A method and system for generating expansion section of connecting rod

Abstract

The present invention discloses a method and system for generating a connecting rod expansion profile. The method comprises: classifying historical expansion-fractured connecting rods based on the inner diameter of the connecting rod's big-end hole, the connecting rod's thickness, and the expansion-section area, and establishing a classification profile database containing material grades and heat treatment parameters; performing fracture mechanics analysis on the classification profile database, using a theoretical expansion profile as a reference plane, and employing a sphere envelope algorithm to fit the actual profile, outputting a set of fracture characteristic parameters; invoking the fracture characteristic parameter set and performing a dynamic fracture process simulation in combination with the material's fracture toughness value, generating an initial expansion-fracture profile surface containing randomly distributed protrusions and pits; and performing finite element contact analysis on the initial expansion-fracture profile surface, outputting an optimized expansion-fracture profile surface. Using embodiments of the present invention, accurate modeling and optimization of the expansion-section profile morphology can be achieved, ensuring the mechanical properties of the expansion section and improving the overall reliability of the connecting rod.

Description

Technical Field

[0001] The invention belongs to the technical field of connecting rods for internal combustion engines, and in particular to a method and system for generating an expansion cross-section of a connecting rod. Background Art

[0002] In modern internal combustion engines, the connecting rod is a key component connecting the piston and the crankshaft. Its design and manufacturing quality directly affect the performance, reliability and service life of the engine. The traditional connecting rod expansion and breaking process is usually based on empirical design, which makes it difficult to accurately control the expansion surface morphology, resulting in uneven contact stress distribution on the big head hole mating surface, affecting assembly accuracy and fatigue life. In addition, the existing methods lack a systematic analysis of the fracture mechanics properties and cannot effectively optimize the microstructure of the expansion surface, making it difficult for the stiffness and strength of the connecting rod big head hole to meet the reliability requirements under high load conditions. As the engine develops towards high power density, higher requirements are placed on the manufacturing accuracy and mechanical properties of the connecting rod. The randomness of the traditional expansion and breaking process may lead to local stress concentration and reduce the overall load-bearing capacity of the connecting rod. This is a problem that needs to be solved urgently. Summary of the Invention

[0003] The purpose of the present invention is to provide a connecting rod expansion section generation method and system to address the deficiencies in the prior art, enable accurate modeling and optimization of the expansion section morphology, ensure the mechanical properties of the expansion section, and improve the overall reliability of the connecting rod.

[0004] One embodiment of the present application provides a method for generating an expansion cross-section of a connecting rod, the method comprising:

[0005] Historically broken connecting rods were classified based on the inner diameter of the connecting rod's big end hole, connecting rod thickness, and expanded cross-sectional area. 3D scanning was used to obtain actual contour point cloud data of the expanded cross-sections of various connecting rods, and a classification contour database containing material grades and heat treatment parameters was established.

[0006] Performing fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set;

[0007] The fracture characteristic parameter set is called, and a dynamic fracture process simulation is performed in combination with the fracture toughness value of the material to generate an initial fracture contour surface including randomly distributed protrusions / pits;

[0008] Finite element contact analysis is performed on the initial fracture profile surface. With the bolt preload as the boundary condition, the geometric parameters of the protrusions / pits are iteratively adjusted until the contact stress uniformity reaches a preset threshold, and the optimized fracture profile surface is output.

[0009] Optionally, the historical broken connecting rods are classified according to the inner diameter of the connecting rod big end hole, the connecting rod thickness and the expanded cross-sectional area, and the actual contour point cloud data of the expanded cross-sections of various connecting rods are obtained by 3D scanning to establish a classification contour database including material grades and heat treatment parameters, including:

[0010] A laser interferometer 3D scanner is used to scan the cross section of the historically broken connecting rod with sub-micron accuracy, and outputs raw point cloud data containing surface topography details.

[0011] Based on the process parameter combination of the inner diameter, thickness and expanded cross-sectional area of ​​the connecting rod's big end hole, a three-dimensional parameter space grid is constructed, the original point cloud data is mapped to the corresponding grid cells, and a parameterized classified point cloud cluster is output;

[0012] Add material grade and heat treatment process labels to each parameterized point cloud cluster, align them to a unified coordinate system through a point cloud registration algorithm, and output a standardized cross-section contour point cloud set;

[0013] The standardized cross-section contour point cloud set is imported into the NoSQL database, and a classified contour database is established with process parameters as index keys and point cloud data and material properties as values.

[0014] Optionally, the classification profile database is subjected to fracture mechanics analysis, with the theoretical expansion fracture surface as the reference plane, and the actual profile is fitted using a sphere envelope algorithm, the number of protrusions / pits, center coordinates, and sphere radius distribution are counted, and a fracture feature parameter set is output, including:

[0015] Read the cross-section point cloud corresponding to the target process parameters from the classification profile database, fit the theoretical expansion reference plane using the least squares method, and output the normal deviation field from the point cloud to the reference plane;

[0016] Perform multi-scale curvature analysis on the normal deviation field, identify local extreme points as concave-convex feature seed points, and output a set of candidate concave-convex region coordinates;

[0017] Taking the candidate concave-convex area as the center, the adaptive radius sphere envelope algorithm is used for surface fitting, and the envelope sphere radius and the spatial coordinates of the sphere center of each concave-convex area are output;

[0018] The gamma distribution parameters of the envelope sphere radius are statistically analyzed, and combined with the spatial autocorrelation characteristics of the sphere center coordinates, a spatial distribution topology map of the concave and convex features is generated;

[0019] The gamma distribution parameters are integrated with the spatial topology map to construct a fracture feature parameter set including the number, location, and size statistics of asperities.

[0020] Optionally, calling the fracture characteristic parameter set and performing a dynamic fracture process simulation in combination with the fracture toughness value of the material to generate an initial fracture contour surface containing randomly distributed protrusions / pits includes:

[0021] Analyze the gamma distribution parameters in the fracture characteristic parameter set, generate a random concave-convex radius sequence that conforms to actual statistical laws, and output an initial concave-convex size sample pool;

[0022] Based on the initial concave-convex size sample pool and the coordinate association rules of the spatial topology graph, random concave-convex is spread on the theoretical expansion section to ensure that the spatial distribution of concave-convex conforms to the historical statistical characteristics, and an initial geometric model with random concave-convex is output;

[0023] Load the material fracture toughness value into the explicit dynamics solver, control the timing of bump generation in the initial geometric model with the crack growth rate, perform high strain rate fracture process simulation, and output the dynamic crack path;

[0024] The cross-sectional morphology of the crack in the stable propagation stage is extracted to generate an initial fracture contour surface containing randomly distributed protrusions / pits.

[0025] Optionally, performing finite element contact analysis on the initial fracture profile surface, taking bolt preload as a boundary condition, iteratively adjusting geometric parameters of the protrusions / pits until the contact stress uniformity reaches a preset threshold, and outputting an optimized fracture profile surface, includes:

[0026] Apply the bolt preload boundary condition on the initial expansion and fracture contour surface, calculate the local contact pressure field using the Hertz contact theory, and output the initial contact stress distribution cloud map;

[0027] Calculate the stress uniformity index based on the current contact stress distribution cloud map. If the stress uniformity is lower than the preset threshold, locate the stress concentration area and output the high stress area coordinate set and stress gradient vector.

[0028] Based on the high stress area coordinate set and stress gradient vector, the concave and convex geometric parameters of the corresponding position are dynamically adjusted, and the expanded fracture contour surface with updated parameters is output;

[0029] The finite element contact analysis is performed again on the expanded fracture contour surface after the parameter update, a new contact stress contour is generated, and the step of calculating the stress uniformity index based on the current contact stress distribution contour is returned to execute until the stress uniformity reaches the preset threshold, and finally the optimized expanded fracture contour surface is output.

