Analysis method, system and device for internal flow structure of flow field

Abstract

The invention discloses an analysis method, system and device for an internal flow structure of a flow field. The analysis method comprises the following steps: acquiring characteristic quantity information of the flow field; converting the flow field characteristic quantity information into an atlas changing along with frequency, and obtaining discrete peak frequency in flow field pulsation from the atlas; obtaining a nonlinear relationship between discrete peak frequencies, and confirming a first characteristic frequency corresponding to the unsteady flow; processing the characteristic quantity information of the flow field by using an intrinsic orthogonal decomposition method, obtaining energy distribution and a flow structure of the flow field, and determining a dominant frequency and a corresponding dominant unsteady flow structure in combination with the first characteristic frequency; and performing frequency domain transformation on the characteristic quantity information of the flow field to obtain a two-dimensional distribution cloud picture corresponding to the dominant frequency, obtaining space distribution of the dominant frequency in the flow field, and combining energy distribution of the flow field to obtain an area where the dominant unsteady flow structure is located. According to the method, through multi-angle analysis, the dominant unsteady flow structure is clearly determined, and the characteristic frequency and the action area of the dominant unsteady flow structure are clarified.

Description

Technical Field

[0001] The present invention relates to the technical field of flow field analysis, and in particular to a method, system and device for analyzing the internal flow structure of a flow field. Background Art

[0002] At present, the analysis of unsteady flow energy loss inside rotating machinery, such as the study of energy loss inside pump turbine, is still mainly based on entropy generation analysis. In previous studies, fluid mechanics and thermodynamics were almost regarded as two independent disciplines. The introduction of entropy generation method to evaluate the hydraulic loss of turbine can determine the amount of energy dissipation and the location where the dissipation occurs. In fluid machinery such as pump turbine, entropy generation analysis helps to understand the impact of irreversible factors such as friction loss, turbulence effect, flow separation, etc. on the efficiency of component working conditions.

[0003] Through the entropy generation analysis of the impeller area, it was found that under large flow conditions, the entropy generation inside the impeller is the largest and the energy loss is the most serious, followed by small flow conditions, and the energy loss is the smallest under the highest efficiency conditions; the entropy generation rate in areas with larger values ​​is larger, so the vortex flow is an important cause of energy loss; the area where the impeller impacts has a higher rate, so the impact phenomenon is also the main cause of energy loss; the area with larger surface velocity has a higher rate, so the friction phenomenon is also an important cause of energy loss.

[0004] By using entropy generation analysis to study the spatial distribution characteristics of hydraulic loss of pump-turbine under different flow conditions in pump mode, it is found that with the increase of flow rate, the total entropy generation is consistent with the hydraulic loss of the pressure difference method, which first decreases significantly and then gradually increases. The change trend of indirect entropy generation, direct entropy generation and wall entropy generation is consistent with the total entropy generation, and indirect entropy generation and direct entropy generation are dominant. The research of many scholars has pointed out that the specific location and state of internal energy loss of pump-turbine can be clearly detected by entropy generation method. However, there are still great limitations in using entropy generation to analyze the internal energy loss of pump-turbine. For example, the internal flow of pump-turbine is very complex, including three-dimensional effect, rotation effect, boundary layer separation and secondary flow, etc. These complex flow characteristics are difficult to be fully described by entropy generation alone.

[0005] In complex fluid dynamics environments, unsteady flow patterns are ubiquitous. Identifying and optimizing these patterns is critical to the design of rotating fluid machinery, but it is also extremely challenging. In the analysis and diagnosis of unsteady flow inside rotating machinery, the flow field is usually very complex and changeable. Therefore, it is particularly important to establish an efficient and accurate internal flow structure analysis method. Summary of the invention

[0006] In order to overcome the defects in the prior art, the present invention provides a method, system and device for analyzing the internal flow structure of a flow field. The method clearly determines which unsteady flow structure is the main flow field action structure under the working condition through multi-angle analysis from three aspects: energy contribution, frequency domain characteristics and spatial distribution characteristics, and clarifies its characteristic frequency and action area, thereby providing a clear goal and direction for the subsequent optimization design and improving the optimization efficiency.

