Method for generating three-dimensional external channel sound adiabatic parameter of aero-engine

Abstract

The application discloses an aero-engine three-dimensional outer channel sound attenuation parameter generation method, comprising the following steps: constructing a three-dimensional fan post-transmission noise physical simplified calculation model according to a frequency domain convection wave control equation; and obtaining an optimal solution of the model with high precision based on a high-order space discretization and time advancing algorithm to obtain parameters corresponding to aero-engine three-dimensional outer channel sound attenuation. The application solves the problem that a traditional two-dimensional sound attenuation design method can only reduce noise at a specific frequency under one working condition or at most two working conditions, has no inhibitory effect on noise under other working conditions, solves the defect of a narrow sound absorption frequency band of the traditional two-dimensional sound attenuation design method, proposes a three-dimensional fan post-transmission noise calculation method, a gradient guide, a multi-island genetic algorithm and a secondary optimization combination technology with consideration of precision and efficiency, and develops a three-dimensional multi-section outer channel sound attenuation design method, which can inhibit noise at three noise airworthiness examination points of approach, sideline and flyover, significantly expands a sound absorption frequency band, and has a broadband sound attenuation effect.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of aero-engine noise control, and particularly relates to a three-dimensional outer channel sound lining parameter generation method of an aero-engine. BACKGROUND

[0002] The Federal Aviation Regulation (FAR) Part 36, the European Joint Aviation Standard Part 36 and the International Civil Aviation Convention (ICAO) Annex 16 all make very strict airworthiness noise regulations for commercial aircraft. The bypass ratio of contemporary civil turbofan engines is getting larger and larger, even exceeding 10, and the exhaust airflow speed is greatly reduced. Compared with compressor, jet and turbine noise, fan noise has become the main noise source of modern high-bypass-ratio engines. Figure 1 A 150-seat aircraft is shown in the noise source distribution at each noise observation point. Among them, the fan forward transmission noise, the fan backward transmission noise, the turbine noise and the jet noise are shown; wherein 1-4 correspond to the sideline operating condition, 6-9 correspond to the flyover operating condition, and 11-14 correspond to the approach operating condition. It can be known from the above that for a modern civil aircraft, fan noise is the main noise source during the sideline, flyover and approach of the aircraft. As a major component of fan noise, fan backward transmission noise is the most important noise source of the aero-engine flyover and approach. The physical process that the noise is propagated and radiated to the far field through the channel is particularly worth paying attention to.

[0003] Since the emergence of jet engines, sound lining has been the most important means of aero-engine noise control, and all civil aircraft in operation today rely on this technology to meet the noise specifications of international airworthiness regulations. With the gradual increase in the size of aero-engines, the problem of low-frequency broadband noise is becoming increasingly prominent, and the increase in the diameter and length ratio of the noise reduction nacelle also makes the design of sound lining more difficult. In addition, there are high-speed non-uniform airflow and high sound pressure level environment, complex multi-modal composite sound field, variable cross-section, pipe radiation, wall shear flow and other complex conditions in the outer channel, which all cause great difficulty in the design of aero-engine noise reduction sound lining.

[0004] Bypass duct is one of the airflow ducts of an aero-engine. Fan noise can be propagated forward through the inlet duct and backward through the bypass duct to the atmosphere. Sound propagation is a physical property of noise. After aerodynamic noise is generated at a sound source, it will spread and propagate in all directions. Sound propagation refers to this physical process. Acoustic liner is an effective noise reduction means in engine noise control. A typical single-layer acoustic liner is mainly composed of a panel, a honeycomb core and a rigid back plate. The panel is often a perforated plate, a wire mesh or a perforated sandwich panel. The honeycomb core is connected behind the panel to form a core cavity. The rigid back plate connected to the honeycomb core seals the core cavities and isolates them from each other, thereby forming a single resonator. The conventional acoustic liner is a two-dimensional axisymmetric acoustic liner.

[0005] In the prior art, the transfer unit method based on the analytical prediction model can meet the engineering needs in terms of calculation efficiency and prediction accuracy, but can only carry out two-dimensional sound propagation analysis. The sound propagation calculation based on the sound ray theory can meet the engineering needs in terms of calculation efficiency, but the prediction accuracy is not suitable for the prediction of bypass duct sound propagation. The finite element method and the aerodynamic acoustic method based on numerical methods can meet the engineering requirements in terms of calculation accuracy and can also consider three-dimensional sound propagation problems, but the calculation efficiency is low and a large amount of computing resources are required.

