A method for constructing load spectra of aviation turbojet and turbofan engines considering randomness

Abstract

The present invention belongs to the field of aviation engine technology, and specifically relates to a method for constructing a load spectrum of aviation turbojet and turbofan engines taking into account randomness, including collecting engine field use data to obtain load spectrum characteristic parameters; performing random distribution fitting on the load spectrum characteristic parameters to obtain the random distribution law of the load spectrum characteristic parameters; randomly extracting the load spectrum characteristic parameters to generate characteristic parameter samples; using the characteristic parameter samples to construct a parameter spectrum related to the engine operating environment and a high-pressure rotor speed spectrum, which together constitute the engine load spectrum. Based on field flight data, the present invention takes into account the most important load related to the durability of the engine rotor structure, proposes a reasonable relationship between load characteristics and a preferred distribution law, and provides a reasonable basis for the durability evaluation of the engine structure.

Description

Technical Field

[0001] The invention belongs to the technical field of aviation engines, and in particular relates to a method for constructing a load spectrum of aviation turbojet and turbofan engines taking randomness into consideration. Background Art

[0002] Aviation turbojet and turbofan engines are the main power units of advanced aircraft. The durability of their rotor structures constrains the safety and service life of aircraft. The engine load spectrum contains the engine operating status history, reflects the specific loads during the engine's mission execution, and is an important basis for engine structural service damage assessment, structural durability testing, and service mission design. The engine load spectrum specifically includes the parameter spectrum related to the engine's operating environment and the engine's operating status load. In the actual service process, affected by the mission objectives, pilot operations, and local air conditions, the engine is subjected to a strong time-varying random load process within a wide range of working conditions. With the improvement of the mechanical properties of aviation engines and materials, the richness of engine mission types and the randomness of service loads are constantly increasing, which has aggravated the dispersion of the actual service life of the engine, and should be paid attention to in the engine load spectrum work. By establishing an engine load spectrum that takes randomness into account, it is possible to evaluate the load severity of the design service mission, design engine structure or whole machine durability tests, and conduct structural damage assessment in combination with the actual use of the engine during the use phase.

[0003] The main factors affecting the durability of engine structures involve fatigue damage caused by structural stress cycles and creep damage caused by long-term residence in high-speed states. Traditionally, the life determination of engines and their main components relies on the engine design load spectrum. On this basis, the fatigue life and creep life of the engine structure are managed in combination with parameters such as the number of engine take-offs and landings and the use time. This method obviously cannot represent the complexity of the actual use load of the engine. At present, no technical research has been found to carry out the design of load spectra for aviation turbojets and turbofan engines by considering the randomness of durability-related load spectra.

[0004] At present, the main methods for constructing aircraft engine load spectra are: Patent CN112115787A "A method for dividing aircraft engine load spectrum mission segments based on actual flight actions" divides the original measured spectrum into take-off, middle and landing segments, summarizes relevant flight actions, selects characteristic parameters, performs rain flow filtering and data processing, thereby extracting flight mission segments representing basic and complex flight actions, and realizing effective mission segment division of the middle segment of the speed spectrum. However, this method mainly focuses on the deterministic logical relationship between each flight action in the mission, and does not take into account the random changes in flight loads caused by many service factors, so it cannot provide random load input for durability research.

[0005] Patent CN117763445A "Method for compiling a comprehensive mission spectrum for aircraft engines based on load dispersion characteristics" extracts fatigue loads, creep loads and thermal shock loads from the mission profile, establishes a load database and gives its probability distribution, then extracts various types of loads based on random sampling, and finally compiles a comprehensive mission spectrum. Although this invention takes into account the load randomness corresponding to the three types of damage forms, it only models specific flight actions and does not establish a complete engine multi-parameter load spectrum from starting to stopping, and has large application limitations. At the same time, the load randomness characterization method provided by this invention cannot take into account the mutual influence and interference relationship between different loads, and is not suitable for a wider load range and complex load combinations.

[0006] Therefore, under the above background, it is necessary to establish a method for constructing aero-engine load spectrum taking into account the randomness of load. Summary of the invention

[0007] In order to solve the above technical problems, the present invention provides a method for constructing a load spectrum of aviation turbojet and turbofan engines taking into account randomness. Based on the statistics of the actual service history of the engine, the randomness of the engine load is characterized and an engine random load spectrum is generated to provide input for the durability test and reliability assessment of the engine and its structure.