[0030] Another embodiment of the present application provides a connecting rod expansion cross-section generation system, the system comprising:

[0031] The acquisition module is used to classify historical broken connecting rods based on the inner diameter of the connecting rod big end hole, the connecting rod thickness, and the expanded cross-sectional area. The actual contour point cloud data of the expanded cross-sections of various connecting rods is obtained through 3D scanning, and a classification contour database containing material grades and heat treatment parameters is established.

[0032] An analysis module is used to perform fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set;

[0033] A generation module is used to call the fracture characteristic parameter set, perform dynamic fracture process simulation in combination with the fracture toughness value of the material, and generate an initial fracture contour surface containing randomly distributed protrusions / pits;

[0034] The optimization module is used to perform finite element contact analysis on the initial fracture profile surface, use the bolt preload as the boundary condition, iteratively adjust the geometric parameters of the protrusion / pit until the contact stress uniformity reaches a preset threshold, and output the optimized fracture profile surface.

[0035] Yet another embodiment of the present application provides a storage medium, wherein the storage medium stores a computer program, wherein the computer program is configured to execute any of the above methods when run.

[0036] Yet another embodiment of the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any of the above methods.

[0037] Compared with the prior art, the present invention provides a method for generating a connecting rod expansion section, which classifies historical expanded and broken connecting rods according to the inner diameter of the connecting rod big end hole, the connecting rod thickness and the expanded section area, and establishes a classification contour database containing material brand and heat treatment parameters; performs fracture mechanics analysis on the classification contour database, takes the theoretical expanded section as the reference plane, adopts the sphere envelope algorithm to fit the actual contour, and outputs a fracture feature parameter set; calls the fracture feature parameter set, combines the material fracture toughness value to perform dynamic fracture process simulation, and generates an initial expanded fracture contour surface containing randomly distributed protrusions / pits; performs finite element contact analysis on the initial expanded fracture contour surface, and outputs an optimized expanded fracture contour surface, thereby enabling accurate modeling and optimization of the expanded section morphology, ensuring the mechanical properties of the expanded section, and improving the overall reliability of the connecting rod. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A hardware structure block diagram of a computer terminal for a method for generating a connecting rod expansion cross section provided by an embodiment of the present invention;

[0039] Figure 2 A schematic flow chart of a method for generating an expansion cross section of a connecting rod provided in an embodiment of the present invention;

[0040] Figure 3 A front view of an expansion-breaking connecting rod provided by an embodiment of the present invention;

[0041] Figure 4 A schematic diagram of an expanded cross section provided by an embodiment of the present invention;

[0042] Figure 5 A magnified scan of an actual expansion section provided by an embodiment of the present invention;

[0043] Figure 6 A schematic structural diagram of a connecting rod expansion cross-section generation system provided in an embodiment of the present invention.

[0044] Description of reference numerals:

[0045] 1. Small end of connecting rod; 2. Connecting rod body; 3. Expansion section; 4. Big end of connecting rod; 5. Connecting rod bolt; 6. Raised ball; 7. Concave ball. DETAILED DESCRIPTION

[0046] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limiting the present invention.

[0047] The embodiment of the present invention first provides a method for generating an expansion cross-section of a connecting rod. The method can be applied to electronic devices, such as computer terminals, specifically ordinary computers, etc.

[0048] The following describes it in detail by taking running on a computer terminal as an example. Figure 1 The hardware structure block diagram of a computer terminal for a connecting rod expansion section generation method provided by an embodiment of the present invention. Figure 1 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.

[0049] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the methods for generating the expansion cross-section of the connecting rod.

[0050] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0051] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any method for generating the expansion cross section of the connecting rod.

[0052] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 1The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0053] It should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0054] See also Figure 2-Figure 5 The connecting rod is composed of a small connecting rod head 1, a connecting rod body 2, a big connecting rod head 4, and a connecting rod bolt 5. The big connecting rod head is broken at position 3. A broken surface is as follows Figure 5 An embodiment of the present invention provides a method for generating a connecting rod expansion cross section, which may include the following steps:

[0055] S201, classifying historically broken connecting rods based on the inner diameter of the connecting rod big end hole, connecting rod thickness, and expanded cross-sectional area, obtaining actual contour point cloud data of the expanded cross-sections of various connecting rods through 3D scanning, and establishing a classified contour database containing material grades and heat treatment parameters;

[0056] Specifically, a laser interferometer 3D scanner can be used to scan the cross section of the historically broken connecting rod with submicron precision, and output original point cloud data containing surface morphology details;

[0057] This step requires a high-precision laser interferometric 3D scanner. The core principle of this device is laser interferometry: a laser beam (typically with a wavelength of 632.8 nanometers, corresponding to a helium-neon laser) is emitted by the instrument and irradiated onto the expanded surface of the connecting rod. The reflected light interferes with the reference light at the detector. When the surface height of the cross section varies, the optical path difference (OPD) of the reflected light changes accordingly, causing the interference fringes to shift. A highly sensitive charge-coupled device (CCD, a type of image sensor) captures these fringe displacements, and a phase unwrapping algorithm is used to accurately calculate the three-dimensional coordinates (X, Y, Z) of each point on the cross section relative to the scanner's reference. To achieve submicron accuracy (typically defined as measurement accuracy better than 1 micron, for example, 0.5 micron), the scanner is equipped with a precision temperature control system (TCS) to eliminate the effects of environmental thermal deformation and is mounted on an active vibration isolation platform (AVIP) to isolate ground vibrations. Before scanning, the connecting rod fracture surface is cleaned and dried to remove oil and debris, and a thin layer of matte developer (such as magnesium oxide powder) is sprayed to enhance laser scattering, ensuring effective detection of deep holes and steep edges. The scanning process uses a helical path to cover the entire cross-section, with a point spacing (PS) set to 10 microns to ensure that all surface morphological details, including microcracks, dimples, and cleavage steps, are captured.

[0058] During the scanning process, the system performs real-time data verification. The laser interferometer's built-in Optical Calibration Module (OCM) regularly (e.g., every 5 minutes of scanning) performs in-situ calibration using a standard gauge block (GB, with an accuracy of 0.1 micron) to compensate for laser wavelength drift and optical system errors. Simultaneously, the Environmental Sensor (ES) continuously monitors and records temperature (T), humidity (H), and pressure (P) data. These data are input into the Environmental Compensation Model (ECM) to correct the Raw Measurement Value (RMV) in real time. After scanning, the system performs preliminary point cloud processing: applying the Outlier Removal Algorithm (ORA) to remove noise points (NP) caused by dust or stray light; and performing a Data Integrity Check (DIC) to ensure that no areas were missed. The final output raw point cloud data (RPCD) is a collection of millions to tens of millions of three-dimensional coordinate points (X_i, Y_i, Z_i), each of which is accompanied by reflectivity intensity information (RI). This dataset fully records the microscopic geometric characteristics of the cross-section and is stored in an open format (such as ASCII XYZ format or PLY format) for subsequent processing.

[0059] To ensure data reliability and traceability, a metadata (MD) log is generated for each scan. This log includes information such as the scanner serial number (SSN), scan timestamp (TS), environmental parameters (T, H, P), calibration record (CR), total point count (TPC), and bounding box dimensions (BBD). This log is stored in conjunction with the original point cloud data (RPCD) to form a complete scan archive. This step forms the foundation for database construction, and the resulting submicron-precise point cloud provides high-fidelity geometric input for subsequent fracture feature extraction and simulation.

[0060] Based on the process parameter combination of the inner diameter, thickness and expanded cross-sectional area of ​​the connecting rod's big end hole, a three-dimensional parameter space grid is constructed, the original point cloud data is mapped to the corresponding grid cells, and a parameterized classified point cloud cluster is output;

[0061] Process Parameter Combination (PPC) is the core basis for classification, including:

[0062] Big End Bore Diameter (BEBD): Take multiple measurements at a specified location on the connecting rod (usually the parting surface) and take the average value with an accuracy of 0.01 mm.

