One-dimensional network automatic construction method for internal cooling of turbine blade of gas turbine

Abstract

The invention provides a one-dimensional network automatic construction method for internal cooling of turbine blades of a gas turbine. The method comprises the following steps: S1, constructing a gas turbine integrated design platform; s2, analyzing the geometric model; s3, cooling unit division and identification; s4, generating a one-dimensional network node and a connection relationship; s5, constructing a one-dimensional network model for internal cooling of the turbine blade of the gas turbine in the integrated design platform of the gas turbine; and S6, network model verification and optimization: performing comparison verification and optimization on the generated one-dimensional network model and the detailed three-dimensional geometric model of the turbine blade. According to the method, on the basis of a gas turbine integrated design platform, one-key modeling and subsequent thermodynamic calculation preparation of the complex cooling network are achieved through the steps of structural topology recognition, automatic flow path classification, thermodynamic parameter correlation mapping and the like, multi-working-condition quick response and repeated iterative calculation can be conducted, the accuracy rate is high, data consistency is good, and the method is suitable for large-scale popularization and application. And the design and simulation efficiency of the cooling system is obviously improved.

Description

Technical Field

[0001] This invention relates to the field of gas turbine cooling structure modeling and thermodynamic calculation technology, and in particular to a method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades. Background Technology

[0002] Heavy-duty gas turbine blades operate under high-temperature, high-pressure gas environments for extended periods, requiring precise internal cooling channels for effective cooling to ensure their thermal stability and structural safety. Currently, blade internal cooling channel designs often employ a combination of heat transfer enhancement methods, such as fin reinforcement, impact / re-impact, and turbine bore cooling, forming complex branched network structures. Traditional one-dimensional thermal network modeling methods for these structures suffer from the following shortcomings: 1. Low modeling efficiency, requiring manual definition of cooling branches, connection nodes, and boundary conditions segment by segment, resulting in high time consumption and error rates; 2. Strong geometric dependence, making it difficult to adapt to variations in blade cooling topologies and lacking general processing capabilities; 3. Lack of automatic judgment mechanisms, making it difficult to intelligently distinguish complex flow characteristics such as main branches, sub-branches, blind ends, or return channels; 4. Decoupling from three-dimensional geometric models, failing to quickly extract structural parameters from CAD models, leading to poor data consistency. Therefore, there is an urgent need to develop a cooling network construction method that is practically applicable to engineering applications and possesses automatic identification and rapid modeling capabilities to support practical engineering needs such as heat load calculation, heat transfer optimization, and rapid response under multiple operating conditions. Summary of the Invention

[0003] The main objective of this invention is to provide an automatic one-dimensional network construction method for the internal cooling of gas turbine blades, in order to solve the problems in the background art.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An automatic one-dimensional network construction method for internal cooling of gas turbine blades includes the following steps: S1: Construct an integrated design platform for gas turbines; S2: Geometric model analysis: The three-dimensional geometric model of the turbine blade of a heavy-duty gas turbine is read through the gas turbine integrated design platform, and structural feature data is extracted. S3: Cooling Unit Division and Identification: The internal cooling structure of the blades is divided into units, and different types of cooling units and their connection relationships are identified. S4: Generation of one-dimensional network nodes and connections. Based on the partitioned cooling units and their characteristic parameters, determine the location and number of nodes in the one-dimensional network and establish the connection relationships between nodes. S5: Construct a one-dimensional network model of the internal cooling of the turbine blades of the gas turbine in the gas turbine integrated design platform, calculate the parameters of each node and connection path in the one-dimensional network, calculate the heat transfer parameters involved through heat transfer correlation, and assign the calculation results to the corresponding nodes or connection paths of the one-dimensional network to generate a one-dimensional network model. S6: Network Model Verification and Optimization. The generated one-dimensional network model is compared and optimized with the detailed three-dimensional geometric model of the turbine blade.

[0005] Furthermore, in S1, the gas turbine integrated design platform includes: The data reading module reads data from CSV format files and parses it into a processable dictionary format. The XML structure building module creates the root element and basic information nodes of an XML file in "FMDNA" format, and creates nodes such as components, fittings, nodes and links based on CSV data, and sets the corresponding attributes and text content. The relationship mapping module constructs upstream and downstream relationship mappings between components, as well as mapping relationships between components and nodes; The coordinate calculation module calculates the coordinate information of the component based on information such as the component type, channel number, and channel column number. The file generation module formats the constructed XML tree structure and writes it into the specified XML file.

[0006] Furthermore: In S2, the geometric model analysis includes the following steps: S21: The gas turbine integrated design platform uses a computer program to automatically read the three-dimensional geometric model of a heavy-duty gas turbine blade. This model includes the blade's shape and detailed structure of the internal cooling channels, such as film cooling holes, fins, and diffuser sections. By analyzing the topology and geometric parameters of the geometric model, it extracts key information about the shape, size, and position of each part of the blade and automatically identifies the main edges and rounded edges. It calculates and outputs parameters such as cross-sectional area, perimeter, aspect ratio, and side length distribution, providing basic data support for the subsequent construction of a one-dimensional network model. S22: Automatically preprocess the extracted geometric parameters, including data cleaning, format conversion, and unit unification, to ensure data consistency and accuracy so that they can be smoothly input into the automatic construction algorithm for processing.

