A method for aerodynamic design of a cooling fan driven by a vortex

Abstract

This invention discloses an aerodynamic design method for a vortex-driven cooling fan. First, initial parameters are given according to design requirements, and the rotor outlet tangential velocity vortex is adaptively selected. Based on radial balance and axial flow assumptions, the axial velocity distribution along the blade height is solved. The actual total pressure ratio is calculated using the Euler turbine equation and isentropic relations. The work coefficient and flow coefficient are iteratively updated through relaxation factors until the pressure ratio and flow rate converge. The rotor and guide vane geometric angles are calculated using velocity triangles and empirical blade models. After aerodynamic constraint checks, a multi-parameter mapping configuration relationship for vortex-driven cooling fans is established, the accumulation law is automatically selected, a meridional channel is constructed, and the geometric parameters are parsed into three-dimensional blade coordinates, achieving rapid conversion from one-dimensional design to three-dimensional modeling. This method couples radial balance, vortex design, closed-loop iteration, and empirical correction, achieving fully automatic and fast convergence. It is suitable for axial flow fans under wide operating conditions, especially for high-flow-rate, high-load cooling fans, and has broad applicability.

Description

Technical Field

[0001] This invention belongs to the field of turbomachinery aerodynamic design technology, specifically relating to an aerodynamic design method for a vortex-driven cooling fan. Background Technology

[0002] As a key heat dissipation component in systems such as internal combustion engines, generators, and intelligent power units, the aerodynamic design efficiency of cooling fans determines the development cycle of the heat dissipation system and the operational reliability of the entire machine.

[0003] Existing aerodynamic design technologies for vortex-driven cooling fans mainly fall into two categories: one is an aerodynamic design method based on velocity triangles, which starts from aerodynamic requirements such as flow rate and pressure ratio, and determines blade geometric parameters through one-dimensional / two-dimensional analysis; the other is an optimization design method centered on CFD flow field calculations and optimization algorithms, which automatically seeks optimization based on parametric geometric models. The former has clear physical meaning and high computational efficiency, making it a classic approach in engineering design; the latter does not require a pre-defined detailed velocity distribution, allowing for exploration of a larger design space, but significantly increases computational costs and relies on existing models. For aerodynamic design methods, after the one-dimensional aerodynamic analysis, subsequent steps still heavily rely on human experience and lack a direct connection with three-dimensional geometry. Parameters such as geometric angles and chord lengths output from one-dimensional analysis need to be manually transcribed into three-dimensional modeling software. Furthermore, there is a lack of feedback correction mechanisms between the one-dimensional empirical model and the three-dimensional CFD verification in the aforementioned steps. While optimization design methods can avoid some manual steps, they come at the cost of repeatedly solving CFD forward problems, resulting in long calculation cycles, sensitivity to initial values, and optimization results that often deviate from aerodynamic constraints such as radial balance. This leads to insufficient engineering controllability and makes it difficult to meet the high-load, short-cycle development requirements of cooling fans.

[0004] In summary, existing aerodynamic design methods lack effective integration in areas such as coordinate generation after one-dimensional analysis, one-dimensional to three-dimensional geometric transformation, and deviation correction, thus failing to directly support the rapid development of cooling fans. Therefore, this patent applies aerodynamic design theory to establish a mechanism for the configuration of vortex and blade stacking patterns, proposing a coupled one-dimensional and three-dimensional aerodynamic design method. This method is of great significance for shortening the design cycle of cooling fans and improving engineering reliability. Summary of the Invention

[0005] To address the technical problems mentioned above, this invention provides an aerodynamic design method for a vortex-driven cooling fan, comprising the following steps: S1. Generate a one-dimensional aerodynamic layout and initial parameters based on the design input and vortex type selection; S2. Based on the one-dimensional aerodynamic layout and radial balance equation, generate a convergent rotor inlet and outlet flow field; S3. Based on the empirical correlation between the converged rotor inlet and outlet flow fields and the blade cascade, generate the two-dimensional geometric angles of the rotor and guide vanes; S4. Based on the two-dimensional geometric angles of the guide vane and the aerodynamic constraints, generate blade cross-section parameters that satisfy the constraints; S5. Generate a three-dimensional blade geometry file based on the blade cross-section parameters and three-dimensional stacking control; S6. Based on the three-dimensional blade geometry file and CFD diagnostics, generate the final design after closed-loop correction.

[0006] Preferably, S1 specifically includes: Define dimensionless design parameters based on the initialization parameters input by the design, and select the vortex type based on the target total pressure ratio and hub ratio; Based on the vortex type, the target aspect ratio initial value, superposition method, and constraint limit set are automatically determined through a preset mapping relationship. Determine whether guide vanes are needed based on the target pressure ratio and the preset pressure ratio threshold, and determine whether to use a single rotor or dual-channel layout. The number of blades for the rotor and guide vanes is estimated based on the current geometry and the target aspect ratio.

[0007] Preferably, S2 specifically includes: Based on the selected vortex type, specify the rotor outlet tangential velocity distribution and establish the radial equilibrium equation; An iterative method was used to solve for the axial velocity at the hub, and the axial velocity distribution along the blade height was obtained by numerical integration. Adjust the axial speed of the wheel hub by comparing the integral of the mass flow rate with the target value until the mass conservation is satisfied; The Euler work and pressure ratio are calculated based on the converged velocity triangle, and the flow coefficient and work coefficient are updated using a relaxation factor. The iteration continues until the pressure ratio error and mass flow rate error are both less than the preset tolerance.

[0008] Preferably, S3 specifically includes: The relative airflow angle and bending angle of each section are calculated based on the converged velocity triangle, and the consistency and chord length are determined by the empirical correlation of the blade cascade. The angle of attack and lag angle models of the Lieblein zero-curvature blade cascade were used to determine the angle of attack and lag angle of each cross section. Calculate the rotor inlet and outlet axial installation angles based on the angle of attack and the angle of retreat; When determining the need for guide vanes, the Howell model is used to calculate the guide vane lag angle and the guide vane inlet and outlet geometric installation angles based on the airflow angles at the guide vane inlet and outlet.

[0009] Preferably, S4 specifically includes: The set of constraint limits is dynamically adjusted according to the selected vortex type, and multi-dimensional aerodynamic constraint checks are performed on the rotor and guide vanes. Constraint checks include constraints on equivalent diffusion factor, counterforce, relative Mach number, De Haller number, aspect ratio, and blade geometry. If any constraint fails, adjust the corresponding section parameters and regenerate until all constraints are satisfied; Output the chord length, mounting angle, and bending angle parameters of representative radial sections of the rotor and guide vanes.