[0007] To achieve the above object, the technical solution adopted by the present invention is: a method for analyzing the internal flow structure of a flow field, comprising the following steps:

[0008] Obtaining flow field characteristic quantity information;

[0009] Converting the flow field characteristic quantity information into a spectrum that changes with frequency, and obtaining the discrete peak frequency in the flow field pulsation from the spectrum;

[0010] Establishing a relationship model between the discrete peak frequencies to obtain a nonlinear relationship between the discrete peak frequencies and confirming a first characteristic frequency corresponding to an unsteady flow;

[0011] Intrinsic orthogonal decomposition, using the intrinsic orthogonal decomposition method to decompose the flow field characteristic quantity information into a number of POD modes with different energy contributions, so as to obtain the energy distribution and flow structure of the flow field, and combine the first characteristic frequency to confirm the dominant frequency and the corresponding dominant unsteady flow structure;

[0012] Frequency domain visualization, for the dominant frequency, the pulsation information in the flow field characteristic information is transformed into the frequency domain to obtain a two-dimensional distribution cloud map corresponding to the dominant frequency, obtain the spatial distribution of the dominant frequency in the flow field, and combine the energy distribution of the flow field to obtain the area where the dominant unsteady flow structure is located.

[0013] Through the above technical scheme, the discrete peak frequencies in the flow field pulsation are obtained by spectral analysis. The nonlinear relationship between the discrete peak frequencies is analyzed to eliminate the influence of the linear superposition of the axial frequencies, and the first characteristic frequency corresponding to the unsteady flow is obtained. Then, the intrinsic orthogonal decomposition is combined with the two-dimensional frequency domain visualization method to obtain the energy contribution and frequency domain spatial distribution characteristics of the flow structure of the flow field, so as to deeply analyze the flow structure strength characteristics, distribution and development laws corresponding to different characteristic frequencies of the flow field.

[0014] Furthermore, the step of acquiring flow field characteristic quantity information includes: arranging monitoring points and monitoring surfaces in the flow area to be analyzed to collect the flow field characteristic quantity information, and the flow field characteristic quantity information includes two-dimensional time domain characteristics.

[0015] Further, the step spectrum analysis includes:

[0016] The flow field characteristic quantity information is subjected to Fourier transformation to obtain a spectrum of the pressure at the monitoring point changing with frequency, from which the discrete peak frequency can be obtained. The discrete peak frequency usually corresponds to the dominant periodic structure in the flow field. That is, the discrete peak frequency can reflect the key dynamic behavior in the flow field.

[0017] Further, the step of obtaining the nonlinear relationship between the discrete peak frequencies includes:

[0018] The discrete peak frequency is subjected to three Fourier transforms to calculate the phase coupling relationship between each frequency component, obtain the nonlinear relationship between different frequencies in the frequency domain, and eliminate the influence of the linear superposition of the shaft frequency, thereby obtaining the first characteristic frequency. Since the shaft frequency of the rotating parts in the flow field is usually consistent with the discrete peak frequency in the spectrum analysis, the spectrum analysis cannot distinguish whether these discrete peak frequencies are mainly derived from the linear superposition of the shaft frequency or from other complex flows. Therefore, bispectral analysis is used to further verify and distinguish whether these discrete peak frequencies have a linear superposition relationship with the shaft frequency in order to screen out the first characteristic frequency.

[0019] Further, the steps of eigenorthogonal decomposition include:

[0020] The pressure data in the flow field characteristic quantity information is organized into a data matrix, and the data matrix is ​​decoupled and calculated to obtain the POD mode and the corresponding POD eigenvalue. The POD mode is sorted in descending order according to the size of the POD eigenvalue, and the POD features corresponding to each POD mode are accumulated from large to small to obtain a cumulative value. When the cumulative value is greater than or equal to 50%, the accumulation is stopped, and the corresponding POD mode in the cumulative value is taken as the high-energy mode. POD modes with a proportion greater than or equal to 50% can cover significant flow structures and changes in the flow field. Therefore, it is only necessary to use high-energy modes to analyze the dominant flow structure characteristics in the flow field in the future. Through the POD method, it is expanded from a specific monitoring point to the entire cross-section, which provides strong support for studying the energy pulsation characteristics of the flow field.