[0063] Connecting Rod Thickness (CRT): Measured in a designated area in the middle of the shaft with an accuracy of 0.01 mm.

[0064] Fracture Surface Area (FSA): obtained by calculating the projected area of ​​the original point cloud data (RPCD) on the theoretical fracture plane, with an accuracy of 0.1 square millimeter.

[0065] The system constructs a three-dimensional parameter space (3DPS) based on historical data ranges (e.g., BEBD: 50.00-80.00 mm, CRT: 5.00-15.00 mm, FSA: 200.0-600.0 mm²). This space is divided into discrete grid cells (GC). Granularity (G) is a key parameter, balancing classification accuracy and computational efficiency. For example, BEBD is binned into 0.5 mm, CRT into 0.2 mm, and FSA into 10 mm². Each grid cell is uniquely identified by its center coordinates (BEBD_c, CRT_c, FSA_c), representing a specific range of process parameter combinations (e.g., BEBD: 60.25-60.75 mm, CRT: 8.9-9.1 mm, FSA: 350-360 mm²).

[0066] The mapping process categorizes the scanned raw point cloud data (RPCD) into corresponding grid cells (GC). The system reads each connecting rod's measured process parameters (measured BEBD, M_BEBD; measured CRT, M_CRT; calculated FSA, C_FSA) and calculates the Euclidean distance (ED) to all grid cell centers, or uses nearest neighbor matching (NNM) to find the target grid cell (TGC) with the minimum distance. The connecting rod's raw point cloud data (RPCD) and its metadata (MD) are then associated with the TGC. To improve efficiency, the mapping process utilizes a spatial index acceleration algorithm (such as KD-Tree). The system also maintains a mapping relation table (MRT), which records the connecting rod ID, the grid cell ID to which it belongs, and the process parameter values.

[0067] After mapping is complete, the raw point cloud data (RPCD) of all connecting rods belonging to the same grid cell (GC) are grouped together to form a parameterized classified point cloud cluster (PCPCC). Each PCPCC represents a collection of connecting rod sections that fractured within a specific process parameter combination range (defined by that grid cell). When the system outputs these PCPCCs, it appends the grid cell's parameter range label (PRL). This step enables the automatic classification and organization of historically fractured connecting rods based on key process parameters, laying a structured data foundation for the subsequent extraction of common fracture features.

[0068] Add material grade and heat treatment process labels to each parameterized point cloud cluster, align them to a unified coordinate system through a point cloud registration algorithm, and output a standardized cross-section contour point cloud set;

[0069] Each connecting rod not only has geometric parameters, but its material properties and heat treatment history also have a decisive influence on the fracture morphology. Therefore, a material grade and heat treatment process label (MGHTL) must be attached to each connecting rod point cloud in each parameterized point cloud cluster (PCPCC). The material grade (e.g., C70S6, 42CrMo4) is extracted from the production batch records. The heat treatment process label contains key parameters: quenching temperature (QT), tempering temperature (TT), holding time (HT), and cooling medium (CM) (e.g., oil quenched (OQ) or water-based polymer (WBP)). This information is linked to the corresponding point cloud data file using structured fields (e.g., XML format). The labeling process ensures data consistency by establishing a link between the production database and the point cloud database using a connecting rod unique identifier (URI).

[0070] Because the connecting rods are positioned in different positions during scanning, each point cloud is in a different local coordinate system (LCS). For effective comparison and analysis, they must be aligned to the unified coordinate system (UCS) using a point cloud registration algorithm. The Iterative Closest Point (ICP) algorithm is used as the core registration method. The specific process is as follows:

[0071] Select a reference point cloud: In each parameterized point cloud cluster (PCPCC), select a typical link point cloud with high point cloud quality (such as low noise and complete coverage) as the reference point cloud (RPC).

[0072] Coarse registration: For other point clouds in the cluster (called floating point clouds, FPCs), the initial rotation-translation matrix (IRTM) is estimated through principal component analysis (PCA) based on the normal vector of its theoretical expansion plane and the center of the connecting rod big head hole, so that the FPC is roughly aligned with the RPC.

[0073] Fine registration: Apply ICP algorithm for iterative optimization:

[0074] Find corresponding point pairs: Find the nearest neighbor point (NNP) for each point in RPC on FPC.

[0075] Compute the transformation: Based on the found point pairs, calculate the optimal rigid body transformation (rotation matrix R and translation vector T) in the least squares sense to minimize the root mean square error (RMSE) between the corresponding point pairs.

[0076] Apply Transformation: Apply the calculated R and T to the FPC.

[0077] Iteration: Repeat the above steps until the change in RMSE is less than a preset threshold (such as 0.1 microns) or the maximum number of iterations (such as 100) is reached.

[0078] The registration accuracy was evaluated by calculating the final RMSE, which was ensured to be at the submicron level (e.g., <1 μm).

[0079] After registration, the point clouds of all links in the cluster are transformed into a unified coordinate system (UCS) defined by the location of the reference point cloud (RPC). Standardization is then performed: the registered point clouds are resampled to ensure that all point clouds have the same point density (e.g., 100 points per square millimeter). Portions outside the common bounding box (CBB) are cropped to ensure that all point clouds cover the same cross-sectional area. The resulting output, a Standardized Fracture Surface Profile Point Cloud Set (SFSPPCS), contains point cloud data in the same UCS, with the same point cloud density and coverage, and additional material grade and heat treatment process label (MGHTL).

[0080] This standardized point cloud set eliminates the effects of pose differences and sampling inconsistencies, allowing direct quantitative comparison and statistical analysis of the cross-sectional morphologies of different connecting rods.

[0081] The standardized cross-section contour point cloud set is imported into the NoSQL database, and a classified contour database is established with process parameters as index keys and point cloud data and material properties as values.

[0082] Standardized Section Profile Point Cloud Sets (SFSPPCS) are large in data volume (a single point cloud can reach GB levels) and complex in structure (including coordinates, intensities, labels, and more), making them difficult for traditional relational databases to efficiently process. Therefore, a NoSQL database (not only SQL database), particularly a document-oriented database (such as MongoDB) or a key-value database (such as Redis), is used for storage. NoSQL databases offer schema-less flexibility, good horizontal scalability, and are suitable for storing unstructured or semi-structured data (such as point clouds). Database instances are deployed on high-performance servers or distributed clusters, equipped with large-capacity solid-state drive (SSD) arrays to meet high-speed read and write requirements.

[0083] The database's index key (PPC) is designed as a process parameter combination (PPC), specifically formatted as a string that uniquely identifies the corresponding grid cell. For example, the index key (IK) might be designed as: BEBD_60.5_CRT_9.0_FSA_355. This key precisely corresponds to the center parameter value (or range identifier) ​​of a specific grid cell (GC) in the previously constructed three-dimensional parameter space grid, representing a specific combined range of connecting rod big end bore inner diameter, connecting rod thickness, and expanded cross-sectional area. This index key ensures fast data retrieval, allowing users or applications to quickly locate the corresponding data entry by specifying a specific process parameter value or range.

[0084] The value associated with an index key (IK) is a structured document or object consisting of two core parts:

[0085] Point Cloud Data (PCD): stores the point clouds of all connecting rods belonging to the process parameter grid cells in the standardized point cloud set (SFSPPCS). The storage method can be:

[0086] Directly stores a file path (FP) pointing to a point cloud file (such as PLY or LAS format) stored in a distributed file system (such as HDFS) or object storage (such as Amazon S3).

[0087] Alternatively, the point cloud data can be serialized (e.g., using Protobuf or MessagePack format) and stored in the database as a binary large object (BLOB).