[0006] Since the acoustic liner design needs to calculate a large number of examples, and the calculation of a single three-dimensional noise calculation is time-consuming, the acoustic liner design in the prior art is mostly two-dimensional axisymmetric design.

[0007] Two-dimensional acoustic liner design has many assumptions and cannot consider real three-dimensional duct configurations, variable cross-section, finite length ducts and wall shear flow, etc. The real bypass duct structure of a civil aviation engine is complex and has non-uniform flow, boundary reflection and other phenomena. The method based on the analytical prediction model can meet the engineering needs in terms of calculation efficiency and prediction accuracy, but can only carry out two-dimensional sound propagation analysis. The finite element method and the CAA method based on numerical methods can meet the engineering requirements in terms of calculation accuracy and can also consider three-dimensional sound propagation problems, but the calculation efficiency is low and the cost of three-dimensional acoustic liner design calculation is unbearable. The three-dimensional noise reduction acoustic liner design calculation of fan backward transmission noise is difficult and costly. In order to solve this technical problem, an efficient fan backward transmission noise calculation model and technology must be developed.

[0008] The published aero-engine acoustic liners are mostly two-dimensional axisymmetric single-degree-of-freedom inlet acoustic liners, and are single-section uniform acoustic liners, which are designed for a single main propagation mode at the noise frequency of a working condition. The newly served new engine updates the acoustic liner technology and develops a double-degree-of-freedom acoustic liner, which can be designed for two working conditions of noise frequency. The acoustic liner obtained in this way can only reduce the noise at a specific frequency under one working condition or at most two working conditions, and has no inhibitory effect on the noise under other working conditions. With the increasing of the noise specification of airworthiness regulations, the indexes of approach, sideline and flyover three noise airworthiness test points are becoming more and more stringent. The noise control problem of the aero-engine under the three working conditions becomes a challenging and urgent problem. In order to cope with this technical challenge, it is necessary to develop a noise reduction acoustic liner technology that can reduce the noise of approach, sideline and flyover three airworthiness working conditions at the same time.

[0009] The conventional optimization method first calculates a large amount of sample point data, then establishes a surrogate model based on this, obtains optimal data by optimizing the surrogate model, and repeatedly repeats this process to obtain the optimal point. The conventional method needs many initial samples, long iteration period and long optimization calculation time. SUMMARY

[0010] The technical problem to be solved by the present application is to overcome the defects in the prior art that the two-dimensional acoustic liner cannot consider complex conditions such as variable cross-section, finite length pipeline and wall shear flow, and to provide an aero-engine three-dimensional bypass acoustic liner parameter generation method.

[0011] The present application solves the above technical problems by the following technical solutions:

[0012] The present application provides an aero-engine three-dimensional bypass acoustic liner parameter generation method, comprising the following steps:

[0013] S1, constructing a three-dimensional fan post-transmission noise physical simplified calculation model according to a frequency domain convection wave control equation;

[0014] S2, obtaining an optimal solution of the three-dimensional fan post-transmission noise calculation model based on a high-order spatial discretization and time advancing algorithm to obtain parameters corresponding to the aero-engine three-dimensional bypass acoustic liner. Preferably, the spatial discretization includes spatial discretization according to a 4th-order dispersion relationship preserving scheme or a compact scheme.

[0015] Preferably, the time advancing algorithm includes any one of a 4th-order 4 / 6th-level low-dissipation, low-dispersion Runge-Kutta method and a leapfrog scheme.

[0016] Preferably, the step S2 comprises:

[0017] S201, determining the value range of the resistance and the reactance of the aero-engine three-dimensional bypass acoustic liner;

[0018] S202, determine sample point data;

[0019] S203, calculate the noise target value of the initial sample point by using a three-dimensional fan post-transmission noise physical simplified calculation model;

[0020] S204, extract gradient information of the design variable;

[0021] S205, establish a proxy model guided by the gradient information;

[0022] S206, search for an optimization point on the proxy model by using a multi-island genetic algorithm;

[0023] S207, perform two rounds of optimization near the optimization point;

[0024] S208, judge whether the convergence condition is met, if yes, take the result of the two rounds of optimization as the parameters corresponding to the aero-engine three-dimensional bypass duct sound lining, if not, return to step S202.

