Digital twinning body and digital twinning method of hydraulic test system of turbine pump of liquid rocket engine

Abstract

The invention discloses a digital twinborn body and a digital twinborn method for a hydraulic test system of a turbine pump of a liquid rocket engine, and solves the technical problem of high production and test resource consumption caused by research and development of engine products. The basic architecture of the digital twinning body comprises a digital twinning engine, a physical entity and a digital virtual body, and dynamic unification of digital virtual body prediction performance and object actual measurement performance is realized through virtual-real interaction and model training and optimization; finally, the display effect that a digital virtual body completely replaces a liquid rocket engine turbine pump hydraulic test system entity can be achieved, and the contradiction between increasing turbine pump hydraulic test tasks and limited physical test resources is greatly relieved.

Description

Technical Field

[0001] This invention relates to digital twins and twinning methods, specifically to digital twins and digital twinning methods for a liquid rocket engine turbopump hydraulic test system. Background Technology

[0002] The development of liquid rocket engines is a complex and multifaceted systems engineering project. The turbopump is the heart of a liquid rocket engine, and its work capacity directly affects the engine's performance and reliability. The traditional turbopump development model involves design, simulation verification, hydraulic testing, hot-fire testing, and design optimization, which is time-consuming and costly.

[0003] Meanwhile, with the development of my country's aerospace industry, the number of space launch missions has been increasing year by year, and the demand for engine products has also increased accordingly. This has led to the problem of high consumption of production and testing resources. Therefore, the guiding role of digital design methods in engine development has become increasingly important. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problem of high consumption of production and testing resources due to the research and development of engine products, and to provide a digital twin of a liquid rocket engine turbopump hydraulic test system and a digital twin method.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A digital twin of a liquid rocket engine turbopump hydraulic test system is characterized by comprising a digital twin engine, a physical entity, and a digital virtual entity. The physical entity includes a control module, a measurement module, a data acquisition module, and a data interface. The control module and the measurement module are connected to the digital twin engine through the data interface. The control module receives the operating parameters input from the digital twin engine and converts them into control commands, which are then input to the drive components of the liquid rocket engine turbopump hydraulic test system. The measurement module is used to measure the key position parameters of the liquid rocket engine turbopump hydraulic test system and transmit them to the digital twin engine. The digital virtual entity includes a one-dimensional dynamic simulation model in which each component of the liquid rocket engine turbopump hydraulic test system is embedded in the form of nodes, as well as a turbopump hydraulic model, a cavitation model, a real-time flow field display ROM model, a UI interface, and a built-in algorithm module. The UI interface is used to input the operating parameters of the liquid rocket engine turbopump hydraulic test system. The one-dimensional dynamic simulation model, turbopump hydraulic model, and cavitation model are used to predict the turbopump inflow conditions, critical cavitation margin, turbopump head, and power. The real-time flow field display ROM model is used for flow field decomposition and reconstruction, and displays the turbopump hydraulic performance, cavitation performance, and three-dimensional flow field information in real time. The built-in algorithm module is used to post-process the reconstructed flow field. The digital twin engine includes a function and algorithm module, an artificial intelligence algorithm module, a communication component connecting the physical entity and the digital virtual entity, and an auxiliary node component. The function and algorithm module is used to verify the one-dimensional dynamic simulation model. The communication component connects to the data interface of the physical entity, and is used to send operating parameters to the control module and receive key position parameters transmitted by the measurement module. The artificial intelligence algorithm module is used to optimize and train the turbine pump hydraulic model and cavitation model based on the key position parameters, and to construct and calibrate the real-time flow field display ROM model. The auxiliary node component connects to the UI interface, and is used to receive operating parameters input from the UI interface and transmit them to the control module of the physical entity through the communication component.

[0006] Furthermore, the operating parameters include the motor output speed of the inducer in the turbopump, the pressure reduction parameters of the vacuum pump, the pressure increase parameters of the booster pump, and the output flow parameters of the control valve. The key location parameters include the pressure within the turbopump test section, the motor output torque, the temperatures upstream and downstream of the turbopump test section, the flow rate in the booster pump output pipe, and images of the turbopump test section.

[0007] Furthermore, the measurement module includes a data acquisition board for a pressure pulsation sensor installed in the turbopump test section, a torque meter connected to a motor, temperature sensors installed upstream and downstream of the turbopump test section, an electromagnetic flow meter connected to the output pipe of the booster pump, and a high-speed camera positioned directly opposite the turbopump test section.

[0008] Furthermore, the digital twin engine is a B / S architecture based on the Industrial Internet of Things.

