Gas-liquid two-phase jet flow condensation flow field reconstruction method and system based on modal decomposition

Abstract

The invention discloses a gas-liquid two-phase jet flow condensation flow field reconstruction method and system based on modal decomposition. The method comprises the following steps that a plurality of dominant flow structures with large advantages are selected for flow field reconstruction; respectively constructing an original flow field gray matrix and a reconstructed flow field gray matrix based on the original flow field snapshot-based reconstructed flow field snapshot; gray gradients of vectors of the original flow field gray matrix and the reconstructed flow field gray matrix are calculated, abrupt change points of the gray gradients are extracted, and the position of the gas-liquid interface is recognized according to the abrupt change points of the gray gradients; the gas-liquid interface error between the reconstructed flow field and the original flow field is calculated, and if the gas-liquid interface error is smaller than 5%, reconstruction is completed; if the gas-liquid interface error is larger than 5%, the steps are repeated until the error requirement is met, and then a gas-liquid two-phase jet flow condensation flow field is obtained; unified representation of the vortex structure is achieved, and the modal aliasing limitation of a single decomposition method is broken through through collaborative application of intrinsic orthogonal decomposition and dynamic modal decomposition.

Description

Technical Field

[0001] The present invention belongs to the field of jet flow field structure extraction and reconstruction, and specifically relates to a gas-liquid two-phase jet condensation flow field reconstruction method and system based on modal decomposition. Background Art

[0002] Gas-liquid two-phase jet condensation, a typical multiphase flow phenomenon, occurs when a high-speed gas jet and a liquid crossflow, under intense shear, induce a rapid vapor phase transition and form a dynamic condensation shock wave structure through direct contact condensation. Due to its exceptional heat and mass transfer rates achieved through the coupling of microscopic interfacial instability with macroscopic flow, and its high system reliability due to its lack of moving parts and the absence of mechanical agitation, this phenomenon has important applications in energy and power equipment, including emergency core cooling of nuclear reactors, regenerative cooling of liquid rocket engine thrust chambers, fluid mixing in underwater propulsion systems, and supersonic mixing in chemical ejectors. However, in high-thrust liquid oxygen-kerosene rocket engines, the intense interphase interaction between the gas jet and liquid crossflow at supercritical pressure can induce shear layer instability and vortex ring shedding, forming a cross-scale vortex system encompassing boundary layer microvortices, Kelvin-Helmholtz main vortices, and large-scale coherent structures. During their unsteady evolution, these vortex structures excite broadband pressure pulsations through vortex-acoustic coupling. Their low-frequency components easily couple with the axial acoustic modes of the propellant delivery pipeline, forming standing wave resonances. High-frequency turbulent pulsations, on the other hand, induce structural fatigue damage, leading to instabilities such as thrust oscillations and pogo pinhole vibrations in the engine, seriously threatening system safety. Therefore, uncovering the multiscale interactions between condensation shock wave oscillations, vortex transport, and acoustic energy transfer, and establishing a cross-scale mapping relationship between vortex dynamics characteristic frequencies and acoustic modes, has become a core scientific issue for overcoming the bottleneck of engine dynamic stability and developing active flow control strategies.

[0003] In actual engineering environments, high-thrust liquid oxygen-kerosene rocket engines face extreme operating conditions such as high flow rates, wide temperature ranges, and variable operating conditions, which lead to significant non-steady-state and non-uniform phase distribution characteristics in gas-liquid interactions. This is specifically manifested in the following core challenges: (1) The diversity of vortex structures. The vortex structure generation mechanism under variable operating conditions presents multimodal characteristics, and its structural type (ring / hairpin / horseshoe), characteristic scale (micrometer to centimeter scale) and evolution law (merger / breakup / collapse) show significant differences; (2) The complexity of full-condition analysis: the cost of obtaining full-condition flow field information using traditional methods increases exponentially with the model dimension; (3) There is an essential contradiction between model construction accuracy and efficiency. Model construction needs to break through the bottleneck of computational efficiency while ensuring prediction accuracy. To this end, it is urgent to develop a reduced-order model of the gas-liquid two-phase jet condensation flow field that takes into account both accuracy and efficiency. Summary of the Invention

[0004] In order to solve the problems existing in the prior art, the present invention provides a gas-liquid two-phase jet condensation flow field reconstruction method and system based on modal decomposition, which realizes the unified characterization of vortex structure and breaks through the modal aliasing limitation of single decomposition method through the coordinated application of intrinsic orthogonal decomposition and dynamic modal decomposition.

