High-pressure cylinder flow field analysis method based on computational fluid dynamics

Abstract

The invention discloses a high-pressure cylinder flow field analysis method based on computational fluid dynamics, and particularly relates to the field of high-pressure cylinder flow field analysis, and the method comprises the steps: constructing a flow field model through a cross-manufacturer design knowledge base, recognizing a sensitive area according to a flow loss criterion, carrying out the meshing, carrying out the partitioning parallel solving of a fluid equation set, and dynamically terminating the iteration based on the entropy yield change. And finally, packaging the analysis result. According to the high-pressure cylinder flow field analysis method based on computational fluid dynamics, by automatically recognizing a secondary flow loss sensitive area and combining gridding processing, intelligence and light weight of grid generation are achieved, the manual intervention time is shortened, and the grid generation efficiency is improved; a computational domain is divided into a mainstream region, a boundary layer region and a vortex core region, and a corresponding fluid control equation set is matched, so that the utilization rate of computational resources is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of high-pressure cylinder flow field analysis, and more particularly to a high-pressure cylinder flow field analysis method based on computational fluid dynamics. Background Art

[0002] With the increasing demand for efficiency improvement and transformation of thermal power units, refined simulation of the flow design of the high-pressure cylinder of the turbine is particularly important. Traditional design relies on the empirical formula system of a single manufacturer and achieves a step-by-step distribution of enthalpy drop by simplifying the flow field equations. Although traditional technology can complete basic design, it has fundamental flaws: it is limited by the two-dimensional model simplification mechanism and manual grid discretization method, and cannot accurately capture the three-dimensional secondary flow vortex system structure and boundary layer separation effect.

[0003] To overcome the defects of traditional designs, existing technologies introduce computational fluid dynamics three-dimensional full-channel simulation methods, use parametric modeling tools to construct a million-level grid model, optimize blade profiles through hybrid loading flow pattern algorithms and controllable vortex design, and use multi-condition transient solvers to achieve dynamic distribution of enthalpy drop, thereby improving the accuracy of high-pressure cylinder efficiency prediction.

[0004] However, in actual use, it still has some shortcomings, such as low mesh generation efficiency, and the construction of a single model with a million-level mesh requires long-term manual intervention; the consumption of computing resources has increased dramatically, and the full three-dimensional transient solution causes the simulation of a single working condition to take too long; the optimization iteration cost is too high, and multiple parameter iterations require weeks of computing cycles; which ultimately leads to delayed engineering response and out-of-control hardware costs. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a high-pressure cylinder flow field analysis method based on computational fluid dynamics, which solves the problems raised in the above-mentioned background technology through the following scheme.

[0006] To achieve the above object, the present invention provides the following technical solutions: A high-pressure cylinder flow field analysis method based on computational fluid dynamics, comprising: S1: Based on the enthalpy drop distribution rules and blade profile parameters in the preset cross-manufacturer design knowledge base, generate an inter-stage constraint matrix and blade cascade geometric topology to construct a first flow field analysis model of the high-pressure cylinder; S2: In response to a preset flow loss design criterion, identifying a secondary flow loss sensitive area in the first flow field analysis model, and generating a second flow field analysis model through meshing; S3: Based on the second flow field analysis model, perform region division and match the corresponding fluid control equations to perform parallel solutions in the partitions to obtain a first flow field data set of the high-pressure cylinder; S4: using the first flow field data set, monitoring the entropy generation rate change of the dynamic and static blade cascade intersection area in real time; when the entropy generation rate change is lower than the convergence threshold, terminating the iterative calculation, and identifying the second flow field data set based on the reverse direction; S5: Encapsulate the analysis process including S1 to S4 and the second flow field data set output by S4 into a lightweight container, and deploy it to the distributed computing node.

[0007] Preferably, the step S1, constructing a first flow field analysis model of the high-pressure cylinder, specifically includes: The low-level load is quantified as an inter-stage constraint matrix, and the low-level load is specifically manifested as small enthalpy drop and multiple stages; Obtaining the cascade geometry topology of the cascade channel based on the characteristics of low hub ratio and high aspect ratio ,in, Expressed as the blade root diameter, It is expressed as relative blade height, and the characteristics of low hub ratio and high aspect ratio are specifically manifested as low root diameter and large relative blade height; Output the full flow channel geometric topology model structure of the high-pressure cylinder, that is, the first flow field analysis model.