[0009] Furthermore, the one-dimensional dynamic simulation model is established by establishing one-dimensional flow sub-models of each component in the hydraulic test system of liquid rocket engine turbopump based on the fluid dynamics equations of mass conservation, momentum conservation, and energy conservation, and the one-dimensional flow sub-models of each component are connected in series or in parallel to form the model. The expression for the hydraulic model of the turbine pump is: ; ; Where H and P represent the head and power of the turbopump, respectively; A1, A2, A3, B1, B2, and B3 all represent empirical coefficients; and Q and n represent the flow rate and speed of the turbopump, respectively. The expression for the cavitation model is: ; Where NPSHr represents the critical net positive suction head (NPSH), v a1 u m1Here, λ represents the axial velocity of the incoming flow to the turbopump and the circumferential velocity of the impeller, respectively; ζ is the loss coefficient related to the turbopump inlet structure; g is the acceleration due to gravity; and λ is the cavitation coefficient. a0 and b0 are empirical coefficients.

[0010] Meanwhile, the present invention also provides a digital twin method for a liquid rocket engine turbopump hydraulic test system. Based on the aforementioned digital twin of the liquid rocket engine turbopump hydraulic test system, its special feature is that it includes the following steps: Step 1: Perform geometric measurements on each component of the liquid rocket engine turbopump hydraulic test system to obtain the true dimensions of each component; Step 2: Conduct flow resistance characteristic tests on the hydraulic test system of the liquid rocket engine turbopump to obtain flow resistance characteristic data, and assign values ​​to each node in the one-dimensional dynamic simulation model based on the actual dimensions of each component. Step 3: Connect the measurement module to the key position of the liquid rocket engine turbopump hydraulic test system, connect the control module to the drive component of the liquid rocket engine turbopump hydraulic test system, input the operating parameters of the liquid rocket engine turbopump hydraulic test system through the UI interface, control the operation of the drive component of the liquid rocket engine turbopump hydraulic test system through the control module, and use the measurement module to collect the parameters of the key position. Step 4: Using the operating parameters input from the UI interface, the key location parameters collected by the measurement module, and the turbine pump structural parameters as inputs, the function and algorithm modules are used to verify and calibrate the one-dimensional dynamic simulation model. The artificial intelligence algorithm module is used to optimize and train the turbine pump hydraulic model and cavitation model, construct the real-time flow field display ROM model, and calibrate the real-time flow field display ROM model. Step 5: Input the operating condition parameters to be displayed into the verified and calibrated one-dimensional dynamic simulation model through the UI interface to obtain the predicted turbopump inflow conditions. Input the predicted turbopump inflow conditions into the real-time flow field display ROM model for flow field reconstruction. Use the built-in algorithm module to post-process the reconstructed flow field. At the same time, input the operating condition parameters to be displayed into the optimized turbopump hydraulic model and cavitation model to obtain the real-time displayed three-dimensional flow field information of turbopump hydraulic performance and cavitation performance.

[0011] Furthermore, in step 3, the key position of connecting the measurement module to the hydraulic test system of the liquid rocket engine turbopump is specifically as follows: connecting the pressure pulsation sensor to the turbopump test section, connecting the torque meter to the motor, setting the temperature sensors upstream and downstream of the turbopump test section respectively, connecting the electromagnetic flow meter to the output pipe of the booster pump, and setting the high-speed camera relative to the turbopump test section. The specific method of connecting the control module to the drive component of the liquid rocket engine turbopump hydraulic test system is as follows: the output end of the control module is connected to the motor, vacuum pump, booster pump and control valve of the liquid rocket engine turbopump hydraulic test system.

[0012] Furthermore, step 4 specifically involves: Step 4.1: Input the operating parameters from the UI interface into the one-dimensional dynamic simulation model. The function and algorithm modules use the one-dimensional finite volume method to numerically discretize and solve the one-dimensional dynamic simulation model, preliminarily predict the inflow conditions of the turbopump, and verify and calibrate the one-dimensional dynamic simulation model in combination with key position parameters. The inflow conditions of the turbopump include the speed, pressure and flow rate of the turbopump. Step 4.2: Input the operating condition parameters input from the UI interface into the verified and calibrated one-dimensional dynamic simulation model to obtain the predicted turbine pump inflow conditions. Input the predicted turbine pump inflow conditions into the turbine pump hydraulic model. The artificial intelligence algorithm module dynamically adjusts the empirical coefficients in the turbine pump hydraulic model in combination with key position parameters to obtain the optimized turbine pump hydraulic model. Step 4.3: Input the operating condition parameters and turbine pump structural parameters input from the UI interface into the cavitation model to obtain the predicted critical cavitation margin. Use 3D CFD simulation software to simulate the cavitation performance under different operating conditions to obtain the cavitation performance curve. The artificial intelligence algorithm module combines the cavitation performance curve to dynamically adjust the empirical coefficients in the cavitation model to obtain an optimized cavitation model. Step 4.4: Given the boundary conditions of the turbopump operating parameters, the sample flow field is simulated using 3D CFD simulation software to obtain the full flow field grid node calculation results for each sample flow field. The artificial intelligence algorithm module performs intrinsic orthogonal decomposition on the full flow field grid node calculation results to obtain flow field information modes of different orders. An AI model of flow field information mode coefficients is then established based on these different orders of flow field information modes. A real-time flow field display ROM model is constructed based on the AI ​​model. The specified turbopump operating parameters are input into the real-time flow field display ROM model, and the 3D flow field under the specified operating parameters is reconstructed, rendered, and displayed as an image. The artificial intelligence algorithm module then calibrates the real-time flow field display ROM model using the test section images of the turbopump within the key position parameters and the displayed images.