[0010] Preferably, S5 specifically includes: Based on the stacking method corresponding to the selected vortex type, the rotor is stacked in three dimensions to determine the axial coordinates of the leading and trailing edges of each section; When it is determined that guide vanes are needed, the guide vanes are stacked in three dimensions; Perform a tip clearance loss check; if it fails, reselect the stacking parameters. The middle arc is generated using a parabola, and the thickness distribution is generated based on the NACA65 series reference thickness distribution. The semi-thickness distribution is offset along the normal of the mid-arc line to generate the coordinates of the suction surface and the pressure surface; Rotate the local coordinates to the global axial-circumferential plane by the tilt angle, and export the 3D blade geometry file; Verify the consistency between the leading edge axial position and geometric angles in the exported 3D geometry file and the 1D design output.

[0011] Preferably, S6 specifically includes: Import the 3D blade geometry file into simulation software for mesh generation and numerical calculation to obtain 3D flow field information; In the post-processing of 3D CFD, total pressure loss, secondary flow loss and tip leakage loss are extracted, and a loss correction factor is established. The correction factor is fed back to the empirical correction factor and variable efficiency in the one-dimensional input parameters to reconstruct the two-dimensional cross section and three-dimensional geometry, and the closed-loop iteration is repeated. The final design is output only when the deviation between the three-dimensional calculation results and the one-dimensional aerodynamic target falls within the preset margin range.

[0012] Preferably, the selection of the vortex type is determined based on the target total pressure ratio, hub ratio, and blade tip circumferential speed, specifically including: When the target total pressure ratio is less than 1.15 and the hub ratio is greater than or equal to 0.55, select free vortex; When the target total pressure ratio is greater than or equal to 1.8 or the hub ratio is greater than 0.75, select the vortex with equal reaction degree; When the blade tip circumferential velocity is greater than 450 m / s and the target total pressure ratio is greater than 1.5, select the isovorticity; If none of the above conditions are met, choose the exponential vortex.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention establishes a mapping relationship between vortex type and associated parameters, automatically adjusting the target aspect ratio, stacking method, and constraint limits by employing different vortex types when target requirements change. During the iterative solution process, the number of blades, work coefficient, and flow coefficient are automatically updated by simultaneously matching the target pressure ratio and mass flow rate, ensuring design accuracy and convergence stability. Furthermore, by introducing aerodynamic constraint limits, constraint checks are made more closely aligned with actual flow field characteristics, improving design safety and reliability.

[0014] This invention reduces human error by establishing a conversion method between one-dimensional aerodynamic design parameters and three-dimensional coordinates, ensuring a high degree of consistency between the one-dimensional design objective and the three-dimensional geometry. By feeding back the loss decomposition of the three-dimensional simulated flow and the lag angle correction factor to the one-dimensional design, the inherent deviation between the empirical model and the actual flow field is gradually converged, further improving design accuracy. The entire design process is fully automated and converges quickly, making it suitable for rapid aerodynamic design of cooling fans with a wide operating range. It can also flexibly adapt to different hub ratios, load levels, manufacturing constraints, and three-dimensional modeling formats through various modifications, exhibiting wide applicability and scalability. Attached Figure Description

[0015] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating the overall design method of the present invention. Figure 2 Logic diagram for selecting vortex type; Figure 3 The diagrams show the distribution of inlet and outlet velocity components along the blade height in Example 2; (a) is a schematic diagram of the distribution of inlet and outlet velocity components along the blade height in Example 3. Figure 4 This is a schematic diagram showing the convergence of pressure ratio error and work coefficient with the number of iterations; where (a) is the pressure ratio-iteration number curve; and (b) is the work coefficient-iteration number curve. Figure 5 The diagram shows the parameterized construction of the blade cross section; where (a) is the blade cross section shape diagram of Example 2, (b) is the half-thickness distribution diagram along the chord direction of Example 2, (c) is the blade cross section shape diagram of Example 3, and (d) is the half-thickness distribution diagram along the chord direction of Example 3. Figure 6The diagram shows the meridional channel of the blade; where (a) is the meridional channel diagram of Example 2; and (b) is the meridional channel diagram of Example 3. Figure 7 The diagrams are three-dimensional geometric shapes; (a) is a three-dimensional geometric shape diagram of Example 2; and (b) is a three-dimensional geometric shape diagram of Example 3. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] Example 1: As attached Figure 1 The diagram shows the logical flow of the design method in this embodiment: First, the design input is determined and parameters are initialized based on system requirements; then, the vortex type is automatically selected and design constraints are configured in conjunction with a preset mapping relationship; next, it is determined whether guide vanes are needed and the layout of the rotor and stator vanes is determined; based on this, design parameters are set and the radial equilibrium equation is iteratively solved; after convergence, the geometric parameters of the rotor and guide vanes are calculated respectively; finally, three-dimensional stacking, geometric generation, and one-dimensional and three-dimensional closed-loop correction are performed. The specific implementation steps include: S1. Generate a one-dimensional aerodynamic layout and initial parameters based on the design input and vortex type selection.

[0020] This step includes design input initialization, dimensionless parameter initialization, vortex selection, guide vane determination and axial layout, and preliminary estimation of the number of blades.

[0021] S101, Given the design input parameter mass flow rate ṁ Rotation speed N Imported total pressure P 0 Imported total temperature T 0. Target total pressure ratio PR target Meanwhile, the hub radius is given. r hub Initial radius of the rim r shroud_init Blade parameters include relative thickness t / c Location of maximum curvature a / c.

[0022] S102. Define dimensionless design parameters: φ = C z / U Work coefficient ψ = ΔC θ / U ,in U = ωr , U The blade circumferential velocity, C z For axial velocity, ΔC θ The difference in tangential velocity between the rotor inlet and outlet. Subsonic axial flow stage. φ Typically, the value is taken as 0.45~0.65. ψ The initial value is derived from the target pressure ratio and the estimated variable efficiency and iteratively corrected. For low pressure ratio fans (target pressure ratio less than 1.15), the initial value of the power factor is 0.25; if the target pressure ratio is large, the initial value of the power factor is 0.40.

[0023] S103, Select Vortex Type: PR target <1.15 and wheel hub ratio r hub / r shroud ≥0.55, select free vortex. At this point, the fan power is small, and the distribution of free vortex reduces the tip reaction force, which is beneficial for controlling the relative Mach number at the tip; if PR target ≥1.8 or wheel hub ratio r hub / r shroud >0.75, select vortex with equal reaction degree; if the tip velocity is... U tip >450m / s and PR target If the aspect ratio is greater than 1.5, select the equal vortex; if none of the above conditions are met, i.e., under normal load conditions, select the exponential vortex as the default compromise. The initial value of the target aspect ratio, the stacking method of the rotor and guide vanes, and the set of constraint limits are automatically determined through the preset mapping relationship table (Table 1).

