A method for calculating the aerodynamic force of tail rotor in the airflow of helicopter rotor based on a double momentum source model

Abstract

The application discloses a kind of based on double momentum source model's helicopter rotor airflow in tail rotor aerodynamic force calculation method, using the momentum source model based on CFD to carry out flow field calculation, including: establishing isolated tail rotor momentum source model and rotor-tail rotor double momentum source model, using isolated model to calculate out tail rotor convergence total pitch based on tail rotor target thrust, and it is used as the input of tail rotor momentum source in double momentum source model, respectively solve two models to obtain tail rotor induced velocity;Based on element theory, induced velocity is converted into distributed aerodynamic force, and the distributed aerodynamic force caused by rotor wake interference is obtained by difference of distributed aerodynamic force purely by tail rotor.The application uses momentum source model to replace entity blade modeling, without complex nested grid, and the result is accurate with high calculation efficiency.

Description

Technical Field

[0001] This invention belongs to the field of aerodynamics calculation technology, specifically relating to a numerical simulation method for calculating the aerodynamic forces of rotor wake on tail rotor when a helicopter is hovering or flying forward. Background Technology

[0002] Thanks to their vertical takeoff and landing (VTOL) and hovering capabilities, helicopters play a vital role in cargo delivery and civilian rescue operations. The tail rotor is a critical component in helicopters, responsible for balance and directional control. During low-speed forward flight or hovering, the strong, non-uniform wake generated by the helicopter rotor directly impacts the tail rotor, causing drastic changes in the tail rotor's blade element angle of attack and generating significant aerodynamic interference. This interference not only reduces the tail rotor's aerodynamic efficiency but can also lead to tail rotor failure due to vibration. Therefore, accurately and rapidly establishing a method for calculating the aerodynamic forces of rotor wake interference on the tail rotor has become an urgent research task in helicopter technology development.

[0003] Currently, research methods for tail rotor aerodynamics under rotor wake interference mainly include experimental methods and numerical calculation methods. Among them, computational fluid dynamics (CFD) methods based on nested or sliding grids have become the mainstream numerical research approach because they can realistically simulate the relative motion of the blades and the evolution of unsteady wakes, and better preserve complex flow field structures such as tip vortices. For example, Fan Feng et al. from Nanjing University of Aeronautics and Astronautics used the nested grid method to simultaneously consider the flow field information exchange between the rotor, tail rotor, and vertical tail, and conducted numerical simulations of typical rotor and tail rotor configurations. Qiu Fengchang et al. from the China Helicopter Design Institute also used a similar method to calculate the rotor wake interference characteristics under hovering and low-speed forward flight states, and analyzed the influence of parameters such as tail rotor rotation direction and position on the interference effect.

[0004] Existing CFD methods based on nested or sliding meshes, while offering high accuracy, require complex and time-consuming processes to generate full-size 3D models and calculate the collective pitch-thrust curves of the rotor / tail rotor when generating nested or sliding meshes in engineering applications. Furthermore, the number of meshes generated ranges from tens of millions to hundreds of millions, requiring weeks to months to compute on a civilian 64-core computer. Therefore, supercomputers are commonly used for calculations, resulting in long computation times and high costs. This makes it difficult to directly apply these methods to helicopter aerodynamic layout design optimization and helicopter rotor / tail rotor aerodynamic optimization, limiting their application in the model development process. Summary of the Invention

[0005] This invention addresses the problems of high computational cost and low computational efficiency of existing CFD methods based on nested or sliding meshes when solving for the aerodynamic forces of the rotor wake on the tail rotor. It proposes a method for calculating the aerodynamic forces of the tail rotor in the helicopter rotor airflow based on a dual momentum source model.

[0006] To achieve the above objectives, the technical solution provided by this invention is:

[0007] A method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum-source model includes the following steps:

[0008] Step 1: Establish a momentum source model and perform CFD flow field solution to extract the tail rotor disk induced velocity; including the following sub-steps:

[0009] Step 1.1: Establish an isolated tail rotor momentum source model and a rotor-tail rotor dual momentum source model that simultaneously includes a rotor momentum source model and a tail rotor momentum source model;

[0010] Step 1.2: Using the target thrust of the tail rotor determined by the overall parameters of the helicopter as input, perform CFD flow field solution on the isolated tail rotor momentum source model, and back-calculate to obtain the tail rotor collective pitch and the full flow field information of the tail rotor under the isolated tail rotor momentum source model.

[0011] Step 1.3: Use the collective pitch of the tail rotor convergence as the input collective pitch of the tail rotor momentum source model in the rotor-tail rotor dual momentum source model, and use the rotor target thrust determined by the overall parameters of the helicopter as the input of the rotor momentum source model in the rotor-tail rotor dual momentum source model. Perform CFD flow field solution on the rotor-tail rotor dual momentum source model to obtain the full flow field information of the tail rotor under the rotor-tail rotor dual momentum source model.

[0012] Step 1.4: Extract the corresponding tail rotor disk induced velocity from the tail rotor full flow field information of the isolated tail rotor momentum source model and the tail rotor full flow field information of the rotor-tail rotor dual momentum source model, respectively.

[0013] Step 2: Based on the blade element theory, the tail rotor disk induced velocities obtained in Step 1 are converted into corresponding tail rotor distributed aerodynamic forces.