[0013] Furthermore, in step 4.4, the formula for performing eigenorthogonal decomposition on the calculation results of the entire flow field grid nodes is as follows: ; In the formula, a represents the eigenorthogonal decomposition result of the full flow field grid node calculation at point x at time t. i (t) represents the i-th order flow field information modal coefficient obtained from the intrinsic orthogonal decomposition, φi (x) represents the i-th flow field information mode obtained from the intrinsic orthogonal decomposition.

[0014] Furthermore, in step 4.3, the turbine pump structural parameters include the loss coefficient related to the turbine pump inlet structure, the turbine pump inlet axial velocity and impeller circumferential velocity, and the cavitation coefficient.

[0015] The beneficial effects of this invention are: 1. The digital twin and digital twin method of the liquid rocket engine turbopump hydraulic test system provided by the present invention have a basic architecture consisting of three parts: digital twin engine, physical entity and digital virtual entity. Through virtual-real interaction, model training and optimization, the predicted performance of digital virtual entity and the measured performance of physical entity are dynamically unified. In the end, the digital virtual entity can completely replace the physical entity in the display effect, which greatly alleviates the contradiction between the increasing number of turbopump hydraulic test tasks and the limited physical test resources.

[0016] 2. The digital twin and digital twin method for the liquid rocket engine turbopump hydraulic test system provided by this invention can collect key position parameters based on the physical entity, and dynamically calibrate the one-dimensional dynamic simulation model in the digital virtual body by combining functions and algorithms in the digital twin engine, maintaining a high degree of consistency with the key position parameters of the liquid rocket engine turbopump hydraulic test system; and improve the accuracy of hydraulic performance and cavitation performance prediction by training the turbopump hydraulic model and cavitation model through artificial intelligence algorithms. Attached Figure Description

[0017] Figure 1 This is a schematic diagram showing the positional relationship of the torque meter, flow meter, and high-speed camera in the hydraulic test system and measurement module of the liquid rocket engine turbopump in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of an embodiment of the liquid rocket engine turbopump hydraulic test system of the present invention; Figure 3 This is a schematic diagram illustrating the result of assigning values ​​to each node in a one-dimensional dynamic simulation model based on the flow resistance characteristic data and the actual dimensions of each component in an embodiment of the present invention. Figure 4 This is a flowchart of the three-dimensional flow field display in an embodiment of the digital twin method for the hydraulic test system of a liquid rocket engine turbopump of the present invention; Figure 5 This is a schematic diagram of the internal interior of the turbine pump test section in an embodiment of the digital twin method for the liquid rocket engine turbine pump hydraulic test system of the present invention, wherein (a) is a real internal schematic diagram of the turbine pump test section in the liquid rocket engine turbine pump hydraulic test system, and (b) is a schematic diagram of the internal interior of the turbine pump test section obtained from the real-time flow field display ROM model.

[0018] The attached figures are labeled as follows: 1-Boost pump, 2-High-speed camera, 3-Heat exchanger, 4-Control valve, 5-Water tank, 6-Filter, 7-Motor, 8-Torque meter, 9-Test section of turbine pump, 10-Flow meter. Detailed Implementation

[0019] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] The structure of the liquid rocket engine turbopump hydraulic test system in this embodiment is as follows: Figure 1 As shown, the system includes a booster pump 1, a control valve 4, a water tank 5, a filter 6, a heat exchanger 3, a test section 9 of a turbopump, and the booster pump 1, which are connected in sequence to form a closed loop. A flow meter 10 is connected to the pipe at the output end of the booster pump 1, and a vacuum pump is also connected to the water tank 5. A torque meter 8 is connected to a motor 7, and the output shaft of the motor 7 is connected to the inducer wheel of the test section of the turbopump 9. The motor provides the required speed to drive the inducer wheel, and the torque meter 8 measures the torque of the test piece during operation. In this embodiment, the outer shell of the test section of the turbopump 9 is made of transparent plexiglass, which can be used in conjunction with the high-speed camera 2 in the measurement module to dynamically observe the morphology of the internal cavitation zone.

