic algorithm and simulation results. The goal of the iterative optimization is to improve the quality of the inlet flow field of the aero-engine test component, and the constraints are geometric size limitations, operating condition requirements, etc.

[0102] The visualization and data analysis module is used for monitoring the simulation process and visualizing the results, including real-time display of the geometric model of the test section, the mesh, and the intermediate and final results of the flow field simulation.

[0103] This embodiment proposes an integrated platform for modeling, simulation, and optimization of an aero-engine icing wind tunnel test section. It includes a geometric parametric modeling module, a mesh generation and boundary condition setting module, an aero-engine performance calculation proxy model module, a CFD simulation module, an iterative optimization module, and a visualization and data analysis module. Compared with existing technologies, this embodiment demonstrates significant advantages in the following aspects:

[0104] On the one hand, this embodiment realizes the integration and automation of the entire process of experimental section modification design from data input to scheme output, solving the problems of independent geometric modeling, mesh generation, CFD simulation and optimization design, cumbersome data transmission and large workload of repetitive operations in the traditional process. By integrating multiple links into the same platform and automatically connecting them, the work efficiency of experimental section modification design is significantly improved.

[0105] On the other hand, this embodiment introduces an aero-engine performance calculation proxy model to participate in the flow field simulation calculation of the test section, making up for the deficiency of existing methods in failing to analyze the aerodynamic coupling characteristics of the test section and the aero-engine test object at a lower computational cost. This embodiment introduces an aero-engine performance calculation proxy model embedded in the flow field simulation of the test section, realizing real-time bidirectional coupling between the operating conditions of the aero-engine test object and the flow field of the test section, which can significantly improve the accuracy and reliability of the flow field simulation results of the test section without increasing computational costs.

[0106] Furthermore, by incorporating an iterative optimization module, this embodiment constructs a simulation-driven automatic optimization closed loop, reducing reliance on human experience and enabling intelligent adjustment of design parameters. Compared to the traditional approach of repeated manual calculations, this automatic optimization mechanism can explore a wider design space in a shorter time and quickly converge to an optimized solution that meets the test requirements. Therefore, this application can significantly shorten the design cycle for modifying the wind tunnel test section of an aero-engine and reduce the risks and costs of aero-engine icing tests.

[0107] Preferably, the initial values ​​of the dimensions and layout of each component in the test section include the rear compartment of the test cabin, corner section, engine air intake duct, free jet nozzle, test bench, test platform, exhaust diffuser, aero-engine outline, and the dimensions and relative positions of the drive shaft. When generating the geometric model of the test section, in order to ensure the accuracy and efficiency of geometric modeling, spline curves or NURBS surfaces (non-uniform rational B-spline surfaces) are used to describe the surface profiles of components such as the engine air intake duct and free jet nozzle. After automatically generating the geometric model of the test section, based on the preset parameters, the automated tool is called to pass the geometric model of the test section to the mesh generation and boundary condition setting module.

[0108] Preferably, when dividing the three-dimensional mesh of the internal flow region of the test section, the mesh type is selected as structured, unstructured, or hybrid mesh (see...). Figure 3 The test chamber's rear compartment uses a structured grid to ensure an orderly grid distribution, while complex components use an unstructured grid. In order to accurately capture the flow details at the inlet and outlet of the aero-engine test object, key areas such as the free jet nozzle outlet, aero-engine inlet and outlet, and exhaust diffuser inlet are locally densified or unstructured grids are used to capture the flow details of the airflow field and cloud field. The grid quality is ensured by optimizing indicators such as element distortion, orthogonality, and volume distribution.

[0109] When setting the boundary conditions of the computational domain of the test section, the inlet boundary of the test section is set as a pressure inlet or a mass flow inlet, and the outlet boundary is set as a pressure outlet; the solid surface inside the test section is set as a wall boundary condition; since a spray device is installed in the free jet nozzle, the initial parameters of the atomized water droplets need to be defined at the outlet of the spray device, and a large number of simulated particles are released at the nozzle position through the discrete phase model (DPM) to represent the atomized water droplets.