[0008] Preferably, the step S2, identifying the secondary flow loss sensitive area, specifically includes: Quantitative analysis of vortex core strength and calculation of secondary flow intensity factor , specifically expressed as: , in, Expressed as the velocity component of the fluid cluster in the tangential direction, Expressed as the velocity component of the fluid cluster in the axial direction, It is expressed as the radial position of the fluid cluster in the flow channel; Preferably, the step S2, identifying the secondary flow loss sensitive area, further includes: By calculating the entropy risk index Quantitative analysis locates the entropy increase risk area, specifically expressed as: , in, Expressed as the blade root diameter, Expressed as relative leaf height, Expressed as a proportionality coefficient, Expressed as a function of flow channel curvature.

[0009] Preferably, the specific steps of the gridding process in S2 include: executing, based on the secondary flow loss sensitive area identified by the secondary flow intensity factor; Secondary flow intensity factor The vortex core area of ​​>5% is encrypted to 0.1mm grid resolution, and the grid resolution that satisfies the wall distance is generated near the wall. Boundary layer grid of <1; Entropy risk index The non-critical area of ​​<0.1% is sparsely distributed to a 2mm grid.

[0010] Preferably, the step S3 of obtaining the first flow field dataset specifically includes: Dividing the calculation domain of the second flow field analysis model into a mainstream region, a boundary layer region and a vortex core region; The vortex core region is determined by the secondary flow intensity factor >5% decision, boundary layer area, based on the wall distance of the near-wall grid The rest is divided into mainstream areas; Solve the Euler equation in the mainstream region, solve the Prandtl boundary layer equation in the boundary layer region, and solve the vortex transport equation in the vortex core region; By building a data interface, data exchange between the three zones is transmitted in real time.

[0011] Preferably, the convergence threshold in S4 is defined as the rate of change of entropy production rate <0.1%.

[0012] Preferably, the reverse identification in S4 is specifically: performing directional optimization on a key design parameter set of the high-pressure cylinder flow field based on the extraction steam regulation accuracy of ±2%.

[0013] The technical effects and advantages of the present invention are as follows: 1. The present invention automatically identifies the secondary flow loss sensitive areas through S2 and combines it with grid processing to achieve intelligent and lightweight grid generation, reduce manual intervention time, and improve grid generation efficiency; 2. The present invention divides the computational domain into the mainstream region, boundary layer region, and vortex core region through S3 and matches the corresponding fluid control equations, thereby improving the utilization of computing resources, shortening the time consumption of full three-dimensional transient solution, and alleviating the problem of sharp increase in computing resource consumption; 3. The present invention uses S4 to monitor the entropy production rate changes in the transition area between the moving and static blades in real time and dynamically terminate the iteration, which greatly reduces the number of iterations and the dimension of optimization parameters, and reduces the calculation cycle and cost of the optimization iteration. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flowchart of a method for analyzing the flow field of a high-pressure cylinder based on computational fluid dynamics according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] 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 the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0016] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "said", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more of the listed items.

[0017] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0018] As attached Figure 1 The computational fluid dynamics-based high-pressure cylinder flow field analysis method shown here constructs a flow field model using a cross-manufacturer design knowledge base, identifies and meshes sensitive areas based on flow loss criteria, solves the fluid equations in parallel using partitions, dynamically terminates iterations based on changes in entropy production, and ultimately packages the analysis results. Specifically, the method includes the following steps: S1: Based on the enthalpy drop distribution rules and blade profile parameters in the preset cross-manufacturer design knowledge base, generate an inter-stage constraint matrix and blade cascade geometric topology to construct a first flow field analysis model of the high-pressure cylinder; S2: In response to a preset flow loss design criterion, identifying a secondary flow loss sensitive area in the first flow field analysis model, and generating a second flow field analysis model through meshing; S3: Based on the second flow field analysis model, perform region division and match the corresponding fluid control equations to perform parallel solutions in the partitions to obtain a first flow field data set of the high-pressure cylinder; S4: using the first flow field data set, monitoring the entropy generation rate change of the dynamic and static blade cascade intersection area in real time; when the entropy generation rate change is lower than the convergence threshold, terminating the iterative calculation, and identifying the second flow field data set based on the reverse direction; S5: Encapsulate the analysis process including S1 to S4 and the second flow field data set output by S4 into a lightweight container, and deploy it to the distributed computing node.