[0021] This embodiment provides a digital twin of a liquid rocket engine turbopump hydraulic test system, such as... Figure 2 As shown, it includes a digital twin engine, physical entities, and digital virtual entities.

[0022] The physical entity includes a control module, a measurement module, a data acquisition module, and a data interface. The control module and measurement module are connected to the digital twin engine via the data interface. The control module receives the operating parameters input from the digital twin engine and converts them into control commands, which are then input to the drive components of the liquid rocket engine turbopump hydraulic test system (including motor 7, vacuum pump, booster pump 1, and control valve 4). Through the combined action of the drive components, the necessary inflow pressure and flow rate are provided to the test section. The operating parameters include the output torque of the motor controlling the inducer in the turbopump, the decompression parameters of the vacuum pump, the pressurization parameters of the booster pump, and the output flow of the control valve. The measurement module includes a data acquisition board, a pressure pulsation sensor installed in the turbopump test section, a torque meter connected to the motor, temperature sensors installed upstream and downstream of the turbopump test section, an electromagnetic flow meter connected to the booster pump output pipe, and a high-speed camera positioned directly opposite the turbopump test section. The measurement module is used to measure key position parameters of the liquid rocket engine turbopump hydraulic test system and transmit them to the digital twin engine. The key position parameters include the pressure in the turbopump test section, the temperature upstream and downstream of the turbopump test section, the flow rate in the booster pump output pipe, and images of the turbopump test section.

[0023] The digital virtual entity includes a one-dimensional dynamic simulation model that embeds each component of the liquid rocket engine turbopump hydraulic test system in the form of nodes, as well as a turbopump hydraulic model, a cavitation model, a real-time flow field display ROM model, a UI interface, and a built-in algorithm module. The UI interface is used to input the operating parameters of the liquid rocket engine turbopump hydraulic test system. The one-dimensional dynamic simulation model, turbopump hydraulic model, and cavitation model are used to predict the turbopump inflow conditions, critical cavitation margin, turbopump head, and power. The real-time flow field display ROM model is used for flow field decomposition and reconstruction, and displays the turbopump hydraulic performance, cavitation performance, and three-dimensional flow field information in real time. The built-in algorithm module is used for post-processing the reconstructed flow field.

[0024] The model is structured as follows: (1) The one-dimensional dynamic simulation model is established based on the fluid dynamics equations of mass conservation, momentum conservation, and energy conservation to create one-dimensional flow sub-models for each component in the hydraulic test system of the liquid rocket engine turbopump. The one-dimensional flow sub-models of each component are connected in series or in parallel. Finally, the one-dimensional dynamic simulation model is verified by the function and algorithm modules in the digital twin engine.

[0025] mass conservation equation: (1) In the formula, ρ is the fluid density and u is the velocity.

[0026] Momentum conservation equation: (2) In the formula, p is the pressure and f is the source term of the momentum equation.

[0027] Equations of state: (3) If temperature changes are ignored, density is only a function of pressure: (4) The above equations were numerically discretized and solved using the one-dimensional finite volume method. Key location parameters acquired from physical entities were used to evaluate, verify, and calibrate the one-dimensional fluid network model. The calibrated one-dimensional fluid network model was then integrated into a digital virtual volume, and one-dimensional dynamic simulations were used to provide the necessary inflow conditions for the test section.

[0028] (2) The expression for the hydraulic model of the turbine pump is: (5) (6) Where H and P represent the head and power of the turbine pump, respectively; A1, A2, A3, B1, B2, and B3 are all empirical coefficients; and Q and n represent the flow rate and speed of the turbine pump, respectively.

[0029] In this embodiment, the turbine pump hydraulic model is based on a traditional hydraulic mechanical performance model, such as H=f(Q,n), where H is the head, Q, and n are the flow rate and rotational speed, respectively. Typically, the head is a quadratic function of the flow rate and rotational speed. The empirical coefficients in the turbine pump hydraulic model are dynamically adjusted based on the data predicted by the one-dimensional dynamic model in the digital virtual entity, obtaining model prediction values ​​that are closer to the measured results. This achieves high-precision prediction of the turbine pump's hydraulic performance in the digital virtual entity. Simultaneously, the artificial intelligence algorithm built into the digital twin engine can train the turbine pump hydraulic model. When enough training samples are accumulated, the hydraulic performance of the turbine pump can be accurately predicted using the digital virtual entity, achieving the effect of replacing the physical entity.