[0088] Material Properties (MP): Stores the material grade and heat treatment process label (MGHTL) for all connecting rods within the grid cell. Usually stored as structured data, such as a JSON (JavaScript Object Notation) document:

[0089] {

[0090] "rods": [

[0091] {

[0092] "rod_id": "URI_001",

[0093] "material_grade": "C70S6",

[0094] "heat_treatment": {

[0095] "quenching_temp_c": 880,

[0096] "tempering_temp_c": 600,

[0097] "holding_time_min": 30,

[0098] "cooling_medium": "OQ"

[0099] }

[0100] },

[0101] {

[0102] "rod_id": "URI_002",

[0103] }...

[0104] } ]

[0106] }.

[0107] In addition, the value object can also contain metadata, such as the total rod count (TRC), average point cloud size (APCS), and creation timestamp (CTS) of the grid cell. The resulting Classified Profile Database (CPD) is an efficient and scalable data warehouse that supports rapid query and retrieval of connecting rod cross-section point clouds with similar fracture characteristics and their material heat treatment information based on process parameters, providing solid data support for subsequent fracture mechanics analysis, feature extraction, and profile generation. The database provides standard interfaces (such as RESTful APIs) for upper-level applications to call.

[0108] The method first scientifically classifies historical fractured connecting rods based on key process parameters, acquires cross-sectional micromorphology data through high-precision 3D scanning technology, and constructs a database containing material properties and process parameters. The classification dimensions select the inner diameter of the large head hole, thickness, and cross-sectional area, three parameters that directly affect the fracture quality, to ensure the engineering relevance of the data classification. The establishment of a classification profile database provides a structured data foundation for subsequent analysis, converting discrete case experience into quantifiable process knowledge. By correlating material grades and heat treatment parameters, the cross-analysis capability of fracture morphology and material properties is achieved, laying a data foundation for intelligent process optimization.

[0109] S202, performing fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set;

[0110] Specifically, the cross-section point cloud corresponding to the target process parameters can be read from the classification profile database, and the theoretical expansion reference plane can be fitted by the least squares method to output the normal deviation field from the point cloud to the reference plane.

[0111] The system first calculates the target process parameter combination specified by the user (for example, the inner diameter of the connecting rod big end hole ID=60mm, thickness THK=12mm, and the expansion section area A=450mm). 2), and then searches the established Classification Contour Database (CCDB). This database utilizes a NoSQL architecture (non-relational database) and uses process parameters as index keys (IK), enabling rapid location of matching Standardized Contour Point Cloud Sets (SCPCS). The retrieved SCPCS contains hundreds of thousands to millions of three-dimensional (X, Y, Z) coordinates, acquired through high-precision 3D scanning, with a density of up to 500 points per square millimeter. After reading the point cloud, the system performs a least squares method (LSM) fitting operation: using the point cloud's Z coordinate (perpendicular to the cross section) as the dependent variable and the X and Y coordinates as the independent variables, the plane equation Z=aX+bY+c is constructed. The algorithm determines the optimal plane coefficients a (slope in the X direction), b (slope in the Y direction), and c (plane intercept) by minimizing the sum of squared residuals (SSR) of the perpendicular distances from all points to the plane. This plane is the theoretical fracture reference plane (TFRP), which represents the completely flat fracture morphology under ideal conditions.

[0112] After the reference plane is determined, the system calculates the normal distance (ND) from each scan point in the point cloud to the TFRP point by point. The normal distance is defined as the difference between the point cloud coordinates and the projection of the reference plane along its normal direction (defined by the plane normal vector N=(a,b,-1)). When calculating, the shortest distance ND from the point to the plane is first solved by vector operations. i , where the subscript i represents the i-th point. The ND values ​​of all points form a normal deviation field (NDF) that corresponds one-to-one with the original point cloud. This field data is stored in matrix form, with each element containing the position coordinates (X_i, Y_i) and its corresponding deviation value ND_i (unit: micrometer μm). A positive value indicates that the point is above the reference plane (convex), while a negative value indicates that it is below (concave). Precision control in this step is crucial, and the fitting residual (SSR) must be less than 0.5μm to ensure the reliability of subsequent concave-convex feature analysis.

[0113] To optimize computational efficiency, the system employs a spatial grid acceleration strategy: the XY plane is divided into 1mm×1mm grid cells (GC). The ND values ​​of points within each cell are interpolated using bilinear interpolation (BI) to generate a continuous field. Simultaneously, the system calculates global deviation statistics: mean deviation μ_ND (reflecting overall flatness), standard deviation σ_ND (indicating fluctuations), and skewness skew_ND (indicating concave and convex asymmetry). These statistics serve as quality control indicators. If σ_ND exceeds the 95% confidence interval (CI) of historical data, a data review process is triggered. The final output NDF is not only a point set, but also structured field data containing spatial distribution, statistical characteristics, and quality control labels, providing input for subsequent curvature analysis.

[0114] Perform multi-scale curvature analysis on the normal deviation field, identify local extreme points as concave-convex feature seed points, and output a set of candidate concave-convex region coordinates;

[0115] The normal deviation field (NDF) is essentially a height field representation of a three-dimensional surface. The system uses multi-scale curvature analysis (MCA) to detect its local geometric features. The core is to calculate the Gaussian curvature (K) and mean curvature (H) of each point at different spatial scales. The curvature calculation is based on the moving least squares (MLS) method: with the current point P as the center, a neighborhood sphere (NS) with a radius of R is defined, and a local quadratic surface Z=c1X is fitted within NS. 2 +c2XY+c3Y 2 +c4X+c5Y+c6, and then derive the curvature tensor. Multi-scale is reflected in the gradient change of R: R_min=0.1mm (capturing micron-level defects), R_max=2.0mm (recognizing millimeter-level features), step size ΔR=0.2mm, a total of 10 scale levels.

[0116] At each scale R, the system scans all points in the NDF and marks candidate points that meet the extreme value conditions:

[0117] Local Maximum Point (LMP): H < 0 and K > 0 (convex peak);

[0118] Local Minimum Point (LmP): H>0 and K>0 (valley).

[0119] The extreme value determination must meet two conditions simultaneously: ① The curvature value ranks in the top 5% (Percentile Rank, PR) within the NS; ② The curvature difference with the adjacent point exceeds the threshold ΔK_min=0.01μm⁻ 1 To prevent noise interference, the system performs non-maximum suppression (NMS): if the distance between two points is less than R / 2 and their curvature is similar (|K1-K2| < 0.05K_max), only the one with greater curvature is retained. A multi-scale fusion strategy is used to determine the seed points: if a point is detected as an extreme value at ≥3 consecutive scales, it is included in the Candidate Feature Coordinate Set (CFCS), which records the (X, Y) position of each seed point and its dominant scale R_dom (the scale with the strongest curvature response).

[0120] To verify the validity of the seed points, the system adds an energy gradient criterion: the Strain Energy Density Gradient (SEDG) of each seed point is calculated. The deformation energy density is approximately SED∝ND 2 / R, the gradient modulus |∇SED| must be greater than the material-dependent threshold (e.g., the steel connecting rod threshold = 1.5 J / m 3 / mm). Points that do not meet the standard are considered as pseudo features and are eliminated. The final output CFCS contains two types of data:

[0121] The Core Seed List (CSL) records the center coordinates (X,Y) of the multiple-filtered bumps and asperities. The Feature Scale Map (FSM) stores the R_dom and maximum |K| value for each seed point. This step ensures that only the macroscopic and microscopic bump features that significantly influence fracture behavior are captured.

[0122] Taking the candidate concave-convex area as the center, the adaptive radius sphere envelope algorithm is used for surface fitting, and the envelope sphere radius and the spatial coordinates of the sphere center of each concave-convex area are output;

[0123] For each seed point in the candidate concave-convex region coordinate set (CFCS), the system initializes an envelope sphere (ES) centered at its location (X_s, Y_s). The initial value of the sphere's radius, R_es, is set to the point's dominant scale, R_dom (derived from the FSM table). The goal is to optimally cover the actual surface of the concave-convex region with a single sphere. The core of the algorithm is Adaptive Radius Adjustment (ARA): Through iterative optimization, the sphere is guaranteed to contain a specified proportion of neighboring points (target coverage ratio, TCR, = 99%) while minimizing the distance to points outside the sphere.