[0014] Step 3: Subtract the tail rotor distributed aerodynamic force based on the rotor-tail rotor dual momentum source model from the tail rotor distributed aerodynamic force based on the isolated tail rotor momentum source model to obtain the tail rotor distributed aerodynamic force purely caused by rotor airflow interference:

[0015]

[0016]

[0017] in, and These are the thrust and drag coefficients of the upper blade element of the tail rotor, obtained based on the isolated tail rotor momentum source model. and These are the thrust coefficient and drag coefficient of the upper blade element of the tail rotor, obtained based on the rotor-tail rotor dual momentum source model. for and The difference represents the tail rotor distributed thrust caused by rotor wake interference; for and The difference represents the tail rotor distributed drag caused by rotor wake interference.

[0018] Furthermore, in step 1.1, the isolated tail rotor momentum source model and the rotor-tail rotor dual momentum source model adopt the same mesh topology and node distribution, and a mesh refinement area is set in the spatial region corresponding to the expected position of the rotor. In the mesh refinement area of ​​the isolated tail rotor momentum source model, no momentum source is applied to ensure the numerical consistency when performing the difference in step 3.

[0019] Furthermore, step 1.2 involves back-calculating the tail rotor collective pitch and the tail rotor full flow field information under the isolated tail rotor momentum source model, including the following process:

[0020] The target thrust of the tail rotor, determined by the overall parameters of the helicopter, is used as the input to the isolated tail rotor momentum source model. The flow field of the isolated tail rotor momentum source model is solved by a CFD solver to calculate the thrust corresponding to the current tail rotor collective pitch.

[0021] Based on the difference between the current tail rotor collective pitch and the target tail rotor thrust, an automatic feedback adjustment algorithm is used to dynamically calculate and adjust the tail rotor collective pitch.

[0022] Repeat the CFD solution steps and tail rotor collective pitch calculation and adjustment steps until convergence occurs when the difference between the thrust output by the isolated tail rotor momentum source model and the target tail rotor thrust is less than or equal to the preset error threshold. Record the converged collective pitch as the tail rotor converged collective pitch. The tail rotor converged collective pitch represents the collective pitch value required for the isolated tail rotor to generate the target tail rotor thrust in an environment without rotor interference.

[0023] Further, in step 1.4, the tail rotor disk-induced velocity is extracted from the tail rotor full flow field information of the isolated tail rotor momentum source model and the tail rotor full flow field information of the rotor-tail rotor dual momentum source model, respectively, including the following process:

[0024] Determine the extraction area, which is located in a plane parallel to the tail rotor disk plane and 0.2 to 0.5 times the tail rotor radius downstream of the tail rotor disk plane. The area of ​​the extraction area is equal to the area of ​​the disk.

[0025] Within the extraction region, the velocity perpendicular to the disk plane is extracted as the tail rotor disk induced velocity; and the tail rotor disk induced velocity extracted from the tail rotor full flow field information of the isolated tail rotor momentum source model is denoted as V1, and the tail rotor disk induced velocity extracted from the tail rotor full flow field information of the rotor-tail rotor dual momentum source model is denoted as V2.

[0026] Furthermore, based on blade element theory, the tail rotor disk-induced velocity obtained in step 1 is converted into the corresponding tail rotor distributed aerodynamic force, including the following steps:

[0027] Step 2.1: Using the extracted tail rotor disk induced velocity, calculate the actual angle of attack corresponding to the true incoming flow direction at each blade element position of the tail rotor. and Mach number The calculation formula is:

[0028]

[0029]

[0030]

[0031]

[0032]

[0033]

[0034]

[0035] In the formula, It is the incoming flow velocity; It is the local speed of sound; It is the specific heat ratio; It is the gas constant; For absolute temperature, the specific heat ratio is relative to air. The value is 1.4, and the gas constant is... It is 287 J / (kg·K). This invention uses a standard atmospheric pressure environment and an absolute temperature of 287 J / (kg·K). The calculated local speed of sound is 288.15K. It is 340.3 m / s; It is the tail rotor speed; It is the installation angle of the leaf element section; It is the leaf element azimuth angle; It is the angle between the incoming airflow velocity and the plane of the propeller disk; The tail rotor radius; For leaf element radial positioning; For induced velocity, where, when calculated based on the isolated tail rotor momentum source model, The tail rotor disk-induced velocity V1 is extracted from the full flow field information of the tail rotor in the isolated tail rotor momentum source model; when calculating based on the rotor-tail rotor dual momentum source model... Take the tail rotor disk induced velocity V2 extracted from the full flow field information of the rotor-tail rotor dual momentum source model; The axial velocity of the leaf element; The circumferential velocity of the leaf element; The radial velocity of the leaf element; The actual induction angle of the leaf element;

[0036] Step 2.2, radial positioning of leaf elements And the actual angle of attack of the leaf element was calculated. and Mach number The input is fed into a pre-built and trained neural network model to predict the lift coefficient at the position of each blade element in the wind coordinate system. and drag coefficient ;

[0037] Step 2.3: Calculate the lift coefficient at all blade element positions in the wind coordinate system. and drag coefficient Switch to propeller disk

[0038] In the coordinate system, the aerodynamic characteristics of the blade elements in the propeller disk coordinate system are obtained; the conversion formula is:

[0039]

[0040] In the formula, This represents the thrust coefficient of the blade element in the propeller disk coordinate system; This represents the drag coefficient of the blade element in the propeller disk coordinate system.

[0041] Step 2.4: Based on the processes in Steps 2.1-2.3, calculate the tail rotor distributed aerodynamic force based on the isolated tail rotor momentum source model and the tail rotor distributed aerodynamic force based on the rotor-tail rotor dual momentum source model.