[0030] (3) The expression for the cavitation model is: ; Where NPSHr represents the critical net positive suction head (NPSH), v a1 u m1 ζ represents the axial velocity of the incoming flow to the turbopump and the circumferential velocity of the impeller, respectively; ζ is the loss coefficient related to the turbopump inlet structure; λ is the cavitation coefficient; and g is the gravitational acceleration. ; In the formula, a0 and b0 are empirical coefficients, with a0 recommended to be 0.01~0.05 and b0 recommended to be 0.115.

[0031] In this embodiment, the cavitation model is calculated based on the critical net positive suction head (NPSHr) determined according to the turbine pump structure and operating parameters. The critical NPSH is the NPSH corresponding to a 2.5% decrease in turbine pump head. Extensive 3D CFD simulations were performed to obtain cavitation performance curves under different operating conditions, and the cavitation model was calibrated accordingly. During the experiment, the degree of cavitation risk was assessed based on the proximity of the predicted critical NPSH, and operational suggestions for the test system were provided. For example, when the incoming flow NPSH is greater than the critical point, the cavitation risk is considered low, and the test system can operate normally; when the incoming flow NPSH is lower than the critical point, the cavitation risk is considered high, and the digital virtual system provides decision suggestions such as reducing the speed, increasing the flow rate, or increasing the incoming flow pressure to reduce the cavitation risk.

[0032] (4) The real-time flow field display ROM model can effectively avoid the problems of insufficient real-time performance of 3D CFD and huge computational and memory resource consumption. Its main idea is to decompose the physical field and project it into an orthogonal space, where the influencing factors with higher weights will be applied to the approximate physical field. Its main advantage is high computational efficiency, which can meet the real-time requirements of digital twin flow field display. In the early stage, 3D CFD calculations were performed on the flow field of the turbopump under different speed, flow rate and pressure conditions to obtain sample flow fields. On this basis, the ROM model was trained using Proper Orthogonal Decomposition (POD) to quickly output the flow field information under given boundary conditions. When the boundary conditions are input, the ROM model reconstructs the 3D flow field under the specified conditions by optimizing the modal coefficients to obtain flow field information consistent with the 3D CFD results. The main advantage of POD technology is that it significantly shortens the calculation time while meeting the accuracy requirements of flow field analysis, and achieves the effect of real-time display of 3D flow field.

[0033] The digital twin engine is a B / S (browser / server) architecture based on the Industrial Internet of Things (IIoT), developed using the latest version of the Chrome browser (including but not limited to). It includes function and algorithm modules, artificial intelligence algorithms, communication components connecting physical entities and digital virtual entities, and auxiliary node components. The function and algorithm modules are used to verify the one-dimensional dynamic simulation model. The communication component connects to the data interface of the physical entity, sending operating parameters to the control module and receiving key position parameters transmitted by the measurement module. The artificial intelligence algorithm module optimizes and trains the turbine pump hydraulic model and cavitation model based on the key position parameters, and constructs and calibrates a real-time flow field display ROM model. The auxiliary node component connects to the UI interface, receiving operating parameters input from the UI and transmitting them to the physical entity's control module via the communication component.

[0034] This embodiment also provides a digital twin method for a liquid rocket engine turbopump hydraulic test system. Using the digital twin of the liquid rocket engine turbopump hydraulic test system, the method includes the following steps: Step 1: Perform geometric measurements on each component of the liquid rocket engine turbopump hydraulic test system to obtain the true dimensions of each component.

[0035] Step 2: Conduct flow resistance characteristic tests on the liquid rocket engine turbopump hydraulic test system to obtain flow resistance characteristic data. Then, assign values ​​to each node in the one-dimensional dynamic simulation model based on the actual dimensions of each component. The results are as follows: Figure 3 As shown.

[0036] Step 3: Connect the measurement module to the key position of the liquid rocket engine turbopump hydraulic test system, connect the control module to the drive component of the liquid rocket engine turbopump hydraulic test system, input the operating parameters of the liquid rocket engine turbopump hydraulic test system through the UI interface, control the operation of the drive component of the liquid rocket engine turbopump hydraulic test system through the control module, and use the measurement module to collect the key position parameters.