[0024] Table 1 .

[0025] S104. Determine if a guide vane is needed: The preset pressure ratio threshold is 1.15. PR target If ≤1.15, guide vanes are not needed; if PR targetA pressure ratio greater than 1.15 indicates the need for guide vanes. When the system has strict requirements for upstream and downstream pressures, guide vanes can be forcibly retained even if the pressure ratio is below the threshold; if the structure is constrained, guide vanes can be forcibly skipped. The single-rotor channel layout places the rotor's leading edge at the axial zero point, and the channel includes an inlet extension and an outlet extension, generating only the rotor's 3D geometry file. The dual-channel layout places the guide vanes in the negative axial region and the rotor in the positive axial region, with an axial gap between them, generating separate 3D geometry files for the guide vanes and the rotor.

[0026] S105. Preliminary estimate of blade count: Calculate the number of rotor and guide vane blades based on the current geometry and target aspect ratio. (Based on blade height...) Target chord length The average radius of the grid is determined by the average radius of the grid density. With the target chord length determined, that is The number of blades is estimated from the average radius and grid pitch. ,in, r shroud is the rim radius. r hub refers to the radius of the wheel hub. AR target is the aspect ratio of the target object and its average radius. The number of rotor blades is limited to between 12 and 30, and the number of guide vane blades is limited to between 10 and 25.

[0027] S2. Based on the one-dimensional aerodynamic layout and radial balance equation, generate a convergent rotor inlet and outlet flow field.

[0028] This step includes establishing the radial equilibrium equation, numerical integration and hub axial velocity iteration, determining the tangential velocity distribution at the vortex outlet, pressure ratio calculation and design parameter iteration, and convergence criteria.

[0029] S201. Establish the radial equilibrium equation: Specify the rotor outlet tangential velocity distribution based on the selected vortex type. C θ2 ( r Given the hub radius, initial rim radius, and rotor inlet tangential velocity distribution. C θ1 ( r For operating conditions without inlet guide vanes or with guide vanes in a zero pre-rotation state, C θ1 ( r C = 0; For operating conditions with guide vanes and inlet pre-swirl, C θ1 Determined by the airflow angle at the upstream guide vane exit. Ignoring streamline curvature, a simple radial equilibrium equation is established: in, r Represents the radial position coordinates along the blade height direction. C θ This indicates the tangential velocity.

[0030] When there is no pre-rotation at the rotor inlet C θ1 =0, its radial gradient is zero, and the equation simplifies to If there is an imported pre-spinning, then we obtain information about... The first-order ordinary differential equation.

[0031] S202, Numerical Integration and Hub Axial Velocity Iteration: Due to the hub... Cz The unknown value is solved using an iterative method. First, the initial values ​​of Cz,hub and the flow coefficient are estimated from the mass flow rate and the annular area. ,in , ,but The fourth-order Runge-Kutta method was used to obtain numerical integration from the hub to the rim. Cz ( r )distributed.

[0032] Average tip velocity , Then the average axial velocity The rim radius is iteratively corrected by the mass flow rate and the average axial velocity: Calculated using isentropic relationship T ( r ), P ( r ) and density ρ ( r ): Mass flow rate is calculated using the following integral expression: Comparison with target value ṁ target Error adjustment C z,hub : The mass flow rate is recalculated until the error is less than the preset tolerance. This process ensures that the inlet flow field satisfies mass conservation and radial balance.

[0033] S203. Determine the tangential velocity distribution at the vortex outlet: The rotor outlet tangential velocity distributions corresponding to the four main vortex types are as follows.

[0034] Free vortices satisfy circulation conservation, that is: , in, Let be the tangential velocity at the average radius, and const denote a constant. r mean represents the average radius. ω It represents angular velocity.

[0035] The reaction force R of the constant reaction vortex remains constant along the blade height, with no inlet pre-swirl. )hour, ,in The target counterforce is calculated. After the calculation is complete, the above distribution is normalized to its mean, so that its value at the average radius is equal to... .

[0036] Isovortices satisfy It is a constant, i.e., the rigid body rotational distribution: .

[0037] The tangential velocity of the exponential vortex is distributed radially according to a power function: , where the index n The vortex type configuration mapping table is determined based on the load level, typically... .

[0038] S204. Pressure Ratio Calculation and Design Parameter Iteration: Calculate Euler work based on the velocity triangle at the average radius. Based on the velocity triangle at the average radius, the Euler work is calculated as follows: in, This represents the inlet tangential velocity of the average radius section.

[0039] Calculate the isentropic outlet total temperature: in, Indicates the total temperature at the outlet. Indicates the total import temperature. This indicates the specific heat capacity at constant pressure.

[0040] Calculate the isentropic temperature rise using the Euler-Republican exit total temperature formula described above: The isentropic pressure ratio is: Where γ represents the specific heat ratio.

[0041] Considering flow losses, the actual pressure ratio is: in, η poly For variable efficiency.

[0042] Define pressure ratio error Flow error ,in For the target pressure ratio, The actual mass flow rate obtained from the aforementioned integral formula, The target mass flow rate is used. The flow coefficient and work coefficient are updated using a relaxation factor: in, k Indicates the number of iterations. The relaxation factor is set to 0.6–0.9; based on engineering experience, the upper limit of the single-step correction amplitude for the design power coefficient and flow coefficient is set. , They are respectively: , And set the variable range constraint interval: .

[0043] S205. Convergence Criterion: Repeat the above radial balance, outlet flow field calculation and design parameter iteration until the following conditions are met simultaneously: pressure ratio error and mass flow rate error are both less than the preset tolerance, and the number of blades is stable, then terminate the iteration.

[0044] S3. Based on the empirical correlation between the converged rotor inlet and outlet flow fields and the blade cascade, generate the two-dimensional geometric angles of the rotor and guide vanes.

[0045] S301. Calculate rotor blade geometric parameters: Calculate the relative airflow angle of each cross section based on the converged velocity triangle. β 1. β 2, , Leaf-shaped bend θ camber = β 2- β 1. The density is independently selected based on the vortex type and radial position using an empirical correlation formula for the blade cascade. The cascade pitch is calculated from the number of blades and the cross-sectional radius after convergence of the S2 iteration. Based on this, the chord length is determined. .