[0019] It should be noted that this embodiment takes a 330MW imported Alstom steam turbine as an example. It aims to, without changing the size of the turbine body, apply the full-process algorithm from S1 to S5, comprehensively consider the performance optimization of various components of the turbine, and comprehensively improve the overall performance of the turbine by modifying the high-pressure cylinder flow and the intermediate-pressure cylinder rotating baffle, as well as improving the condenser-related technologies, thereby achieving a dual improvement in the unit's flow efficiency and industrial steam supply capacity. Ultimately, the advancement and effectiveness of the algorithm of the present invention are verified with actual operating data.

[0020] Specifically, in S1, structured processing of input data is achieved by pre-building a cross-manufacturer design knowledge base.

[0021] It should be noted that the cross-manufacturer design knowledge base is not a simple data set, but is constructed through deep integration and fusion of proprietary design rules and parameter sets of multiple manufacturers.

[0022] In this embodiment, the cross-manufacturer design knowledge base integrates the enthalpy drop distribution rule library of Factory A and the blade profile parameter data set of Factory B. The enthalpy drop distribution rule library defines the recommended value range, constraint conditions and adjustment strategy of the enthalpy drop ratio between each stage in the high-pressure cylinder under different working conditions. The blade profile parameter data set contains key parameters in more than 20 dimensions, including but not limited to blade installation angle, chord length, inlet and outlet airflow angles, etc.

[0023] In one possible implementation, generating the inter-stage constraint matrix and the cascade geometry topology includes: quantizing the low-stage load into an inter-stage constraint matrix [M] n×n ; Obtain the cascade geometry topology of the cascade channel based on the characteristics of low hub ratio and high aspect ratio ,in, Expressed as the blade root diameter, It is expressed as relative blade height; the full flow channel geometric topology model structure of the output high-pressure cylinder is the first flow field analysis model.

[0024] In this embodiment, the first flow field analysis model adopts the B-rep+NURBS hybrid representation in STEP P242 format, and only stores the control point coordinates and the constraint matrix [M]. n×n , reducing the file size; and through the inter-level constraint matrix [M] n×n The design variable dimension is compressed from traditional full parameter optimization to only adjusting the non-zero elements of the matrix, which reduces the upper limit of the number of S4 iterations.

[0025] Furthermore, the low-stage load is specifically represented by the classic layout of “small enthalpy drop, multiple stages”, with the goal of making each flow stage operate near its highest efficiency point and making it a computable constraint condition, specifically quantified as the inter-stage constraint matrix [M] n×n Implementation, specifically expressed as: , in, and Respectively expressed as Level and The enthalpy drop of the stage, and It is expressed as the constraint boundary of the enthalpy drop value, and the value is dynamically loaded by the enthalpy drop allocation rule library in the cross-manufacturer design knowledge base according to the specific design conditions.

[0026] Furthermore, the characteristics of low hub ratio and high aspect ratio significantly reduce the axial flow velocity by realizing the geometric topology of "low root diameter and large relative blade height", thereby effectively suppressing the secondary flow loss and ultimately achieving an improvement in flow efficiency. Construct NURBS control points of the parameterized template. The construction method of the control points is specifically expressed as follows: , in, , , , , Expressed as the blade chord length, Expressed as the blade root diameter, It is expressed as relative leaf height; it should be noted that the Represented as the leading edge point, specifically embodied as a low root diameter constraint; Represented as the leaf basin control point; It is represented as the leaf back control point, which is specifically reflected in the large relative leaf height constraint; It is represented by the trailing edge point, which is specifically reflected in the low root diameter constraint.