[0110] Preferably, when selecting the flow field model and governing equations, a numerical model is constructed by coupling the discrete phase model and the component transport model to numerically simulate the motion, evaporation, and heat and mass transfer processes of atomized water droplets in cold air. The governing equations for the flow field in the test section include:

[0111] Continuity equation: ;

[0112] in, Density (kg / m³) 3 ), For time (s), Flow velocity (m / s);

[0113] Momentum equation: ;

[0114] in, Spatial coordinate vector The i-th (i=1,2,3) component (m). For the first Velocity vector components in each direction (m / s) Pressure (Pa). The value is the dynamic viscosity (Pa·s). For the first Volume force vector components (N) in each direction;

[0115] Energy equation: ;

[0116] in, Temperature (K) Thermal conductivity (W / m·K) Specific heat capacity (J / kg·K) Internal heat source (W / m 3 );

[0117] Component equation: ;

[0118] in, For matter Concentration (kg / m³) 3 ), For matter diffusion coefficient (m) 2 / s), For matter The source item;

[0119] Equations of state for each component: , ;

[0120] in, , The densities (kg / m³) of components a and b in the mixture are respectively. 3 ), These are the partial pressures (Pa) of components a and b, respectively. It is the ideal gas constant (8.314 J / (mol·k));

[0121] Pressure division equation: ;

[0122] The turbulence model uses the RNG k-ε model, and its governing equations are:

[0123] ;

[0124] ;

[0125] in, Turbulent kinetic energy (m 2 / s 2 ), Turbulent dissipation rate (m 2 / s 3 ), Let be the model constants of equation k. The turbulent viscosity is (Pa·s). Effective viscosity (Pa·s). The strain rate is 1 / s. For the model constants of the generated terms, for The model constants of the equation, For the dissipation term, the model constants are... Additional terms for analyzing flow problems under rapid strain;

[0126] ;

[0127] in, , airflow velocity vector The i-th and j-th components (m / s). Spatial coordinate vector The i-th and j-th components (m).

[0128] Discrete Phase Initial Conditions: At the initial stage of the simulation, it can be assumed that there is no ice layer in the flow field of the test section, and that the atomized water droplets are uniformly distributed in the nozzle. Numerically, the process of a spray device injecting atomized water droplets into the airflow can be simulated by releasing a certain concentration of water droplet particles at the nozzle location. To reduce computational load, calculations are first performed under pure air conditions, and atomized water droplets are introduced after the airflow stabilizes. The initial diameter distribution of the atomized water droplets can be simulated using the Rosin-Rammler distribution.

[0129] In the cloud and fog field of the ice wind tunnel test section, the volume fraction of water droplets (i.e., discrete phase particles) is very small, typically 10%. -6 For two-phase flows with such small volume fractions, the interactions between discrete phase particles can be neglected, and the influence of the discrete phase on the continuous phase flow field can also be ignored. The DPM model in Fluent is used to simulate this sparse two-phase flow with a volume fraction less than 10%. The DPM model first calculates the continuous phase as a single-phase flow, and then numerically solves for the discrete phase in Laplace coordinates based on the equations of motion of the discrete phase. It can consider the heat and mass transfer processes of the discrete phase, the coupling between the two phases, and the influence of the coupling results on the discrete phase trajectory and the continuous phase flow.

[0130] The motion of a water droplet depends on the magnitude and direction of the forces acting on it. For a water droplet suspended in an airflow, the forces acting on it include: aerodynamic drag, gravity and buoyancy, added mass force, pressure gradient force, Basset force, etc., as well as Magnus force, Saffman force, thermophoretic force, and Brownian force, electrophoretic force, and photophoretic force that may appear in non-uniform flow fields, and Brownian force, electrophoretic force, and photophoretic force that appear on submicron particles.