[0037] Step 4: Using the operating parameters input from the UI interface, key location parameters collected by the measurement module, and turbine pump structural parameters as inputs, the one-dimensional dynamic simulation model is verified and calibrated using function and algorithm modules. The artificial intelligence algorithms in the digital twin engine are used to optimize and train the turbine pump hydraulic model and cavitation model. Finally, the artificial intelligence algorithms in the digital twin engine are used to construct a real-time flow field display ROM model and to calibrate the real-time flow field display ROM model. Specifically: Step 4.1: Input the operating parameters from the UI interface into the one-dimensional dynamic simulation model. Use the one-dimensional finite volume method in the function and algorithm module to numerically discretize and solve the one-dimensional dynamic simulation model to preliminarily predict the turbine pump inflow conditions, which include the turbine pump speed, pressure, and flow rate. Combine the key position parameters to verify and calibrate the one-dimensional dynamic simulation model. Step 4.2: Input the operating condition parameters input from the UI interface into the verified and calibrated one-dimensional dynamic simulation model to obtain the predicted turbine pump inflow conditions. Input the predicted turbine pump inflow conditions into the turbine pump hydraulic model. Use artificial intelligence algorithms combined with key position parameters to dynamically adjust the empirical coefficients in the turbine pump hydraulic model to obtain an optimized turbine pump hydraulic model. Step 4.3: Input the operating condition parameters and turbine pump structural parameters input from the UI interface into the cavitation model to obtain the predicted critical cavitation margin. Use artificial intelligence algorithms combined with key location parameters collected by the measurement module to dynamically adjust the empirical coefficients in the cavitation model to obtain an optimized cavitation model. Among them, the turbine pump structural parameters include the loss coefficient related to the turbine pump inlet structure, the read turbine pump inflow axial velocity and impeller circumferential velocity, and the cavitation coefficient.

[0038] Step 4.4: Given the boundary conditions of the turbopump operating parameters, simulate the sample flow field using 3D CFD simulation software to obtain the full flow field grid node calculation results for each sample flow field. This process is called sampling. Perform intrinsic orthogonal decomposition on the full flow field grid node calculation results to obtain flow field information modes of different orders. Combine the artificial intelligence algorithm in the digital twin engine to establish an AI model of flow field information mode coefficients. Based on the AI ​​model, construct a real-time flow field display ROM model to complete the flow field decomposition. Input the specified turbopump operating parameters into the real-time flow field display ROM model to reconstruct the 3D flow field under the specified operating parameters, completing the flow field reconstruction. Render and display the reconstructed flow field to complete the flow field output. Finally, verify the displayed image using the test section image of the turbopump within the key position parameters.

[0039] In this embodiment, the workflow of the real-time flow field display ROM model is as follows: Figure 4 As shown, the process mainly consists of sampling, flow field decomposition, flow field reconstruction, and output. For the turbopump display, at least 5 flow rates, 5 rotational speeds, and 12 pressures are simulated for cavitation in the early stages, resulting in a total of 5×5×12=120 sample flow fields (the more samples, the closer the model prediction results are to the 3D CFD simulation results, but the longer the initial sampling period; at least 100 sample points are recommended). The full flow field grid node calculation results for each sample flow field are output. (Note that the more grid nodes in the early calculation process, the more detailed the flow field information, but the longer the 3D CFD simulation calculation time and the longer the time required for POD-ROM model flow field reconstruction. To reduce the digital virtual body response time, the number of grid nodes should be reasonably controlled according to the complexity of the test specimen structure. For example, for a separate induced impeller flow field, it is recommended that the number of grid nodes be between 500,000 and 2,000,000.)

[0040] The formula for performing intrinsic orthogonal decomposition of the sample flow field is: ; In the formula, a represents the eigenorthogonal decomposition result of the sample flow field at point x at time t. i (x) represents the weight coefficient of the i-th basis function obtained from the eigenorthogonal decomposition, φ i (x) represents the i-th basis function obtained from the eigenorthogonal decomposition.

[0041] Step 5: Input the operating condition parameters to be displayed into the verified and calibrated one-dimensional fluid network model through the UI interface to obtain the predicted turbine pump inflow conditions. Input the predicted turbine pump inflow conditions into the real-time flow field display ROM model for flow field reconstruction. Use the built-in algorithm module to post-process the reconstructed flow field. At the same time, input the operating condition parameters to be displayed into the optimized turbine pump hydraulic model and cavitation model to obtain the three-dimensional flow field information of the real-time display of the turbine pump's hydraulic performance and cavitation performance.

[0042] The display results of the turbopump test section using a digital twin of a liquid rocket engine turbopump hydraulic test system and a digital twin method in this embodiment are as follows: Figure 5 As shown in (b), the actual internal display results of the turbopump test section are as follows. Figure 5 As shown in (a); by comparing (a) and (b), it can be seen that the simulation accuracy of the twin in (b) is very high.