[0007] Furthermore, S21 includes the following steps: S211: The user selects the geometric entity through the plug-in interface and sets the number of segments n along the height direction; S212: The plugin generates n parallel cutting surfaces at equal intervals along the Z-axis at the starting position at the bottom of the cube solid, dividing the solid into m blocks of equal height; at the same time, corresponding cross sections are created on the n cutting surfaces for subsequent contour recognition and parameter extraction. S213: Each section is automatically assigned a unique number and uniformly grouped into a "section group" for batch management and processing; S214: The user manually selects a representative edge line on a certain cross section as the starting line of the "long side" of the cross section profile; S215: Starting from this edge line, the plugin automatically identifies the complete closed contour according to the topological relationship. All boundary lines that make up the closed contour are set m, and the boundary lines are topologically sorted. The length values ​​of all edge lines are extracted to form a "set d". In "set d", the shorter edges are removed in order of topological order. These edges are called "short edges". The remaining edges form a "list Q" which still maintains the original topological order. The edge adjacent to the "long edge" is the "wide edge", and the edge opposite to the "long edge" is still the "long edge". S216: The plugin automatically identifies the main edge line and extracts the channel boundary, cross-sectional shape, and flow path geometric features based on edge length characteristics and relative position; in addition, the plugin provides unified settings for the display attributes of the cross-section. S217: The plugin exports all the above cross-section parameter data in CSV file format. Each row corresponds to a cross-section, and the fields include cross-section number, perimeter, area, length of the longer side and the width of the shorter side, and aspect ratio.

[0008] Furthermore, in S3, the division and identification of cooling units includes the following steps: S31: Based on the geometric features and flow characteristics of the internal cooling channels of the blade, a set of cooling unit division criteria is established. According to these criteria, the internal cooling structure of the blade is automatically divided into units, and different types of cooling units are identified, such as straight channel units, curved channel units, finned cooling units, and film cooling units. The cooling structure is divided into several functional units (such as main channels, branches, orifices, chambers, etc.) to facilitate the identification of their physical properties and connection relationships. S32: For each type of cooling unit, define its corresponding characteristic parameters and connection rules, including the parameters of length, diameter, and roughness of the straight channel unit, as well as the connection method between adjacent straight channel units; by dividing and defining the characteristics of the cooling units, the complex internal cooling system of the blade is decomposed into several relatively simple basic units, providing a modular basis for the construction of a one-dimensional network model.

[0009] Furthermore, in S4, the generation of one-dimensional network nodes and connections includes the following steps: S41: Based on the divided cooling units and their characteristic parameters, automatically determine the position and number of nodes in the one-dimensional network, set nodes at the beginning, end and key feature positions of the cooling units, and assign a unique identifier and coordinate information to each node; S42: Establish the connection relationship between nodes. According to the flow path of the cooling medium inside the blade, connect adjacent nodes according to the actual physical connection situation to form directional flow paths. Each connection path has corresponding attributes, including parameters such as flow direction, flow length, and equivalent diameter. These parameters are derived from the geometric features and physical properties of the corresponding cooling unit.

[0010] Furthermore, S42 is achieved through the following steps: S421: Create XML root element and basic information nodes: Create the root element "FMDNA" of the FMDNA format XML file, and add a Logon information node under the root element, including information such as data source, database, login project, username, and password; at the same time, add basic information nodes such as Network, Project, EyePointLocation, and ViewPointLocation, and set the corresponding text content. S422: Read CSV data: Use a CSV reading tool to read data from a CSV file, parse each row of data into a dictionary format, and store it in a data list; S423: Construct component relationship mapping: Traverse the read CSV data, construct the mapping relationship between upstream component and current component, and downstream component and current component respectively, based on the upstream component number and downstream component number, and store them in the corresponding mapping dictionary; Bidirectional mapping construction: Through independent upstream / downstream component mapping tables, clearly establish the connection relationship between components to form a directed graph structure. Each node can quickly find upstream and downstream connected objects through the mapping table; S424: Generate node mapping relationships: Based on the upstream component mapping relationship, generate a unique node number for each different upstream component and establish a mapping relationship between the component number and the node number; based on the downstream component mapping relationship, combined with the generated upstream node mapping relationship, establish a mapping relationship between the downstream component and the node number; the node coordinates directly inherit the coordinates of the associated components to ensure that the geometric position is consistent with the topological relationship. S425: Calculate component coordinates: Traverse the information of each component in the CSV data, calculate the X, Y, and Z coordinates of the component according to the component type, channel number, and channel column number, and store the coordinate information in the coordinate mapping dictionary; the coordinate calculation results of all components are uniformly stored in coordinate_map, and nodes and links directly reference this mapping table to ensure the geometric consistency of the entire network layout. Special components are arranged vertically with upstream components through the Y coordinate incremental mapping table. S426: Create component and pipe nodes: Iterate through the CSV data again and determine whether to create a component node or a pipe node based on the component type; for pipes, create an "FMPipe" node under the "Pipes" node and set its name, network number, start coordinates, and end coordinates; for other components, create an "FMComponent" node under the "Components" node and set its name, network number, and location coordinates. S427: Create Nodes and Link Nodes: Based on the node mapping relationship, create an "FMNode" node under the "Nodes" node, and set the node name, network number, and location coordinates; based on the component and node mapping relationship, create upstream and downstream "FMLink" nodes under the "Links" node respectively, and set the link type, component, starting arm, node, and whether it is orthogonal; dynamically set the FromArm attribute according to whether it is an endpoint component, uniformly set the LinkType, and link directly associates the component ID and node number to form a complete topology unit of "component-link-node"; S428: Generate a formatted XML file: Convert the constructed XML tree structure into a string, use an XML parsing tool to parse the string into a DOM object, format the DOM object, remove the XML declaration line, and finally write the formatted XML content into the specified XML file.