[0027] Specifically, in S2, the geometric topology model structure of the entire flow channel output by S1 is inherited, and at the same time, preset design criteria containing expert knowledge, namely, flow loss design criteria, are integrated to establish an intelligent identification mechanism for areas sensitive to secondary flow losses, thereby constructing a lightweight grid model, namely, the second flow field analysis model.

[0028] In this embodiment, the flow loss design criterion is derived from the design criterion for hybrid loading blade shaping and the controllable vortex design parameters in the external configuration file.

[0029] In one possible embodiment, the intelligent identification of the secondary flow loss sensitive area includes: quantitative analysis of the vortex core strength, calculation of the secondary flow intensity factor , to evaluate the secondary flow loss risk at each point in the flow channel, specifically expressed as: , in, Expressed as the velocity component of the fluid cluster in the tangential direction, Expressed as the velocity component of the fluid cluster in the axial direction, Indicates the radial position of the fluid micro-group in the flow channel; in this embodiment, when the calculated When the value exceeds the preset threshold of 5%, the area is automatically marked as a secondary flow loss sensitive area.

[0030] In a possible embodiment, the intelligent identification of the secondary flow loss sensitive area further includes: constructing an entropy increase prediction model based on the design principle of hybrid loading blade molding, and calculating the entropy increase risk index Quantitative analysis locates the entropy increase risk area, specifically expressed as: , in, Expressed as the blade root diameter, Expressed as relative leaf height, Expressed as a proportional coefficient, it is determined according to the design criteria for mixed loading blade forming. It is expressed as a flow channel curvature function and is used to evaluate the entropy increase risk caused by flow channel curvature.

[0031] In one possible embodiment, the gridding process includes: identifying the secondary flow loss sensitive area based on the secondary flow intensity factor; performing The vortex core area of ​​>5% is encrypted to 0.1mm grid resolution, and the grid resolution that satisfies the wall distance is generated near the wall. <1 boundary layer grid to ensure the accuracy of boundary layer flow solution; The non-critical area of ​​<0.1% is sparsely distributed to a 2mm grid.

[0032] In this embodiment, in order to achieve global lightweighting, an Octree space partitioning topology optimization algorithm is also used to minimize the total number of grid nodes while ensuring solution accuracy.

[0033] Specifically, in S3, the lightweight grid model generated by S2 is received, and the preset optimization objectives are integrated, the calculation domain is divided, and the corresponding fluid control equations are matched to the divided areas for partitioned and parallel solution. At the same time, real-time data exchange is performed between the areas through the preset flux matching interface, so as to obtain the transient flow field data set of the high-pressure cylinder, that is, the first flow field data set.

[0034] In this embodiment, the optimization target is derived from the dynamic-static matching optimization parameters loaded in the configuration file and the enthalpy drop distribution data inherited from the original design solution.

[0035] In a possible embodiment, the matching of the region division and the fluid control equations includes: dividing the calculation domain of the second flow field analysis model into a mainstream region, a boundary layer region and a vortex core region, wherein the vortex core region is determined by the secondary flow intensity factor >5%, the boundary layer area is determined by the wall distance of the near-wall grid The value is determined by the equation of mass flow, and the rest is divided into the mainstream area; the Euler equation is solved in the mainstream area, the Prandtl boundary layer equation is solved in the boundary layer area, and the vortex transport equation is solved in the vortex core area; the data exchange between the three areas is transmitted in real time by building a data interface.

[0036] It should be noted that the physical markers carried by each grid in the second flow field analysis model generated by S2 are automatically divided into regions; in this embodiment, the wall distance The grid with a value of <5 is identified as the boundary layer area; the Euler equation directly loads the enthalpy drop distribution data of the original design scheme as the initial condition for the solution; in the boundary layer area with severe flow gradient but extremely thin scale, the Prandtl boundary layer equation is used for accurate calculation, and the key parameters of the turbulence model are combined with the inherited S1 In the vortex core region where the flow structure is most complex, the vortex transport equation that captures the vortex motion is adopted, and its boundary conditions are coupled with the design criteria of hybrid loading blade forming inherited from the S2 input.