[0124] The specific iteration process is as follows:

[0125] Neighborhood point extraction: With the current sphere center C=(X_s, Y_s, Z_c) as the center and R_es as the radius, extract all points in the original point cloud that fall within this sphere to form a subset S_sub.

[0126] Sphere Fitting: Perform Least Squares Sphere Fitting (LSSF) on S_sub. The objective function is to minimize the sum of squared distances from the point to the sphere, and then solve for the sphere's center coordinates O = (X_0, Y_0, Z_0) and radius R.

[0127] Containment ratio check: Calculate the inclusion ratio CR (Containment Ratio) of the new sphere to the original neighborhood points (the range is extended to 1.5×R_dom).

[0128] Radius adjustment: If CR < TCR, increase R_es = R_es × (1 + α) (step factor α = 0.05); if CR > TCR, decrease R_es = R_es × (1 - α).

[0129] Convergence judgment: Stop when |CR - TCR| < 0.5% or the number of iterations > 20 times, and output the final sphere parameters.

[0130] After the fit is complete, the system records two key outputs:

[0131] Envelope Sphere Radius (ESR): Unit: mm, represents the size of concave and convex features;

[0132] Sphere Center Coordinate (SCC): (X_0, Y_0, Z_0), where Z_0 is transformed from the reference plane: Z_0 = aX_0 + bY_0 + c + δZ, where δZ is the normal offset from the sphere center to the reference plane.

[0133] A Fitting Quality Report (FQR) is also generated, including: Root Mean Square Error (RMSE) (required to be <3μm); Maximum Residual (MaxRes) (required to be <10μm); and the actual value of the Coverage Rate (CR). Abnormal fitting points (such as RMSE exceeding the standard) will trigger a manual review process to ensure data reliability.

[0134] The gamma distribution parameters of the envelope sphere radius are statistically analyzed, and combined with the spatial autocorrelation characteristics of the sphere center coordinates, a spatial distribution topological map of the concave and convex features is generated. The gamma distribution parameters and the spatial topological map are integrated to construct a fracture feature parameter set containing the statistics of the number, location, and size of the concave and convex.

[0135] After obtaining the enveloping sphere radius (SR_i) of all concave and convex features, the statistical distribution characteristics of these radii need to be analyzed. In engineering, the gamma distribution is often used to describe positively skewed random variables with values ​​greater than zero, such as defect size, crack length, and concave and convex size. The gamma distribution is determined by two parameters: the shape parameter (k > 0) and the scale parameter (θ > 0). The system collects all SR_i values ​​(i = 1 to N) to form a sample dataset. It then uses maximum likelihood estimation (MLE) to fit the gamma distribution. MLE finds the parameter values ​​(k_MLE, θ_MLE) that maximize the probability of observing the current sample data as the fitting parameters of the gamma distribution. The system then uses a statistical computing library (e.g., one based on a numerical optimization algorithm) to complete the fitting and outputs goodness-of-fit metrics (such as the P value from the Kolmogorov-Smirnov test) and the final gamma distribution parameters (k, θ). These parameters (gamma distribution parameters) quantitatively describe the statistical laws of the feature sizes of bumps and concavities on the expanded surface under the process parameters. For example, a small k value and a large θ value indicate a dispersed size distribution and more large bumps and concavities.

[0136] Knowing the size distribution alone is not enough. The spatial distribution pattern of the concave and convex features on the fracture surface (e.g., random, clustered, or uniform) is also crucial, as it influences the fracture path and subsequent assembly contact. To this end, the spatial autocorrelation characteristics of the spherical center coordinates (SCC_i) need to be analyzed. Spatial autocorrelation measures whether the attribute values ​​(primarily existence in this case, but also size) of spatially close features are similar (positively correlated) or dissimilar (negatively correlated). Common methods include calculating Ripley's K-function or Moran's index (I). The system calculates the Euclidean distance (Distance) between each pair of SCC_i and statistically calculates the ratio or difference between the number of observed "point pairs" and the expected number under a completely random spatial distribution (Complete Spatial Randomness, CSR) at different distance thresholds (DistanceThreshold, DT). The spatial distribution pattern of the concave-convex features is determined by analyzing the deviation of the K function curve from the CSR reference line (e.g., above the CSR line indicates clustering, below the CSR line indicates uniformity). Based on this analysis, the system generates a spatial distribution topology map of the concave-convex features. This map can be a two-dimensional or three-dimensional heat map, using color depth to represent the point density (number of concave-convex features per unit area) or average size of the concave-convex features in different regions. It can also be a network graph (if the algorithm detects that the concave-convex features tend to be distributed in a chain-like manner along a specific direction (such as the direction of crack propagation)), or a schematic diagram that indicates the location and range of major clusters. This map intuitively and quantitatively displays the arrangement of the concave-convex features on the cross-section.

[0137] Finally, the system fuses all the above analysis results to construct a comprehensive and structured Fracture Characteristic Parameter Set. This dataset contains quantitative information in multiple dimensions:

[0138] Number of Features (NoF): The total number N of identified bumps and pits.

[0139] Location Information: A list of the center coordinates of all bumps and bumps (Xc_i, Yc_i, Zc_i). This is the most basic location data. More important are key location features extracted from the spatial topology, such as the center coordinates and range of major clusters (cluster center coordinates X_cluster, Y_cluster, cluster radius R_cluster), or spatial location descriptions of high-density and low-density areas.

[0140] Size Statistics: The core of these statistics is the fitted gamma distribution parameters (shape parameter k, scale parameter θ). These statistics also include sample statistics such as the mean radius (MR), standard deviation (SD), maximum radius (MaxR), minimum radius (MinR), and median radius (MedR) of all SR_i values, providing a more intuitive description of the size distribution. The average size of regions reflected in the spatial topology map is also integrated here.

[0141] Spatial Distribution Characteristics: The spatial topology map itself is an important visualization result, and quantitative indicators need to be extracted from it, such as the Cluster Intensity Index (CII), the Uniformity Index (UI), or the Directionality Parameter (DP, if there is obvious directionality).

[0142] This fracture characteristic parameter set is no longer simply raw point cloud data; instead, it is a highly condensed set of quantitative indicators derived through in-depth processing and analysis, capable of characterizing the core features of the microscopic morphology of the connecting rod's expanded cross section under specific process parameters. It accurately describes the presence of irregularities (size distribution), their number (number), and their location (position and spatial distribution), providing key input parameters for subsequent fracture simulation and profile optimization.

[0143] Guided by fracture mechanics theory, the actual cross-section is compared and analyzed with an idealized plane, and the geometric abstraction of sphere envelope is employed to quantify surface concavity and convexity. The algorithm not only calculates macroscopic parameters but also records the distribution patterns of microscopic features, forming a multi-dimensional fracture characterization system. This transforms complex cross-sectional morphology into computable mathematical features, breaking through the limitations of traditional qualitative evaluation. The sphere envelope algorithm preserves the geometric essence of the morphology while enabling quantitative representation of features through parameterization, providing precise input for subsequent simulations.