[0008] The present invention is implemented by providing a method for constructing a load spectrum of an aviation turbojet or turbofan engine taking into account randomness, comprising the following steps:

[0009] S1: Collect the engine field use data, obtain the flight history records within the scope of the engine's flight mission in a use cycle, extract the engine flight altitude spectrum, flight Mach number spectrum and high-pressure rotor speed spectrum from the flight history records as the engine operation load parameters, divide each flight history record into flight stages, and extract the load spectrum characteristic parameters in each flight stage based on the engine operation load parameters;

[0010] S2: Perform random distribution fitting on the characteristic parameters of the load spectrum to obtain the random distribution law of the characteristic parameters of the load spectrum;

[0011] S3: Randomly extract characteristic parameters of the load spectrum to generate characteristic parameter samples;

[0012] S4: Using the characteristic parameter samples, construct a parameter spectrum related to the engine operating environment;

[0013] S5: Using the characteristic parameter samples, a high-pressure rotor speed spectrum is constructed according to the flight phase sequence of the flight history recorded in S1. The parameter spectrum related to the engine operating environment and the high-pressure rotor speed spectrum together constitute the engine load spectrum.

[0014] Preferably, in S1, the flight phases recorded in the flight history are divided into: a ground phase before takeoff, an air phase and a ground phase after landing, and the air phase is divided into a takeoff phase, a mission phase and a landing phase.

[0015] Further preferably, the load spectrum characteristic parameters that need to be counted in the ground stage before takeoff include: ground ambient temperature, slow running duration before takeoff, and warm-up duration;

[0016] The load spectrum characteristic parameters that need to be counted in the air stage include: flight altitude, flight Mach number, take-off state duration, high-pressure rotor speed cycle peak value, air stage duration, high-speed state duration, speed range-time ratio in the high-speed state, high-pressure rotor speed three-type cycle number and its peak-valley value, high-pressure rotor speed four-type cycle number and its peak-valley value, high-pressure rotor speed three-type cycle ratio in the landing stage to the total three-type cycle of high-pressure rotor speed, and peak-valley value of the three-type cycle of high-pressure rotor speed in the landing stage;

[0017] The load spectrum characteristic parameters in the ground phase after landing include: the duration of slow vehicle after landing.

[0018] Further preferably, the high speed state refers to the high pressure rotor speed N2 exceeding a certain threshold value N 2,th The engine status at the time;

[0019] The speed range-time ratio in the high speed state is: the speed range N2>N 2,th Divide it into several continuous small intervals, assuming that one of the intervals is 99%<N2≤100%, and count the ratio of the time in this speed range to the duration of the high speed state in this flight mission.

[0020] Further preferably, the t-Copula function is used to establish the joint distribution relationship between the characteristic parameters of two load spectra with different but coupled random distribution laws:

[0021] ;

[0022] ;

[0023] in, and Represents the characteristic parameters of two different load spectra and The marginal distribution function of is the t-Copula function, u and v are the random variables of the t-Copula function, and is the distribution parameter, is the cumulative distribution function of the t distribution, r and s are the integrands.

[0024] Further preferably, for two load spectrum characteristic parameters having an influencing relationship with random distribution laws, the distribution of the affected load spectrum characteristic parameter is described as a function of the value of the influencing load spectrum characteristic parameter, and this is implemented by dividing the value of the influencing load spectrum characteristic parameter into intervals.

[0025] Further preferably, in S4, the parameter spectrum related to the engine operating environment includes a flight altitude spectrum and a flight Mach number spectrum.

[0026] More preferably, in S2, the fitting method is maximum likelihood estimation method.

[0027] The present invention provides a method for generating an engine load spectrum by considering the actual flight load characteristics in the field. On this basis, the temperature and stress spectrum of the life assessment position of the structural component corresponding to the load spectrum can be analyzed to provide input for the durability test and reliability evaluation of the engine and its structure based on actual use, which has the following beneficial effects:

[0028] (1) The present invention fully considers the typical damage forms and random characteristics of load spectra related to engine structural durability. Based on the constructed load spectrum, it is possible to carry out engine durability assessment work that is more in line with actual field use;

[0029] (2) The present invention fully considers the correlation between the engine mission characteristics and the load spectrum characteristic parameters, and provides different probability distribution characterization methods for different types of characteristic parameters, providing a more reasonable basis for constructing the engine random load spectrum. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is the realization of the influence of three types of cycle numbers on the distribution of three types of cycle peak and valley values ​​and four types of cycle number distribution;