[0042] Furthermore, in step 2.2, the neural network model adopts a multilayer perceptron neural network, whose input layer nodes correspond to the angle of attack, Mach number and blade radial position that affect the aerodynamic forces of the airfoil, and the output layer nodes correspond to the lift coefficient, drag coefficient and moment coefficient; a regularization algorithm is used to train and optimize the multilayer perceptron neural network.

[0043] Furthermore, it also includes obtaining the total thrust of the tail rotor. and total tail rotor drag The steps involve calculating the tension coefficients of all blade elements on the propeller disk obtained in step 2.3. and drag coefficient Integrating, we get:

[0044]

[0045]

[0046] In the formula, It is the leaf element azimuth angle, and its value range is: ; It is the radial position of leaf element, and its value range is: .

[0047] The advantages of this invention are:

[0048] 1. This invention uses a momentum source model based on computational fluid dynamics (CFD) to replace the solid blade model. By constructing an isolated tail rotor momentum source model and a rotor-tail rotor dual momentum source model, the collective pitch of the tail rotor, calculated from the isolated tail rotor momentum source model, is transferred to the tail rotor momentum source model in the rotor-tail rotor dual momentum source model. Finally, the aerodynamic distribution forces of the tail rotor obtained from the two models are subtracted. Within the momentum source model framework, the accurate separation and calculation of the rotor wake interference aerodynamic forces are achieved, providing a basis for studying the influence of different flight parameters on interference. This invention does not require the generation of complex nested or sliding meshes, nor does it require the establishment of complex meshes near the blades. Practical verification shows that, compared with the traditional nested mesh CFD method, the computational efficiency of this invention is improved by 26 times under the same computational resources.

[0049] 2. Traditional methods rely solely on blade element theory and empirical formulas, making it difficult to accurately obtain the tail rotor induced velocity, resulting in large errors in tail rotor aerodynamic calculations. This invention, however, provides an accurate wake-induced velocity field through a CFD-based momentum source model, accurate airfoil aerodynamic response data based on an airfoil database, and combines blade element theory with neural network model predictions. This improves computational efficiency while ensuring the accuracy of tail rotor aerodynamic characteristics (e.g., tail rotor thrust and drag). Verification shows that the error of the method obtained by this invention compared to experimental methods (taking tail rotor thrust as an example) is 2.31%–4.50%, and the error compared to unsteady flow field simulation methods is 1.53%–9.43%, meeting the requirements of practical engineering applications.

[0050] 3. In the method of this invention, the establishment of the momentum source model and the setting of parameters (thrust, collective pitch, rotational speed, etc.) are all based on parametric input, eliminating the need to construct a complex tail rotor geometry model. Combined with a CFD solver, fully automated calculations from parameter setting to result output can be easily achieved. Attached Figure Description

[0051] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0052] Figure 1 This is a flowchart of the method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual momentum source model, according to the present invention.

[0053] Figure 2 This is a schematic diagram of the encrypted mesh generated by CFD for momentum source model simulation in this invention;

[0054] Figure 3 This is a schematic diagram of the rotor-tail rotor dual momentum source model in this invention;

[0055] Figure 4 This is a schematic diagram of the velocity and aerodynamic decomposition of the blade element cross section in this invention;

[0056] Figure 5 This is a schematic diagram of the blade element azimuth angle and velocity components in this invention;

[0057] Figure 6 The total thrust of the tail rotor under different methods in the embodiments of the present invention. With sideslip angle A diagram showing the changes. Detailed Implementation

[0058] The embodiments of the present invention are described in detail below. These embodiments are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0059] This invention addresses the high computational cost and low efficiency of existing helicopter rotor / tail rotor interference aerodynamic calculation methods (such as nested mesh CFD). This embodiment proposes a highly efficient method for calculating tail rotor aerodynamics in helicopter rotor airflow based on a dual-momentum source model. The following section combines... Figure 1 The process shown below provides a detailed explanation of the specific implementation steps of this invention.

[0060] Step 1: Establish an isolated tail rotor momentum source model and a rotor-tail rotor dual momentum source model, and calculate the induced velocity at the tail rotor disk respectively.

[0061] Step 1.1: Based on the overall design parameters of the target helicopter, determine the operating state to be analyzed (hovering state or low-speed forward flight state). Identify the rotor thrust and tail rotor thrust under this operating state.

[0062] Step 1.2: Establish an isolated tail rotor momentum source model and solve for the flow field information under this model. The specific process is as follows:

[0063] First, in the CFD software, a 3D mesh for the momentum source model is generated. This 3D mesh is a polyhedral mesh used for local refinement at the location of the tail rotor virtual disk and its wake region. Simultaneously, a rotor refinement zone is set in the spatial region corresponding to the expected rotor position according to the incoming flow direction. Especially in the wake region, the mesh is extended outwards downstream of the rotor and tail rotor according to the incoming flow direction. The purpose is to enhance the mesh resolution in the wake region, effectively reduce the influence of numerical dissipation on the rotor wake vortex, and ensure that the rotor wake can act more completely on the tail rotor region during propagation, thereby more accurately capturing the aerodynamic interference effect of the main tail rotor and reducing calculation errors caused by insufficient mesh resolution. However, no rotor momentum source is applied to the rotor refinement zone of the isolated tail rotor momentum source model to ensure that the mesh topology and node distribution of the finally established isolated tail rotor momentum source model are completely consistent with the rotor-tail rotor dual momentum source model. The mesh generated by CFD for the momentum source model is shown below. Figure 3 As shown.