[0021] Further, 5. The method for analyzing the internal flow structure of the flow field as described in claim 4 is characterized in that each of the POD modes has a corresponding POD modal base and POD modal coefficients, and the second characteristic frequency of each mode can be obtained by applying Fourier transform to the POD modal coefficients, and the spatial distribution of the POD mode can be drawn according to the coordinate points corresponding to the POD modal base. The second characteristic frequency is the frequency extracted from the POD mode, representing the main oscillation characteristics of each mode in time. It can further refine or confirm the first characteristic frequency, and may also reveal more complex or secondary dynamic behaviors in the flow. Use MATLAB or other drawing software to draw the spatial distribution according to the coordinate points corresponding to the POD modal base to achieve visualization of the modal distribution.

[0022] Furthermore, the spatial distribution of the POD mode is compared with the spectrum of the pressure at the monitoring point changing with the frequency, and the frequency corresponding to both the high-energy mode and the first characteristic frequency is the dominant frequency.

[0023] Further, the frequency domain visualization of the steps includes:

[0024] The velocity pulse or the pressure pulse is transformed in the frequency domain to obtain a two-dimensional velocity distribution cloud map or a two-dimensional pressure distribution cloud map corresponding to the dominant frequency. The spatial distribution of the dominant frequency in the flow field and the source of instability in the flow can be directly observed through the two-dimensional distribution cloud map. The source of instability in the flow is the area where the dominant unsteady flow structure is located.

[0025] Furthermore, the frequency domain visualization of the steps also includes:

[0026] The pulsation amplitudes in the two-dimensional distribution cloud map are sorted in descending order, and the first 70% are taken as the high pulsation amplitude area. The high pulsation amplitude area represents the part of the flow field that is most strongly affected by the frequency, and the spatial distribution of the dominant frequency in the flow field can be clarified to understand how the dominant frequency affects the flow structure in a specific area;

[0027] Combining the two-dimensional distribution cloud map and the POD modal basis, the relationship between the dominant frequency and the spatial distribution of each POD mode is obtained, and the characteristic frequency components of different flow structures are distinguished. If it is observed that the characteristic morphology of some POD modes is more prominent in the cloud map of a specific frequency component, it indicates that these modes contribute more to the frequency component;

[0028] According to the spatial distribution of the dominant frequency in the two-dimensional distribution cloud map, the dominant unsteady flow structure in the flow field is identified. By visualizing the flow field distribution of different axial sections, the spatial distribution under the dominant frequency can be observed. If the flow field distribution and the dominant unsteady flow structure corresponding to the dominant frequency show consistent characteristics and spatial distribution on different sections, it can be confirmed that the dominant frequency is indeed caused by these dominant unsteady flow structures.

[0029] A system for analyzing flow structure inside a flow field, comprising:

[0030] An acquisition module is used to obtain flow field characteristic quantity information of the flow area to be analyzed;

[0031] A two-dimensional spectrum module, used for performing spectrum analysis and bispectral analysis on the flow field characteristic quantity information, so as to obtain a first characteristic frequency;

[0032] A POD module is used to perform intrinsic orthogonal decomposition on the flow field characteristic quantity information to obtain the energy distribution and flow structure of the flow field, and combine the first characteristic frequency to confirm the dominant frequency and the corresponding dominant unsteady flow structure;

[0033] The visualization module is used to perform frequency domain transformation on the pulsation information in the flow field characteristic quantity information obtained from the monitoring surface to obtain a two-dimensional distribution cloud map corresponding to the dominant frequency, obtain the spatial distribution of the dominant frequency in the flow field from the two-dimensional distribution cloud map, and obtain the area where the dominant unsteady flow structure is located in combination with the energy distribution of the flow field.

[0034] An analysis device for the internal flow structure of a flow field, comprising:

[0035] Memory for storing computer programs;

[0036] A processor, when executing the computer program, implements the above-mentioned method for analyzing the internal flow structure of the flow field.