[0005] To achieve the above objectives, the present invention first provides a gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition, comprising the following steps: S1, select several dominant flow structures with greater advantages to reconstruct the flow field; S3, constructing the original flow field grayscale matrix and the reconstructed flow field grayscale matrix based on the original flow field snapshot and the reconstructed flow field snapshot; S4, calculating the grayscale gradient of the original flow field grayscale matrix and the reconstructed flow field grayscale matrix vector, extracting the grayscale gradient mutation point, and identifying the gas-liquid interface position according to the grayscale gradient mutation point; S5. Based on the calculated gas-liquid interface error between the reconstructed flow field and the original flow field, if the gas-liquid interface error is less than 5%, the reconstruction is completed; if the gas-liquid interface error is greater than 5%, the above steps are repeated until the error requirement is met, and the gas-liquid two-phase jet condensation flow field is obtained.

[0006] Furthermore, several advantageous flow structure extraction methods with greater advantages include the following steps: S11, construct the flow field grayscale matrix of the original condensation grayscale field according to the original flow field data, reconstruct the grayscale matrix of the original condensation flow field into a column vector, and record it in chronological order n The grayscale vector of the instantaneous flow field forms a two-dimensional matrix; S12, using intrinsic orthogonal decomposition and dynamic mode decomposition methods to decompose the gray field two-dimensional matrix, and separate the original flow field data into time-averaged flow field and pulsating flow field; S13, performing fast Fourier transform analysis on the time coefficient of POD and performing decay rate analysis on the time coefficient of DMD; S14, by comparing the FFT main frequency of the POD mode with the single frequency of the DMD mode, the similarities and differences between the POD mode and the DMD mode are obtained; combined with the decay rate of the DMD mode time coefficient, the persistence of the mode in the flow field is determined; S15, based on the similarities and differences between POD and DMD modes and flow field snapshots, identify the dominant flow structure and its causes and characteristics.

[0007] Furthermore, in S11, the original flow field data includes obtaining the real grayscale field or velocity field data of the gas-liquid two-phase jet condensation by using a high-speed camera or a particle image velocimetry system.

[0008] Furthermore, in S11, the original flow field data includes the use of CFD simulation software, with a multi-flow field model as the basic framework, a comprehensive group balance model and a large eddy simulation method to capture the evolution behavior of the vortex structure, and setting boundary conditions according to actual working parameters to obtain the simulation results of the gas-liquid two-phase jet condensation grayscale field or velocity field. By comparing the simulation results with the experimental data, the simulation model is adjusted, and after verifying the accuracy of the simulation model, the simulation results of the grayscale field or velocity field under different working conditions are output based on the simulation model.

[0009] Furthermore, in S12, POD decomposition provides three pieces of information about the flow field: (1) eigenvectors, i.e., POD modes, which constitute an orthogonal basis and provide spatial information about the coherent structure in the flow field; (2) eigenvalues, which reflect the eigenvectors in L The degree of extraction of the original data under the 2-norm, each eigenvalue gives the energy contribution captured by the respective POD mode; (3) the time coefficient, which retains the time information related to each POD mode. When decomposing POD, a finite-dimensional data vector set is used to represent the flow field, and the optimal vector basis that can replace the given data is found, that is, the eigenvectors and eigenvalues ​​of the given data are calculated.

[0010] Furthermore, in S3, the grayscale matrix of the original condensation flow field is reconstructed into a column vector. n The grayscale vectors of the instantaneous flow field are recorded in a two-dimensional matrix in sequence according to time. After the main modes are determined, the flow field can be approximately expressed by a finite series as follows: r The sum of the products of the main modes and their time coefficients.