[0064] Then, the momentum source model (corresponding to the virtual disk model) is activated in the CFD solver. A tail rotor virtual disk is built on the generated 3D mesh, and the geometric and operational parameters of the tail rotor virtual disk are set to form an isolated tail rotor momentum source model. The geometric parameters include the disk center position coordinates, the inner and outer radii of the disk, the disk thickness, and the number of blades; the kinematic parameters include the airfoil section, chord length distribution, blade torsion distribution, rotation direction, and rotation speed.

[0065] Next, the target tail rotor thrust determined in step 1.1 is used as input to the isolated tail rotor momentum source model. A CFD solver is used to calculate the flow field of the isolated tail rotor momentum source model, yielding the thrust corresponding to the current tail rotor collective pitch. Based on the difference between the thrust corresponding to the current tail rotor collective pitch and the target tail rotor thrust, an automatic feedback adjustment algorithm is used to dynamically calculate and adjust the tail rotor collective pitch. The CFD solving step and the tail rotor collective pitch calculation and adjustment step are repeated until the difference between the thrust output by the isolated tail rotor momentum source model and the target tail rotor thrust is less than or equal to a preset error threshold, at which point convergence occurs. The converged collective pitch is recorded as the tail rotor converged collective pitch. The tail rotor converged collective pitch represents the collective pitch value required for the isolated tail rotor to generate the target tail rotor thrust in an environment without rotor interference. In this application, the tail rotor collective pitch refers to the angle value at which the installation angles of all blades in the tail rotor change simultaneously, used to control the magnitude of the total thrust generated by the tail rotor.

[0066] The specific process of solving based on the CFD solver is as follows:

[0067] First, set the CFD solver parameters and related initial parameters, including: setting the far field to a free-flow boundary condition, which treats the far field as an infinite flow region far from the object, almost undisturbed, where the fluid can naturally enter or leave the computational domain; setting airflow parameters such as pressure and temperature, as well as the inflow velocity and direction in the far field, according to the actual environmental conditions. In this embodiment, the environmental conditions are set to standard atmospheric pressure by default; setting the maximum time step, generally the minimum number of steps required to meet the convergence conditions. For the first calculation, the time step is usually set to 1000 steps or more to ensure sufficient calculation stability. Since the collective pitch of a real tail rotor is generally in the range of 0° to 20°, the initial collective pitch of the tail rotor is set to 0°.

[0068] Based on the parameters set above, the CFD solver begins iterative calculation. During the calculation, the isolated tail rotor momentum source model calculates the corresponding thrust based on the initial collective pitch of the tail rotor. It then determines whether the obtained thrust and the target thrust of the tail rotor satisfy the convergence criterion. In this embodiment, the convergence criterion is that the absolute value of the difference between the thrust and the target thrust of the tail rotor is less than or equal to a preset threshold. If the convergence criterion is not met, the collective pitch value is adjusted until the convergence criterion is met. After the iterative calculation converges, the tail rotor collective pitch value and the tail rotor full flow field information in this convergence state are recorded. The tail rotor collective pitch value in this convergence state represents the collective pitch required for the isolated tail rotor to generate the target thrust in an environment without rotor interference.

[0069] After the convergence calculation is completed, the process also includes a step to check the stability of the tail rotor collective pitch obtained during the iteration. The purpose is to check whether the tail rotor collective pitch value is stable during the iteration. The criterion for stability is usually: within a certain number of consecutive iteration steps, the fluctuation range of the tail rotor collective pitch narrows and remains within the set allowable tolerance threshold (the allowable tolerance threshold set in this invention is an absolute fluctuation of less than 0.5°), and does not show a diverging trend. This indicates that the tail rotor collective pitch value is stable during the iteration, and the tail rotor collective pitch value in the current convergence state is retained. Conversely, if the tail rotor collective pitch continues to diverge with the number of iteration steps, or oscillates violently within a large range and cannot converge to a certain value for a long time, it is considered that the tail rotor collective pitch iteration is unstable. If the tail rotor collective pitch cannot converge stably, it is usually considered that the number of airfoil sections selected as the CFD momentum source model is insufficient or the accuracy of the airfoil sections themselves is insufficient to meet the calculation requirements. It is necessary to increase the number of input airfoil sections or improve the accuracy of the airfoil sections before performing iterative calculations through the CFD solver. In this invention, airfoil section data (including angle of attack, Mach number, lift coefficient, and drag coefficient) can be selected from the subsequently constructed airfoil database to expand the number of airfoil section inputs when the tail rotor collective pitch cannot converge stably. It should be noted that those skilled in the art understand that setting the parameters of the CFD solver and checking the iterative stability of the tail rotor collective pitch are conventional techniques in the field. Therefore, the specific operational procedures will not be elaborated here.

[0070] Step 1.3: Establish a rotor-tail rotor dual momentum source model and solve for the flow field information under this model.

[0071] This step directly uses the 3D mesh generated when establishing the isolated tail rotor momentum source model, maintaining complete consistency in mesh topology and node distribution. This 3D mesh has been locally refined at the location of the tail rotor virtual disk and its wake region. Simultaneously, a rotor refinement zone has been set within the spatial region corresponding to the expected rotor position, based on the incoming flow direction.