[0037] By means of the above technical solution, the beneficial effects of the present invention are as follows:

[0038] In the traditional single mode analysis, there are usually multiple frequency components acting together. This method has certain limitations in capturing the influence of a single and detailed dominant frequency on the flow field. Therefore, in order to overcome this limitation, the intrinsic orthogonal decomposition method and the two-dimensional frequency domain visualization method complement and verify each other to determine the energy contribution, area and intensity of the dominant unsteady flow structure corresponding to each dominant frequency, so as to have a more comprehensive understanding of the dynamic characteristics of the flow structure and its contribution to the energy distribution of the flow field.

[0039] In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 is a flow chart of a method for analyzing the internal flow structure of a flow field in an embodiment of the present invention;

[0042] Figure 2 is a schematic diagram of selecting a propeller monitoring surface in an embodiment of the present invention;

[0043] Figure 3 is a schematic diagram of the pressure pulsation spectrum analysis results of some monitoring points in an embodiment of the present invention;

[0044] Figure 4 Schematic diagram of the pulsation bispectral analysis results of some monitoring points in an embodiment of the present invention;

[0045] Figure 5 Schematic diagram of the energy pulsation ratio of the first 20-order POD modes in different cross sections in an embodiment of the present invention;

[0046] Figure 6 1 is a schematic diagram of spectrum analysis of the first four-order POD modal coefficients of a cross section in an embodiment of the present invention;

[0047] Figure 7 Schematic diagram of spatial distribution of high-energy POD modes in radial and axial sections in the field according to an embodiment of the present invention;

[0048] Figure 8 is St in the embodiment of the present invention 0.6618 and St 0.80 Schematic diagram of the spatial visualization of the velocity pulsation amplitudes of the two dominant frequencies. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and to distinguish similar objects. There is no order of precedence between the two, and they cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0051] Example: Combination Figure 1-8 As shown, embodiment 1 discloses a method for analyzing the internal flow structure of a flow field, characterized in that it includes the following steps:

[0052] Monitoring points and detection surfaces are set up according to the complex unsteady flow area monitored by numerical simulation or experiment, or the flow area that actually needs to be analyzed, to collect the flow field characteristic quantity information, which includes two-dimensional time domain characteristics, pulsation information, snapshot data, etc. The flow field characteristic quantity information is used to analyze the time periodicity characteristics, energy intensity and specific spatial distribution of the unsteady flow structure.

[0053] Taking the unsteady flow of a propeller as an example, since the unsteady flow of a propeller has a strong three-dimensional nature, when analyzing it, Figure 2 As shown, multiple monitoring surfaces are selected in the propeller x-direction section to analyze the flow characteristics in different areas.

[0054] Spectral analysis is to perform Fourier transform on the pressure information in the flow field characteristic quantity information obtained at the monitoring point to obtain a spectrum of the flow field characteristic quantity changing with frequency. The spectrum reveals the amplitudes corresponding to different frequencies, that is, the pulsation information corresponding to each frequency. Among them, the larger the amplitude, the higher the energy of the frequency in the flow field, and the stronger the pulsation effect. The discrete peak frequency is obtained from the spectrum, and the discrete peak frequency usually corresponds to the dominant periodic structure in the flow field. That is, the discrete peak frequency can reflect the key dynamic behavior in the flow field.

[0055] Optionally, in the spectrum analysis, it is also possible to perform Fourier transform on the velocity information in the flow field characteristic quantity information.

[0056] Taking the unsteady flow of a propeller as an example, when analyzing the monitoring points of the propeller flow field data, all frequencies are dimensionless. The frequency values ​​in the following analysis are expressed by the Strouhal number (St), which is defined as:

[0057] St=f / f BPF

[0058] Where, f represents the frequency, f BPF Represents the blade passing frequency. Note: This step is not necessary and can also be performed without quantification.

[0059] like Figure 3 As shown in the figure, the results of the spectrum analysis of the pressure at different monitoring points on the monitoring surface in this example are shown. It can be clearly seen from the figure that the discrete peak frequencies of the pressure pulsation in the flow field of the monitoring surface are: St 0.6618 、St 0.70 and St 0.80 .

[0060] However, as a rotating part, the shaft frequency of the impeller in the propeller is usually consistent with the discrete peak frequency in the spectrum analysis, and the spectrum analysis cannot distinguish whether these discrete peak frequencies are mainly derived from the linear superposition of the shaft frequency or from other complex flows. Therefore, bispectral analysis is used to further verify and distinguish whether there is a linear superposition relationship between these discrete peak frequencies and the shaft frequency.