[0046] S302. Calculate the rotor inlet angle of attack: Use the Lieblein zero-angle blade cascade angle of attack model. Zero-angle reference angle of attack. Due to the relative airflow angle at the inlet β 1 With cascade consistency σ The empirical function is determined as follows: in, Thickness correction Superimposed dimensionless slope terms proportional to the curvature. mTypically, a value of 0.1 to 0.3 is used to obtain the angle of attack. S303. Calculate the rotor lag angle: Similarly, the Lieblein zero-curvature blade cascade lag angle model is used. Zero-curvature reference lag angle. Determined by an empirical function of the inlet relative airflow angle and consistency: exist Thickness correction term is added on the basis The slope term is proportional to the curvature, where the coefficient of the slope term is taken as... n =0.25, thus obtaining the backward angle. : Finally, the axial installation angles of the rotor inlet and outlet are obtained. and : .

[0047] S304. Calculate the guide vane geometry angle: This is performed only when S1 determines that a guide vane is needed. The guide vane is located upstream of the rotor, and its outlet airflow directly serves as the rotor inlet condition. The absolute airflow angle at the guide vane inlet is typically taken as the axial inflow angle. If there is a vortex upstream, then The absolute airflow angle α2 at the outlet is determined by the downstream rotor inlet demand; in this embodiment's free vortex and equal reaction degree vortex scheme, the rotor inlet adopts a no-pre-swirl design, therefore... α 2 = 0; if the specified inlet pre-spinning scheme is adopted, then Corner θ = α 2- α 1.

[0048] S305. Calculate the guide vane lag angle: using the Howell model. in, a / c The location of maximum curvature. δ 0，10 = 0.5°. The guide vane design point is typically based on a zero-angle-of-attack design, thus yielding the guide vane inlet and outlet geometric installation angles: in, Indicates the guide vane inlet installation angle. This indicates the guide vane outlet installation angle.

[0049] S4. Based on the two-dimensional geometric angles of the guide vane and the aerodynamic constraints, generate blade cross-sectional parameters that satisfy the constraints.

[0050] This step performs a multi-dimensional aerodynamic constraint check, and the set of constraint limits is dynamically adjusted according to the vortex type selected in S1. Different vortex types correspond to different upper limits for the equivalent diffusion factor, upper and lower limits for the reaction force, and upper limits for the Mach number.

[0051] Specifically, constraint checks include: (1) To avoid boundary layer separation, the equivalent diffusion factor Deq should satisfy: rotor Deq≤2.0, guide vane Deq≤1.8.

[0052] (2) To control the load distribution, the reaction force The blade root is greater than 0.1, the blade tip is not higher than the upper limit corresponding to the vortex type, the free vortex is less than 0.7, the vortex with equal reaction degree is less than 0.65, and the exponential vortex is less than 0.85.

[0053] (3) To avoid excessive shock wave loss, the relative Mach number Mrel at the rotor inlet should be ≤1.15.

[0054] (4) To prevent excessive diffusion, the De Haller number should satisfy: rotor W2 / W1≥0.68, guide vane C3 / C2≥0.68.

[0055] (5) Balancing structural strength and losses, aspect ratio .

[0056] (6) Ensure processing feasibility and throat area, and the absolute value of blade geometric angle |κ|≤75°.

[0057] If any of the above constraint checks fail, the corresponding section parameters are adjusted and regenerated until all constraints are satisfied; the blade section parameters such as chord length, installation angle, and bend angle of each representative radial section of the rotor and guide vane are output.

[0058] S5. Generate a three-dimensional blade geometry file based on the blade cross-section parameters and three-dimensional stacking control.

[0059] S501, Rotor Stacking: Based on the stacking method configuration corresponding to the selected vortex type, the rotor is stacked in three dimensions. The stacking method is automatically determined by the mapping table between the vortex type and the matching parameter configuration, including torsional stacking, radial stacking, or sweep-free and bend-free stacking. For each representative radial section, the average tilt angle of the blade chord is calculated based on the inlet and outlet geometry angles of that section. Based on the chord length and inclination angle of the cross section, calculate the axial offset from the leading edge reference axial position to the trailing edge, thereby determining the axial coordinates of the leading and trailing edges of each cross section.

[0060] If radial stacking is used, the axial coordinates of the leading edge at each radius remain constant, and the axial coordinates of the trailing edge are determined by the product of the superimposed chord length of the leading edge reference position and the cosine of the installation angle, so that both the leading edge and the trailing edge are located on a pure radial line.

[0061] If torsional stacking is used, the leading edge axial coordinates are allowed to undergo a small axial displacement with the radius to match the inlet airflow angle that varies along the blade height, while the trailing edge coordinates are still determined based on the chord length and the installation angle.

[0062] If a sweep-free and bend-free stacking method is adopted, the leading edge axial coordinates of all sections are explicitly forced to be the same constant, and the trailing edge coordinates are determined based on the chord length and installation angle to ensure that the blade has no sweep angle and no bend angle in the radial direction.

[0063] S502, Guide Vane Stacking: This is executed only when guide vanes are deemed necessary. The stacking method is determined based on the guide vane properties and vortex type configuration, and the axial offset of each section is calculated to ensure outlet uniformity. If the guide vane is an adjustable inlet guide vane, the stacking axis is set to the rotation axis during overall reinstallation.

[0064] S503. Tip clearance loss check: In rotor stacking, two proxy indicators are extracted to evaluate tip leakage loss: one is the relative tip clearance, which is the ratio of clearance height to blade height; the other is tip circulation drop, which is the amount of work capacity loss caused by leakage in the tip region. If the relative tip clearance exceeds the preset threshold, or the tip circulation drop exceeds the preset limit, the tip clearance loss constraint is determined to have failed, and the stacking parameters are reselected.

[0065] S504, 3D geometry generation: The arc is generated using a parabola, and its dimensionless form is as follows: in, The coordinates are dimensionless chordal coordinates.

[0066] When generating the thickness distribution, the NACA65 series reference thickness distribution is used as the benchmark, and the distribution is scaled according to the relative thickness t / c: in, t NACA65 This indicates the reference thickness distribution of the NACA65 series.

[0067] The relative thickness t / c is adaptively selected based on radial position and vortex type, with a larger value chosen at the blade root to control diffusion and meet strength requirements, and a smaller value chosen at the blade tip to control relative Mach number and losses. A minimum thickness threshold is set. If the original thickness at the leading or trailing edge is lower than this threshold, then in the region... n Perform a linear transition within a point, so that the thickness changes from... Gradually return to the original thickness distribution.