[0072] A virtual rotor disk is created on the generated 3D mesh. Its geometric parameters (disk center position, disk radius, disk thickness, inner-outer radius ratio), kinematic parameters (rotation direction, rotation speed), and blade parameters (number of blades, chord length distribution, torsional distribution) are set. A virtual tail rotor disk is also created on the generated mesh. All parameters of the tail rotor virtual disk are strictly consistent with those of the tail rotor virtual disk in the isolated tail rotor momentum source model, including geometric, kinematic, and blade parameters. The rotor and tail rotor virtual disks are placed in the same 3D mesh, and momentum source terms are applied to both virtual disks to form a complete rotor-tail rotor dual momentum source model. In the resulting rotor-tail rotor dual momentum source model, the rotor momentum source model uses the rotor target thrust determined by the overall helicopter parameters as input.

[0073] The difference from step 1.2 is that the tail rotor momentum source model in this step no longer uses the tail rotor target thrust as input, but instead directly fixes it to the tail rotor convergence collective pitch obtained by back-calculation based on the isolated tail rotor momentum source model. This tail rotor convergence collective pitch is then used as the input collective pitch of the tail rotor virtual disk in the rotor-tail rotor dual momentum source model. The reason for adopting this processing measure in this invention is as follows:

[0074] Because helicopter rotor wake interference can change the tail rotor angle of attack, if the target tail rotor thrust is directly assigned to the tail rotor momentum source in the dual momentum source model during the interference flow field calculation, the CFD solver will force the tail rotor target thrust to be met by adjusting relevant parameters. In other words, the control input of the tail rotor is changed when the collective pitch is unknown, resulting in a mismatch between the collective pitch of the tail rotor in the interference flow field and the actual required collective pitch. That is, the collective pitch of the tail rotor in the dual momentum source model and the isolated tail rotor model is inconsistent, so the thrust obtained by subsequent subtraction not only includes the real additional aerodynamic force caused by rotor wake interference, but may also be mixed with the error caused by the mismatch of control variables. Based on this, this application establishes an isolated tail rotor momentum source model and a dual momentum source model including the rotor and tail rotor. First, the required collective pitch value of the tail rotor under the target thrust is calculated iteratively through the isolated tail rotor momentum source model, and the collective pitch value is transferred to the tail rotor momentum source in the dual momentum source model. This ensures that the control input of the tail rotor in the two models is consistent, so that the difference result of the distributed aerodynamic force obtained by the two models is purely the additional aerodynamic force of the tail rotor caused by the rotor wake interference. This improves the calculation efficiency while ensuring the accuracy of the calculation results.

[0075] Next, the CFD solver was started to solve the CFD flow field of the rotor-tail rotor dual momentum source model. The fluid free boundary conditions, time step, convergence criteria, etc. in the CFD solver were exactly the same as those in the isolated tail rotor momentum source model calculation in step 1.2. After calculation, the full flow field information of the tail rotor under the rotor-tail rotor dual momentum source model was obtained.

[0076] Step 1.4: Extract the induction rate.

[0077] After convergence, the induced velocity field of a specific region downstream of the tail rotor disk is extracted from the full flow field information of the two models. Specifically, this specific region is located parallel to the tail rotor disk plane and 0.2-0.5 times the tail rotor radius downstream of the disk plane, and its area is equal to the area of ​​the disk. The reason for selecting this region in this invention is that it avoids the non-physical extrema affected by the momentum source term input when extracting induced velocity close to the tail rotor disk. This is because the momentum source virtual disk in CFD is usually implemented by applying momentum source terms (e.g., volume forces) on a very limited number of grid layers. This leads to huge pressure jumps and numerical oscillations in the region close to the momentum source virtual disk. In contrast, the induced velocity field distribution formed by the downwash flow field generated by the downward push of the rotor rotation and the airflow field discharged backward (or laterally) during tail rotor rotation is stable. The data measured under this condition is more representative and the calculation results are more reliable.

[0078] Extract the velocity components perpendicular to the rotor disk plane on the plane of this specific region, namely the required induced velocities V1 (from the isolated tail rotor momentum source model) and V2 (from the rotor-tail rotor dual momentum source model).

[0079] Step 2: Based on blade element theory and neural network airfoil database, the tail rotor disk induced velocity obtained in Step 1 is converted into the corresponding tail rotor distributed aerodynamic force.

[0080] Step 2.1: Calculate the actual angle of attack and Mach number of the leaf element.

[0081] Blade element theory assumes that the blade is divided into several micro-segments (i.e., blade elements) along the radial and circumferential directions, and that adjacent blade elements do not interfere with each other. The airfoil aerodynamic characteristics at the midpoint of each blade element are taken as the airfoil aerodynamic characteristics of the entire blade element. Then, the aerodynamic characteristics of the entire blade disk are obtained by integrating over all blade elements. For a blade element on the tail rotor disk, its blade element azimuth angle and incoming flow velocity components are as follows: Figure 4 , Figure 5 As shown.