[0131] For moving water droplets, aerodynamic drag must be considered; the effect of gravity must also be considered in the computational domain of the test section, and buoyancy, as a volume force, is analyzed together with the effect of gravity; since the initial velocity of the atomized water droplets ejected from the nozzle decays rapidly, the additional mass force reaches 10. -4The mass force is on the order of N, so the added mass force cannot be ignored. Due to the large pressure gradient in the test section, the influence of the pressure gradient force needs to be considered. The Basset force is a non-constant aerodynamic force generated when there is relative acceleration between particles and fluid in a two-phase flow. Studies have shown that the influence of the Basset force can be ignored for the motion of liquid particles in a gas. The Magnus force is a drag generated by non-spherical particles having angular velocity. Laser holographic studies show that in most regions of the flow field, particles are constrained by fluid viscosity and have no rotational effect. Therefore, the influence of the Magnus force can be ignored. The Saffman force is caused by the transverse velocity gradient (shear layer flow) and has a significant effect only on small water droplets of 1μm to 10μm. Therefore, its effect is only considered when examining the boundary layer effect and is not included in the water droplet motion control equation. The temperature gradient in the flow field of the test section will introduce thermophoretic force, so the influence of thermophoretic force needs to be considered. Since the diameter of the atomized water droplets in the test section is not within the mesoscopic scale, the influence of the Brownian force can be ignored. Under the absence of strong external field influence, other forces such as electrophoresis and photophoresis are also not considered.

[0132] Therefore, considering the effects of aerodynamic drag, gravity and buoyancy, additional mass force, pressure gradient force, and thermophoretic force on the motion of the water droplet, the governing equation for the motion of the water droplet is:

[0133] ;

[0134] in: Let be the acceleration vector of the water droplet. The mass of the water droplet is (kg). Let be the position vector of the water droplet. The dynamic viscosity of the gas is (Pa·s). The diameter of the water droplet is in meters (m). The dynamic drag coefficient of the water droplet. For relative Reynolds number, The velocity vector of the airflow. Let be the velocity vector of the water droplet. The velocity of the water droplet is denoted as m / s. The volume of the water droplet (m 3 ), Water droplet density (kg / m³) 3 ), Gas density (kg / m³) 3 ), It is the gravitational acceleration vector. This is the virtual mass force coefficient (with a value of 0.5). For airflow pressure gradient, For airflow temperature gradient, Thermophoretic force coefficient, Let the gas temperature be K. Rewrite the above equation in a form that does not show the diameter of the water-containing droplet:

[0135] ;

[0136] in, The drag coefficient is calculated using the following formula:

[0137] ;

[0138] Relative Reynolds Number The calculation formula is:

[0139] ;

[0140] in, , The speeds of the airflow and water droplets are respectively (m / s);

[0141] Dynamic drag coefficient of water droplets It is a key factor in accurately simulating the generation and dispersion process of clouds and fog within the test section. Since the physical properties of supercooled water droplets are difficult to determine, summarizing an empirical correlation for the dynamic drag coefficient of water droplets is very challenging. Drawing on methods used by some scholars to calculate the droplet drag coefficient, this embodiment uses linear interpolation between the drag coefficients of spherical and disk-shaped water droplets based on the droplet deformation parameters. Specifically, it combines the drag coefficients of spherical and disk-shaped water droplets (the drag coefficient of disk-shaped water droplets at high Reynolds numbers is taken as 1.54), using the relative deformation of the water droplets as a variable for linear interpolation, to obtain the dynamic drag coefficient of the water droplets. The calculation method is as follows:

[0142] ;

[0143] in, The drag coefficient of a spherical water droplet. This is the linear interpolation result corresponding to the relative deformation of the water droplet;

[0144] in,

[0145] ;

[0146] in, Let the Reynolds number be the number of the water droplet. The Reynolds number is used as a reference to determine the state of the flow.

[0147] Near the wall, consider the Saffman force of the shear layer. Its expression is:

[0148] ;

[0149] in, The lift coefficient of a spherical water droplet in an ideal fluid shear flow is taken as 0.5;

[0150] To account for the influence of turbulence on the droplet trajectory, a Discrete Random Walk (DRW) model is used to describe the effects of turbulence in the flow field. In the DRW model, it is assumed that the magnitude of the fluctuating velocity of the fluid is a piecewise constant function of time. Within the characteristic survival time interval of the fluid vortex, the magnitude of the fluctuating velocity remains constant. Then the instantaneous velocity of the airflow in the water droplet trajectory equation is The turbulent diffusivity of water droplets .in, The average airflow velocity , These are the airflow pulsation velocity vectors. The i-th and j-th components, The integral time scale is used. This takes into account the turbulent diffusion of water droplets. By calculating the trajectories of a large number of water droplets using this method, the random influence of turbulence on the water droplets can be reflected.