[0011] Furthermore, in S5, the steam jet plume interface is divided into upper and lower halves with the nozzle centerline as the axis, and the gas-liquid interface positions of the original flow field and the reconstructed flow field are obtained respectively. For each column of the gas-liquid interface, the error of the upper and lower halves of the interface is calculated by the following formula:

[0012] in, p ' is the position of the gas-liquid interface in the upper or lower half of the reconstructed flow field, p is the gas-liquid interface position in the upper or lower half of the corresponding original flow field.

[0013] Secondly, the present invention provides a gas-liquid two-phase jet condensation flow field reconstruction system based on modal decomposition, including a flow field reconstruction module, a gray matrix construction module, a gas-liquid interface position identification module, a gas-liquid interface error calculation module and a judgment cycle module; The flow field reconstruction module is used to select several advantageous flow structures with greater advantages for flow field reconstruction; The grayscale matrix construction module constructs the original flow field grayscale matrix and the reconstructed flow field grayscale matrix based on the original flow field snapshot and the reconstructed flow field snapshot respectively; The gas-liquid interface position identification module is used to calculate the grayscale gradient of the original flow field grayscale matrix and the reconstructed flow field grayscale matrix vector, extract the grayscale gradient mutation point, and identify the gas-liquid interface position according to the grayscale gradient mutation point; The gas-liquid interface error calculation module is used to calculate the gas-liquid interface error between the reconstructed flow field and the original flow field; The judgment cycle module determines that the gas-liquid interface error is less than 5% based on the gas-liquid interface error, and then completes the reconstruction; if the gas-liquid interface error is greater than 5%, the above steps are repeated until the error requirement is met, and then the gas-liquid two-phase jet condensation flow field is obtained.

[0014] Furthermore, the present invention also provides a computer device, including a processor and a memory, the memory is used to store computer executable programs, the processor reads part or all of the computer executable programs from the memory and executes them, and when the processor executes part or all of the computer executable programs, it can realize the above-mentioned gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition.

[0015] A computer-readable storage medium may also be provided, in which a computer program is stored. When the computer program is executed by a processor, the above-mentioned gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition can be implemented.

[0016] Compared with the existing technology, the present invention has at least the following advantages: unified characterization of vortex structures, using coherent structures as a unified characterization method for complex transient flow prediction and control, and as ordered structures with spatiotemporal coherence in complex flow systems, coherent structures can analyze the inherent flow mechanism of complex transient condensation flow fields, laying the foundation for establishing a universal model under complex variable working conditions; Multi-angle analysis of the condensation flow field, through the coordinated application of proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD), effectively integrates the dual advantages of spatial energy optimization (POD method extracts the dominant energy mode based on spatial orthogonalization) and temporal dynamic characteristics (DMD method separates the characteristic frequency modes through temporal orthogonalization), breaking through the modal aliasing limitations of a single decomposition method; Precise reduced-order model construction, based on the spatiotemporal decoupling characteristics of coherent structures, constructs a reduced-order subspace containing dominant modes, significantly reducing computational dimensionality while maintaining the accuracy of key physical processes. Adaptive mode selection enables dynamic construction of optimal basis functions under varying operating conditions, significantly improving the model's adaptability to changes in parameter space. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of the gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition; Figure 2This is the curve showing the variation of the reconstruction error of the gas-liquid two-phase jet condensation flow field with the modal order based on modal decomposition; Figure 3 This is a flow chart of the gas-liquid two-phase jet condensation flow field structure extraction method based on modal decomposition; Figure 4 Comparison between the FFT analysis results of the POD modal time coefficient and the DMD modal frequency; Figure 5 This is the decay rate analysis of the DMD modal time coefficient. DETAILED DESCRIPTION

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] like Figure 1 As shown, a gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition includes the following steps: Before S1 adoption r The flow field is reconstructed using a dominant flow structure. In S1, the energy of the POD mode is generally calculated by the following formula.

[0020] (1) It can be seen that as the modal order n As the value increases, the eigenvalue gradually decreases, and the corresponding eigenmode's contribution to the flow field decreases. Therefore, it is generally necessary to compare the first few eigenmodes with high energy contributions, as they represent the main structure of the flow field. After determining the main POD modes, the flow field can be approximated using a finite series.