[0037] Furthermore, to achieve regional load balancing optimization, computing resources are dynamically allocated according to the computational complexity of different regions, that is, the number of GPU cores is intelligently allocated according to the grid proportion and estimated computational weight of each region. In this embodiment, the vortex core area with a grid proportion of 12% and a computational weight of 58% is allocated to the six GPU cores with the strongest performance, while the mainstream area with the smallest computational workload, with a grid proportion of 65% and a computational weight of 10%, is allocated only one GPU core.

[0038] In this embodiment, in the mainstream region, since the viscosity effect is relatively weak, the Euler equation is used to simplify the model, and its control equation is specifically expressed as follows: , in, Represented as at the current moment , the mass of the fluid per unit volume at the center of the grid cell, represents the transient solution time step, Expressed as the absolute velocity vector at the center of the grid cell; Represented as a unit tensor, Expressed as static pressure.

[0039] It should be noted that 、 、 The initial value of is directly inherited from S1, and no additional calibration is required; viscous dissipation is ignored in the mainstream area. The change of is determined only by mass conservation; the transient solution time step Synchronous with the monitoring frequency of the entropy production rate change in S4; the absolute velocity vector Depend on and Determine, as the initial condition of Euler equation; the static pressure , which is used to calculate the total enthalpy of the mainstream area.

[0040] In this embodiment, in the boundary layer region, the viscosity effect and shear force are the dominant factors, and an accurate calculation is performed based on the Prandtl boundary layer equation, which is specifically expressed as: , in, Expressed as the time-averaged velocity component along the tangential direction of the blade wall, Expressed as the time-averaged velocity component along the wall normal, Expressed as streamline coordinates along the wall, Expressed as the wall normal coordinate, Represented as at the current moment , the mass of the fluid per unit volume at the center of the grid cell, Expressed as static pressure, Expressed as kinematic viscosity.

[0041] It should be noted that in the high pressure cylinder boundary layer, Directly related to large relative leaf height The load distribution on the blade surface is determined by the thin boundary layer assumption. ≪ ,and The magnitude is resolved by the near-wall grid generated by S2 with y⁺<5; the starting point of the streamline coordinates along the wall corresponds to the leading edge of the blade derived by S1, and the end point is the trailing edge, which is used to integrate the friction loss of the entire blade; the wall normal coordinates =0 means the blade surface, = is the outer edge of the boundary layer, Depend on Determined together with the Reynolds number; if >0, boundary layer separation will occur, which will directly affect the entropy generation monitoring of S4; kinematic viscosity It automatically adjusts the turbulence model parameters according to the "large relative blade height" so that the viscosity changes with the blade geometric characteristics, thereby improving the accuracy of boundary layer solution.

[0042] In this embodiment, in the vortex core region, the vortex transport equation is used to solve the problem, and is coupled with the boundary conditions of the design criteria for hybrid loading blade forming in S2, specifically expressed as: , in, It is expressed as the rotation intensity and direction of the microclusters in the vortex core region, It represents the transient solution time step, Expressed as the absolute velocity vector at the center of the grid cell, Expressed as kinematic viscosity.

[0043] It should be noted that It is expressed as the diffusion rate of vorticity in the vortex core to the surrounding due to viscosity; It is expressed as the stretching and bending effect of vortex lines in the velocity gradient field; It is expressed as the instantaneous rate of change of the vorticity vector with time when it moves with the fluid particles in the vortex core region.

[0044] Specifically, in S4, the first flow field data set output by S3 is loaded, and at the same time, the preset stability requirements for steam supply pressure fluctuations are read, and the convergence entropy production rate change threshold is calculated, so as to reversely identify the key design parameter set that has the greatest impact on the target based on the flow field data in the converged state, that is, the second flow field data set.

[0045] In one possible implementation, the convergence threshold is the rate of change of entropy production <0.1%, and the reverse identification is specifically: performing directional optimization on the key design parameter set of the high-pressure cylinder flow field based on the extraction steam regulation accuracy of ±2%.