[0031] Figure 2 It is the realization of the influence of high speed state time on speed time distribution;

[0032] Figure 3 are the flight altitude spectrum and the flight Mach number spectrum;

[0033] Figure 4 is the rotation speed spectrum during the ground phase before takeoff;

[0034] Figure 5 is the speed spectrum during takeoff;

[0035] Figure 6 It is the process of generating the speed spectrum during the mission phase and the landing phase;

[0036] Figure 7 is the speed spectrum during the mission phase and the landing phase;

[0037] Figure 8There are three types of cyclic speed spectra;

[0038] Fig. 9 There are four types of cyclic speed spectra;

[0039] Fig.10 is the speed spectrum during the landing phase;

[0040] Fig.11 is an example of a random speed spectrum constructed by the present invention. DETAILED DESCRIPTION

[0041] The present invention is now further described in conjunction with the accompanying drawings and specific implementation examples. It should be understood that the specific implementation examples described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] The method for constructing a load spectrum of an aviation turbojet or turbofan engine taking into account randomness of the present invention comprises the following steps:

[0043] S1: Extensively collect the engine external load spectrum and count the characteristic parameters of the load spectrum. For a type of engine, collect part or all of the engine use records to reflect the load range and randomness in the engine use cycle. Extract parameters such as the engine flight altitude spectrum, flight Mach number spectrum and high-pressure rotor speed spectrum from the flight history records, and count the characteristic parameters of the load spectrum on this basis. The parameters include: ground ambient temperature, slow-speed duration before takeoff, warm-up duration, flight altitude, flight Mach number, takeoff state duration, high-pressure rotor speed category one cycle peak, air stage duration, high-speed state duration, speed range-time ratio in high-speed state, three types of high-pressure rotor speed cycle numbers and their peak-valley values, four types of high-pressure rotor speed cycle numbers and their peak-valley values, the ratio of the three types of high-pressure rotor speed cycles in the landing stage to the total three types of high-pressure rotor speed cycles, and the peak-valley values ​​of the three types of high-pressure rotor speed cycles in the landing stage, and the slow-speed duration after landing.

[0044] In this step:

[0045] The ground ambient temperature is the average value of the intake duct temperature before the engine takes off; the slow engine duration refers to the time range from the engine starting to the time before entering the warm-up procedure; the flight altitude and flight Mach number refer to the average altitude and Mach number of the aircraft in the mission segment; the main cycle peak refers to the maximum speed of the engine in a single takeoff and landing. Those skilled in the art should understand that this speed usually occurs during takeoff; a type of cycle, whose speed range is 0-N 2,max -0, the speed range of the three types of cycles is N 2,idle -N 2,max -N 2,idle , the speed range of the four types of cycles is N 2,cruise -N 2,max -N 2,cruise Here N2,idle 、N 2,cruise and N 2,max They represent the high-pressure speed ranges in the slow-running state, cruising state, and intermediate thrust state, and are generally marked in the engine maintenance manual. The criterion for determining the high-speed state is that the high-pressure rotor speed N2 exceeds a certain threshold value N 2,th For this type of engine, if there is an engine status with similar purpose, please refer to the engine maintenance manual; change the engine speed from 100% to N 2,th The speed intervals were divided into 1% intervals and the speed interval-time ratio relationship was statistically analyzed. The rain flow counting method was used to extract the number of three-type cycles and four-type cycles in each takeoff and landing, and the peak and valley values ​​were statistically analyzed in the matrix form given in Table 1.

[0046] Table 1 Cycle peak and valley count matrix of a certain takeoff and landing

[0047]

[0048] S2: Fitting the probability distribution of the characteristic parameters of the load spectrum in S1. This method provides a preferred distribution function of the characteristic parameters, as shown in Table 2.

[0049] Table 2 Probability distribution function of load spectrum characteristic parameters

[0050]

[0051] In this step, the number of three types of cycles in a single takeoff and landing is divided into intervals, and the flight records in each interval are counted and fitted with the probability distribution of the number of four types of cycles and the peak and valley values ​​of three types of cycles, such as Figure 1 ; The duration of the high speed state of a single takeoff and landing is divided into intervals, and the flight records in each interval are counted and fitted with the probability distribution of the speed interval-time ratio relationship, such as Figure 2 For the remaining characteristic parameters, probability distribution fitting is performed within the entire flight record range.

[0052] The maximum and minimum values ​​of the statistical feature parameters are used as boundary constraints when generating parameters.