[0025] Preferably, the dimensionless resistance value ranges from -5 to 5, and the dimensionless resistance value ranges from -5 to 5.

[0026] Preferably, the sample point data includes no less than 60 groups of data.

[0027] Preferably, the target value includes transmission loss, insertion loss, and sound pressure level at far field.

[0028] Preferably, step S1 includes:

[0029] In the curvilinear coordinate system, the Laplacian operator is used to transform the frequency domain convection wave control equation to obtain a first equation

[0030] An operator expression is introduced in the first equation to obtain a second equation

[0031] The second equation is factorized to obtain a third equation

[0032] The coupling problem of two-way sound propagation is approximated as a one-way sound propagation problem to obtain a fourth equation

[0033] Define λ = λ0(1 + s) 0.5 , and define the operator (1 + s) 0.5 is expanded by using Taylor series, and the fourth equation is transformed into a fifth equation

[0034] Write the fifth equation in the form to get the three-dimensional fan back noise physical simplified calculation model

[0035] Wherein, x1 represents Cartesian coordinates, p represents sound pressure, i represents frequency domain space, k0 is sound wave number, M a is the Mach number of the incoming flow, is the Mach number related operator. is the pressure related part value in Laplace operator. Subscript gamma represents the cross section;

[0036] is the coordinate transformation coefficient, and xi represents the curve coordinate system; sigma, V and T are operator expressions;

[0037] k0 is the wave number k0 and Mach number related operator;

[0038] a n , n = 0, 1…∞ is the coefficient of Taylor series expansion formula.

[0039] The positive progress effect of the application is that the application solves the problem that the traditional two-dimensional sound lining design method can only reduce the noise at a specific frequency under one working condition or at most two working conditions, and has no inhibitory effect on the noise under other working conditions, solves the defect that the sound absorption frequency band of the traditional two-dimensional sound lining design method is narrow, proposes a three-dimensional fan back noise calculation method considering accuracy and efficiency, and a technology combining gradient guidance and secondary optimization, develops a three-dimensional multi-section outer channel sound lining design method, can inhibit the noise at approach, sideline and flyover three noise airworthiness test points, significantly expands the sound absorption frequency band, and has a wide frequency sound absorption effect. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 It is a schematic diagram of noise source distribution of a 150-seat passenger plane at each noise airworthiness observation point.

[0041] Figure 2 It is a flowchart of the parameter generation method of the three-dimensional outer channel sound lining of the aircraft engine of a preferred embodiment of the application.

[0042] Figure 3 It is a flowchart of step S1 of the parameter generation method of the three-dimensional outer channel sound lining of the aircraft engine of a preferred embodiment of the application.

[0043] Figure 4 It is a flowchart of step S2 of the parameter generation method of the three-dimensional outer channel sound lining of the aircraft engine of a preferred embodiment of the application. DETAILED DESCRIPTION

[0044] The application will be further described below by a preferred embodiment, but the application is not limited in the scope of the described embodiment.

[0045] The embodiment provides a method for generating a three-dimensional outer channel sound adiabatic parameter of an aero-engine. Figure 2 The method for generating the three-dimensional outer channel sound adiabatic parameter of the aero-engine comprises the following steps: step S1, constructing a three-dimensional fan post-transmission noise physical simplified calculation model according to a frequency domain convection wave control equation;

[0046] Step S2, obtaining an optimal solution of the three-dimensional fan post-transmission noise calculation model based on a high-order space discretization and time advancing algorithm to obtain a parameter corresponding to the three-dimensional outer channel sound adiabatic parameter of the aero-engine.

[0047] The technical principle of the method for generating the three-dimensional outer channel sound adiabatic parameter of the aero-engine lies in: developing a three-dimensional high-efficiency noise calculation model (i.e., the three-dimensional fan post-transmission noise physical simplified calculation model) to reduce the evaluation time of a single noise example, developing an efficient optimization technology to reduce the optimization search time, and reducing the three-dimensional sound adiabatic optimization time. A multi-section sound adiabatic design method is developed to suppress the noise at the approach, sideline and flyover three noise airworthiness test points.