[0082] Based on the tail rotor disk induced velocities V1 and V2 extracted from the isolated tail rotor momentum source model in step 1, calculate the actual angle of attack corresponding to the true inflow direction of the blade element at different blade radial positions. and Mach number The calculation formula is:

[0083]

[0084]

[0085]

[0086]

[0087]

[0088]

[0089]

[0090] In the formula, It is the incoming flow velocity; It is the local speed of sound; It is the specific heat ratio; It is the gas constant; For absolute temperature, the specific heat ratio is relative to air. The value is 1.4, and the gas constant is... It is 287 J / (kg·K). This invention uses a standard atmospheric pressure environment and an absolute temperature of 287 J / (kg·K). The calculated local speed of sound is 288.15K. It is 340.3 m / s; It is the tail rotor speed; It is the installation angle of the leaf element section; It is the leaf element azimuth angle; It is the angle between the incoming airflow velocity and the plane of the propeller disk; The tail rotor radius; For leaf element radial positioning; For induced velocity, where, when calculated based on the isolated tail rotor momentum source model, The tail rotor disk-induced velocity V1 is extracted from the full flow field information of the isolated tail rotor momentum source model; when calculating based on the rotor-tail rotor dual momentum source model... Take the tail rotor disk induced velocity V2 extracted from the full flow field information of the rotor-tail rotor dual momentum source model; The axial velocity of the leaf element; The circumferential velocity of the leaf element; The radial velocity of the leaf element; This represents the actual induction angle of the leaf element.

[0091] Step 2.2: Construct a neural network model.

[0092] Collect the airfoil used for the tail rotor at different Mach numbers Aerodynamic data for different angles of attack and different blade radial positions, including at least lift and drag coefficient data. This aerodynamic data can be obtained through wind tunnel testing or calculated using CFD methods. In this embodiment, the Mach number ranges from 0 to 0.8, and the angle of attack ranges from -30° to 30°. An airfoil database is constructed using the corresponding Mach number, angle of attack, blade radial position, and lift and drag coefficient data.

[0093] Construct a multilayer perceptron (MLP) neural network model. The input layer of the neural network structure has three input nodes to receive input features, namely the key parameters affecting the airfoil's aerodynamic forces, which correspond to the angle of attack. ,Mach number and blade radial positioning The output layer has two output nodes, which are used to predict the output target features (i.e., aerodynamic coefficients), corresponding to the lift coefficient and drag coefficient, respectively. The hidden layer is located between the input layer and the output layer and is used to extract features, map and perform nonlinear transformations on the input data. Since the number of neurons in the hidden layer affects the fitting effect, too few neurons result in insufficient learning ability of the model and it cannot fit the nonlinear relationship of the data well. Too many neurons completely memorize the training data and lack generalization ability. Therefore, in this embodiment, the hidden layer is set to 50 neurons.

[0094] Training the Neural Network: From the airfoil database established above, 70% of the dataset was used as the training set, 15% as the validation set, and 15% as the test set. The constructed Multilayer Perceptron (MLP) neural network model was trained using a regularization algorithm on the partitioned datasets. After training, a neural network capable of accurately predicting the lift, drag, and moment coefficients of airfoil aerodynamic forces was obtained. This network accurately expresses the generalized functions of the aerodynamic coefficients of the airfoil under different operating conditions (i.e., different lift and drag conditions), expressed as:

[0095]

[0096]

[0097] In the formula, The lift coefficient; This is the drag coefficient; For the angle of attack; It is the Mach number; This refers to the radial position of the blade.

[0098] Step 2.3: Calculate the blade element aerodynamic force and transform it to the propeller disk coordinate system to obtain the distributed aerodynamic force of the entire propeller disk.

[0099] First, the angle of attack of the blade element at different blade radial positions is obtained according to the calculation formula in step 2.1. and Mach number The radial position of the blades and the calculated actual angle of attack for each blade element. and Mach number As input, the neural network model trained in step 2.2 is substituted into it, and the corresponding aerodynamic coefficients, including lift coefficient and drag coefficient, are output.

[0100] In this invention, the aerodynamic coefficients predicted by the neural network model are applied to the wind coordinate system, i.e. Figure 4 middle and The coordinate system formed Indicates the direction of lift. This indicates the direction of resistance. It needs to be transformed to the propeller disk coordinate system, i.e. Figure 4 middle and The coordinate system formed Indicates the direction of thrust. The formula for transforming between the two coordinate systems, representing the direction of the backward force (i.e., the direction parallel to the plane of the propeller disk and pointing backward), is:

[0101]

[0102] Then, according to the coordinate system transformation formula, the lift coefficient corresponding to each blade element in the wind coordinate system is... and drag coefficient The transformation yields the aerodynamic characteristics of the blade element in the propeller disk coordinate system, i.e., the thrust coefficient of the blade element in the propeller disk coordinate system. and drag coefficient This represents the distributed aerodynamic forces across the entire propeller disk. The conversion formula is:

[0103]

[0104] Alternatively, the tension coefficients of all blade elements can be analyzed separately according to actual analytical needs. and drag coefficient Integrate to obtain the total thrust of the tail rotor. and total tail rotor drag :

[0105]

[0106]

[0107] In the formula, It is the leaf element azimuth angle, and its value range is: ; It is the radial position of leaf element, and its value range is: .

[0108] Step 2.4: Based on the process in Steps 2.1-2.3, obtain the tail rotor distributed aerodynamic force based on the isolated tail rotor momentum source model and the tail rotor distributed aerodynamic force based on the rotor-tail rotor dual momentum source model.

[0109] Step 3: Subtract the tail rotor aerodynamic forces from the two models obtained in Step 2 to obtain the tail rotor aerodynamic forces purely caused by rotor airflow interference. The subtraction formula is:

[0110]

[0111]

[0112] in, and These are the thrust and drag coefficients of the upper blade element of the tail rotor, obtained based on the isolated tail rotor momentum source model. and These are the thrust coefficient and drag coefficient of the upper blade element of the tail rotor, obtained based on the rotor-tail rotor dual momentum source model. for and The difference represents the tail rotor distributed thrust caused by rotor wake interference; for and The difference represents the tail rotor distributed drag caused by rotor wake interference.