[0151] The calculation of water droplet turbulent diffusion uses an integral time scale. The concept, This represents the time it takes for the water droplet's trajectory to be in a state of turbulent motion:

[0152] ;

[0153] in, The selected reference time, Indicates the trajectory of the same water droplet, relative to time. Lag time At that moment, The magnitude of the water droplet pulsation velocity (m / s) This represents the mean square value of the water droplet pulsation velocity;

[0154] In the calculation of actual water droplet turbulent diffusion, Taken as the vortex survival time T L Time of water droplets passing through vortices T cross Minimum value:

[0155] ;

[0156] Where τ is the characteristic relaxation time of the water droplet (s). Scale for vortex length (m);

[0157] Magnitude of airflow fluctuations in turbulent vortices Given by the turbulent kinetic energy k:

[0158] ;

[0159] in: This represents the mean square value of the airflow pulsation velocity. These are random numbers that follow a normal distribution.

[0160] Preferably, the aero-engine performance calculation proxy model module is specifically used for:

[0161] Based on mathematical models or experimental data of aero-engines, a proxy model for aero-engine performance calculation is constructed using surrogate modeling technology. This model is used to calculate the inlet and outlet parameters of the aero-engine in real time. The inputs to the proxy model include parameters such as the aero-engine's total intake pressure, total intake temperature, exhaust static pressure, and engine speed or fuel flow rate. The outputs are the engine's inlet and outlet flow rates, temperature, and pressure. The construction process of the aero-engine performance calculation proxy model is described in [link to documentation]. Figure 4 ;

[0162] For the selected aero-engine test object, firstly, sample data of surrogate models are collected through aero-engine whole-machine test or engine mathematical model. Then, multiple candidate surrogate models are established using surrogate model modeling methods, including polynomial response surface method, radial basis function method, Kriging model method, and orthogonal polynomial method.

[0163] The optimal surrogate model is selected as the surrogate model for engine performance calculation by means of surrogate model evaluation indicators. Preferably, the surrogate model evaluation indicators include the coefficient of determination, root mean square error, and maximum absolute error.

[0164] When the test section flow field simulation system specifies the engine intake and exhaust environment parameters (total intake temperature, total pressure, and static exhaust pressure) and engine speed or fuel flow rate for the aero-engine test object, the engine inlet and outlet flow rates, temperature, and pressure are predicted within milliseconds by the aero-engine performance calculation proxy model. This is used to update the aerodynamic interface parameters of the intake and exhaust devices of the aero-engine test object and the rear compartment of the test cabin in real time during the test section CFD simulation, thus solving the problem of balancing the calculation accuracy and calculation efficiency in the test section CFD simulation.

[0165] Preferably, the coupling process between the aero-engine performance calculation proxy model and the CFD solver includes:

[0166] The program initializes the data and starts the coupled module;

[0167] During the iteration process, the coupling module establishes connections and neighborhood search links between code, each code handles its own computational tasks, and exchanges data at predetermined times;

[0168] Finally, disconnect the code and stop all calculations;

[0169] In this process, while the CFD simulation solver solves the airflow field, the simulation module considers multiphase flow and phase change processes such as the movement of water droplets, deposition, and ice formation. It uses the Lagrange discrete phase model to solve the force and motion equations of atomized water droplets in the airflow field and simulates their transport process.

[0170] Preferably, the iterative optimization module is specifically used for:

[0171] The aerodynamic layout parameters of the test section are automatically adjusted based on the simulation results. The iterative optimization objective is to improve the quality of the inlet flow field of the aero-engine. The constraints include geometric size restrictions and operating condition requirements. The geometric size restrictions include the size of the free jet nozzle, the size of the exhaust diffuser, and the axial distance between the spray device and the aero-engine. The operating condition requirements include the area and location of the core area of ​​the cloud and fog field. Specifically, improving the quality of the inlet flow field of the aero-engine involves reducing the non-uniformity of the pressure field, temperature field, velocity field, and cloud and fog field of the aero-engine inlet flow field, ensuring that they are less than or equal to the set values.