[0011] Furthermore, in S5, constructing a one-dimensional network model for the internal cooling of gas turbine blades within the gas turbine integrated design platform includes the following steps: S51: Based on the geometric parameters and physical properties of each cooling unit, and combined with the corresponding thermodynamic and fluid dynamic models, automatically calculate the parameter values ​​of each node and connection path in the one-dimensional network. S52: For the heat transfer parameters involved, the corresponding heat transfer correlation is used for calculation, and the calculation results are assigned to the corresponding one-dimensional network nodes or connection paths. S53: All calculated parameter values ​​are automatically stored in the data structure of the one-dimensional network model, forming a complete one-dimensional network model together with nodes and connections.

[0012] Furthermore, S6 includes the following steps: S61: The automatically generated one-dimensional network model is compared and verified with the detailed three-dimensional geometric model of the blade. The internal cooling process of the blade is numerically simulated using computational fluid dynamics simulation to obtain detailed temperature distribution, pressure distribution and flow field information. These simulation results are compared and analyzed with the prediction results of the one-dimensional network model to evaluate the accuracy of the one-dimensional network model. S62: Based on the results of the comparative verification, the one-dimensional network model is optimized and adjusted. The optimization methods include refining or simplifying the division of cooling units, recalculating and calibrating the network parameters, and correcting the node positions and connection relationships; repeated verification and optimization iterations are carried out.

[0013] Compared with existing technologies, this invention achieves automatic identification of complex cooling channel topology and efficient reconstruction of one-dimensional network models, significantly improving the automation and efficiency of the modeling process. Through topology extraction and geometric feature recognition mechanisms, it can quickly analyze the flow paths, branch structures, and node connection relationships within the cooling channels, automatically generating a physically meaningful one-dimensional network model. This invention is easy to integrate into existing gas turbine integrated design platforms, enabling cross-platform deployment and engineering applications, effectively improving the consistency, accuracy, and reusability of the model, and reducing errors and uncertainties caused by differences in modelers. It significantly reduces reliance on manual modeling experience, improves the stability and robustness of the network topology construction process, and is suitable for batch rapid modeling scenarios of multi-blade and multi-configuration cooling structures. It has good scalability and promotional value, and is particularly suitable for engineering design and optimization analysis of complex cooling systems such as heavy-duty gas turbines, significantly improving the efficiency of cooling system design and simulation. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the entity module segmentation process of the plug-in in this invention.

[0015] Figure 2 This is a schematic diagram of the operation of the cross-sectional information extraction module of the present invention.

[0016] Figure 3 This is a schematic diagram illustrating the classification and labeling of the cross-sectional edge lines of the present invention.

[0017] Figure 4 This is a flowchart of the data conversion algorithm of the present invention. Detailed Implementation

[0018] To make the objectives, technical means and advantages of the present invention readily understood, the present invention will be further described below in conjunction with specific embodiments.

[0019] like Figures 1 to 4As shown, this invention provides an automatic one-dimensional network construction method for internal cooling of gas turbine blades, the steps of which are as follows: S1: Construct an integrated design platform for gas turbines; S2: Geometric model analysis: The three-dimensional geometric model of the turbine blade of a heavy-duty gas turbine is read through the gas turbine integrated design platform, and structural feature data is extracted. S3: Cooling Unit Division and Identification: The internal cooling structure of the blades is divided into units, and different types of cooling units and their connection relationships are identified. S4: Generation of one-dimensional network nodes and connections. Based on the partitioned cooling units and their characteristic parameters, determine the location and number of nodes in the one-dimensional network and establish the connection relationships between nodes. S5: Construct a one-dimensional network model of the internal cooling of the turbine blades of the gas turbine in the gas turbine integrated design platform, calculate the parameters of each node and connection path in the one-dimensional network, calculate the heat transfer parameters involved through heat transfer correlation, and assign the calculation results to the corresponding nodes or connection paths of the one-dimensional network to generate a one-dimensional network model. S6: Network Model Verification and Optimization. The generated one-dimensional network model is compared and optimized with the detailed three-dimensional geometric model of the turbine blade. Based on a gas turbine integrated design platform, this invention achieves one-click modeling and subsequent thermal calculation preparation for complex cooling networks through steps such as structural topology identification, automatic flow path classification, and thermodynamic parameter correlation mapping. It can perform rapid response under multiple operating conditions and iterative calculations with high accuracy, intelligently distinguish complex flow characteristics, and has good data consistency.