[0021] (2) At this point, the high-dimensional condensation flow field is effectively simplified to r The sum of the POD modes.

[0022] S2, obtain the original flow field snapshot and the reconstructed flow field snapshot.

[0023] S3, constructs the grayscale matrices of the original flow field and the reconstructed flow field respectively.

[0024] Gray matrix of the original flow field: Reconstruct the gray matrix of the original condensation flow field into a column vector, n The grayscale vectors of the instantaneous flow field are recorded in a two-dimensional matrix sequentially according to time.

[0025] Reconstruct the gray matrix of the flow field: After determining the main modes, the flow field can be approximated by a finite series as follows: r The sum of the products of the main modes and their time coefficients.

[0026] S4, calculate the grayscale gradient of the column vector of the grayscale matrix of the original flow field and the reconstructed flow field respectively.

[0027] Grayscale gradient: Calculate the differential of each column vector of the grayscale matrix of the original flow field and the reconstructed flow field.

[0028] S5, determining the mutation points of the grayscale gradients of the original flow field and the column vectors of the reconstructed flow field matrix.

[0029] Because the grayscale values ​​of the gas phase and the liquid phase are significantly different, the differential result of each grayscale vector (i.e., the grayscale gradient) will undergo a sudden change at the gas-liquid interface. Specifically, the grayscale gradient value is lowest when the liquid phase transitions to the gas phase, and highest when the gas phase transitions to the liquid phase.

[0030] S6, the gray gradient mutation point is the gas-liquid interface position of this column.

[0031] The maximum and minimum values ​​of the grayscale gradient are the gas-liquid interface positions of the column. Therefore, after extracting the maximum and minimum values ​​of the grayscale gradient of each column respectively, the maximum and minimum values ​​of the grayscale gradient of each column are connected to obtain the gas-liquid interface position of the steam jet plume.

[0032] S7, calculating the error between the gas-liquid interface of the reconstructed flow field and the gas-liquid interface of the original flow field.

[0033] The steam jet plume interface is divided into two halves, the upper and lower halves, with the nozzle centerline as the axis. The gas-liquid interface positions of the original flow field and the reconstructed flow field are obtained in S6. For each column of the gas-liquid interface, the error of the upper and lower halves is calculated using the following formula: (3) in, p ' is the position of the gas-liquid interface in the upper or lower half of the reconstructed flow field, p is the gas-liquid interface position in the upper or lower half of the corresponding original flow field.

[0034] S8 determines whether the error is less than 5%. If so, the calculation is complete. If the error is greater than 5%, then r = r + 1, and steps S1-S8 are repeated until the error meets the requirement.

[0035] In S8, the reconstruction interface error is as follows Figure 2 As shown in the figure, the error in reconstructing the gas-liquid interface decreases with increasing modal order. The first 10 dominant modes can reduce the interface error to below 5%, achieving efficient reduced-order representation of high-dimensional flow fields.

[0036] like Figure 3 As shown in FIG, a gas-liquid two-phase jet condensation flow field structure extraction method based on modal decomposition includes the following steps: S11, obtaining original flow field data.

[0037] In S11, there are two main methods for obtaining the original flow field data. One is based on experimental measurement, mainly using a high-speed camera or a particle image velocimetry system to obtain the real grayscale field or velocity field data of the gas-liquid two-phase jet condensation. The second is based on simulation. Mainly using CFD simulation software such as Fluent, with a multi-flow field model as the basic framework, a comprehensive group balance model and a large eddy simulation method to capture the evolution behavior of the vortex structure, and setting boundary conditions according to actual working parameters, etc., to preliminarily obtain the simulation results of the grayscale field or velocity field of the gas-liquid two-phase jet condensation. By comparing the simulation results with the experimental data, adjusting the simulation model, and verifying the accuracy of the simulation model, the simulation results of the grayscale field or velocity field under different working conditions are output.

[0038] S12, constructing a flow field parameter matrix of the original condensation grayscale field. Reconstructing the grayscale matrix of the original condensation flow field into a column vector, and recording the grayscale vectors of the n instantaneous flow fields in a two-dimensional matrix in sequence according to time.