[0068] Constructing the coordinates of the suction and pressure surfaces: Offset the fused half-thickness distribution to both sides along the normal of the mid-arc line to generate the coordinates of the suction and pressure surfaces in the local coordinate system. (Based on the derivative of the mid-arc line...) Calculate the local tilt angle, and then determine the normal vector. , Then the coordinates of the suction surface ( ) and pressure surface coordinates ( This can be represented as: in, x For dimensionless chord coordinates, c For the chord length, This represents the relative curvature distribution, which is the ratio of the ordinate of the mid-arc line to the chord length.

[0069] S505, Coordinate Transformation and Derivation: Inclination angle determined according to S501 γ Rotate the local coordinates to the global axial-circumferential plane and superimpose the axial offsets of the leading edges of each section. To obtain the global coordinates: in, X , Y , Z These are the global Cartesian coordinates corresponding to the circumferential, radial, and axial directions.

[0070] Export the coordinates into a 3D blade geometry file using a common 3D geometry format, including the GEOMTURBO format, which contains the meridional channel definition, blade section coordinates, superposition lines, and rotational speed.

[0071] S506, Geometric Consistency Verification: Verifies the consistency between the leading edge axial position and leading and trailing edge geometric angles in the exported 3D geometry file and the 1D design output; if the deviation exceeds the preset tolerance, the export is deemed to have failed and a correction prompt is given.

[0072] S6. Based on the 3D blade geometry file and CFD diagnostics, generate the final design after closed-loop correction.

[0073] S601, 3D CFD Diagnosis: Import 3D geometry into simulation application software for mesh generation and numerical calculation to obtain detailed flow field information.

[0074] S602, Correction Factor Extraction: Extract the total pressure loss, secondary flow loss and tip leakage loss of the blade surface and end region in the three-dimensional CFD post-processing, and establish the loss correction factor.

[0075] S603, Feedback Iteration: The correction amount is fed back to the empirical correction factor and variable efficiency in the one-dimensional input parameters, the two-dimensional cross section and three-dimensional geometry are reconstructed, and the closed-loop iteration from one-dimensional design to three-dimensional CFD is repeated until the deviation between the three-dimensional calculation result and the one-dimensional aerodynamic target falls within the preset margin range. .

[0076] S604, Performance Margin Reserved: In engineering practice, the total pressure ratio of 3D CFD is usually 2% to 5% lower than that of 1D design, and the isentropic efficiency is 1 to 3 percentage points lower; the target pressure ratio and efficiency are adjusted upward according to this margin to ensure that the actual performance after 3D implementation meets the original requirements.

[0077] Optionally, the choice of vortex type is not limited to the above five types, but can be extended to any analytical vortex distribution that satisfies radial balance. All practices that adaptively select vortex type and perform one-dimensional design according to design objectives are within the scope of protection of this patent.

[0078] Optionally, the decision boundaries in the automatic vortex type selection logic can be scaled according to specific application requirements, including pressure ratio boundaries, hub ratio boundaries, and tip velocity boundaries.

[0079] Optionally, the pressure ratio threshold in the automatic determination of guide vane demand is not limited to 1.15, but can be adjusted according to the fan application field, the tolerance of downstream pipelines to swirl, and structural constraints.

[0080] Optionally, the determination of guide vane demand can also take into account multiple factors such as rotational speed, hub ratio, blade height and downstream loss sensitivity, rather than relying solely on the pressure ratio.

[0081] Optionally, the radial equilibrium equation can include streamline curvature terms and adopt a full radial equilibrium model to improve the calculation accuracy under large curvature flow channels.

[0082] Optionally, the angle of attack and lag angle correction factors in the rotor blade geometry design can be designed as piecewise constants, linear functions, or nonlinear functions that vary with blade height, to be adjusted separately for the blade root, blade middle, and blade tip; or they can be replaced by a data-driven model based on neural networks.

[0083] Optionally, the airfoil database can be expanded to include controllable diffusion airfoils, double circular arc airfoils, or other specialized airfoils, and the correction factors and thickness scaling rules of the angle of attack / lag angle model can be adjusted accordingly.

[0084] Optionally, three-dimensional stacking control can introduce bend stacking, sweep stacking, or combined stacking to further improve radial load distribution and suppress secondary flow losses.

[0085] Alternatively, the 3D geometry export format is not limited to GEOMTURBO, but can be extended to a universal format that can be read by IGES, STEP, Parasolid, or any CAD / CAE software.

[0086] Optionally, the performance margin range in one-dimensional and three-dimensional closed-loop corrections can be adjusted according to the number of fan stages, Reynolds number level, and machining accuracy.

[0087] Optionally, the iterative relaxation factor can be adaptively adjusted during the iteration process, dynamically increasing or decreasing according to the error change trend or the convergence residual, in order to accelerate convergence or suppress oscillations.

[0088] Optionally, endwall boundary layer clogging correction can be introduced during the design process, taking into account the clogging factor when estimating the rim radius, so that the geometry is closer to reality.

[0089] Alternatively, the variable efficiency can be selected from different empirical values ​​based on the design point operating conditions, or calculated iteratively through a loss model, rather than being a fixed constant.

[0090] Optionally, if the constraint check fails, the design parameters can be adjusted in reverse, including increasing the hub ratio, adjusting the initial value of the flow coefficient, changing the vortex type, or reducing the target pressure ratio, in order to achieve adaptive design that satisfies the constraints.

[0091] Example 2: This embodiment focuses on the aerodynamic design of a single-rotor cooling fan. This fan is used in the cooling system of intelligent electronic devices and requires a simple structure, light weight, low cost, and insensitivity to residual swirl in the downstream flow field; therefore, a single-rotor, guide vane-less layout is adopted. The design follows the steps outlined in this embodiment, with the following design objectives: rotational speed 10000 rpm, mass flow rate 5 kg / s, target total pressure ratio 1.05, and polytropic efficiency 0.86. Given an inlet total pressure of 101325 Pa, an inlet total temperature of 288.15 K, a hub radius of 0.105 m, an initial rim radius of 0.2 m, an initial flow coefficient of 0.55, and an initial work coefficient of 0.25. (See attached...) Figure 2 As shown, the vortex type is determined based on the operating conditions. Since the target total pressure ratio of 1.05 is lower than 1.15 and the initial hub ratio is close to 0.55, a free vortex is selected for subsequent design. The automatic linkage configuration is achieved through a preset mapping relationship table: the rotor target aspect ratio is 1.5, the rotor stacking mode is determined to be torsional stacking, the upper limit of the blade tip reaction force is 0.75, the lower limit of the blade root reaction force is 0, the upper limit of the rotor inlet relative Mach number is 1.25, and the upper limit of the blade geometry angle is 75°.