[0043] 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 changes or substitutions within the technical scope disclosed in the present invention should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A digital twin of a hydraulic test system for a liquid rocket engine turbopump, characterized in that: This includes digital twin engines, physical entities, and digital virtual entities; The physical entity includes a control module, a measurement module, and a data interface. The control module and the measurement module are connected to the digital twin engine through the data interface. The control module receives the operating parameters input from the digital twin engine and converts them into control commands, which are then input to the drive components of the liquid rocket engine turbopump hydraulic test system. The measurement module is used to measure the key position parameters of the liquid rocket engine turbopump hydraulic test system and transmit them to the digital twin engine. The digital virtual entity includes a one-dimensional dynamic simulation model in which each component of the liquid rocket engine turbopump hydraulic test system is embedded in the form of nodes, as well as a turbopump hydraulic model, a cavitation model, a real-time flow field display ROM model, a UI interface, and a built-in algorithm module. The UI interface is used to input the operating parameters of the liquid rocket engine turbopump hydraulic test system. The one-dimensional dynamic simulation model, turbopump hydraulic model, and cavitation model are used to predict the turbopump inflow conditions, critical cavitation margin, turbopump head, and power. The real-time flow field display ROM model is used for flow field decomposition and reconstruction, and displays the turbopump hydraulic performance, cavitation performance, and three-dimensional flow field information in real time. The built-in algorithm module is used to post-process the reconstructed flow field. The digital twin engine includes a function and algorithm module, an artificial intelligence algorithm module, a communication component connecting the physical entity and the digital virtual entity, and an auxiliary node component. The function and algorithm module is used to verify the one-dimensional dynamic simulation model. The communication component connects to the data interface of the physical entity, and is used to send operating parameters to the control module and receive key position parameters transmitted by the measurement module. The artificial intelligence algorithm module is used to optimize and train the turbine pump hydraulic model and cavitation model based on the key position parameters, and to construct and calibrate the real-time flow field display ROM model. The auxiliary node component connects to the UI interface, and is used to receive operating parameters input from the UI interface and transmit them to the control module of the physical entity through the communication component.

2. The digital twin of the liquid rocket engine turbopump hydraulic test system according to claim 1, characterized in that: The operating parameters include the motor output speed of the inducer in the turbopump, the pressure reduction parameters of the vacuum pump, the pressure increase parameters of the booster pump, and the output flow parameters of the control valve. The key location parameters include the pressure within the turbopump test section, the motor output torque, the temperatures upstream and downstream of the turbopump test section, the flow rate in the booster pump output pipe, and images of the turbopump test section.

3. The digital twin of the liquid rocket engine turbopump hydraulic test system according to claim 2, characterized in that: The measurement module includes a data acquisition board for a pressure pulsation sensor installed in the turbopump test section, a torque meter connected to a motor, temperature sensors installed upstream and downstream of the turbopump test section, an electromagnetic flow meter connected to the output pipe of the booster pump, and a high-speed camera positioned directly opposite the turbopump test section.

4. The digital twin of the liquid rocket engine turbopump hydraulic test system according to claim 1, characterized in that: The digital twin engine is based on a B / S architecture of industrial IoT.

5. The digital twin of the liquid rocket engine turbopump hydraulic test system according to claim 1, characterized in that: The one-dimensional dynamic simulation model is established by establishing one-dimensional flow sub-models of each component in the hydraulic test system of liquid rocket engine turbopump based on the fluid dynamics equations of mass conservation, momentum conservation, and energy conservation, and the one-dimensional flow sub-models of each component are connected in series or in parallel to form the model. The expression for the hydraulic model of the turbine pump is: ； ； Where H and P represent the head and power of the turbopump, respectively; A1, A2, A3, B1, B2, and B3 all represent empirical coefficients; and Q and n represent the flow rate and speed of the turbopump, respectively. The expression for the cavitation model is: ； Where NPSHr represents the critical net positive suction head (NPSH), v a1 u m1 Here, λ represents the axial velocity of the incoming flow to the turbopump and the circumferential velocity of the impeller, respectively; ζ is the loss coefficient related to the turbopump inlet structure; g is the acceleration due to gravity; and λ is the cavitation coefficient. a0 and b0 are empirical coefficients.

6. A digital twin method for a liquid rocket engine turbopump hydraulic test system, using the digital twin of the liquid rocket engine turbopump hydraulic test system according to any one of claims 1-5, characterized in that, Includes the following steps: Step 1: Perform geometric measurements on each component of the liquid rocket engine turbopump hydraulic test system to obtain the true dimensions of each component; Step 2: Conduct flow resistance characteristic tests on the hydraulic test system of the liquid rocket engine turbopump to obtain flow resistance characteristic data, and assign values ​​to each node in the one-dimensional dynamic simulation model based on the actual dimensions of each component. Step 3: Connect the measurement module to the key position of the liquid rocket engine turbopump hydraulic test system, connect the control module to the drive component of the liquid rocket engine turbopump hydraulic test system, input the operating parameters of the liquid rocket engine turbopump hydraulic test system through the UI interface, control the operation of the drive component of the liquid rocket engine turbopump hydraulic test system through the control module, and use the measurement module to collect the parameters of the key position. Step 4: Using the operating parameters input from the UI interface, the key location parameters collected by the measurement module, and the turbine pump structural parameters as inputs, the function and algorithm modules are used to verify and calibrate the one-dimensional dynamic simulation model. The artificial intelligence algorithm module is used to optimize and train the turbine pump hydraulic model and cavitation model, construct the real-time flow field display ROM model, and calibrate the real-time flow field display ROM model. Step 5: Input the operating condition parameters to be displayed into the verified and calibrated one-dimensional dynamic simulation model through the UI interface to obtain the predicted turbopump inflow conditions. Input the predicted turbopump inflow conditions into the real-time flow field display ROM model for flow field reconstruction. Use the built-in algorithm module to post-process the reconstructed flow field. At the same time, input the operating condition parameters to be displayed into the optimized turbopump hydraulic model and cavitation model to obtain the real-time displayed three-dimensional flow field information of turbopump hydraulic performance and cavitation performance.