[0039] S13, POD and DMD decomposition of grayscale field. The POD and DMD methods are used to decompose the grayscale field matrix of the gas-liquid jet condensation, respectively, and the original flow field data is decomposed into the time-averaged flow field and the pulsating flow field. The modal decomposition method can identify and extract the coherent structure in the flow field. As an ordered structure with spatiotemporal coherence in a complex flow system, the coherent structure has a dual significance for realizing flow prediction and control. On the one hand, the coherent structure can analyze the intrinsic flow mechanism of the complex transient condensation flow field, laying the foundation for establishing a universal model under complex variable working conditions. On the other hand, the modal decomposition method can be used to achieve model order reduction, which significantly improves the computational efficiency while ensuring accuracy.

[0040] Furthermore, proper orthogonal decomposition (POD) and dynamic modal decomposition (DMD) approach flow field decomposition from different perspectives. The former focuses on spatial orthogonalization, effectively identifying the most energetic flow structures, while the latter focuses on temporal orthogonalization, more effectively decomposing complex flow fields into uncoupled coherent structures with specific dynamic modes and corresponding frequencies. The modal decomposition method combining POD and DMD avoids the incomplete decomposition limitations of a single decomposition method.

[0041] In S13, the POD decomposition method uses a finite-dimensional data vector set to represent the flow field: (4) in v ( ξ , t) is the original flow field matrix minus its time mean q ̅( ξ The purpose of POD analysis is to find the optimal vector basis that can replace the given data, that is, to calculate the eigenvectors and eigenvalues ​​of the given data: (5) in R is a vector v ( ξ , t ) covariance matrix, R It can be expressed as: (6) The eigenvectors calculated by formula (2) are also called POD modes. POD modes are mutually orthogonal and are q ( ξ , t ) provides a complete basis. This means that the following relations are satisfied between POD modes: (7) Eigenvalue λ k Reflects the eigenvector ϕ k exist L The degree of extraction of the original data under the 2-norm. Each eigenvalue gives the average energy captured by the respective POD mode.

[0042] The time coefficient is also called the POD modal coefficient. It retains the time information associated with each POD mode and is generally calculated using the following formula.

[0043] (8) The DMD decomposition method assumes continuous snapshots v i and v i+1 There is a linear mapping between A : (9) In any case, assuming the snapshot v i The mapping between is constant, which allows us to express the flow field sequence as: (10) When the number of snapshots is large enough, further flow fields are added to the data sequence. v i It will not V 1 N The spanned vector space has an impact. iWhen this limit is reached, the vector v N It can be expressed as v i A linear combination of .

[0044] (11) Or in matrix form: (12) in a i is the coefficient of determination, r is the residual vector. Formula (11) means V 1 N Captures the main characteristics of the underlying stream. Data sequence V 2 N It can be expressed as: (13) in e N-1 ∈ yes N - 1 unit vector, S is the adjoint matrix, expressed as: (14) In fact, S The eigenvalues ​​of A The eigenvalues ​​of are approximated. The adjoint matrix is ​​calculated using QR decomposition S .

[0045] (15) (16) in Q H yes Q The complex transpose of 。S The eigenvalues ​​and corresponding eigenvectors can be obtained by eigenvalue decomposition: (17) Based on this, the dynamic mode of the condensation flow field is defined as (18) Furthermore, by applying the logarithmic mapping of the eigenvalues, the corresponding frequencies f i and growth rate ω i as follows: (19) S14, POD and DMD decomposition of grayscale field. The grayscale field matrix is ​​decomposed using the proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD) methods, respectively, to separate the original flow field data into the time-averaged flow field and the pulsating flow field.

[0046] S15, POD and DMD time coefficient analysis: Perform FFT analysis on the POD time coefficient and perform decay rate analysis on the DMD time coefficient.