[0048] In the embodiment, in step S1, a high-efficiency calculation model (i.e., the three-dimensional fan post-transmission noise physical simplified calculation model) is developed, which takes into account the efficiency and accuracy. The fan post-transmission noise has the physical characteristic of mainly propagating downstream. By using curve coordinate transformation, factor decomposition, Taylor expansion and other mathematical derivations, the two-way sound propagation problem in the pipe is decoupled and dimensionally reduced to a one-way sound propagation problem, and a high-efficiency sound propagation model is established.

[0049] In specific implementation, referring to Figure 3 , step S1 comprises the following steps:

[0050] Step S101, in a curve coordinate system, a Laplace operator is used to transform a frequency domain convection wave control equation to obtain a first equation.

[0051] The frequency domain convection wave control equation is as follows:

[0052]

[0053] Wherein, x1 represents a Cartesian coordinate, p represents sound pressure, i represents a frequency domain space, k0 is a sound wave number, M a is a Mach number of a flow, is an operator related to the Mach number. is a part value related to pressure in the Laplace operator. The subscript γ represents a cross section.

[0054] In the curve coordinate system, the Laplace operator can be written as:

[0055]

[0056] where, is the Laplace operator, is the coordinate transformation coefficient. ξ1 represents the curvilinear coordinate system.

[0057] Bringing these coordinate transformation relations into the frequency-domain convection wave equation, the first equation can be obtained:

[0058] Step S102, introducing the operator expression in the first equation to obtain the second equation.

[0059] In the implementation, three operators are introduced:

[0060]

[0061] where, σ, V, T are the operator expressions, and the meanings of other parameters are as shown above.

[0062] Therefore, the first equation can be expressed as the second equation:

[0063]

[0064] Step S103, factorizing the second equation to obtain the third equation.

[0065] The second equation is a simple quadratic equation of p, and the solution of p can be written as:

[0066]

[0067] where, the positive and negative signs respectively represent two roots,

[0068]

[0069] For the above equation, the third equation can be obtained through the factorization technique:

[0070]

[0071] Step S104, approximating the coupling problem of bidirectional sound propagation into a unidirectional sound propagation problem to obtain the fourth equation.

[0072] For the fan back noise problem, the sound wave propagates along the negative direction, so the coupling problem of bidirectional sound propagation is approximated into a unidirectional sound propagation problem to obtain the fourth equation:

[0073]

[0074] Step S105, using the Taylor series expansion method to transform the fourth equation into the fifth equation.

[0075] where, λ is defined as λ = λ0(1 + s) 0.5In comparison with the lambda expression, we have

[0076]

[0077] where λ0is an operator related to wave number k0and Mach number.

[0078] Define the operator Using the Taylor series expansion method, we have

[0079] where a n ,n=0,1…∞ are the coefficients of the Taylor series expansion formula.

[0080] Substitute the fourth equation into the fifth equation to obtain the sixth equation:

[0081]

[0082] Step S106, write the fifth equation in a display form to obtain a three-dimensional fan back-transmission noise physical simplified calculation model.

[0083] Write the fifth equation in a display form to obtain a three-dimensional fan back-transmission noise physical simplified calculation model:

[0084] The equation is a first-order partial differential equation, which can be solved by using a high-order numerical method. The meanings of various parameters in the formula are shown above.

[0085] Compared with the aerodynamic problem, for the aeroacoustic problem, it is not enough to only consider the stability problem, and the dissipation and dispersion errors of the sound wave also need to be considered. The control equation of the noise problem is usually unsteady and has parabolic characteristics. In step S2, the high-order spatial discretization can be performed by using the 4th-order dispersion relation preserving (DRP) scheme or the compact scheme proposed by Tam and Webb. The time advancing algorithm can be performed by using the 4th-order 4 / 6 low-dissipation and low-dispersion Runge-Kutta method (LDDRK) or the leapfrog scheme proposed by Hu.