[0020] Preferred: In S1, the gas turbine integrated design platform includes: The data reading module reads data from a CSV file and parses it into a processable dictionary format. The implementation principle is as follows: using Python's "csv" module, it reads the data line by line from the CSV file, determines the field name of each data item based on the file's header row, and then parses each line of data into a dictionary. The dictionary's keys are the field names, and the values ​​are the corresponding data content. Finally, all dictionaries are stored in a list for subsequent processing. Input: CSV file path. Output: A list of dictionaries containing component information.

[0021] The XML structure building module creates the root element and basic information nodes for an "FMDNA" formatted XML file. It creates component, pipe, node, and link nodes based on CSV data, setting their corresponding attributes and text content. The implementation principle is as follows: using Python's "xml.etree.ElementTree" or "lxml" module to build the XML tree, it first creates the root element, then adds basic information nodes according to the FMDNA format requirements, iterates through the component information dictionary list obtained from the data reading module, creates corresponding component, pipe, node, and link elements for each component, sets their attributes and text content, and adds them to the appropriate positions in the XML tree. Input: The component information dictionary list obtained from the data reading module. Output: The constructed XML tree structure.

[0022] The relationship mapping module constructs upstream and downstream relationship mappings between components, as well as mappings between components and nodes. The implementation principle is as follows: It analyzes the elements in the component information dictionary list, extracts data representing the connection relationships between components (such as component input ports, output ports, etc.), and constructs a dictionary of upstream and downstream relationship mappings between components based on this data. The key is the component identifier, and the value is a list of component identifiers with upstream and downstream relationships. Simultaneously, it constructs a mapping dictionary between components and nodes to correctly establish the association between components and nodes in the XML tree. Input: Component information dictionary list obtained from the data reading module. Output: Dictionary of upstream and downstream relationship mappings between components and dictionary of component-node mappings.

[0023] The coordinate calculation module calculates the coordinate information of components based on their type, channel number, and channel column number. The implementation principle is as follows: according to predefined coordinate calculation rules, using component type, channel number, and channel column number as parameters, a series of mathematical formulas are used to determine the coordinate position of each component in the FMDNA format XML file. For example, the channel number and channel column number can be used as a basis, combined with a specific offset of the component type, to calculate the X and Y coordinate values ​​of the component. Input: A dictionary list of component information (containing component type, channel number, channel column number, etc.). Output: A dictionary or data structure containing component coordinate information.

[0024] The file generation module formats the constructed XML tree structure and writes it to a specified XML file. The implementation principle is as follows: the constructed XML tree structure is formatted to conform to the requirements of the FMDNA format, including indentation and line breaks. Then, the write function of the "xml.etree.ElementTree" module or the "lxml" module is used to write the formatted XML tree structure to the XML file at the specified path. Input: The constructed XML tree structure, the specified XML file path. Output: An XML file conforming to the FMDNA format requirements.

[0025] Through the collaborative work of the above modules, automatic conversion from CSV data to FMDNA format XML files can be achieved, enabling rapid construction of one-dimensional network models, improving work efficiency, reducing human error, and meeting the data processing and model building needs in complex industrial scenarios.

[0026] Preferred method: In S2, the geometric model analysis includes the following steps: S21: The gas turbine integrated design platform utilizes a computer program to automatically read the 3D geometric model of a heavy-duty gas turbine blade. This model includes the blade's external shape and detailed internal cooling channel structure, such as film cooling holes, fins, and diffuser sections. By analyzing the topology and geometric parameters of the geometric model, it extracts key information about the shape, size, and location of each part of the blade, automatically identifies principal edges and fillet edges, and calculates and outputs parameters such as cross-sectional area, perimeter, aspect ratio, and side length distribution, providing fundamental data support for the subsequent construction of a one-dimensional network. The extracted structural feature data is not only used for spatial geometric statistics but also serves as a basis for deriving the resistance and heat transfer parameters of each cooling section in the one-dimensional network, especially providing direct support for modeling the thermodynamic performance of film cooling holes or finned channels. All parameters are ultimately exported as a unified CSV file, with section numbers identified and associated with the structure for easy integration. S22: Automatically preprocess the extracted geometric parameters, including data cleaning, format conversion, and unit unification, to ensure data consistency and accuracy so that they can be smoothly input into the automatic construction algorithm for processing.

[0027] Preferred, such as Figures 1 to 4 As shown, S21 includes the following steps: S211: The user selects a geometric entity through the plug-in interface and sets the number of segments n along the height direction (Z-axis); S212: The plugin generates n parallel cutting surfaces at equal intervals along the Z-axis at the starting position at the bottom of the cube solid, dividing the solid into m blocks of equal height; at the same time, corresponding cross sections are created on the n cutting surfaces for subsequent contour recognition and parameter extraction. S213: Each section is automatically assigned a unique number and uniformly grouped into a "section group" for batch management and processing; S214: The user manually selects a representative edge line on a certain cross section as the starting line of the "long side" of the cross section profile; S215: Starting from this edge line, the plugin automatically identifies the complete closed contour according to the topological relationship. All boundary lines that make up the closed contour are set m, and the boundary lines are topologically sorted. The length values ​​of all edge lines are extracted to form a "set d". In "set d", the shorter edges are removed in order of topological order. These edges are called "short edges". The remaining edges form a "list Q" which still maintains the original topological order. The edge adjacent to the "long edge" is the "wide edge", and the edge opposite to the "long edge" is still the "long edge". S216: The plugin automatically identifies the main edge line and extracts geometric features such as channel boundaries, cross-sectional shape, and flow path based on edge length characteristics and relative position. In addition, the plugin provides unified settings for the display attributes of the cross-section, such as setting the cross-section color to blue, transparency to 50%, and display name to "Section_001". S217: The plugin exports all the above cross-sectional parameter data in CSV file format. Each row corresponds to one cross-section, and the fields include cross-section number, perimeter, area, length of the longer side and the wider side, aspect ratio, etc. Edge line recognition and sorting combine topological relationships and geometric features to ensure stable and reliable results.