[0053] S3: Use the Monte Carlo method to randomly sample the characteristic parameters. Obviously, this step is similar to step S2: first extract the characteristic parameters that affect other characteristics, the number of three types of cycles and the duration of the high speed state, and extract the number of four types of cycles, the peak and valley values ​​of three types of cycles and the speed interval-time ratio based on the intervals of the above two parameters. The remaining characteristic parameters are directly extracted according to the fitting results in S2. When there are characteristic parameters that exceed the boundaries of the field records, the characteristic parameters are re-extracted.

[0054] S4: Construct a parameter spectrum related to the engine operating environment according to the extracted characteristic parameters.

[0055] In this step, if Figure 3 ,According to the engine flight mission characteristics, the engine flight altitude spectrum and engine flight Mach number spectrum are constructed respectively.

[0056] S5: Constructing the engine high-pressure rotor speed spectrum according to the extracted characteristic parameters.

[0057] Step S5 specifically includes:

[0058] S51: Figure 4 , the rotation speed spectrum of the ground phase before takeoff is constructed based on the manual parameters and the extracted characteristic parameters.

[0059] S52: Figure 5 , construct the takeoff stage speed spectrum based on the extracted characteristic parameters. Preferably, except for the speed change process, the engine is in a high speed state during the takeoff stage. After the high-pressure rotor speed reaches the peak speed, it is arranged in order from high to low according to the speed range of the high-speed state. For the duration of each speed interval in the takeoff stage: ① In the high-speed state of the overall mission, the part between the maximum value of the secondary cycle peak speed and the main cycle speed peak is included in the takeoff stage. ② For the remaining part, preferably, the high-speed states are numbered in descending order of speed in seconds, and the high-speed state of the takeoff stage time is extracted using a uniform distribution method.

[0060] S53: Construct the speed spectrum of the mission phase and the landing phase according to the extracted characteristic parameters. The detailed steps are as follows. Figure 6 Specifically:

[0061] ① Since the landing phase lasts for a short time, the speed spectrum of the landing phase is constructed first. Count all the sub-cycle peaks and evenly distribute the high speed state time of the sub-cycle, taking the 97%-98% speed range as an example:

[0062]

[0063] in, Indicates high speed state time, is the number of three types of cycles with peak values ​​above 98%, Indicates the time when the speed is 98% in these three types of cycles. The sub-cycles include three types of cycles and four types of cycles. Figure 7 , construct the sub-speed spectrum of the three types of cycles in the landing stage. Preferably, Figure 8, the sub-speed spectrum starts from the cruising speed and increases at a speed of 3% / s to the high speed state. The high speed state interval is symmetrical with the speed peak. It first gradually increases to the peak speed, and then gradually decreases until it exits the high speed state, and then decreases to the valley speed at a speed of 1% / s before pulling back to the cruising speed. In this step, if the high speed state time of the landing stage exceeds the total landing stage time in a certain three-type cycle, the high speed state will not be included from the beginning of the three-type cycle. Finally, the three-type cycle is inserted in the cruising speed in the landing stage according to the uniformly distributed extraction position.

[0064] ② Count the peak values ​​of the mission segment cycles and evenly distribute the high speed state time of the mission segment cycles. Construct three types of cycle sub-speed spectra in the mission segment, and insert three types of cycles in the cruise speed in the landing segment according to uniformly distributed extraction positions.

[0065] ③ Insert four types of loops in the mission phase and the landing phase. The positions of the four types of loops are uniformly drawn from any position in the mission phase and the landing phase. Preferably, Fig. 9 Construct a sub-speed spectrum of four types of cycles. The speed of the extracted position is used as the initial speed of each sub-speed spectrum. The sub-speed spectrum starts from the initial speed, decreases to the speed valley at a rate of 1% / s, and then rises through the high speed state and then reaches the peak speed. After reaching the peak speed, the speed drops back to the initial speed. In particular, if the initial speed is not within the speed range of the current four types of cycles, the four types of cycles are skipped. After completing a round of implantation of the four types of cycles, the remaining positions are extracted in the cruise speed positions of the mission phase and the landing phase according to a uniform distribution and implanted with the four types of cycles skipped in the first round.

[0066] The above steps can ensure that the characteristic parameters of the finally generated speed spectrum are consistent with the characteristic parameters extracted by the Monte Carlo method.

[0067] S54: Constructing a rotation speed spectrum of the ground segment after landing. Preferably, the rotation speed spectrum of the landing segment is as follows: Fig.10 Build. Reduce the cruising speed to the idling speed by 3% / s and maintain it until the idling time is over.