[0047] In S15, the POD time coefficient is analyzed by FFT and compared with the frequency corresponding to the DMD mode with a single frequency. The results are as follows: Figure 4 As shown. It can be found that the main frequencies corresponding to mode 1 and mode 2 of POD correspond to the first main frequency of DMD mode, which is f =362Hz. The main frequencies of POD mode 3 and mode 5 correspond to the second main frequency of DMD mode, which is f =724Hz. In addition, the decay rate analysis of the time coefficient of the DMD mode is carried out, and the results are as follows: Figure 5 It is found that the time coefficients of DMD mode 1 and mode 2 run through the entire time series, indicating that the contribution of these modes to the flow field does not change with time; while DMD mode 3 decays to 0 after 20ms, indicating that this mode has no effect on the flow field after 20ms.

[0048] S16, POD and DMD mode comparison. By comparing the FFT analysis results of the POD modal time coefficient with the single frequency corresponding to the DMD mode, the similarities and differences between the POD and DMD modes are determined. By analyzing the decay rate of the DMD modal time coefficient, it is determined whether the mode exists in the condensation flow field for a long time.

[0049] like Figure 5 As shown, the attenuation rate of mode 1 of DMD is at the lowest level, which is 0.61. Observing the time coefficient evolution curve of mode 1, it can be seen that the effect of the dynamic information of mode 1 on the flow field runs through the entire time series and basically does not change with time. Secondly, the attenuation rate of mode 2 is 10.38. The time coefficient curve of mode 2 shows that although the effect of the dynamic information of mode 2 on the flow field runs through the entire time series, it is getting weaker and weaker. The attenuation rate of mode 3 is 839.20. Within 4.5ms, the time coefficient of mode 3 decays rapidly from -135 to near 0, indicating that the structure only affects the flow field before 4.5ms and is not representative in the entire time series.

[0050] S7, determination of the dominant flow structure. After the above comparison, combined with the flow field snapshot, determine the cause of the flow structure and its characteristics. Different condensation flow patterns will present different condensation characteristics. For example, the typical feature of the Chugging condensation flow pattern is that steam intermittently enters and exits the nozzle, while the most typical feature of the Oscil-I condensation flow pattern is that bubbles intermittently fall off from the end of the steam jet plume. For different condensation flow patterns, their flow structures also present different characteristics. For example, under the Oscil-I condensation flow pattern, the bubble shedding frequency is 362Hz, which corresponds to the main frequency of DMD or POD mode 1. Therefore, it is determined that the dominant flow structure under the Oscil-I condensation flow pattern is the result of intermittent bubble shedding.

[0051] In summary, the method for reconstructing the gas-liquid two-phase jet condensation flow field based on modal decomposition provided by the present invention includes the following steps: selecting several dominant flow structures with relatively large advantages for flow field reconstruction; constructing the original flow field grayscale matrix and the reconstructed flow field grayscale matrix based on the original flow field snapshot and the reconstructed flow field snapshot; calculating the grayscale gradient of the original flow field grayscale matrix and the reconstructed flow field grayscale matrix vector, extracting the grayscale gradient mutation points, and identifying the gas-liquid interface position according to the grayscale gradient mutation points; calculating the gas-liquid interface error between the reconstructed flow field and the original flow field, if the gas-liquid interface error is less than 5%, the reconstruction is completed; if the gas-liquid interface error is greater than 5%, the above steps are repeated until the error requirement is met, thereby obtaining the gas-liquid two-phase jet condensation flow field. The unified characterization of vortex structure constructs a mathematical framework with coherent structure as the core, and realizes the hierarchical extraction of multi-scale coherent structures by introducing the phase averaging identification method. This characterization system relies on the intrinsic orthogonal decomposition and dynamic modal decomposition methods to establish a quantitative mapping relationship between vortex core morphology and energy transport and characteristic frequency, providing a physically interpretable dynamic primitive for flow stability analysis. In transonic condensation flows, this characterization method successfully correlates the bidirectional coupling mechanism between shock wave / heat transfer deterioration-induced vortex collapse and nonequilibrium condensation latent heat release. A multi-angle analysis of the condensation flow field employs an improved multi-fidelity data assimilation strategy, fusing the orthogonal basis functions of the POD with the Koopman operator of the DMD via a tensor product to form a hybrid decomposition paradigm with spatiotemporal coupling. Specifically, during the evolution of unsteady condensation wave systems, this method dynamically tracks the intermittent collapse of condensation bubbles and successfully isolates the characteristic frequency components corresponding to the bubble collapse process. A precise jet condensation reduction model with spatiotemporal parameter adaptability is constructed using a gas-liquid interface tracking algorithm based on grayscale gradient mutation detection. A scalable Hilbert space can be constructed based on the basis function cluster of the source condition. Combined with multi-parameter coupling analysis such as decay rate and power spectral density, this reduced-order model can achieve online correction of the optimal basis vectors under target conditions during the migration of typical condensation flow patterns, ultimately achieving an optimal balance between computational efficiency and accuracy in the prediction of steam nonequilibrium condensation flows.