[0172] When there are multiple optimization objectives, the weighted aggregation method or Pareto front method is used to solve them. For weighted objectives, including simultaneously reducing the non-uniformity of the engine inlet pressure field and the non-uniformity of the temperature field, a comprehensive objective is constructed, the weights are adjusted to reflect the emphasis, and then a single-objective optimization method is used to solve it.

[0173] The iterative optimization module, combined with the visualization and data analysis module, allows users to intuitively understand the optimization process. Specifically, it plots the optimal value and population diversity index curves during the genetic algorithm iteration process in real time to monitor the convergence trend. After optimization, users can view the sensitivity analysis results of the optimization objective to each design parameter, and analyze the impact of the design parameters on the optimization objective through surrogate models or response surface methodology, thereby obtaining engineering design guidance.

[0174] Through the above process, the entire system achieves an automated closed-loop process for the ice tunnel test section of aero-engines, from geometric parametric modeling to flow field simulation analysis and aerodynamic layout optimization, providing technical support for obtaining a test section modification design scheme that meets the test requirements. This intelligent optimization based on simulation data significantly improves the intelligence level and quality of the ice tunnel test section modification design.

[0175] Preferably, the visualization and data analysis module is developed based on a B / S architecture, which displays the geometric model, mesh, intermediate state and final results of the flow field simulation of the test section in real time on the browser, including cloud maps of parameters such as pressure, temperature, velocity, and liquid water content, as well as droplet size distribution, so that users can analyze the simulation results and evaluate the optimized design scheme of the test section.

[0176] In summary, based on the overall architecture of the integrated modeling, simulation, and optimization platform for aero-engine icing wind tunnel test sections, this application integrates the geometric parametric modeling, flow field simulation analysis, and aerodynamic layout optimization tools required for the modification design of aero-engine icing wind tunnel test sections, constructing an integrated modeling, simulation, and optimization platform for aero-engine icing wind tunnel test sections. The workflow of this platform is described in [link to documentation]. Figure 5 .

[0177] This application addresses the shortcomings of existing technologies by providing the following two key innovations to achieve an automated closed loop from geometric parameterization modeling to flow field simulation analysis and aerodynamic layout optimization in the ice wind tunnel test section of an aero-engine.

[0178] (1) Integrated parametric modeling and simulation: An integrated simulation method is proposed that deeply couples geometric parametric modeling, aero-engine performance calculation proxy model and CFD numerical simulation. This simulation process can realize the automatic generation of the geometric model of the test section according to the parameters, and embed the aero-engine performance calculation proxy model into the flow field simulation of the test section, realizing the real-time bidirectional coupling between the working condition of the aero-engine test object and the flow field of the test section, which greatly improves the automation level and accuracy of the simulation calculation of the ice wind tunnel test section.

[0179] (2) Automatic optimization strategy based on simulation feedback: An automatic optimization strategy is proposed that uses the results of flow field simulation to directly drive design optimization. This strategy automatically adjusts the key geometric parameters of the test section based on the flow field uniformity and other indicators obtained from the simulation through the built-in optimization algorithm. It realizes the closed-loop automatic iteration of the aerodynamic layout of the test section from simulation analysis to parameter optimization, and can quickly obtain the ice wind tunnel test section modification design scheme that meets the requirements of aero-engine icing test.

[0180] This application proposes an integrated platform for modeling, simulation, and optimization of ice tunnel test sections for aero-engines. This platform organically couples geometric parametric modeling, CFD simulation, aero-engine performance calculation, and multidisciplinary optimization techniques to achieve an automated closed-loop process for ice tunnel test sections, from geometric modeling to flow field simulation analysis and aerodynamic layout optimization. On one hand, this application deeply integrates geometric parametric modeling and flow field simulation, overcoming the problems of independent geometric modeling and CFD simulation, complex data interfaces, and numerous repetitive operations in traditional design processes. On the other hand, this application introduces a proxy model for aero-engine performance calculation embedded in flow field simulation, achieving real-time bidirectional coupling between the operating conditions of the aero-engine test object and the flow field of the test section. This overcomes the shortcomings of existing methods, such as idealized boundary conditions and simulation results deviating from real operating conditions due to a lack of coupling. Through the integration of these technologies, this application significantly improves the automation level of geometric modeling and flow field simulation in ice tunnel test sections, enhancing the efficiency and physical realism of simulation calculations.