[0061] Bispectral analysis performs three Fourier transforms on the discrete peak frequency to calculate the phase coupling relationship between the frequency components, obtain the nonlinear relationship between different frequencies in the frequency domain, eliminate the influence of the linear superposition of the axis frequency, and obtain the first characteristic frequency.

[0062] The first characteristic frequency refers to the frequency directly extracted from the original flow field data by combining spectrum analysis or bispectral analysis and other methods. These frequencies are often related to significant periodicity or nonlinear behavior in the flow field.

[0063] Taking the unsteady flow of a propeller as an example, in order to further explore Figure 3 The relationship between the three peak frequencies shown in the spectrum diagram obtained at the monitoring point in the middle. The corresponding pressure data is processed using the bispectral analysis method. Figure 4 is the result of the bispectral analysis of the pressure at the monitoring point. The high amplitude frequencies on the diagonal line in the figure are nonlinear characteristic frequencies. Therefore, it can be determined that the first characteristic frequencies are St 0.6618 and St 0.80 .

[0064] at the same time, Figure 3 Another peak frequency St 0.70 ≈5×(St 0.80 -St 0.6618 ), so St 0.70 It is St 0.80 and St 0.6618 The linear superposition of these two nonlinear frequencies produces the physical meaning of St 0.80 and St 0.6618 It may be the characteristic frequency of the periodic flow structure in the flow field, the frequency St 0.70 The peak at is caused by the interaction of these two flow structures. The peak generated by this linear superposition reflects the complex dynamic process of the periodic flow structure in the flow field.

[0065] The intrinsic orthogonal decomposition (POD) method is used to study fluid flow phenomena, such as turbulence. The intrinsic orthogonal decomposition can be used to extract the main structures in the flow field. For example, for a complex turbulent pipe flow, the dominant vortex structure mode can be found through POD. The eigenvalues ​​corresponding to these modes reflect their contribution to energy. The modes corresponding to larger eigenvalues ​​contain most of the energy in the flow field, while the modes corresponding to smaller eigenvalues ​​contain less energy and may represent some small-scale fluctuations or noise. Specifically, the intrinsic orthogonal decomposition includes:

[0066] Arrange the velocity or pressure data in the flow field characteristic information into a data matrix:

[0067] Taking the velocity data in the flow field characteristic information as an example, it is organized into a data matrix. At a certain moment, the velocity values ​​of each point on the cross section are arranged in a certain spatial order to form a column vector v(x,t i ), which is the flow field snapshot. The snapshot matrix V composed of m flow field snapshots can be expressed as the formula:

[0068]

[0069] Where: v is the speed; x is the coordinate point; t i is the instantaneous moment; m is the number of flow field snapshots. The instantaneous flow field at any moment can be expressed as the basic flow and pulsation part The superposition of

[0070]

[0071] The POD method converts the pulsating part of the flow field into Decomposed into modal coefficients α j and the modal basis Φ j An orthogonal system:

[0072]

[0073] Where: α j is a row vector arranged in time order, reflecting the time domain characteristics of the jth POD mode; Φ j It is a column vector arranged according to the position of each point, reflecting the spatial characteristics of the j-th POD mode.

[0074] The following formula describes the core part of the POD method, namely the eigenvalue decomposition process of the correlation matrix D:

[0075] D=V T V

[0076]

[0077] DA j =λ j A j

[0078] j=1,2,3,…,m

[0079] Where: D is the correlation matrix; the matrix V is the matrix composed of the snapshots of the flow field fluctuation at m moments; λ j and A j They represent the eigenvalue of the j-th POD mode and its corresponding eigenvector respectively.

[0080] Each POD mode has a corresponding modal basis Φ j and modal coefficient α j , respectively calculated by the following formulas:

[0081]

[0082] α j =(Φ j ) T V

[0083] The characteristic value λ of the POD j It represents the contribution of each mode to the pulsating energy of the flow field. By sorting the modes according to their size, the first few high-energy POD modes of the flow field can be extracted. j Represents the distribution of each mode in the flow field space. By visualizing it in combination with the coordinate points, the spatial distribution characteristics of the mode can be obtained. POD modal coefficient α j It represents the fluctuation information of each mode in time, and the characteristic frequency of the mode can be obtained by Fourier transform.