[0052] Example 2: Based on the technical concept of the above method, the present invention also provides a gas-liquid two-phase jet condensation flow field reconstruction system based on modal decomposition, including a flow field reconstruction module, a gray matrix construction module, a gas-liquid interface position identification module, a gas-liquid interface error calculation module, and a judgment cycle module; The flow field reconstruction module is used to select several advantageous flow structures with greater advantages for flow field reconstruction; The grayscale matrix construction module constructs the original flow field grayscale matrix and the reconstructed flow field grayscale matrix based on the original flow field snapshot and the reconstructed flow field snapshot respectively; The gas-liquid interface position identification module is used to calculate the grayscale gradient of the original flow field grayscale matrix and the reconstructed flow field grayscale matrix vector, extract the grayscale gradient mutation point, and identify the gas-liquid interface position according to the grayscale gradient mutation point; The gas-liquid interface error calculation module is used to calculate the gas-liquid interface error between the reconstructed flow field and the original flow field; The judgment cycle module determines that the gas-liquid interface error is less than 5% based on the gas-liquid interface error, and then completes the reconstruction; if the gas-liquid interface error is greater than 5%, the above steps are repeated until the error requirement is met, and then the gas-liquid two-phase jet condensation flow field is obtained.

[0053] On the other hand, the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can implement the gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition described in the present invention.

[0054] The present invention can also provide a computer device, including a processor and a memory, the memory is used to store a computer executable program, the processor reads the computer executable program from the memory and executes it, and when the processor executes the computer executable program, it can implement the gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition described in the present invention.

[0055] The computer device may be a laptop computer, a desktop computer or a workstation.

[0056] The processor can be a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or an off-the-shelf field-programmable gate array (FPGA).

[0057] The memory of the present invention may be an internal storage unit of a laptop computer, desktop computer or workstation, such as a memory or a hard disk; or an external storage unit, such as a mobile hard disk or a flash memory card.

[0058] Computer-readable storage media may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Computer-readable storage media may include: read-only memory (ROM), random access memory (RAM), solid-state drives (SSD) or optical disks, etc. Among them, random access memory may include resistance random access memory (ReRAM) and dynamic random access memory (DRAM).

[0059] The above content is only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition, characterized in that: The following steps are involved: S1, select several dominant flow structures with greater advantages to reconstruct the flow field; S3, constructing the original flow field grayscale matrix and the reconstructed flow field grayscale matrix based on the original flow field snapshot and the reconstructed flow field snapshot; S4, calculating the grayscale gradient of the original flow field grayscale matrix and the reconstructed flow field grayscale matrix vector, extracting the grayscale gradient mutation point, and identifying the gas-liquid interface position according to the grayscale gradient mutation point; S5. Based on the calculated gas-liquid interface error between the reconstructed flow field and the original flow field, if the gas-liquid interface error is less than 5%, the reconstruction is completed; if the gas-liquid interface error is greater than 5%, the above steps are repeated until the error requirement is met, and the gas-liquid two-phase jet condensation flow field is obtained.