[0181] This application introduces an integrated and automated design process for the modification of aero-engine icing wind tunnel test sections, achieving a closed-loop operation of the entire process from modeling and simulation to optimization. Starting with the input of geometric parameters, the system automatically generates a 3D model, generates meshes, and sets boundary conditions. It embeds engine performance calculations into the CFD solution process using a proxy model, performs simulation calculations and result analysis of the flow field in the test section, and automatically adjusts the geometric parameters of the test section based on the simulation output to perform optimization iterations. This integrated closed-loop solution allows the modification design process of the test section to be carried out continuously on the same platform without human intervention in multiple software / tools, significantly reducing data transmission and repeated trial and error. Using the technical solution of this application, a test section modification design scheme that meets the requirements of aero-engine icing tests can be obtained efficiently, shortening the icing test cycle. At the same time, the simulation conditions more closely resemble the aero-engine icing test conditions, improving the reliability and accuracy of the simulation results. This provides a basis for the optimized design of aero-engine icing wind tunnel test sections and engine icing protection, significantly reducing the risks and costs of aero-engine icing tests.

[0182] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0183] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. An integrated platform for modeling, simulation, and optimization of an aero-engine ice wind tunnel test section, characterized in that, include: The geometric parametric modeling module is used to obtain the initial geometric parameters required for the modification design of the ice wind tunnel test section. By setting control parameters and geometric relationships, it automatically generates the geometric model of the test section. The initial geometric parameters include the size and initial layout values ​​of each component of the test section. The mesh generation and boundary condition setting module is used to discretize the computational domain of the test section, generate a three-dimensional mesh of the flow region inside the test section, set the boundary conditions of the computational domain of the test section, and select the flow field model and governing equations. The aero-engine performance calculation proxy model module is used to construct an aero-engine performance calculation proxy model, calculate the inlet and outlet parameters of the aero-engine in real time, and bidirectionally couple with the test section flow field simulation system. The CFD simulation module is used to launch the CFD simulation solver and perform time-progressed numerical solutions for the flow field inside the test section. During the solution process, the aero-engine performance calculation proxy model and the CFD solver are tightly integrated in a bidirectional coupling manner. They interact in real time through a mechanism that exchanges data before the start of each iteration step. Specifically, the aero-engine performance calculation proxy model module is used for: Based on mathematical models or experimental data of aero-engines, a proxy model for calculating aero-engine performance is constructed using proxy modeling technology. This model is used to calculate the inlet and outlet parameters of aero-engines in real time. The inputs of the proxy model are the total intake pressure, total intake temperature, static pressure of the exhaust environment, and speed or fuel flow parameters of the aero-engine. The outputs are the inlet and outlet flow rates, temperature, and pressure of the engine. For the selected aero-engine test object, firstly, sample data of surrogate models are collected through aero-engine whole-machine test or engine mathematical model, and then multiple candidate surrogate models are established using surrogate model modeling method; The optimal surrogate model is selected as the surrogate model for engine performance calculation by using surrogate model evaluation indicators; When the test section flow field simulation system specifies the engine intake and exhaust environment parameters and engine speed or fuel flow rate for the aero-engine test object, the engine inlet and outlet flow rates, temperature and pressure are predicted in milliseconds through the aero-engine performance calculation proxy model. This is used to update the aerodynamic interface parameters of the aero-engine test object's intake and exhaust devices and the rear compartment of the test cabin in real time during the test section CFD simulation. The iterative optimization module is used to automatically adjust the aerodynamic layout parameters of the test section based on the genetic algorithm and simulation results. The goal of the iterative optimization is to improve the quality of the inlet flow field of the aero-engine test component, and the constraints are geometric size limitations and operating condition requirements. The visualization and data analysis module is used for monitoring the simulation process and visualizing the results, including real-time display of the geometric model of the test section, the mesh, and the intermediate and final results of the flow field simulation.