[0144] S203, calling the fracture characteristic parameter set, combining the material fracture toughness value to perform dynamic fracture process simulation, and generating an initial fracture contour surface including randomly distributed protrusions / pits;

[0145] Specifically, the gamma distribution parameters in the fracture characteristic parameter set can be analyzed to generate a random concave-convex radius sequence that conforms to actual statistical laws, and output an initial concave-convex size sample pool;

[0146] The system first extracts gamma distribution parameters from the set of fracture characteristic parameters. These parameters consist of two key values: shape parameter k (describing the degree of distribution skewness) and scale parameter θ (controlling the distribution range). For example, historical data analysis shows that the radius of a certain type of connecting rod concavity is k = 2.5 and θ = 0.1 mm. Using a Monte Carlo random sampling algorithm, a dedicated mathematical library (such as Boost.Math) is used to generate a sequence of random numbers that conform to this gamma distribution. Each random number represents the theoretical radius value (TRV) of a bump or concavity. The sampling size is typically 1,000-5,000 samples to ensure statistical representativeness. A Box-Muller transform is applied during the generation process to improve sampling efficiency. The final output is an initial concave-convex size sample pool (ICCSSP) containing all TRVs. The sample pool is stored in the form of an array, and each element corresponds to the size data of a concave-convex feature.

[0147] To ensure that the generated random radius complies with actual physical constraints, the system performs a Physical Feasibility Check (PFC) on ICCSP. The verification rules include:

[0148] Radius lower limit constraint: For example, the pit radius must not be less than 0.01 mm (to avoid numerical noise interference);

[0149] Radius upper limit constraint: for example, the protrusion radius must not be greater than 15% of the connecting rod thickness (to prevent structural interference);

[0150] Process matching constraints: For example, the average pit radius of high carbon steel connecting rods is usually 20% lower than that of ductile iron.

[0151] TRVs that fail verification are discarded and supplemented with new samples through Importance Resampling (IR). The final output ICCSP satisfies two characteristics: mathematically conforming to the statistical laws of the gamma distribution and physically conforming to the material processing technology limitations.

[0152] ICCSSP's storage structure utilizes a hierarchical indexing mechanism (HIM). The first layer categorizes the data by bump type (bump / pit), while the second layer buckets the data by radius range (e.g., 0-0.05mm, 0.05-0.1mm, etc.). Each bucket stores the radius value and its probability density weight (PDW), calculated using the gamma distribution probability density function. This structure accelerates size matching queries during the subsequent spatial distribution phase. A Radius Distribution Validation Report (RDVR) is also generated, including the Kolmogorov-Smirnov test results comparing the actual distribution to the theoretical gamma distribution, ensuring no significant difference between the two at a significance level of α = 0.05.

[0153] Based on the initial concave-convex size sample pool and the coordinate association rules of the spatial topology graph, random concave-convex is spread on the theoretical expansion section to ensure that the spatial distribution of concave-convex conforms to the historical statistical characteristics, and an initial geometric model with random concave-convex is output;

[0154] The system reads a spatial topology map (STM), which stores the spatial association rules of concave and convex features in a graph structure. It contains two types of key data:

[0155] Spatial Autocorrelation Matrix (SAM): describes the probability of other concavities and convexities appearing within a radius R around a concavity and convexity (for example, the probability of adjacent concavities and convexities is 70% when R = 1 mm).

[0156] Regional Density Heatmap (RDH): Mesh the expansion section into 0.5mm×0.5mm units and mark the average number of historical bumps in each unit (e.g., central area density = 8 / mm 2 , edge area = 2 / mm 2 ).

[0157] The spreading process first divides the theoretical fracture plane (TFP) into high / low density areas according to RDH, and then determines the initial position candidate points of each bump through the controlled random walk algorithm (CRWA).

[0158] Extract the concave and convex size data from ICCSP in sequence, and perform four-stage matching based on the position distribution rules:

[0159] Stage 1 - Density Matching: Prioritize large-sized bumps (radius > 0.1mm) in high-density areas (such as around connecting rod bolt holes);

[0160] Phase 2 - Spatial Exclusion: An exclusion zone (EZ) with a radius of r (r = 2 × current concave-convex radius) is established with the new scattering point as the center, prohibiting other concave-convex centers from falling into it.

[0161] Stage 3 - Association Strengthening: If SAM shows that the location needs to be "clustered distribution", a secondary bump is forcibly added 0.2mm outside the exclusion zone;

[0162] Stage 4 - Boundary Correction: The radius of bumps and depressions <0.3mm from the edge of the expanded section is automatically reduced to 50% (to avoid edge stress singularity).

[0163] After every 100 bumps are spread, Delaunay triangulation (DT) is performed to check spatial uniformity and to add small bumps to the void areas.

[0164] After spreading is complete, the system maps topological relationships to geometric entities:

[0165] The pit is modeled as a spherical cap surface (SCS), with depth = radius × 0.7 (based on empirical coefficients from historical data);

[0166] The bulge was modeled as a Gaussian Curvature Transition Surface (GCTS), with base diameter = radius × 1.5;

[0167] Curvature continuity constraints (G2 continuity) are imposed on the concave-convex junction area.

[0168] The final output is an Initial Geometric Model with RandomConcave-Convex Features (IGMRC) in STEP or Parasolid format, which can be directly imported into CAE software.

[0169] Load the material fracture toughness value into the explicit dynamics solver, control the timing of bump generation in the initial geometric model with the crack growth rate, perform high strain rate fracture process simulation, and output the dynamic crack path;

[0170] The material fracture toughness value (MFTV) is entered as J-Integral or KIC (plane strain fracture toughness), for example, KIC for 42CrMo steel is 90 MPa·m 1 / 2 . Create a finite element model in an explicit dynamics solver (EDS, such as LS-DYNA or Abaqus / Explicit):

[0171] Meshing: Adaptive meshing (minimum size = 0.05mm) is used in the expansion section area, and the concave and convex features are locally refined to 0.01mm;

[0172] Material model: Define the Johnson-Cook Plasticity Model and the Strain Rate Sensitivity Coefficient C;

[0173] Loading conditions: A radial displacement load (DL) is applied to the connecting rod big end hole at a rate of 10 m / s to simulate hydraulic expansion and fracture impact.

[0174] The crack propagation rate (CPR) is controlled by the Extended Finite Element Method (XFEM):

[0175] The initial crack is set at the preset weak point (such as the bottom of the pit), with a length of 0.1 mm;

[0176] CPR calculation formula: da / dt=C×(K / KIC)^n (da / dt is the crack growth rate, K is the stress intensity factor, and n=4 is the material constant);

[0177] When K>KIC, crack propagation is triggered, and the propagation direction is determined by the Maximum Circumferential Stress Criterion (MCSC).

[0178] Dynamic coupling of bump generation sequence and crack growth:

[0179] When the crack tip is less than 0.5 mm from a concave or convex position, the local mesh refinement of the concave or convex position is activated;

[0180] If the crack passes through a pit, the bottom of the pit is considered as the new crack source; if it encounters a bulge, the expansion path is corrected according to the curvature of the bulge.

[0181] The simulation performs a Multiphysics Coupling Analysis (MCA):

[0182] Mechanical fields: calculation of Mises stress distribution and principal stress directions;

[0183] Damage field: Based on the Accumulated Plastic Strain Criterion (APSC), fracture is determined when the element equivalent plastic strain is greater than 0.8;

[0184] Temperature field: Considering the plastic work-heat conversion (90% converted into heat energy), update the material softening effect.

[0185] The Dynamic Crack Path (DCP) is output in two forms:

[0186] Geometric path: crack tip coordinate sequence (time step 0.1 μs);

[0187] Physical characteristics: path fractal dimension (FD, typical value 1.2-1.5) and local deflection angle histogram.

[0188] The cross-sectional morphology of the crack in the stable propagation stage is extracted to generate an initial fracture contour surface containing randomly distributed protrusions / pits.

[0189] Criteria for determining the stable crack growth stage (SCGS):

[0190] Time domain: excludes the initial cracking transient (the first 5 μs) and the final instability stage (after the load drops to 80% of the peak value);

[0191] Spatial domain: the crack length accounts for 60%-90% of the total length of the expansion section;

[0192] Physical indicators: Crack growth rate fluctuation <±10% (da / dt=2-3mm / ms).