[0084] Sort the POD eigenvalues ​​in descending order, accumulate the POD features corresponding to each POD mode from large to small to obtain a cumulative value, stop accumulating when the cumulative value is greater than or equal to 50%, and take the POD mode corresponding to the cumulative value as the high-energy mode. Compare the spatial distribution of the POD mode with the spectrum of the pressure change at the monitoring point with the frequency. In the figure, the part of the high-energy mode corresponding to the first characteristic frequency is the dominant frequency. The flow structure corresponding to the dominant frequency contributes greatly to the energy of the flow field.

[0085] Taking the unsteady flow of a propeller as an example, Figure 5 It is the energy distribution of the first 20 POD modes at different monitoring section pressures. It can be seen from the figure that the energy contribution of the first 4 modes to the flow field accounts for about 70%, which can cover the significant flow structure and changes in the flow field. Therefore, the flow structure characteristics that dominate the flow field will be analyzed in the future by analyzing the first 4 POD modes. Through the POD method, it is expanded from a specific monitoring point to the entire section, which provides strong support for studying the energy pulsation characteristics of the flow field.

[0086] Figure 6 is the spectrum transformation result of the time coefficient of the first four POD modes. The results show that the dominant frequencies of the modes with high energy contribution are concentrated in St 0.6618 、St 0.80 , which is consistent with the first characteristic frequency obtained by step bispectral analysis, indicating that St 0.6618 、St 0.80 The corresponding unsteady flow structure is the structure that mainly contributes to the energy in the flow field.

[0087] A Fourier transform is applied to the POD modal coefficients to obtain the second eigenfrequency of each mode. The second eigenfrequency is the frequency extracted from the POD modes and represents the main oscillation characteristic of each mode in time. It can further refine or confirm the first eigenfrequency and may also reveal more complex or secondary dynamic behavior in the flow.

[0088] Each coordinate point on the surface of the POD has a corresponding modal base value. Use MATLAB or other drawing software to draw a spatial distribution diagram based on the modal base value corresponding to the coordinate point to obtain the spatial distribution of the POD mode, thus realizing the visualization of modal distribution.

[0089] like Figure 7 As shown in the figure, the vorticity cloud diagram of the radial and axial sections and the spatial distribution of the high-energy POD1 mode in the flow field are displayed. Figure 7 It can be seen that although the blade tip vortex is very significant in the flow field, its energy is small and it is not the main flow structure in the flow field. The hub vortex, the shaft wake and the blade tip wake vortex are the main flow structures in the propeller area.

[0090] Frequency domain visualization obtains the two-dimensional time domain information of all grid nodes in the flow field characteristic quantity information detected by the detection surface, and performs frequency domain conversion on the two-dimensional time domain information for the dominant frequency to obtain a two-dimensional distribution cloud map.

[0091] The pulsation amplitudes in the two-dimensional distribution cloud map are sorted in descending order, and the first 70% are taken as the high pulsation amplitude area. The high pulsation amplitude area represents the part of the flow field that is most strongly affected by the frequency, and the spatial distribution of the dominant frequency in the flow field can be clarified to understand how the dominant frequency affects the flow structure in a specific area;

[0092] The corresponding relationship between the frequency distribution and the POD mode is analyzed, and the two-dimensional distribution cloud map is compared with the POD mode to confirm the connection between the dominant frequency and the POD modal basis. It can be observed that the characteristic morphology of some POD modes is more prominent in the cloud map of specific frequency components, indicating that these modes contribute more to the frequency component;

[0093] According to the spatial distribution of the dominant frequency in the two-dimensional distribution cloud map, the dominant unsteady flow structures in the flow field, such as vortices, recirculation areas and separation areas, are identified. These structures usually appear as obvious pulsation amplitude concentration areas in the frequency domain cloud map. By visualizing the flow field distribution of different axial sections, the spatial distribution under the dominant frequency can be observed. If the flow field distribution and the dominant unsteady flow structure corresponding to the dominant frequency show consistent characteristics and spatial distribution on different sections, then it can be confirmed that the dominant frequency is indeed caused by these dominant unsteady flow structures.