[0113] To verify the effectiveness of the method of the present invention, this embodiment performs a calculation verification on the rotor / tail rotor interference of a certain type of helicopter in hovering state according to the above method. The calculation parameters are as follows: incoming flow velocity 18 m / s, sideslip angle range 0°~210°, tail rotor speed 828.11 rad / s, rotor speed 186 rad / s, and one rotor rotation cycle corresponds to 4.45 tail rotor rotation cycles. The sideslip angle refers to the angle between the incoming flow wind speed and the forward direction of the rotor disk.

[0114] The calculation results of this invention are compared with CFD results based on slip grids and known wind tunnel test data. The comparison results are as follows: Figure 6 As shown. From Figure 6 As can be seen, within the sideslip angle ranges of 0°–30° and 150°–210°, the error of the tail rotor thrust coefficient calculated by the method of this invention (based on the CFD momentum source model) relative to the experimental value is 2.31%–4.50%, while the error of the unsteady flow field simulation method based on the slip grid relative to the experimental value is -4.50%–1.20%. The result error of the method of this invention (based on the CFD momentum source model) relative to the unsteady flow field simulation method based on the slip grid is 1.53%–9.43%. Therefore, within this sideslip angle range, both the method of this invention and the unsteady flow field simulation method based on the slip grid have high computational accuracy. Within the sideslip angle range of 30° to 150°, the close proximity of the tail rotor to the wind tunnel wall enhances wall interference, effectively creating a confined flow field that weakens the free development and energy dissipation of the wake, leading to a deviation in experimental results compared to actual free flight. Unsteady flow field simulation methods, employing free-flow far-field boundary conditions and neglecting wall interference, result in predictions lower than experimental results affected by wall interference. This invention, however, does not directly analyze the actual wake structure but instead averages it, weakening wake diffusion and causing a slower induced velocity decay compared to unsteady flow field simulations. Therefore, the calculated total tail rotor thrust is... The higher value makes the calculation results of the method of this invention closer to the experimental results affected by the tunnel wall interference, that is, it shows the total thrust of the tail rotor, just like the experimental value. The increasing trend makes it more suitable for simulating the aerodynamic distribution characteristics of the tail rotor in a wind tunnel environment.

[0115] In terms of computational efficiency, the method of this invention takes about 12 hours (64 cores in parallel), while the unsteady flow field simulation method based on sliding mesh takes about 13 days. The former is about 26 times more efficient than the latter (that is, the computation time of the method of this invention is only about 3.85% of that of the unsteady flow field simulation method).

[0116] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the scope of the technology disclosed in the present invention, and such modifications or substitutions should all be covered within the scope of protection of the present invention.

Claims

1. A method for calculating the aerodynamic force of tail rotor in the flow of helicopter rotor based on a double momentum source model, characterized in that, Includes the following steps: Step 1: Establish a momentum source model and perform CFD flow field solution to extract the tail rotor disk induced velocity; Includes the following sub-steps: Step 1.1: Establish an isolated tail rotor momentum source model and a rotor-tail rotor dual momentum source model that simultaneously includes a rotor momentum source model and a tail rotor momentum source model; Step 1.2: Using the tail rotor target thrust determined by the overall parameters of the helicopter as input, perform CFD flow field solution on the isolated tail rotor momentum source model, and back-calculate the tail rotor collective pitch and the tail rotor full flow field information under the isolated tail rotor momentum source model. Step 1.3: Use the collective pitch of the tail rotor convergence as the input collective pitch of the tail rotor momentum source model in the rotor-tail rotor dual momentum source model, and use the rotor target thrust determined by the overall parameters of the helicopter as the input of the rotor momentum source model in the rotor-tail rotor dual momentum source model. Perform CFD flow field solution on the rotor-tail rotor dual momentum source model to obtain the full flow field information of the tail rotor under the rotor-tail rotor dual momentum source model. Step 1.4: Extract the corresponding tail rotor disk induced velocity from the tail rotor full flow field information of the isolated tail rotor momentum source model and the tail rotor full flow field information of the rotor-tail rotor dual momentum source model, respectively. Step 2: Based on the blade element theory, the tail rotor disk induced velocity obtained in Step 1 is converted into the corresponding tail rotor distributed aerodynamic force. Step 3: Subtract the tail rotor distributed aerodynamic force based on the rotor-tail rotor dual momentum source model from the tail rotor distributed aerodynamic force based on the isolated tail rotor momentum source model to obtain the tail rotor distributed aerodynamic force purely caused by rotor airflow interference: in, and These are the thrust and drag coefficients of the upper blade element of the tail rotor, obtained based on the isolated tail rotor momentum source model. and These are the thrust coefficient and drag coefficient of the upper blade element of the tail rotor, obtained based on the rotor-tail rotor dual momentum source model. for and The difference represents the tail rotor distributed thrust caused by rotor wake interference; for and The difference represents the tail rotor distributed drag caused by rotor wake interference.

2. The method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum source model according to claim 1, characterized in that, In step 1.1, the isolated tail rotor momentum source model and the rotor-tail rotor dual momentum source model adopt the same grid topology and node distribution, and a grid densification zone is set in the spatial region corresponding to the expected position of the rotor. No momentum source is applied to the grid densification zone of the isolated tail rotor momentum source model to ensure the numerical consistency when performing the difference in step 3.