[0193] During this stage, cross-sectional morphology data were collected every 1 μs, and the instantaneous cross-sectional surface was fitted using the Moving Least Squares Method (MLSM).

[0194] Morphological data processing flow:

[0195] Step 1 - Feature separation: Compare the fitted surface with the theoretical plane and extract the area with deviation > 0.005mm as the effective concave and convex;

[0196] Step 2 - Geometry Reconstruction:

[0197] Pits: Reconstructed using spherical harmonic functions (SHF), order l=5;

[0198] Bump: Non-Uniform Rational B-Spline (NURBS) surface fitting, control point spacing = 0.1mm;

[0199] Step 3 - Topology repair: Eliminate false holes (size < 0.02mm) caused by mesh distortion and merge adjacent bumps with an overlap rate > 30%.

[0200] Finally, the Initial Fracture Contour Surface (IFCS) is generated, and its data characteristics include:

[0201] Geometric accuracy: surface continuity G1 (tangent continuity), maximum deviation <0.01mm;

[0202] Statistical characteristics: KS test value of bump number density, size distribution and input gamma distribution parameters D < 0.05;

[0203] Physical verification: Confocal Microscope (CM) was used to compare the 2D Power Spectral Density (PSD) of the historical cross section at a spatial frequency of 1-100 mm⁻ 1The error within the range is <15%. The output format is encrypted point cloud (point spacing 0.02mm) and STL triangulated mesh for subsequent contact analysis.

[0204] Based on the statistical laws of historical data, the fracture process is recreated through numerical simulation. This method innovatively incorporates the material's intrinsic parameter, fracture toughness, into the simulation, ensuring that the generated cross-sectional morphology conforms to both statistical characteristics and the laws of material mechanics. This completes a closed-loop verification process from empirical data to theoretical models, resulting in an initial profile surface that is both engineering-realistic and physically reasonable. The dynamic simulation process reveals the inherent correlation between morphological characteristics and fracture mechanisms, guiding process optimization.

[0205] S204, performing finite element contact analysis on the initial fracture profile surface, using the bolt preload as a boundary condition, iteratively adjusting the geometric parameters of the protrusions / pits until the contact stress uniformity reaches a preset threshold, and outputting an optimized fracture profile surface.

[0206] Specifically, the bolt preload boundary condition can be applied to the initial expansion and fracture contour surface, the local contact pressure field can be calculated using the Hertz contact theory, and the initial contact stress distribution cloud map can be output;

[0207] Boundary condition loading and mechanical model construction

[0208] The initial fracture profile (containing randomly distributed protrusions and depressions) is imported into finite element analysis software (such as ANSYS or ABAQUS). In this model, the two sides of the connecting rod separation surface are defined as contact pairs: one side is the fracture surface with concave and convex features, and the other side is an ideal plane (simulating the connecting rod cap mating surface). The bolt preload force (BPF) is applied as a concentrated force at the center of the bolt hole, and its value is determined according to engine design standards (for example, a certain diesel engine connecting rod model requires a single bolt preload of 25 kN). The preload force is transmitted to the contact surface through the bolt connection (beam element), forming a normal pressure boundary condition. The degrees of freedom (DOF) of the bottom of the connecting rod body are also constrained to simulate its working condition when fixed to the crankshaft. The material properties are assigned elastic modulus (EM=210 GPa) and Poisson's ratio (PR=0.3) based on the material grade (e.g., C70S6 high carbon steel) in the classification profile database.

[0209] Contact pressure field calculation principle

[0210] Hertzian contact theory is used as the basis for local contact pressure calculation. This theory simplifies concave and convex features to the contact between a sphere and a plane: when a protrusion (equivalent to a spherical cap) contacts the mating surface, the contact area is circular or elliptical, with the maximum contact pressure located at the center. Finite element software uses built-in Hertzian contact algorithms (such as the Augmented Lagrange method) to establish nonlinear contact interactions between contact pairs. The solver calculates the penetration depth (PD) and contact reaction force (CRF) at each contact point based on the Newton-Raphson method, ultimately generating a local contact pressure (CP) distribution field. For example, the top of a protrusion with a radius of 0.5 mm may generate a local pressure of up to 1500 MPa, while the pressure in the concave area is close to 0 MPa.

[0211] Initial contact stress contour generation

[0212] After completing the finite element solution, the system extracts pressure data from all nodes on the contact surface and maps it to a three-dimensional contour coordinate system. Using spatial interpolation algorithms (such as the inverse distance weighted method), the discrete nodal pressures are converted into a continuous pressure field and visualized as an initial contact stress distribution cloud map (CSCM). This cloud map is rendered in pseudo-color: red areas represent high-pressure areas (such as the top of a bump), and blue areas represent low-pressure areas (such as the bottom of a pit). The color scale range is set from 0 to 80% of the material's yield point (e.g., 0-1200 MPa). The cloud map also includes a pressure statistics report, including the maximum (MaxCP), minimum (MinCP), average (AvgCP), and standard deviation (StdDevCP), providing basic data for subsequent stress uniformity assessment.

[0213] Calculate the stress uniformity index based on the current contact stress distribution cloud map. If the stress uniformity is lower than the preset threshold, locate the stress concentration area and output the high stress area coordinate set and stress gradient vector.

[0214] Stress uniformity quantification method

[0215] The Stress Uniformity Index (SUI) is defined as the reciprocal of the coefficient of variation of the contact pressure field (Inverse Coefficient of Variation), and its calculation formula is SUI = Avg CP / Std DevCP. The larger this value is, the more uniform the pressure distribution is (the ideal value is infinity). The Preset Threshold (PT) is set according to engineering experience (for example, for a certain gasoline engine connecting rod, SUI ≥ 8.0 is required). The system reads the Avg CP and Std Dev CP data in the initial CSCM and calculates the current SUI value. If SUI ≥ PT (such as SUI = 10.2), the process terminates; if SUI < PT (such as SUI = 5.3), the stress concentration area location process is triggered.

[0216] Stress Concentration Area Identification Technology

[0217] The dual-threshold method is used to locate the High-Stress Zone (HSZ):

[0218] Primary threshold: Screen areas where the pressure exceeds 80% of the allowable stress of the material (such as > 900 MPa) to avoid local plastic deformation.

[0219] Secondary threshold: Screen areas where the pressure is higher than 2.5 times the standard deviation of the average pressure (such as Avg CP + 2.5 × Std DevCP).

[0220] Cluster analysis (Cluster Analysis) is performed on the grid cells that simultaneously meet the dual thresholds, and adjacent cells are merged to form a continuous high-stress zone (HSZ). Each HSZ records its centroid coordinates (Centroid Coordinate, CC) and the maximum pressure value (Peak Pressure, PP). For example, the HSZ coordinates at the top of a certain protrusion are (35.2, 17.8, 0.0) mm, and PP = 1420 MPa.

[0221] Stress Gradient Vector Generation

[0222] For each HSZ, calculate its Stress Gradient Vector (SGV):

[0223] Direction component: Along the path of the fastest pressure drop, calculate the gradient direction through the pressure difference between adjacent cells (such as from the top of the protrusion to the edge).

[0224] Magnitude component: Define the gradient intensity with the pressure change rate within a unit distance (1 mm) (such as 200 MPa / mm).

[0225] The final output is the high stress zone coordinate set (HSZ Coordinate Set, HCS) and the corresponding SGV data table, for example, "HSZ001: CC=(35.2,17.8,0.0), SGV=(−120, −85, 0) MPa / mm".

[0226] Based on the high stress area coordinate set and stress gradient vector, the concave and convex geometric parameters of the corresponding position are dynamically adjusted, and the expanded fracture contour surface with updated parameters is output;

[0227] Geometric parameter adjustment strategy

[0228] Develop targeted modification plans based on HCS and SGV:

[0229] Bump optimization: If the HSZ is located at the top of a bump, reduce the height (HR) or increase the radius (RI) in the SGV direction. For example, change a sharp bump (0.3mm radius) to a spherical cap (0.8mm radius).