[0028] Preferred: such as Figure 4 As shown, in S3, the division and identification of cooling units includes the following steps: S31: Based on the geometric features and flow characteristics of the internal cooling channels of the blade, a set of cooling unit division criteria is established. According to these criteria, the internal cooling structure of the blade is automatically divided into units, and different types of cooling units are identified, such as straight channel units, curved channel units, finned cooling units and film cooling units. S32: For each type of cooling unit, define its corresponding characteristic parameters and connection rules, such as the length, diameter, roughness, and other parameters of the straight channel unit, as well as the connection method (series or parallel) between adjacent straight channel units; by dividing and defining the characteristics of the cooling units, the complex internal cooling system of the blade is decomposed into several relatively simple basic units, providing a modular basis for the construction of a one-dimensional network.

[0029] Preferably, in S4, the generation of one-dimensional network nodes and connection relationships includes the following steps: S41: Based on the divided cooling units and their characteristic parameters, automatically determine the position and number of nodes in the one-dimensional network, set nodes at the beginning, end and key characteristic positions (such as air film outlet, fin turning point, etc.) of the cooling units, and assign a unique identifier and coordinate information to each node.

[0030] S42: Establish the connection relationship between nodes. According to the flow path of the cooling medium inside the blade, connect adjacent nodes according to the actual physical connection situation to form directional flow paths. Each connection path has corresponding attributes, including parameters such as flow direction, flow length, and equivalent diameter. These parameters are derived from the geometric characteristics and physical properties of the corresponding cooling unit.

[0031] Preferred: such as Figure 4 As shown, S42 is achieved through the following steps: S421: Create XML root element and basic information nodes: Create the root element "FMDNA" of the FMDNA format XML file, and add a Logon information node under the root element, including information such as data source, database, login project, username, and password; at the same time, add basic information nodes such as Network, Project, EyePointLocation, and ViewPointLocation, and set the corresponding text content.

[0032] S422: Read CSV data: Use a CSV reader to read data from a CSV file, parse each row of data into a dictionary format, and store it in a data list.

[0033] S423: Construct component relationship mapping: Traverse the read CSV data, construct the mapping relationship between upstream and current components and downstream and current components respectively according to the upstream component number and downstream component number, and store them in the corresponding mapping dictionary; Bidirectional mapping construction: Through independent upstream / downstream component mapping tables (upstream_mapping / downstream_mapping), clearly establish the connection relationship between components to form a directed graph structure. Each node (component) can quickly find upstream and downstream connected objects through the mapping table.

[0034] S424: Generate Node Mapping Relationships: Based on the upstream component mapping relationship, generate a unique node number for each different upstream component and establish a mapping relationship between component number and node number; based on the downstream component mapping relationship, and combined with the generated upstream node mapping relationship, establish a mapping relationship between downstream components and node numbers. Node Generation Logic: Create a new node only when an upstream component appears for the first time (avoiding duplicate nodes); node coordinates directly inherit the coordinates of the associated component, ensuring that the geometric position is consistent with the topological relationship; establish a bidirectional mapping from component ID to node number through unnode_map / dnnode_map to facilitate subsequent link construction.

[0035] S425: Calculate Component Coordinates: Iterate through the information of each component in the CSV data, calculate the X, Y, and Z coordinates of the component according to the component type, channel number, and channel column number, and store the coordinate information in the coordinate mapping dictionary. Different coordinate calculation methods are used for different types of components (such as fittings, loss components, pressure components, etc.). The coordinate calculation results of all components (including fittings / loss / pressure components) are uniformly stored in the coordinate_map. Nodes and links directly reference this mapping table to ensure the geometric consistency of the entire network layout. Special components (such as loss components) are vertically arranged with upstream components through the Y-coordinate increment mapping table (y_increment_map) to avoid layout overlap.

[0036] S426: Create component and pipe nodes: Iterate through the CSV data again, determining whether to create a component node or a pipe node based on the component type. For pipes, create an "FMPipe" node under the "Pipes" node and set its name, network number, start coordinates, and end coordinates; for other components, create an "FMComponent" node under the "Components" node and set its name, network number, and location coordinates.

[0037] S427: Create Nodes and Link Nodes: Based on the node mapping relationship, create an "FMNode" node under the "Nodes" node, setting the node name, network number, and location coordinates; based on the component-node mapping relationship, create upstream and downstream "FMLink" nodes under the "Links" node, setting attributes such as link type, component, starting arm, node, and orthogonality. Dynamically set the FromArm attribute (single-arm / dual-arm connection) based on whether it is an endpoint component (upstream / downstream ID=0), uniformly set LinkType, and directly associate the link with the component ID and node number to form a complete topology unit of "component-link-node".