2. The integrated platform for modeling, simulation, and optimization of the air-to-air engine ice tunnel test section according to claim 1, characterized in that, The initial dimensions and layout of each component in the test section include the rear compartment of the test cabin, corner section, engine air intake duct, free jet nozzle, test bench, test platform, exhaust diffuser, aero-engine outline, and the dimensions and relative positions of the drive shaft. When generating the geometric model of the test section, spline curves or NURBS surfaces are used to describe the surface profile of the engine air intake duct and free jet nozzle components. After automatically generating the geometric model of the test section, based on the preset parameters, the automated tool is called to pass the geometric model of the test section to the mesh generation and boundary condition setting module.

3. The integrated platform for modeling, simulation, and optimization of the aero-engine ice tunnel test section according to claim 2, characterized in that, When dividing the three-dimensional mesh of the internal flow region of the test section, the mesh type is selected as structured, unstructured or hybrid mesh. The rear compartment of the test chamber uses a structured mesh to ensure the orderly distribution of the mesh, while complex components use unstructured mesh. For key areas including the free jet nozzle outlet, the aero-engine inlet and outlet, and the exhaust diffuser inlet, local densification or unstructured mesh is used to capture the flow details of the airflow field and cloud field. The mesh quality is ensured by optimizing the unit distortion, orthogonality and volume distribution index. When setting the boundary conditions of the computational domain of the test section, the inlet boundary of the test section is set as a pressure inlet or a mass flow inlet, and the outlet boundary is set as a pressure outlet; the solid surface inside the test section is set as a wall boundary condition; the initial parameters of the atomized water droplets need to be defined at the outlet of the spray device, and a large number of simulated particles are released at the nozzle position through a discrete phase model to represent the atomized water droplets.

4. The integrated platform for modeling, simulation, and optimization of the aero-engine ice tunnel test section according to claim 3, characterized in that, When selecting the flow field model and governing equations, a numerical model is constructed within the Eulerian-Lagrange framework by coupling the discrete phase model and the component transport model. This model is used to numerically simulate the motion, evaporation, and heat and mass transfer processes of atomized water droplets in cold air. The governing equations for the experimental flow field of the numerical model include: Continuity equation: ; in, For density, For time, For flow rate; Momentum equation: ; in, Spatial coordinate vector The i-th component, i=(1,2,3), For the first Velocity vector components in each direction, For pressure, For dynamic viscosity, For the first Volume force vector components in each direction; Energy equation: ; in, For temperature, Thermal conductivity, For specific heat capacity, As an internal heat source; Component equation: ; in, For matter concentration, For matter diffusion coefficient, For matter The source item; Equations of state for each component: , ; in, , Let be the densities of components a and b in the mixture, respectively. The partial pressures of components a and b are respectively. It is the ideal gas constant; Pressure division equation: ; The turbulence model uses the RNG k-ε model, and its governing equations are: ； ； in, For turbulent kinetic energy, For turbulent dissipation rate, Let be the model constants of equation k. For turbulent viscosity, For effective viscosity, For strain rate, For the model constants of the generated terms, for The model constants of the equation, For the dissipation term, the model constants are... Additional terms for analyzing flow problems under rapid strain; ； in, , airflow velocity vector The i-th and j-th components, Spatial coordinate vector The i-th and j-th components; Considering the effects of aerodynamic drag, gravity and buoyancy, additional mass force, pressure gradient force, and thermophoretic force on the motion of the water droplet, the governing equation for the motion of the water droplet is: ； in: Let be the acceleration vector of the water droplet. For the mass of the water droplet, Let be the position vector of the water droplet. The dynamic viscosity of the gas. The diameter of the water droplet. The dynamic drag coefficient of the water droplet. For relative Reynolds number, The velocity vector of the airflow. Let be the velocity vector of the water droplet. The speed of the water droplets Let the volume of the water droplet be... For the density of water droplets, For gas density, It is the gravitational acceleration vector. For virtual mass force coefficients, For airflow pressure gradient, For airflow temperature gradient, Thermophoretic force coefficient, Let the gas temperature be the equation, and we can rewrite the above equation in a form that does not show the diameter of the water-containing droplet: ； in, The drag coefficient is calculated using the following formula: ； Relative Reynolds Number The calculation formula is: ； in, , These represent the speeds of the airflow and the water droplets, respectively. By combining the drag coefficients of spherical and disc-shaped water droplets and performing linear interpolation with the relative deformation of the water droplets as the variable, the dynamic drag coefficient of the water droplets is obtained. The calculation method is as follows: ； in, The drag coefficient of a spherical water droplet. This is the linear interpolation result corresponding to the relative deformation of the water droplet; in, ； in, Let the Reynolds number be the number of the water droplet. The Reynolds number is used as a reference to determine the state of the flow. Near the wall, consider the Saffman force of the shear layer. Its expression is: ； in, The lift coefficient of a spherical water droplet in an ideal fluid shear flow is taken as 0.5; A random walk model is used to describe the effect of turbulence in the flow field, and the turbulent diffusivity of the water droplets is... ,in, , These are the airflow pulsation velocity vectors. The i-th and j-th components, The integration time scale; In the calculation of turbulent diffusion of water droplets, Taken as the vortex survival time T L Time of water droplets passing through vortices T cross Minimum value: ； Where τ is the characteristic relaxation time of the water droplet. A scale for vortex length; airflow pulsation velocity in turbulent vortex Given by the turbulent kinetic energy k: ； in: This represents the mean square value of the airflow pulsation velocity. These are random numbers that follow a normal distribution.