[0094] Taking the unsteady flow of a propeller as an example, Figure 8 are the two dominant frequencies St 0.6618 and St 0.80 The spatial visualization results of the velocity pulsation amplitude are shown in the figure. It can be seen from the figure that the hub vortex and the shaft rear vortex contain two characteristic frequencies St 0.6618 and St 0.80The characteristic frequency of the blade tip wake vortex, another main flow structure of the flow field, is St 0.80 .

[0095] By combining POD analysis with frequency domain visualization, the flow field can be comprehensively analyzed from the perspective of energy and frequency. The POD method reveals the energy distribution characteristics of the flow field, while the frequency domain visualization shows the spatial distribution of specific frequency components. The combination of the two can effectively analyze the unsteady characteristics of the flow field and identify key flow structure areas, providing comprehensive support for in-depth understanding of complex flow phenomena.

[0096] The results of comprehensive spectrum analysis, bispectral analysis, frequency domain visualization and POD analysis confirm the dominant frequency and dominant unsteady flow structure in the flow field and their spatial distribution. By comparing these analysis results, we identified the frequencies and flow structures that continue to dominate under different working conditions. Based on these findings, the specific location, morphology and time evolution behavior of the dominant unsteady flow structure in the flow field were determined. Using this information, targeted optimization or control strategies can be formulated.

[0097] Embodiment 2 discloses a system for analyzing the internal flow structure of a flow field, comprising:

[0098] An acquisition module is used to obtain flow field characteristic quantity information of the flow area to be analyzed;

[0099] A two-dimensional spectrum module, used for performing spectrum analysis and bispectral analysis on the flow field characteristic quantity information, so as to obtain a first characteristic frequency;

[0100] A POD module is used to perform intrinsic orthogonal decomposition on the flow field characteristic quantity information to obtain the energy distribution and flow structure of the flow field, and combine the first characteristic frequency to confirm the dominant frequency and the corresponding dominant unsteady flow structure;

[0101] The visualization module is used to perform frequency domain transformation on the pulsation information in the flow field characteristic quantity information obtained from the monitoring surface, obtain a two-dimensional distribution cloud map corresponding to the dominant frequency, obtain the spatial distribution of the dominant frequency in the flow field from the two-dimensional distribution cloud map, and obtain the area where the dominant unsteady flow structure is located in combination with the energy distribution of the flow field. The analysis system for the internal flow structure of the flow field is used to execute the above-mentioned analysis system method for the internal flow structure of the flow field.

[0102] Embodiment 3 discloses a device for analyzing the internal flow structure of a flow field, comprising:

[0103] Memory for storing computer programs;

[0104] A processor, when executing the computer program, implements the above-mentioned method for analyzing the internal flow structure of the flow field.

[0105] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for analyzing the internal flow structure of a flow field, characterized in that: The following steps are involved: Obtaining flow field characteristic quantity information; Converting the flow field characteristic quantity information into a spectrum that changes with frequency, and obtaining the discrete peak frequency in the flow field pulsation from the spectrum; Establishing a relationship model between the discrete peak frequencies to obtain a nonlinear relationship between the discrete peak frequencies and confirming a first characteristic frequency corresponding to an unsteady flow; Intrinsic orthogonal decomposition, using the intrinsic orthogonal decomposition method to decompose the flow field characteristic quantity information into a number of POD modes with different energy contributions, and obtain the energy distribution and flow structure of the flow field according to each POD mode, combined with the first characteristic frequency to confirm the dominant frequency and the corresponding dominant unsteady flow structure; Frequency domain visualization, for the dominant frequency, the pulsation information in the flow field characteristic information is transformed into the frequency domain to obtain a two-dimensional distribution cloud map corresponding to the dominant frequency, obtain the spatial distribution of the dominant frequency in the flow field, and combine the energy distribution of the flow field to obtain the area where the dominant unsteady flow structure is located.