3. The method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum source model according to claim 1, characterized in that, In step 1.2, the collective pitch of the tail rotor and the full flow field information of the tail rotor under the isolated tail rotor momentum source model are obtained through back-calculation. The process includes the following: The target thrust of the tail rotor, determined by the overall parameters of the helicopter, is used as the input to the isolated tail rotor momentum source model. The flow field of the isolated tail rotor momentum source model is solved by a CFD solver to calculate the thrust corresponding to the current tail rotor collective pitch. Based on the difference between the thrust corresponding to the current tail rotor collective pitch and the target thrust of the tail rotor, the tail rotor collective pitch is dynamically calculated and adjusted using an automatic feedback adjustment algorithm. Repeat the CFD solution steps and the tail rotor collective pitch calculation and adjustment steps until the difference between the thrust output by the isolated tail rotor momentum source model and the target thrust of the tail rotor is less than or equal to a preset error threshold and converges; record the converged collective pitch as the tail rotor converged collective pitch; the tail rotor converged collective pitch represents the collective pitch value required for the isolated tail rotor to generate the target thrust of the tail rotor in an environment without rotor interference.

4. The method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum source model according to claim 1, characterized in that, In step 1.4, the tail rotor disk induced velocity is extracted from the tail rotor full flow field information of the isolated tail rotor momentum source model and the tail rotor full flow field information of the rotor-tail rotor dual momentum source model, respectively. The process includes the following: The extraction area is defined as a plane that is parallel to the tail rotor disk plane and located 0.2 to 0.5 times the tail rotor radius downstream of the tail rotor disk plane. The area of ​​the extraction area is equal to the area of ​​the rotor disk. Within the extraction area, the velocity perpendicular to the disk plane is extracted as the tail rotor disk induced velocity; and the tail rotor disk induced velocity extracted from the tail rotor full flow field information of the isolated tail rotor momentum source model is denoted as V1, and the tail rotor disk induced velocity extracted from the tail rotor full flow field information of the rotor-tail rotor dual momentum source model is denoted as V2.

5. The method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum source model according to claim 4, characterized in that, The process of converting the tail rotor disk-induced velocity obtained in step 1 into the corresponding tail rotor distributed aerodynamic force based on blade element theory includes the following steps: Step 2.1: Using the extracted tail rotor disk induced velocity, calculate the actual angle of attack corresponding to the true incoming flow direction at each blade element position of the tail rotor. and Mach number The calculation formula is: In the formula, It is the incoming flow velocity; It is the local speed of sound; It is the specific heat ratio; It is the gas constant; For absolute temperature, the specific heat ratio is relative to air. The value is 1.4, and the gas constant is... It is 287 J / (kg·K). This invention uses a standard atmospheric pressure environment and an absolute temperature of 287 J / (kg·K). The calculated local speed of sound is 288.15K. It is 340.3 m / s; It is the tail rotor speed; It is the installation angle of the leaf element section; It is the leaf element azimuth angle; It is the angle between the incoming airflow velocity and the plane of the propeller disk; The tail rotor radius; For leaf element radial positioning; For induced velocity, where, when calculated based on the isolated tail rotor momentum source model, The tail rotor disk-induced velocity V1 is extracted from the full flow field information of the isolated tail rotor momentum source model; when calculating based on the rotor-tail rotor dual momentum source model... Take the tail rotor disk induced velocity V2 extracted from the full flow field information of the rotor-tail rotor dual momentum source model; The axial velocity of the leaf element; The circumferential velocity of the leaf element; The radial velocity of the leaf element; The actual induction angle of the leaf element; Step 2.2, radial positioning of leaf elements And the actual angle of attack of the leaf element was calculated. and Mach number The input is fed into a pre-built and trained neural network model to predict the lift coefficient at the position of each blade element in the wind coordinate system. and drag coefficient ; Step 2.3: Assign the lift coefficients corresponding to all blade elements in the wind coordinate system. and drag coefficient Switch to propeller disk In the coordinate system, the aerodynamic characteristics of the blade element in the propeller disk coordinate system are obtained; the conversion formula is: In the formula, This represents the thrust coefficient of the blade element in the propeller disk coordinate system. This represents the drag coefficient of the blade element in the propeller disk coordinate system. Step 2.4: Based on the processes in Steps 2.1-2.3, calculate the tail rotor distributed aerodynamic force based on the isolated tail rotor momentum source model and the tail rotor distributed aerodynamic force based on the rotor-tail rotor dual momentum source model.

6. The method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum source model according to claim 5, characterized in that, In step 2.2, the neural network model adopts a multilayer perceptron neural network, whose input layer nodes correspond to the angle of attack, Mach number and blade radial position that affect the aerodynamic forces of the airfoil, and the output layer nodes correspond to the lift coefficient and drag coefficient; a regularization algorithm is used to train and optimize the multilayer perceptron neural network.

7. The method for calculating the aerodynamic forces of the tail rotor in the airflow of a helicopter rotor based on a dual-momentum source model according to claim 5, characterized in that, It also includes obtaining the total thrust of the tail rotor. and total tail rotor drag The steps involve calculating the tension coefficients of all blade elements on the propeller disk obtained in step 2.

3. and drag coefficient Integrating, we get: In the formula, It is the leaf element azimuth angle, and its value range is: ; It is the radial position of leaf element, and its value range is: .

Images