2. The gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to claim 1 is characterized in that: Several advantageous flow structure extraction methods with greater advantages include the following steps: S11, constructing a flow field grayscale matrix of the original condensation grayscale field according to the original flow field data, reconstructing the grayscale matrix of the original condensation flow field into a column vector, and recording n instantaneous flow field grayscale vectors in time sequence to form a two-dimensional matrix; S12, using intrinsic orthogonal decomposition and dynamic mode decomposition methods to decompose the gray field two-dimensional matrix, and separate the original flow field data into time-averaged flow field and pulsating flow field; S13, performing fast Fourier transform analysis on the time coefficient of POD and performing decay rate analysis on the time coefficient of DMD; S14, by comparing the FFT main frequency of the POD mode with the single frequency of the DMD mode, the similarities and differences between the POD mode and the DMD mode are obtained; combined with the decay rate of the DMD mode time coefficient, the persistence of the mode in the flow field is determined; S15, based on the similarities and differences between POD and DMD modes and flow field snapshots, identify the dominant flow structure and its causes and characteristics.

3. The gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to claim 2 is characterized in that: In S11, the original flow field data includes obtaining the real grayscale field or velocity field data of the gas-liquid two-phase jet condensation by using a high-speed camera or a particle image velocimetry system.

4. The gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to claim 2 is characterized in that: In S11, the original flow field data includes the use of CFD simulation software, with a multi-flow field model as the basic framework, a comprehensive group balance model and a large eddy simulation method to capture the evolution behavior of the vortex structure, and setting boundary conditions according to actual working parameters to obtain the simulation results of the grayscale field or velocity field of the gas-liquid two-phase jet condensation. By comparing the simulation results with the experimental data, the simulation model is adjusted, and after verifying the accuracy of the simulation model, the simulation results of the grayscale field or velocity field under different working conditions are output based on the simulation model.

5. The gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to claim 2 is characterized in that: In S12, a finite-dimensional data vector set is used to represent the flow field during POD decomposition, and the optimal vector basis that can replace the given data is found, that is, the eigenvectors and eigenvalues ​​of the given data are calculated.

6. The gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to claim 2 is characterized in that: In S3, the grayscale matrix of the original condensation flow field is reconstructed into a column vector, and the grayscale vectors of n instantaneous flow fields are recorded in a two-dimensional matrix in sequence according to time. After determining the main mode, the flow field is approximated by a finite series as shown in the previous figure. r The sum of the products of the main modes and their time coefficients.

7. The gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to claim 2 is characterized in that: In S5, the steam jet plume interface is divided into upper and lower halves with the nozzle centerline as the axis. The gas-liquid interface positions of the original flow field and the reconstructed flow field are obtained respectively. For each column of the gas-liquid interface, the error of the upper and lower halves of the interface is calculated using the following formula: in, p ' is the position of the gas-liquid interface in the upper or lower half of the reconstructed flow field, p is the gas-liquid interface position in the upper or lower half of the corresponding original flow field.

8. A gas-liquid two-phase jet condensation flow field reconstruction system based on modal decomposition, characterized in that: It includes flow field reconstruction module, gray matrix construction module, gas-liquid interface position recognition module, gas-liquid interface error calculation module and judgment cycle module; The flow field reconstruction module is used to select several advantageous flow structures with greater advantages for flow field reconstruction; The grayscale matrix construction module constructs the original flow field grayscale matrix and the reconstructed flow field grayscale matrix based on the original flow field snapshot and the reconstructed flow field snapshot respectively; The gas-liquid interface position identification module is used to calculate the grayscale gradient of the original flow field grayscale matrix and the reconstructed flow field grayscale matrix vector, extract the grayscale gradient mutation point, and identify the gas-liquid interface position according to the grayscale gradient mutation point; The gas-liquid interface error calculation module is used to calculate the gas-liquid interface error between the reconstructed flow field and the original flow field; The judgment cycle module determines that the gas-liquid interface error is less than 5% based on the gas-liquid interface error, and then completes the reconstruction; if the gas-liquid interface error is greater than 5%, the above steps are repeated until the error requirement is met, and then the gas-liquid two-phase jet condensation flow field is obtained.

9. A computer device, characterized in that: It includes a processor and a memory, the memory is used to store computer executable programs, the processor reads part or all of the computer executable programs from the memory and executes them, and when the processor executes part or all of the computer executable programs, it can implement the gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that A computer program is stored in the computer-readable storage medium. When the computer program is executed by the processor, the gas-liquid two-phase jet condensation flow field reconstruction method based on modal decomposition according to any one of claims 1 to 7 can be implemented.

Images