[0038] S428: Generate a formatted XML file: Convert the constructed XML tree structure into a string, use an XML parsing tool to parse the string into a DOM object, format the DOM object, remove the XML declaration line, and finally write the formatted XML content into the specified XML file.

[0039] This system, based on CSV data, aims to convert CSV data into FMDNA format XML files and generate one-dimensional network nodes and connections. The system addresses the complex need to convert CSV data into XML files in a specific format (such as FMDNA) and to construct a one-dimensional network model of the internal cooling system for heavy-duty gas turbine blades.

[0040] Preferred method: In S5, constructing a one-dimensional network model of the internal cooling of gas turbine blades in the gas turbine integrated design platform includes the following steps: S51: Based on the geometric parameters and physical properties of each cooling unit, and combined with the corresponding thermodynamic and fluid dynamic models, the parameter values ​​of each node and connection path in the one-dimensional network are automatically calculated. For example, the flow resistance coefficient of the cooling channel can be calculated and assigned based on factors such as the channel's shape, size, and surface roughness, using empirical formulas or numerical simulation results; the flow coefficient of the film gas orifice is determined based on parameters such as the orifice diameter, shape, and inlet / outlet pressure difference.

[0041] S52: For the heat transfer parameters involved, such as the convective heat transfer coefficient between the inner wall of the cooling channel and the cooling medium, the film cooling effect, etc., based on the material properties of the blade, the operating temperature, the physical properties of the cooling medium and the flow state, a suitable heat transfer correlation is used for calculation, and the calculation results are assigned to the corresponding one-dimensional network nodes or connection paths.

[0042] S53: All calculated parameter values ​​are automatically stored in the data structure of the one-dimensional network model, forming a complete one-dimensional network model together with nodes and connections, providing an accurate model basis for subsequent cooling performance analysis and optimization.

[0043] Preferably, step S6 includes the following steps: S61: The automatically generated one-dimensional network model is compared and verified with the detailed three-dimensional geometric model of the blade. Computational fluid dynamics (CFD) simulations are used to numerically simulate the internal cooling process of the blade, obtaining detailed temperature distribution, pressure distribution, and flow field information. These CFD simulation results are then compared and analyzed with the prediction results of the one-dimensional network model to evaluate the accuracy of the one-dimensional network model.

[0044] S62: Based on the comparative verification results, the one-dimensional network model is optimized and adjusted. Optimization methods include refining or simplifying the cooling unit division, recalculating and calibrating network parameters, and correcting node positions and connection relationships. Through repeated verification and optimization iterations, the accuracy and reliability of the one-dimensional network model are continuously improved, enabling it to accurately reflect the actual physical process of internal blade cooling.

[0045] The geometric model is parsed to generate a corresponding CSV file, with column names including "Component Number," "Component Type," "Channel Number," "Channel Column Number," "Upstream Component Number," "Downstream Component Number," and "Unique Identifier." First, the data reading module reads the CSV file, parsing each row of data into a dictionary format and storing it in a data list. Next, the XML structure building module creates the root element "FMDNA" in the FMDNA format XML file and adds a Logon information node, setting the data source to "localhost\SQLExpress," the database to "Flomaster," the login project to "Flomaster," the username to "Admin," and the password information to a preset format; it also adds basic information nodes such as Network, Project, EyePointLocation, and ViewPointLocation. Then, the relationship mapping module traverses the data list, constructing mapping relationships between upstream components and the current component, and between downstream components and the current component, based on the "Upstream Component Number" and "Downstream Component Number." For example, if the upstream component number of a component is "1," then that component is added to the mapping list of upstream component number "1." Next, node mapping relationships are generated. Based on the upstream component mapping relationships, a unique node number is generated for each different upstream component, and a mapping relationship between component numbers and node numbers is established. Then, based on the downstream component mapping relationships and the generated upstream node mapping relationships, a mapping relationship between downstream components and node numbers is established. The coordinate calculation module traverses the data list and calculates the component coordinates according to the component type, channel number, and channel column number. For pipe fittings, if the component type is "duct", the start and end coordinates are calculated based on the channel and column number; for other components, the position coordinates are calculated according to the corresponding rules. Then, the XML structure building module traverses the data list again and creates component or pipe fitting nodes according to the component type. For pipe fittings, an "FMPipe" node is created under the "Pipes" node, and its name, network number, start coordinates, and end coordinates are set; for other components, an "FMComponent" node is created under the "Components" node, and its name, network number, and position coordinates are set. Then, based on the node mapping relationship, an "FMNode" node is created under the "Nodes" node, setting its name, network number, and location coordinates. Based on the component-node mapping relationship, upstream and downstream "FMLink" nodes are created under the "Links" node, setting attributes such as link type, component, starting arm, node, and orthogonality. In the file generation module, the constructed XML tree structure is converted into a string, and an XML parsing tool is used to parse the string into a DOM object. After formatting, it is written to the specified XML file.Through the above steps, CSV data was successfully converted into an XML file conforming to the FMDNA format, achieving efficient data conversion and processing. Finally, the FMDNA-compliant XML file was imported into the gas turbine integrated design platform for the automatic construction and calculation of a one-dimensional network model of the internal cooling of gas turbine blades. This invention achieves automatic identification of complex cooling channel topologies and efficient reconstruction of one-dimensional network models, significantly improving the automation and efficiency of the modeling process. Through topology extraction and geometric feature recognition mechanisms, it can quickly analyze the flow paths, branch structures, and node connection relationships within the cooling channels, automatically generating a physically meaningful one-dimensional network model. This invention is easy to integrate into existing gas turbine integrated design platforms, enabling cross-platform deployment and engineering applications. It effectively improves the consistency, accuracy, and reusability of the model, reducing errors and uncertainties caused by differences in modelers. It significantly reduces reliance on manual modeling experience, improves the stability and robustness of the network topology construction process, and is suitable for batch rapid modeling scenarios of multi-blade, multi-configuration cooling structures. It has good scalability and promotional value, and is particularly suitable for engineering design and optimization analysis of complex cooling systems such as heavy-duty gas turbines.