[0046] It should be noted that the entropy production rate field is extracted from the first flow field data set. , and then calculate its derivative in the time dimension , the derivative It is a key indicator to measure the stability and convergence of the flow field. Furthermore, the entropy production monitoring function is continuously called in an iterative loop. Once the monitored When the mean value is lower than the preset convergence threshold of 0.1%, the convergence flag is triggered, indicating that the flow field has stabilized and no further iteration is required.

[0047] In this embodiment, when the steam supply pressure fluctuation is monitored to be greater than 1.5%, the convergence threshold is automatically relaxed to 0.2% to avoid invalid fine iteration under unstable working conditions; otherwise, a stricter threshold of 0.1% is adopted.

[0048] Furthermore, after the convergence flag is triggered, that is, after the iteration is terminated, reverse identification of sensitive parameters is started. Based on the final converged flow field data and combined with the performance indicator of the steam extraction regulation accuracy of ±2%, the sensitivity of each design parameter to the final performance is analyzed by calculating the Jacobian matrix. Through the preset sensitivity threshold, the sensitivity threshold of this embodiment is 0.15, 3 to 5 key parameters with the greatest impact on the performance are automatically screened out as the key design parameter set, that is, the second flow field data set.

[0049] In this embodiment, the extraction steam pressure PG is set as the observed value, and the key design parameters are , then the sensitivity matrix is ,like >0.15, Expressed as the index of the key design parameter, the key design parameter ; Otherwise, remove it.

[0050] The second flow field data set includes velocity field, vorticity distribution of all time steps and boundary layer data.

[0051] Specifically, in S5, the lightweight container is a Docker image, and after deployment, responds to a condenser water supply change operating condition signal to trigger local re-optimization of the high-pressure cylinder.

[0052] In this embodiment, the modifications and changes to the high-pressure cylinder flow passage of the unit are all directly derived from the output results of the algorithm of the present invention; the aerodynamic optimization of the blades is entirely based on the second flow field data set output by S4, that is, a sensitive parameter set containing only 3 to 5 dimensions; the physical transformation strictly implements the small enthalpy drop, multi-stage regulating stage and pressure stage layout scheme defined by the constraint matrix of S1; the specific profile design of the blades follows the flow loss criterion of S2 and adopts an advanced mixed loading flow pattern; the geometric form, by reducing the blade root diameter and increasing the relative blade height at the same time, accurately achieves the optimized geometric topology of low root diameter and large relative blade height defined by S1; through the dimensionality reduction of sensitive parameters of S4 and the topology and flow pattern optimization of S1 and S2, the physical transformation is accurately guided and a significant performance improvement is achieved; Furthermore, the modification of the intermediate pressure cylinder's rotating diaphragm is driven by the algorithm data of this invention. The seal strengthening process is implemented based on the axial thrust balance parameters and material durability constraints of the second flow field data output by S4. Nitriding treatment is performed on the sealing surface to improve surface hardness. To ensure the unit's stability under variable operating conditions, the reserved cooling area is calculated based on the operating condition response requirements of S5's lightweight container. Furthermore, the pipeline configuration plan for large-capacity water makeup to the condenser directly implemented the dynamic deployment plan output by the lightweight container in S5. Under base load conditions, one 100t / h pipeline was activated, and under variable conditions where the demand for makeup water increased sharply, it automatically switched to three 150t / h pipelines running in parallel, ensuring efficient makeup water operation under all conditions. Ultimately, the oxygen content in the condensate was stabilized, successfully achieving the engineering goals set during the S4 optimization iteration and verifying the effectiveness of S5's edge deployment and real-time response.