7. The digital twin method for a hydraulic test system of a liquid rocket engine turbopump according to claim 6, characterized in that, In step 3, the key part of connecting the measurement module to the hydraulic test system of the liquid rocket engine turbopump is as follows: connecting the pressure pulsation sensor to the turbopump test section, connecting the torque meter to the motor, setting the temperature sensors upstream and downstream of the turbopump test section respectively, connecting the electromagnetic flow meter to the output pipe of the booster pump, and setting the high-speed camera relative to the turbopump test section. The specific method of connecting the control module to the drive component of the liquid rocket engine turbopump hydraulic test system is as follows: the output end of the control module is connected to the motor, vacuum pump, booster pump and control valve of the liquid rocket engine turbopump hydraulic test system.

8. The digital twin method for a hydraulic test system of a liquid rocket engine turbopump according to claim 6, characterized in that, Step 4 is as follows: Step 4.1: Input the operating parameters from the UI interface into the one-dimensional dynamic simulation model. The function and algorithm modules use the one-dimensional finite volume method to numerically discretize and solve the one-dimensional dynamic simulation model, preliminarily predict the inflow conditions of the turbopump, and verify and calibrate the one-dimensional dynamic simulation model in combination with key position parameters. The inflow conditions of the turbopump include the speed, pressure and flow rate of the turbopump. Step 4.2: Input the operating condition parameters input from the UI interface into the verified and calibrated one-dimensional dynamic simulation model to obtain the predicted turbine pump inflow conditions. Input the predicted turbine pump inflow conditions into the turbine pump hydraulic model. The artificial intelligence algorithm module dynamically adjusts the empirical coefficients in the turbine pump hydraulic model in combination with key position parameters to obtain the optimized turbine pump hydraulic model. Step 4.3: Input the operating condition parameters and turbine pump structural parameters input from the UI interface into the cavitation model to obtain the predicted critical cavitation margin. Use 3D CFD simulation software to simulate the cavitation performance under different operating conditions to obtain the cavitation performance curve. The artificial intelligence algorithm module combines the cavitation performance curve to dynamically adjust the empirical coefficients in the cavitation model to obtain an optimized cavitation model. Step 4.4: Given the boundary conditions of the turbopump operating parameters, the sample flow field is simulated using 3D CFD simulation software to obtain the full flow field grid node calculation results for each sample flow field. The artificial intelligence algorithm module performs intrinsic orthogonal decomposition on the full flow field grid node calculation results to obtain flow field information modes of different orders. An AI model of flow field information mode coefficients is then established based on these different orders of flow field information modes. A real-time flow field display ROM model is constructed based on the AI ​​model. The specified turbopump operating parameters are input into the real-time flow field display ROM model, and the 3D flow field under the specified operating parameters is reconstructed, rendered, and displayed as an image. The artificial intelligence algorithm module then calibrates the real-time flow field display ROM model using the test section images of the turbopump within the key position parameters and the displayed images.

9. The digital twin method for a hydraulic test system of a liquid rocket engine turbopump according to claim 8, characterized in that, In step 4.4, the formula for performing eigenorthogonal decomposition on the calculation results of the entire flow field grid nodes is as follows: ； In the formula, a represents the eigenorthogonal decomposition result of the full flow field grid node calculation at point x at time t. i (t) represents the i-th order flow field information modal coefficient obtained from the intrinsic orthogonal decomposition, φ i (x) represents the i-th flow field information mode obtained from the intrinsic orthogonal decomposition.

10. The digital twin method for a hydraulic test system of a liquid rocket engine turbopump according to claim 8, characterized in that: In step 4.3, the turbine pump structural parameters include the loss coefficient related to the turbine pump inlet structure, the turbine pump inlet axial velocity and impeller circumferential velocity, and the cavitation coefficient.

Images