[0092] The preset pressure ratio threshold is 1.15. Since the target pressure ratio is lower than this threshold, it is determined that guide vanes are not needed. A single rotor channel layout is automatically selected, and the rotor leading edge is placed at the axial zero point. The channel includes an inlet extension and an outlet extension, and only the rotor three-dimensional geometry file is generated. The number of rotor blades is calculated based on the current geometry and the target aspect ratio. The average radius density is taken as 0.90. The number of blades is estimated from the blade height, target chord length, and average radius pitch. The final number of blades is stable at 14. In this embodiment, there are no inlet guide vanes, and the rotor inlet is in a non-pre-rotating state. After ignoring the streamline curvature, the radial balance equation is simplified to the axial velocity being uniformly distributed along the blade height. The axial velocity at the hub is solved by an iterative method. The initial value is first estimated from the mass flow rate and the annular area. The axial velocity distribution is obtained by numerical integration from the hub to the rim using the fourth-order Runge-Kutta method. Then, the density distribution is calculated through the isentropic relationship. The mass flow rate is integrated and compared with the target value. The hub axial velocity is adjusted until the error is less than the preset tolerance. Finally, the axial velocity of 45.23 m / s is obtained, which is uniformly distributed along the blade height. The rotor outlet tangential velocity distribution is calculated based on the free vortex pattern. The free vortex satisfies circulation conservation. After calculation, the distribution is normalized to its mean, ensuring that its value at the average radius equals the baseline value determined by Euler work. Euler work and isentropic temperature rise are calculated based on the velocity triangle at the average radius, leading to the isentropic pressure ratio and actual pressure ratio, with a polytropic efficiency of 0.86. Pressure ratio error and flow rate error are defined. A relaxation factor of 0.7 is used to update the flow rate coefficient and work coefficient, and upper limits for single-step correction amplitude and variable range constraints are set. The iteration is repeated until both the pressure ratio error and mass flow rate error are less than the preset tolerance and the number of blades stabilizes. (See attached diagram) Figure 3 As shown in (a), in this embodiment, the rotor inlet axial velocity is uniformly distributed along the blade height; after approximately 120 iterations, the flow coefficient converges to 0.2807, as shown in the attached figure. Figure 4 As shown, the work coefficient converged from the initial value of 0.25 to 0.1851; the actual pressure ratio was 1.0501, and the calculated mass flow rate was 5.00 kg / s, all of which met the convergence criterion, and the iteration was terminated.

[0093] In the calculation of rotor blade geometric parameters, this embodiment does not have inlet pre-spin. The Lieblein zero-bend blade cascade angle of attack model and lag angle model are used to determine the angle of attack and lag angle of each section. The angles of attack for the blade root, mid-blade, and tip sections are 10.0°, 4.7°, and 2.61°, respectively, and the lag angles are 6.02°, 3.50°, and 1.84°, respectively, thus obtaining the inlet and outlet axial installation angles. Representative section results are shown in Table 2: Table 2 .

[0094] In this embodiment, guide vanes are not required. The calculation of guide vane geometry angles is skipped, the guide vane result is left blank, and the number of guide vane blades is set to 0.

[0095] The pneumatic constraint checks are as follows: 1. Equivalent diffusion factor Deq: approximately 1.30 at the rotor blade root and approximately 1.60 at the blade tip, both ≤ 2.0; 2. Counterforce R: Approximately 0.22 at the blade root and approximately 0.08 at the blade tip; the blade tip value is 0.75 lower than the upper limit of the free vortex. 3. Relative Mach number Mrl: approximately 0.55 at the blade tip, far below the upper limit of 1.25; 4. De Haller's number: W2 / W1 ratio approximately 0.76~0.89, both ≥ 0.68; 5. The aspect ratio is approximately 1.50, falling within the range of [0.8, 2.5]. 6. The maximum blade geometry angle |κ| is approximately 74.4°, which does not exceed the upper limit of 75° for free vortex.

[0096] The three-dimensional stacking was performed according to the torsional stacking method configured in Example 1. The average tilt angles of each section were -63.8°, -69.6°, and -73.3°, respectively, thereby determining the axial offset from the leading edge to the trailing edge. Torsional stacking allows for a small displacement of the leading edge axial coordinate with the radius to match the inlet airflow angle that varies along the blade height. The trailing edge coordinates were determined based on the chord length and the installation angle. The tip clearance loss proxy check passed, and the blade root diffusion limit check passed based on a comprehensive evaluation of the blade root reaction force, equivalent diffusion factor, and De Haller number.

[0097] During 3D geometry generation, the relative thicknesses of the blade root, mid-blade, and tip sections are set to 0.11, 0.10, and 0.07, respectively. The thickness distribution is generated by scaling relative thicknesses based on the NACA65 series reference thickness distribution. Minimum thickness fusion is performed between the leading and trailing edges, with a minimum thickness threshold of 0.3 mm and 10 fusion points. If the original thickness at the leading or trailing edge is less than 0.3 mm, a linear transition is performed. When constructing the suction and pressure surface coordinates, the fused half-thickness distribution is offset to both sides along the normal of the mid-arc line to generate the suction and pressure surface coordinates in the local coordinate system. The local tilt angle is then calculated using the derivative of the mid-arc line, and the normal vector is determined. The local coordinates are rotated to the global axial-circumferential plane according to the determined tilt angle, and the axial offset of the leading edge of each section is superimposed. In this embodiment, a single-rotor layout is used, and the axial coordinates of the leading edge of each section are uniformly set to 0. (See attached...) Figure 5 (a) Figure 5 As shown in (b), the blade cross-section is parametrically constructed by integrating the mid-curve, thickness distribution, and minimum thickness of the leading and trailing edges to generate the suction and pressure surface coordinates in the local coordinate system. This is then transformed by rotation to obtain the final cross-sectional shape in the global axial-circumferential plane. The meridional channel layout is shown in the attached figure. Figure 6As shown in (a), the coordinates are exported as a three-dimensional blade geometry file in a general three-dimensional geometry format, including the meridional channel definition, blade section coordinates, stacking lines and rotational speed. After export, the consistency between the leading edge axial position and the leading and trailing edge geometric angles and the one-dimensional design output is verified. The deviations are all less than 0.5°, and the verification is passed.