[0053] Secondly: The drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention. Other structures may refer to conventional designs. The same embodiment and different embodiments of the present invention may be combined with each other without conflict. Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A high-pressure cylinder flow field analysis method based on computational fluid dynamics, characterized in that: include: S1: Based on the enthalpy drop distribution rules and blade profile parameters in the preset cross-manufacturer design knowledge base, generate an inter-stage constraint matrix and blade cascade geometric topology to construct a first flow field analysis model of the high-pressure cylinder; S2: In response to a preset flow loss design criterion, identifying a secondary flow loss sensitive area in the first flow field analysis model, and generating a second flow field analysis model through meshing; S3: Based on the second flow field analysis model, perform region division and match the corresponding fluid control equations to perform parallel solutions in the partitions to obtain a first flow field data set of the high-pressure cylinder; S4: using the first flow field data set, monitoring the entropy generation rate change of the dynamic and static blade cascade intersection area in real time; when the entropy generation rate change is lower than the convergence threshold, terminating the iterative calculation, and identifying the second flow field data set based on the reverse direction; S5: Encapsulate the analysis process including S1 to S4 and the second flow field data set output by S4 into a lightweight container, and deploy it to the distributed computing node.

2. The high-pressure cylinder flow field analysis method based on computational fluid dynamics according to claim 1, characterized in that: The S1, constructing a first flow field analysis model of the high-pressure cylinder, specifically includes: The low-level load is quantified as an inter-stage constraint matrix, and the low-level load is specifically manifested as small enthalpy drop and multiple stages; Obtaining the cascade geometry topology of the cascade channel based on the characteristics of low hub ratio and high aspect ratio ,in, Expressed as the blade root diameter, It is expressed as relative blade height, and the characteristics of low hub ratio and high aspect ratio are specifically manifested as low root diameter and large relative blade height; Output the full flow channel geometric topology model structure of the high-pressure cylinder, that is, the first flow field analysis model.

3. The high-pressure cylinder flow field analysis method based on computational fluid dynamics according to claim 1, characterized in that: The step S2, identifying the secondary flow loss sensitive area, specifically includes: Quantitative analysis of vortex core strength and calculation of secondary flow intensity factor , specifically expressed as: ， in, Expressed as the velocity component of the fluid cluster in the tangential direction, Expressed as the velocity component of the fluid cluster in the axial direction, It is expressed as the radial position of the fluid cluster in the flow channel.

4. The method for analyzing the flow field of a high-pressure cylinder based on computational fluid dynamics according to claim 1, characterized in that: The step S2, identifying the secondary flow loss sensitive area, specifically further includes: By calculating the entropy risk index Quantitative analysis locates the entropy increase risk area, specifically expressed as: ， in, Expressed as the blade root diameter, Expressed as relative leaf height, Expressed as a proportionality coefficient, Expressed as a function of flow channel curvature.

5. The method for analyzing the flow field of a high-pressure cylinder based on computational fluid dynamics according to claim 4, characterized in that: The specific steps of the gridding process in S2 include: executing, based on the secondary flow loss sensitive area identified by the secondary flow intensity factor; Secondary flow intensity factor The vortex core area of ​​>5% is encrypted to 0.1mm grid resolution, and the grid resolution that satisfies the wall distance is generated near the wall. Boundary layer grid of <1; Entropy risk index The non-critical area of ​​<0.1% is sparsely distributed to a 2mm grid.

6. The high-pressure cylinder flow field analysis method based on computational fluid dynamics according to claim 1, characterized in that: The step S3 of obtaining the first flow field dataset specifically includes: Dividing the calculation domain of the second flow field analysis model into a mainstream region, a boundary layer region and a vortex core region; The vortex core region is determined by the secondary flow intensity factor >5% decision, boundary layer area, based on the wall distance of the near-wall grid The rest is divided into mainstream areas; Solve the Euler equation in the mainstream region, solve the Prandtl boundary layer equation in the boundary layer region, and solve the vortex transport equation in the vortex core region; By building a data interface, data exchange between the three zones is transmitted in real time.

7. The method for analyzing the flow field of a high-pressure cylinder based on computational fluid dynamics according to claim 1, characterized in that: The S4, convergence threshold is defined as the rate of change of entropy production rate <0.1%.

8. The method for analyzing the flow field of a high-pressure cylinder based on computational fluid dynamics according to claim 1, characterized in that: The S4, reverse identification, is specifically: performing directional optimization on the key design parameter set of the high-pressure cylinder flow field based on the extraction steam regulation accuracy of ±2%.

Images