[0086] The conventional optimization method based on the proxy model adopts a brute force search, and processes the optimization problem as a black box problem, which has a large amount of calculation. The method proposed in the embodiment is as follows, and the method is described with reference to Figure 4 , and step S2 includes the following steps:

[0087] Step S201, determine the value range of the resistance value and the reactance value of the three-dimensional external duct acoustic liner of the aero-engine.

[0088] The design variable (i.e., the parameter of the acoustic liner) of each acoustic liner is the resistance value and the reactance value. First, define the upper and lower limits of the resistance value and the reactance value of each acoustic liner. The dimensionless resistance value R ranges from-5 to 5, and the dimensionless reactance value X ranges from-5 to 5.

[0089] Step S202, determining sample point data.

[0090] In practice, a sample point covering most of the design space is designed by a design of experiment method, generally not less than 60 groups of data.

[0091] Step S203, calculating the noise target value of the initial sample point by using a three-dimensional fan post-transmission noise physical simplified calculation model.

[0092] The target value of the initial sample point is calculated by using a high-efficiency noise calculation model, including transmission loss, insertion loss, sound pressure level size at the far field, etc.

[0093] Step S204, extracting gradient information of the design variable.

[0094] The gradient information of the design variable is extracted to find the internal relationship between the design variable and the objective function (i.e. the noise target value), and to convert the black box optimization problem into a controllable problem with a direction to follow.

[0095] The design variable gradient information is added to the proxy model (i.e. an approximate mathematical model in optimization theory, which is a method of approximating a set of design variable input values and target value output values by a mathematical model. This method is a fast modeling method that uses an approximate mathematical model instead of the original numerical calculation method), and a higher-precision proxy model guided by the gradient information is established.

[0096] Step S206, searching for an optimization point on the proxy model by using a multi-island genetic algorithm.

[0097] The multi-island search technology of the multi-island genetic algorithm avoids the defect that the conventional genetic algorithm is easy to find a local optimal point. The optimization point is quickly searched on the higher-precision proxy model. Compared with the optimization point obtained by the conventional method, the optimization point obtained by this method is closer to the true optimal point.

[0098] The conventional method is based on the initial sample point to perform design of experiment. In this embodiment, two rounds of optimization are performed near the optimization point, and the probability and speed of finding the optimal solution are significantly improved. Based on the optimization point obtained in step S206, a sample point covering most of the design space is designed by a design of experiment method near the optimization point.

[0099] Step S208, judging whether the convergence condition is met, if yes, executing step S209; if no, returning to step S202.

[0100] Step S209, taking the result of the two rounds of optimization as the corresponding parameters of the aero-engine three-dimensional external channel sound lining.

[0101] That is, if the convergence judgment condition is not satisfied, steps S202 to S207 are repeated until the convergence judgment condition is satisfied. The convergence judgment condition is generally that the difference between the value corresponding to the optimization point and the value after two rounds of optimization is less than 0.0001.

[0102] Based on the above search strategy, the optimal solution of the three-dimensional fan back transmission noise physical simplified calculation model can be obtained in general 2-3 rounds of optimization, and the optimization speed is significantly higher than that of the conventional optimization method.

[0103] Then, a three-dimensional multi-section sound lining is constructed based on the above optimal solution. In specific implementation, each section of the sound lining is designed for the approach, sideline and flyover working conditions respectively, and the resistance and reactance of each section of the sound lining are different, and the sound lining as a whole exhibits non-uniformity, achieving the purpose of simultaneously reducing the noise in the three airworthiness conditions.

[0104] The three-dimensional fan back transmission noise efficient calculation method considering precision and efficiency is proposed in the aircraft engine three-dimensional outer channel sound lining parameter generation method, and the gradient guidance and secondary optimization combined technology is proposed to quickly find the global optimal solution and significantly reduce the three-dimensional sound lining design iteration calculation time. Further, the multi-section sound lining design method is proposed for three-dimensional outer channel sound lining design, which can consider the real three-dimensional pipeline configuration and complex flow field.

[0105] The aircraft engine three-dimensional outer channel sound lining parameter generation method solves the problem that the traditional two-dimensional sound lining design method can only reduce the noise at a specific frequency under one working condition or at most two working conditions, and has no inhibitory effect on the noise under other working conditions, solves the defect of narrow sound absorption frequency band of the traditional two-dimensional sound lining design method, proposes a three-dimensional fan back transmission noise calculation method considering precision and efficiency and a gradient guidance and secondary optimization combined technology, and develops a three-dimensional multi-section outer channel sound lining design method, which can suppress the noise at the approach, sideline and flyover three noise airworthiness test points, significantly expands the sound absorption frequency band, and has a broadband noise reduction effect.