[0098] Finally, one-dimensional and three-dimensional closed-loop corrections are performed. The exported single-rotor three-dimensional geometry file is imported into simulation software for mesh generation and numerical calculation. The total pressure loss on the blade surface and tip region is extracted in the three-dimensional post-processing, and decomposed into profile loss, secondary flow loss, and tip leakage loss, with loss correction factors established. The correction amount is fed back to the empirical correction factor and polytropic efficiency in the one-dimensional input parameters, reconstructing the two-dimensional cross-section and three-dimensional geometry, and repeating the closed-loop iteration. With a margin of 2% to 5% for the total pressure ratio and 1 to 3 percentage points for the isentropic efficiency, the three-dimensional calculation results in this embodiment show a total pressure ratio of approximately 1.03 and a polytropic efficiency of approximately 83.5%, meeting the engineering design requirements. The final three-dimensional geometry is attached. Figure 7 As shown in (a).

[0099] Example 3: This embodiment focuses on the aerodynamic design of a cooling fan for a smart power unit. The fan requires a high mass flow rate, and the downstream duct is sensitive to airflow swirl, necessitating that the outlet airflow be as close to the axial direction as possible. Therefore, guide vanes are retained to eliminate rotor outlet swirl. The design objectives are: rotational speed 8500 rpm, mass flow rate 12 kg / s, target total pressure ratio 1.10, and polytropic efficiency 0.88. Given an inlet total pressure of 101325 Pa, an inlet total temperature of 288.15 K, a hub radius of 0.180 m, an initial rim radius of 0.320 m, an initial flow coefficient of 0.60, and an initial power coefficient of 0.25. (See attached...) Figure 2 As shown, based on the operating conditions, the vortex type is determined. The design system, considering the target total pressure ratio of 1.10, hub ratio of approximately 0.5625, and high flow rate conditions, combined with the stringent downstream requirements for low swirl, adopts an equal-reaction vortex pattern. Automatic configuration is achieved through a preset mapping table: the rotor target aspect ratio is 1.2, and the guide vane target aspect ratio is 1.02; the rotor and guide vane stacking methods are automatically determined to be non-swept and non-bending stacking; the constraint limit set is dynamically adjusted to the equal-reaction vortex mode, with an upper limit of 0.65 for tip reaction force, a lower limit of 0.10 for root reaction force, an upper limit of 1.15 for rotor inlet relative Mach number, an upper limit of 0.80 for guide vane inlet relative Mach number, and an upper limit of 75° for blade geometry angle.

[0100] The preset pressure ratio threshold is 1.15. Although the target pressure ratio of 1.10 is lower than this threshold, logically, guide vanes should not be needed. However, in this embodiment, guide vanes are required to meet the strict requirements of the downstream pipeline for low vortex flow, therefore, guide vanes are deemed necessary. A dual-channel layout is automatically selected: the upstream guide vane is placed in the negative axial region, and the downstream rotor is placed in the positive axial region. The axial clearance between them is set to approximately 0.3 times the average radius chord length of the rotor. Subsequently, three-dimensional geometric files for the guide vane and rotor are generated separately. Based on the current geometry and the target aspect ratio, the number of rotor and guide vane blades is calculated. The average radius consistency of the rotor is taken as 0.85, the average radius consistency of the guide vane as 1.10, and the average radius pitch of the guide vane as 125mm, estimating the number of guide vane blades to be approximately 16. A simple radial equilibrium equation is established and iteratively solved as in Embodiment 2, ultimately converging to an axial velocity of 44.93m / s. The tangential velocity distribution at the rotor outlet is calculated based on the isodynamic vortex pattern, with a target reaction force of 0.5. Euler work and isentropic pressure ratio are calculated and iteratively corrected using the same method as in Embodiment 2. (See attached...) Figure 3 As shown in (b), the rotor inlet axial velocity in this embodiment is also uniformly distributed along the blade height; after approximately 70 iterations, the flow coefficient converges to 0.2001, as shown in the attached figure. Figure 4 As shown, the power coefficient converged from the initial value of 0.25 to 0.187; the actual pressure ratio was 1.1001, and the calculated mass flow rate was 11.9999 kg / s, all of which were less than the preset tolerance and met the convergence criterion. Finally, the rotor converged to 12 plates, the stator converged to 16 plates, and the iteration was terminated.

[0101] In the calculation of rotor blade geometric parameters, the calculation methods for velocity triangle and bending angle are the same as in Example 2, and the representative cross-sectional results are shown in Table 3: Table 3 In the calculation of the guide vane geometry angle, the guide vane is placed upstream of the rotor, and the inlet absolute airflow angle is taken as 0° for axial inflow. The Howell model is used to calculate the lag angle, and the representative cross-sectional results are shown in Table 4. Table 4 The pneumatic constraint check was performed according to the method in Example 2, and the results are as follows: 1. Equivalent diffusion factor Deq: Rotor blade root approximately 1.32, blade tip approximately 1.45; Guide vane blade root approximately 1.10, blade tip approximately 1.35, all ≤ 2.0; 2. Counterforce R: Approximately 0.51 for rotor blade tip and 0.46 for guide vane tip, both below the upper limit of 0.65; 3. Relative Mach number Mrl: approximately 0.87 for rotor blade tip and 0.54 for guide vane tip, both below the upper limit of 1.25; 4. De Haller number: rotor approximately 0.76~0.91, stator approximately 0.83~0.89, both ≥ 0.68; 5. The aspect ratio of the rotor is approximately 1.21, and that of the stator is approximately 1.02, falling within the range of [0.8, 2.5]. 6. The blade geometric angle |κ| is approximately 75.8°, which is a physical result of the larger inlet airflow angle under low pressure ratio and large hub ratio conditions; After local consistency correction and tip high load assessment, the system determined that the warning was passed.

[0102] The three-dimensional stacking control uses the stacking method configured in Example 1 to stack the rotor in three dimensions. The average tilt angles of each section are -68.2°, -71.9°, and -73.9°, respectively. The guide vanes are adjustable inlet guide vanes, and the stacking shaft is set as the rotating shaft during overall reinstallation. The stacking is performed without sweeping or bending. The blade tip clearance and blade root diffusion checks are both passed.

[0103] During 3D geometry generation, the methods for constructing the coordinates of the mid-arc line, thickness distribution, suction / pressure surfaces, and coordinate transformation are the same as in Example 2. The coordinates of the suction and pressure surfaces in the local coordinate system and the final cross-sectional shape are as follows: Figure 5 (c) Figure 5 As shown in (d). The meridian passage layout is attached. Figure 6 As shown in (b), the reference axial position of the guide vane leading edge is approximately -0.10m, the reference axial position of the rotor leading edge is approximately +0.02m, and the axial clearance is approximately 20mm. The difference in the contours of the guide vane channel and the rotor channel on the meridional plane is clearly visible. The three-dimensional geometry of the rotor and guide vane was exported in GEOMTURBO 5.9 format to verify the consistency between the leading edge axial position, leading edge and trailing edge geometric angles and the one-dimensional design output. The deviations were all less than 0.5°, and the verification was passed.