[0068] Fig.11 (a), (b), (c), and (d) are examples of speed spectra constructed by step S5. The engine environmental load spectrum and the engine speed spectrum formed by steps S4-S5 together constitute an engine load spectrum that takes randomness into account.

[0069] Through steps S1-S5, the present invention realizes the characterization of random characteristics of engine loads related to loads of field flight missions, and provides a way to reconstruct engine environment and speed spectrum based on Monte Carlo extracted load spectrum characteristic parameters. The present invention considers the most important loads related to the durability of the engine rotor structure, and proposes a reasonable relationship between load characteristics and a preferred distribution law, which is of great significance to the study of engine load and durability.

Claims

1. A method for constructing load spectrum of aviation turbojet and turbofan engines considering randomness, characterized in that: The steps include: S1: Collect the engine field use data, obtain the flight history records within the scope of the engine's flight mission in a use cycle, extract the engine flight altitude spectrum, flight Mach number spectrum and high-pressure rotor speed spectrum from the flight history records as the engine operation load parameters, divide each flight history record into flight stages, and extract the load spectrum characteristic parameters in each flight stage based on the engine operation load parameters; The flight phases recorded in the flight history are divided into: ground phase before takeoff, air phase and ground phase after landing. The air phase is divided into takeoff phase, mission phase and landing phase. The load spectrum characteristic parameters that need to be counted in the ground stage before takeoff include: ground ambient temperature, slow running duration before takeoff, and warm-up duration; The load spectrum characteristic parameters that need to be counted in the air stage include: flight altitude, flight Mach number, take-off state duration, high-pressure rotor speed cycle peak value, air stage duration, high-speed state duration, speed range-time ratio in the high-speed state, high-pressure rotor speed three-type cycle number and its peak-valley value, high-pressure rotor speed four-type cycle number and its peak-valley value, high-pressure rotor speed three-type cycle ratio in the landing stage to the total three-type cycle of high-pressure rotor speed, and peak-valley value of the three-type cycle of high-pressure rotor speed in the landing stage; The characteristic parameters of the load spectrum in the post-landing ground phase include: the duration of slow car after landing; S2: Perform random distribution fitting on the characteristic parameters of the load spectrum to obtain the random distribution law of the characteristic parameters of the load spectrum; Use the t-Copula function to establish the joint distribution relationship between the characteristic parameters of two load spectra with different but coupled random distribution laws: ； ； in, and Represents two different load spectrum characteristic parameters and The marginal distribution function of is the t-Copula function, u and v are the random variables of the t-Copula function, and is the distribution parameter, is the cumulative distribution function of t distribution, r and s are the integrands; S3: Randomly extract characteristic parameters of the load spectrum to generate characteristic parameter samples; S4: Using the characteristic parameter samples, construct a parameter spectrum related to the engine operating environment; S5: Using the characteristic parameter samples, a high-pressure rotor speed spectrum is constructed according to the flight phase sequence of the flight history recorded in S1. The parameter spectrum related to the engine operating environment and the high-pressure rotor speed spectrum together constitute the engine load spectrum.

2. The method for constructing load spectrum of aviation turbojet and turbofan engines considering randomness according to claim 1, characterized in that: The high speed state refers to the high pressure rotor speed N2 exceeding a certain threshold value N 2,th The engine status at the time; The speed range-time ratio in the high speed state is: the speed range N2>N 2,th Divide it into several continuous small intervals, assuming that one of the intervals is 99%<N2≤100%, and count the ratio of the time in this speed range to the duration of the high speed state in this flight mission.

3. The method for constructing load spectrum of aviation turbojet and turbofan engines considering randomness according to claim 1, characterized in that: For two load spectrum characteristic parameters that have an influencing relationship due to random distribution rules, the distribution of the affected load spectrum characteristic parameter is described as a function of the value of the influencing load spectrum characteristic parameter, which is implemented by dividing the value of the influencing load spectrum characteristic parameter into intervals.

4. The method for constructing load spectrum of aviation turbojet and turbofan engines considering randomness according to claim 1, characterized in that: In S4, the parameter spectrum related to the engine operating environment includes a flight altitude spectrum and a flight Mach number spectrum.

5. The method for constructing load spectrum of aviation turbojet and turbofan engines considering randomness according to claim 1, characterized in that: In S2, the fitting method is the maximum likelihood estimation method.

Images