5. The integrated platform for modeling, simulation, and optimization of the aero-engine ice tunnel test section according to claim 1, characterized in that, The proxy modeling methods include polynomial response surface method, radial basis function method, Kriging model method, and orthogonal polynomial method.

6. The integrated platform for modeling, simulation, and optimization of the aero-engine ice tunnel test section according to claim 1, characterized in that, The evaluation metrics for the surrogate model include the coefficient of determination, root mean square error, and maximum absolute error.

7. The integrated platform for modeling, simulation, and optimization of the air-to-air engine ice tunnel test section according to claim 1, characterized in that, The coupling process between the surrogate model for aero-engine performance calculation and the CFD solver includes: The program initializes the data and starts the coupled module; During the iteration process, the coupling module establishes connections and neighborhood search links between code, each code handles its own computational tasks, and exchanges data at predetermined times; Finally, disconnect the code and stop all calculations; In this process, while the CFD simulation solver solves the airflow field, the simulation module considers the motion, deposition, and ice formation of water droplets, as well as the multiphase flow and phase change process. It uses the Lagrange discrete phase model to solve the force and motion equations of atomized water droplets in the airflow field and simulates their transport process.

8. The integrated platform for modeling, simulation, and optimization of the air-to-air engine ice tunnel test section according to claim 1, characterized in that, The iterative optimization module is specifically used for: The aerodynamic layout parameters of the test section are automatically adjusted based on the simulation results. The iterative optimization objective is to improve the quality of the inlet flow field of the aero-engine. The constraints are geometric size restrictions and operating condition requirements. The geometric size restrictions include the size of the free jet nozzle, the size of the exhaust diffuser, and the axial distance between the spray device and the aero-engine. The operating condition requirements include the area and location of the core area of ​​the cloud field. Specifically, improving the quality of the inlet flow field of the aero-engine involves reducing the non-uniformity of the pressure field, temperature field, velocity field, and cloud field of the aero-engine inlet flow field, ensuring that they are less than or equal to the set values. When there are multiple optimization objectives, the weighted aggregation method or Pareto front method is used to solve them. For weighted objectives, including simultaneously reducing the non-uniformity of the engine inlet pressure field and the non-uniformity of the temperature field, a comprehensive objective is constructed, the weights are adjusted to reflect the emphasis, and then a single-objective optimization method is used to solve it. The system can plot the optimal value and population diversity index curves during the iteration process of the genetic algorithm in real time to monitor the convergence trend. After optimization, users can view the sensitivity analysis results of the optimization objective to each design parameter. The system can analyze the impact of the design parameters on the optimization objective through surrogate models or response surface methodology, thereby obtaining engineering design guidance.

9. The integrated platform for modeling, simulation, and optimization of the air-to-air engine ice tunnel test section according to claim 1, characterized in that, The visualization and data analysis module is developed based on a B / S architecture. It displays the geometric model, mesh, intermediate state and final results of the flow field simulation of the test section in real time on the browser, including cloud maps of pressure, temperature, velocity, liquid water content parameters and droplet size distribution, so that users can analyze the simulation results and evaluate the optimized design scheme of the test section.

Images