[0046] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes, equivalent substitutions, and improvements can be made without departing from the spirit and scope of the invention, and all such changes should fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An automatic construction method for a one-dimensional network of internal cooling for gas turbine blades, characterized in that, Includes the following steps: S1: Construct an integrated design platform for gas turbines; S2: Geometric model analysis: The three-dimensional geometric model of the turbine blade of a heavy-duty gas turbine is read through the gas turbine integrated design platform, and structural feature data is extracted. S3: Cooling Unit Division and Identification: The internal cooling structure of the blades is divided into units, and different types of cooling units and their connection relationships are identified. S4: Generation of one-dimensional network nodes and connections. Based on the partitioned cooling units and their characteristic parameters, determine the location and number of nodes in the one-dimensional network and establish the connection relationships between nodes. S5: Construct a one-dimensional network model of the internal cooling of the turbine blades of the gas turbine in the gas turbine integrated design platform, calculate the parameters of each node and connection path in the one-dimensional network, calculate the heat transfer parameters involved through heat transfer correlation, and assign the calculation results to the corresponding nodes or connection paths of the one-dimensional network to generate a one-dimensional network model. S6: Network Model Verification and Optimization. The generated one-dimensional network model is compared and optimized with the detailed three-dimensional geometric model of the turbine blade.

2. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 1, characterized in that, In S1, the gas turbine integrated design platform includes: The data reading module reads data from CSV format files and parses them into a processable dictionary format. The XML structure building module creates the root element and basic information nodes of an XML file in "FMDNA" format, and creates nodes such as components, fittings, nodes and links based on CSV data, and sets the corresponding attributes and text content. The relationship mapping module constructs upstream and downstream relationship mappings between components, as well as mapping relationships between components and nodes; The coordinate calculation module calculates the coordinate information of the component based on information such as the component type, channel number, and channel column number. The file generation module formats the constructed XML tree structure and writes it into the specified XML file.

3. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 1, characterized in that... In S2, the geometric model analysis includes the following steps: S21: The gas turbine integrated design platform uses a computer program to automatically read the three-dimensional geometric model of a heavy-duty gas turbine blade. This model includes the blade's shape and the detailed structure of its internal cooling channels. By analyzing the topology and geometric parameters of the geometric model, it extracts key information about the shape, size, and position of each part of the blade and automatically identifies the main edges and rounded edges. It calculates and outputs parameters such as cross-sectional area, perimeter, aspect ratio, and side length distribution, providing basic data support for the subsequent construction of a one-dimensional network model. S22: Automatically preprocesses the extracted geometric parameters, including data cleaning, format conversion, and unit unification, to ensure data consistency and accuracy.

4. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 3, characterized in that, S21 includes the following steps: S211: The user selects the geometric entity through the plug-in interface and sets the number of segments n along the height direction; S212: The plugin generates n parallel cutting surfaces at equal intervals along the Z-axis at the starting position at the bottom of the cube solid, dividing the solid into m blocks of equal height; at the same time, corresponding cross sections are created on the n cutting surfaces for subsequent contour recognition and parameter extraction. S213: Each section is automatically assigned a unique number and uniformly grouped into a "section group" for batch management and processing; S214: The user manually selects a representative edge line on a certain cross section as the starting line of the "long side" of the cross section profile; S215: The plugin takes this edge line as the starting point, automatically identifies the complete closed contour according to the topological relationship, and forms a set m of all the boundary lines that make up the closed contour. The boundary lines are then topologically sorted. The length values ​​of all the edge lines are extracted to form a set d. In set d, the shorter edges are removed in order of topological order. These edges are called "short edges". The remaining edges form a list Q, which still maintains the original topological order. The edges adjacent to the "long edges" are called "wide edges", and the edges opposite to the "long edges" are also called "long edges". S216: The plugin automatically identifies the main edge line and extracts the channel boundary, cross-sectional shape, and flow path geometric features based on edge length characteristics and relative position; in addition, the plugin provides unified settings for the display attributes of the cross-section. S217: The plugin exports all the above cross-section parameter data in CSV file format. Each row corresponds to a cross-section, and the fields include cross-section number, perimeter, area, length of the longer side and the width of the shorter side, and aspect ratio.

5. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 1, characterized in that, In S3, the division and identification of cooling units includes the following steps: S31: Based on the geometric features and flow characteristics of the internal cooling channels of the blade, a cooling unit division criterion is established. According to these criteria, the internal cooling structure of the blade is automatically divided into units, and different types of cooling units are identified. S32: For each type of cooling unit, define its corresponding characteristic parameters and connection rules, including the parameters of length, diameter, and roughness of the straight channel unit, as well as the connection method between adjacent straight channel units; by dividing and defining the characteristics of the cooling units, the complex internal cooling system of the blade is decomposed into several relatively simple basic units, providing a modular basis for the construction of a one-dimensional network model.

6. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 1, characterized in that, In S4, the generation of one-dimensional network nodes and connections includes the following steps: S41: Based on the divided cooling units and their characteristic parameters, automatically determine the position and number of nodes in the one-dimensional network, set nodes at the beginning, end and key feature positions of the cooling units, and assign a unique identifier and coordinate information to each node; S42: Establish the connection relationship between nodes. According to the flow path of the cooling medium inside the blade, connect adjacent nodes according to the actual physical connection situation to form directional flow paths. Each connection path has corresponding attributes, including parameters such as flow direction, flow length, and equivalent diameter. These parameters are derived from the geometric features and physical properties of the corresponding cooling unit.

7. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 6, characterized in that, S42 is achieved through the following steps: S421: Create XML root element and basic information nodes: Create the root element "FMDNA" of the FMDNA format XML file, and add a Logon information node under the root element, including information such as data source, database, login project, username, and password; at the same time, add basic information nodes for Network, Project, EyePointLocation, and ViewPointLocation, and set the corresponding text content; S422: Read CSV data: Use a CSV reading tool to read data from a CSV file, parse each row of data into a dictionary format, and store it in a data list; S423: Construct component relationship mapping: Traverse the read CSV data, construct the mapping relationship between upstream component and current component, and downstream component and current component respectively, based on the upstream component number and downstream component number, and store them in the corresponding mapping dictionary; Bidirectional mapping construction: Through independent upstream / downstream component mapping tables, clearly establish the connection relationship between components to form a directed graph structure. Each node can quickly find upstream and downstream connected objects through the mapping table; S424: Generate node mapping relationship: Based on the upstream component mapping relationship, generate a unique node number for each different upstream component, and establish a mapping relationship between component number and node number; Based on the downstream component mapping relationship, and combined with the already generated upstream node mapping relationship, establish the mapping relationship between downstream components and node numbers; Node coordinates directly inherit the coordinates of the associated components, ensuring that the geometric position is consistent with the topological relationship; S425: Calculate component coordinates: Traverse the information of each component in the CSV data, calculate the X, Y, and Z coordinates of the component according to the component type, channel number, and channel column number, and store the coordinate information in the coordinate mapping dictionary; the coordinate calculation results of all components are uniformly stored in coordinate_map, and nodes and links directly reference this mapping table to ensure the geometric consistency of the entire network layout. Special components are arranged vertically with upstream components through the Y coordinate incremental mapping table. S426: Create component and pipe nodes: Iterate through the CSV data again and determine whether to create a component node or a pipe node based on the component type; for pipes, create an "FMPipe" node under the "Pipes" node and set its name, network number, start coordinates, and end coordinates; for other components, create an "FMComponent" node under the "Components" node and set its name, network number, and position coordinates. S427: Create Nodes and Link Nodes: Based on the node mapping relationship, create an "FMNode" node under the "Nodes" node, and set the node name, network number, and location coordinates; based on the component and node mapping relationship, create upstream and downstream "FMLink" nodes under the "Links" node respectively, and set the link type, component, starting arm, node, and whether it is orthogonal; dynamically set the FromArm attribute according to whether it is an endpoint component, uniformly set the LinkType, and link directly associates the component ID and node number to form a complete topology unit of "component-link-node"; S428: Generate a formatted XML file: Convert the constructed XML tree structure into a string, use an XML parsing tool to parse the string into a DOM object, format the DOM object, remove the XML declaration line, and finally write the formatted XML content into the specified XML file.

8. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 1, characterized in that, In S5, constructing a one-dimensional network model for the internal cooling of gas turbine blades in the gas turbine integrated design platform includes the following steps: S51: Based on the geometric parameters and physical properties of each cooling unit, and combined with the corresponding thermodynamic and fluid dynamic models, automatically calculate the parameter values ​​of each node and connection path in the one-dimensional network. S52: For the heat transfer parameters involved, the corresponding heat transfer correlation is used for calculation, and the calculation results are assigned to the corresponding one-dimensional network nodes or connection paths. S53: All calculated parameter values ​​are automatically stored in the data structure of the one-dimensional network model, forming a complete one-dimensional network model together with nodes and connections.

9. The method for automatically constructing a one-dimensional network for internal cooling of gas turbine blades according to claim 1, characterized in that, S6 includes the following steps: S61: The automatically generated one-dimensional network model is compared and verified with the detailed three-dimensional geometric model of the blade. The internal cooling process of the blade is numerically simulated using computational fluid dynamics simulation to obtain detailed temperature distribution, pressure distribution and flow field information. These simulation results are compared and analyzed with the prediction results of the one-dimensional network model to evaluate the accuracy of the one-dimensional network model. S62: Based on the results of the comparative verification, the one-dimensional network model is optimized and adjusted. The optimization methods include refining or simplifying the division of cooling units, recalculating and calibrating the network parameters, and correcting the node positions and connection relationships; repeated verification and optimization iterations are carried out.

Images