2. The method for analyzing the internal flow structure of a flow field according to claim 1, characterized in that: The step of converting the flow field characteristic quantity information into a spectrum that varies with frequency includes: performing Fourier transform on the pressure information in the flow field characteristic quantity information to obtain a spectrum of the pressure at the monitoring point that varies with frequency, and the discrete peak frequency can be obtained from the spectrum.

3. The method for analyzing the internal flow structure of a flow field according to claim 2, characterized in that: The step of obtaining the nonlinear relationship between the discrete peak frequencies comprises: With respect to the discrete peak frequencies, a bispectral analysis is performed on the flow field characteristic quantity information to obtain a nonlinear relationship between the discrete peak frequencies, and the influence of linear superposition of axial frequencies is eliminated, thereby obtaining the first characteristic frequency.

4. The method for analyzing the internal flow structure of a flow field according to claim 3, characterized in that: The steps of proper orthogonal decomposition include: The pressure data in the flow field characteristic quantity information is organized into a data matrix, and the data matrix is ​​decoupled and calculated to obtain the POD mode and the corresponding POD eigenvalue. The POD mode is sorted in descending order according to the size of the POD eigenvalue, and the POD features corresponding to each POD mode are accumulated from large to small to obtain a cumulative value. When the cumulative value is greater than or equal to 50%, the accumulation is stopped, and the corresponding POD mode in the cumulative value is taken as the high-energy mode.

5. The method for analyzing the internal flow structure of a flow field according to claim 4, characterized in that: Each of the POD modes has a corresponding POD modal basis and POD modal coefficients. By applying Fourier transform to the POD modal coefficients, the second characteristic frequency of each mode can be obtained. The spatial distribution of the POD mode can be drawn according to the coordinate points corresponding to the POD modal basis.

6. The method for analyzing the internal flow structure of a flow field according to claim 5, characterized in that: The spatial distribution of the POD mode is compared with the spectrum of the pressure change at the monitoring point with the frequency, and the frequency corresponding to both the high-energy mode and the first characteristic frequency is the dominant frequency.

7. The method for analyzing the internal flow structure of a flow field according to claim 6, characterized in that: Steps for frequency domain visualization include: Performing frequency domain transformation on the velocity pulse or pressure pulse in the flow field characteristic quantity information to obtain a velocity two-dimensional distribution cloud map or a pressure two-dimensional distribution cloud map corresponding to the dominant frequency, and obtaining the spatial distribution of the dominant frequency in the flow field and the source of instability in the flow from the two-dimensional distribution cloud map; Sorting the pulsation amplitudes in the two-dimensional distribution cloud map in descending order, and taking the first 70% as the high pulsation amplitude area; By combining the two-dimensional distribution cloud map and the POD mode basis, the relationship between the dominant frequency and the spatial distribution of each POD mode is obtained, so that the characteristic frequency components of different flow structures can be distinguished.

8. The method for analyzing the internal flow structure of a flow field according to claim 1, characterized in that: The step of acquiring flow field characteristic quantity information includes: arranging monitoring points and monitoring surfaces in the flow area to be analyzed to collect the flow field characteristic quantity information.

9. A system for analyzing the internal flow structure of a flow field, characterized in that: include: An acquisition module is used to obtain flow field characteristic quantity information of the flow area to be analyzed; A two-dimensional spectrum module, used for performing spectrum analysis and bispectral analysis on the flow field characteristic quantity information, so as to obtain a first characteristic frequency; A POD module is used to perform intrinsic orthogonal decomposition on the flow field characteristic quantity information to obtain the energy distribution and flow structure of the flow field, and combine the first characteristic frequency to confirm the dominant frequency and the corresponding dominant unsteady flow structure; The visualization module is used to perform frequency domain transformation on the pulsation information in the flow field characteristic quantity information to obtain a two-dimensional distribution cloud map corresponding to the dominant frequency, obtain the spatial distribution of the dominant frequency in the flow field from the two-dimensional distribution cloud map, and obtain the area where the dominant unsteady flow structure is located in combination with the energy distribution of the flow field.

10. An analysis device for the internal flow structure of a flow field, characterized in that: include: Memory for storing computer programs; A processor, when executing the computer program, the processor implements the method for analyzing the internal flow structure of the flow field according to any one of claims 1-8.

Images