[0104] Finally, one-dimensional and three-dimensional closed-loop corrections were performed. The dual-channel three-dimensional geometry was imported into CFD software for mesh generation and numerical calculation. The total pressure loss on the blade surface and tip region was extracted in the three-dimensional post-processing, and decomposed into profile loss, secondary flow loss, and tip leakage loss, with loss correction factors established. The corrections were fed back to the empirical correction factor and polytropic efficiency in the one-dimensional input parameters, reconstructing the two-dimensional cross-section and three-dimensional geometry, and repeating the closed-loop iteration. The three-dimensional CFD results showed a total pressure ratio of approximately 1.085, an isentropic efficiency of approximately 84.1%, an outlet swirl angle of less than 5°, a total pressure ratio deviation of approximately 2.3%, and an efficiency deviation of approximately 3 percentage points, all within the margin range, indicating design convergence. The final three-dimensional geometry is attached. Figure 7 As shown in (b).

[0105] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A vortex-driven aerodynamic design method for a cooling fan, characterized in that, Includes the following steps: S1. Generate a one-dimensional aerodynamic layout and initial parameters based on the design input and vortex type selection; S2. Based on the one-dimensional aerodynamic layout and radial balance equation, generate a convergent rotor inlet and outlet flow field; S3. Based on the empirical correlation between the converged rotor inlet and outlet flow fields and the blade cascade, generate the two-dimensional geometric angles of the rotor and guide vanes; S4. Based on the two-dimensional geometric angles of the guide vane and the aerodynamic constraints, generate blade cross-section parameters that satisfy the constraints; S5. Generate a three-dimensional blade geometry file based on the blade cross-section parameters and three-dimensional stacking control; S6. Based on the three-dimensional blade geometry file and CFD diagnostics, generate the final design after closed-loop correction.

2. The aerodynamic design method for a vortex-driven cooling fan according to claim 1, characterized in that, S1 specifically includes: Define dimensionless design parameters based on the initialization parameters input by the design, and select the vortex type based on the target total pressure ratio and hub ratio; Based on the vortex type, the target aspect ratio initial value, superposition method, and constraint limit set are automatically determined through a preset mapping relationship. Determine whether guide vanes are needed based on the target pressure ratio and the preset pressure ratio threshold, and determine whether to use a single rotor or dual-channel layout. The number of blades for the rotor and guide vanes is estimated based on the current geometry and the target aspect ratio.

3. The aerodynamic design method for a vortex-driven cooling fan according to claim 1, characterized in that, S2 specifically includes: Based on the selected vortex type, specify the rotor outlet tangential velocity distribution and establish the radial equilibrium equation; An iterative method was used to solve for the axial velocity at the hub, and the axial velocity distribution along the blade height was obtained by numerical integration. Adjust the axial speed of the wheel hub by comparing the integral of the mass flow rate with the target value until the mass conservation is satisfied; The Euler work and pressure ratio are calculated based on the converged velocity triangle, and the flow coefficient and work coefficient are updated using a relaxation factor. The iteration continues until the pressure ratio error and mass flow rate error are both less than the preset tolerance.

4. The aerodynamic design method for a vortex-driven cooling fan according to claim 1, characterized in that, S3 specifically includes: The relative airflow angle and bending angle of each section are calculated based on the converged velocity triangle, and the consistency and chord length are determined by the empirical correlation of the blade cascade. The angle of attack and lag angle models of the Lieblein zero-curvature blade cascade were used to determine the angle of attack and lag angle of each cross section. Calculate the rotor inlet and outlet axial installation angles based on the angle of attack and the angle of retreat; When determining the need for guide vanes, the Howell model is used to calculate the guide vane lag angle and the guide vane inlet and outlet geometric installation angles based on the airflow angles at the guide vane inlet and outlet.

5. The aerodynamic design method for a vortex-driven cooling fan according to claim 1, characterized in that, S4 specifically includes: The set of constraint limits is dynamically adjusted according to the selected vortex type, and multi-dimensional aerodynamic constraint checks are performed on the rotor and guide vanes. Constraint checks include constraints on equivalent diffusion factor, counterforce, relative Mach number, De Haller number, aspect ratio, and blade geometry. If any constraint fails, adjust the corresponding section parameters and regenerate until all constraints are satisfied; Output the chord length, mounting angle, and bending angle parameters of representative radial sections of the rotor and guide vanes.

6. The aerodynamic design method for a vortex-driven cooling fan according to claim 1, characterized in that, S5 specifically includes: Based on the stacking method corresponding to the selected vortex type, the rotor is stacked in three dimensions to determine the axial coordinates of the leading and trailing edges of each section; When it is determined that guide vanes are needed, the guide vanes are stacked in three dimensions; Perform a tip clearance loss check; if it fails, reselect the stacking parameters. The middle arc is generated using a parabola, and the thickness distribution is generated based on the NACA65 series reference thickness distribution. The semi-thickness distribution is offset along the normal of the mid-arc line to generate the coordinates of the suction surface and the pressure surface; Rotate the local coordinates to the global axial-circumferential plane by the tilt angle, and export the 3D blade geometry file; Verify the consistency between the leading edge axial position and geometric angles in the exported 3D geometry file and the 1D design output.

7. The aerodynamic design method for a vortex-driven cooling fan according to claim 1, characterized in that, S6 specifically includes: Import the 3D blade geometry file into simulation software for mesh generation and numerical calculation to obtain 3D flow field information; In the post-processing of 3D CFD, total pressure loss, secondary flow loss and tip leakage loss are extracted, and a loss correction factor is established. The correction factor is fed back to the empirical correction factor and variable efficiency in the one-dimensional input parameters to reconstruct the two-dimensional cross section and three-dimensional geometry, and the closed-loop iteration is repeated. The final design is output only when the deviation between the three-dimensional calculation results and the one-dimensional aerodynamic target falls within the preset margin range.

8. The aerodynamic design method for a vortex-driven cooling fan according to claim 2, characterized in that, The selection of the vortex type is determined based on the target total pressure ratio, hub ratio, and blade tip circumferential speed, specifically including: When the target total pressure ratio is less than 1.15 and the hub ratio is greater than or equal to 0.55, select free vortex; When the target total pressure ratio is greater than or equal to 1.8 or the hub ratio is greater than 0.75, select the vortex with equal reaction degree; When the blade tip circumferential velocity is greater than 450 m / s and the target total pressure ratio is greater than 1.5, select the isovorticity; If none of the above conditions are met, choose the exponential vortex.

Images