[0106] Although the specific embodiments of the present application are described above, those skilled in the art should understand that this is only an example, and the protection scope of the present application is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of the present application, and these changes and modifications all fall within the protection scope of the present application.

Claims

1. A method for generating three-dimensional outer bypass duct acoustic liner parameters for an aero-engine, characterized in that, Includes the following steps: S1. Construct a simplified physical calculation model of three-dimensional fan back-transmission noise based on the frequency domain convection wave control equation; S2. Obtain the optimal solution of the three-dimensional fan back-transmission noise calculation model based on the high-order spatial discretization and time propagation algorithm to obtain the parameters corresponding to the three-dimensional bypass duct acoustic liner of the aero-engine; Step S1 includes: In a curvilinear coordinate system, using the Laplacian operator For frequency domain convective wave control equations Transform the equation to obtain the first equation. Introducing operator expressions into the first equation To obtain the second equation Factorize the second equation to obtain the third equation. Based on the physical characteristics of noise propagation downstream within the duct, the coupling problem of bidirectional sound propagation can be transformed into a unidirectional sound propagation problem, leading to the fourth equation. Define the simplified parameter λ = λ0(1+s) 0.5 Define the operator (1+s) 0.5 By using Taylor series expansion, the fourth equation is transformed into the fifth equation. The fifth equation is written in explicit form to obtain the physical simplified calculation model of the three-dimensional fan back-transmission noise. Where x1 represents the Cartesian coordinates, p represents the sound pressure level, i represents the frequency domain space, k0 is the sound wave number, and M... a For the Mach number of the incoming flow, It is an operator related to the Mach number; This is the pressure-related value in the Laplace operator, and the subscript γ indicates the cross section; These are coordinate transformation coefficients, where ξ1 represents the curvilinear coordinate system; σ, V, T are operator expressions; λ0 is an operator related to the wave number k0 and the Mach number; a n , n=0,1…∞ are the coefficients of the Taylor series expansion formula.

2. The method for generating three-dimensional outer bypass duct acoustic liner parameters for aero-engines as described in claim 1, characterized in that, The spatial discretization includes spatial discretization based on a format-preserving or compact format according to a fourth-order dispersion relation.

3. The method for generating three-dimensional outer bypass duct acoustic liner parameters for aero-engines as described in claim 1, characterized in that, The time-progression algorithm includes any one of the following: 4th order 4 / 6 level low dissipation, low dispersion Runge-Kutta method, and frog-jump scheme.

4. The method for generating three-dimensional outer bypass duct acoustic liner parameters for aero-engines as described in claim 1, characterized in that, Step S2 includes: S201. Determine the range of values ​​for the resistance and impedance of the acoustic liner of the three-dimensional bypass duct of the aero-engine; S202. Determine the sample point data; S203. Calculate the target noise value of the initial sample points using the three-dimensional fan back-transmission noise physical simplification calculation model. S204. Extract the gradient information of the design variables; S205. Establish an agent model guided by gradient information; S206. The multi-island genetic algorithm is used to search for the optimal point on the surrogate model; S207. Perform a second round of optimization near the optimization point; S208. Determine whether the convergence condition is met. If yes, use the result of the second round of optimization as the parameter corresponding to the acoustic liner of the three-dimensional bypass duct of the aero-engine. If not, return to step S202.

5. The method for generating three-dimensional outer bypass duct acoustic liner parameters for aero-engines as described in claim 4, characterized in that, The dimensionless resistance value ranges from (-5, 5), and the dimensionless reactance value ranges from (-5, 5).

6. The method for generating three-dimensional outer bypass duct acoustic liner parameters for aero-engines as described in claim 4, characterized in that, The sample point data includes no fewer than 60 sets of data.

7. The method for generating three-dimensional outer bypass duct acoustic liner parameters for aero-engines as described in claim 4, characterized in that, The target values ​​include transmission loss, insertion loss, and sound pressure level in the far field.

Images