[0230] Concave Optimization: If the HSZ is located at the edge of the concave, press SGV to reverse the material filling (MF) or reduce the concave depth (DD). For example, change a V-shaped concave to a U-shaped one.

[0231] The adjustment amplitude (AA) is positively correlated with the degree to which the current SUI deviates from the threshold (e.g., when SUI = 5.3, AA is 150% of the standard adjustment amount).

[0232] Parametric geometry reconstruction

[0233] The concave and convex features in the initial fracture profile are stored parameterized (e.g., convex = sphere center coordinates + radius, concave = center point + depth + opening diameter). The system modifies the corresponding parameters based on the adjustment plan:

[0234] For bumps: New radius*R_new = R_old × (1 + k × |SGV|)*, where k is the gain factor (e.g. k = 0.02 / MPa·mm).

[0235] For pits: new depth D_new = D_old − m × |SGV|, where m is the attenuation coefficient (e.g. m = 0.015 mm 2 / MPa).

[0236] The modified parameter set drives the CAD kernel (such as ACIS or Parasolid) to reconstruct the surface in real time to ensure smooth transition between convex and concave surfaces (G 2continuous).

[0237] Updated contour surface generation and verification

[0238] The reconstructed geometric model is discretized to generate an Updated Fracture Contour Surface (UFCS) with updated parameters. Surface smoothness is verified through a Curvature Consistency Check (CCC). The curvature difference (CD) between the adjusted area and its surroundings is calculated. If the CD exceeds the allowable value (e.g., 0.1 mm), local smoothing (Laplacian smoothing) is initiated until the target is met. Finally, a UFCS model file (e.g., in STEP format) is output that meets geometric quality requirements.

[0239] The finite element contact analysis is performed again on the expanded fracture contour surface after the parameter update, a new contact stress contour is generated, and the step of calculating the stress uniformity index based on the current contact stress distribution contour is returned to execute until the stress uniformity reaches the preset threshold, and finally the optimized expanded fracture contour surface is output.

[0240] Iterative analysis loop construction

[0241] Import the UFCS model back into the finite element environment and repeat the following process:

[0242] The same bolt preload (BPF = 25 kN) and restraint conditions were applied;

[0243] Run the Hertz contact calculation to generate a new contact stress contour map (New CSCM);

[0244] The new SUI value is calculated and compared with the preset threshold (PT=8.0).

[0245] This process is implemented through a loop call using an automated script (e.g., Python encapsulating the ABAQUS command stream), and each iteration takes about 5-10 minutes (depending on the mesh size).

[0246] Convergence conditions and termination mechanism

[0247] The iteration termination condition is SUI_current ≥ PT or reaching the maximum number of iterations (Max Iteration, MI=10). Key data is recorded for each iteration, such as:

[0248] Iterations SUI value Maximum pressure (MPa) Modify the number of features 1 5.3 1420 7 2 6.8 1210 5 3 8.5 980 3

[0252] …

[0253] When the SUI reaches 8.5 (>PT=8.0), the system automatically terminates the cycle.

[0254] Optimization result output and verification

[0255] The final output optimized fracture contour surface (OFCS) meets the following requirements: contact stress uniformity SUI ≥ 8.0 (e.g. 8.5); no overpressure areas (e.g. maximum pressure 980 MPa < material yield strength 1100 MPa); and the number of concave and convex features is reduced by 40% (e.g. from 15 to 9).

[0256] Physical tests have verified that after the bolts of the connecting rod using this profile are pre-tightened, the contact pattern area ratio is increased and the stress unevenness is reduced.

[0257] Using the finite element method to simulate actual assembly conditions, the surface topography is intelligently adjusted with contact stress uniformity as the optimization objective. During the iterative process, the geometric parameters of the concave and convex features are dynamically updated based on the stress distribution, ultimately achieving a cross-sectional topography with optimal mechanical properties. This ensures that the optimized cross-sectional topography not only meets the statistical requirements for topography but also achieves optimal stress distribution during assembly. This approach elevates process design from "similar in form" to "similar in spirit," truly achieving function-oriented intelligent optimization.

[0258] It can be seen that the historical fractured connecting rods are classified according to the inner diameter of the connecting rod big end hole, the connecting rod thickness and the expansion cross-sectional area, and a classification contour database containing material brand and heat treatment parameters is established; the classification contour database is subjected to fracture mechanics analysis, and the theoretical expansion cross-sectional area is used as the reference surface. The sphere envelope algorithm is used to fit the actual contour and output the fracture characteristic parameter set; the fracture characteristic parameter set is called, and the dynamic fracture process simulation is carried out in combination with the fracture toughness value of the material to generate an initial expansion fracture contour surface containing randomly distributed protrusions / pits; the initial expansion fracture contour surface is subjected to finite element contact analysis, and the optimized expansion fracture contour surface is output, so as to achieve accurate modeling and optimization of the expansion cross-sectional morphology, ensure the mechanical properties of the expansion cross-sectional area, and improve the overall reliability of the connecting rod.

[0259] Another embodiment of the present invention provides a connecting rod expansion section generation system, see Figure 6 , the system may include:

[0260] Acquisition module 601 is used to classify historical broken connecting rods based on the inner diameter of the connecting rod big end hole, the connecting rod thickness, and the expanded cross-sectional area, obtain the actual contour point cloud data of the expanded cross-section of each type of connecting rod through 3D scanning, and establish a classification contour database containing material grades and heat treatment parameters;

[0261] An analysis module 602 is configured to perform fracture mechanics analysis on the classified contour database, using a theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a set of fracture characteristic parameters;

[0262] A generation module 603 is configured to call the fracture characteristic parameter set, perform dynamic fracture process simulation in combination with the fracture toughness value of the material, and generate an initial fracture contour surface including randomly distributed protrusions / pits;

[0263] The optimization module 604 is used to perform finite element contact analysis on the initial fracture profile surface, use the bolt preload as the boundary condition, iteratively adjust the geometric parameters of the protrusions / pits until the contact stress uniformity reaches a preset threshold, and output the optimized fracture profile surface.

[0264] An embodiment of the present invention further provides a storage medium storing a computer program, wherein the computer program is configured to execute the steps of any one of the above method embodiments when running.

[0265] Specifically, in this embodiment, the above-mentioned storage medium may be configured to store a computer program for performing the following steps:

[0266] S201, classifying historically broken connecting rods based on the inner diameter of the connecting rod big end hole, connecting rod thickness, and expanded cross-sectional area, obtaining actual contour point cloud data of the expanded cross-sections of various connecting rods through 3D scanning, and establishing a classified contour database containing material grades and heat treatment parameters;

[0267] S202, performing fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set;

[0268] S203, calling the fracture characteristic parameter set, combining the material fracture toughness value to perform dynamic fracture process simulation, and generating an initial fracture contour surface including randomly distributed protrusions / pits;

[0269] S204, performing finite element contact analysis on the initial fracture profile surface, using the bolt preload as a boundary condition, iteratively adjusting the geometric parameters of the protrusions / pits until the contact stress uniformity reaches a preset threshold, and outputting an optimized fracture profile surface.

[0270] An embodiment of the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any one of the above method embodiments.

[0271] Specifically, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0272] Specifically, in this embodiment, the processor may be configured to execute the following steps through a computer program:

[0273] S201, classifying historically broken connecting rods based on the inner diameter of the connecting rod big end hole, connecting rod thickness, and expanded cross-sectional area, obtaining actual contour point cloud data of the expanded cross-sections of various connecting rods through 3D scanning, and establishing a classified contour database containing material grades and heat treatment parameters;

[0274] S202, performing fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set;

[0275] S203, calling the fracture characteristic parameter set, combining the material fracture toughness value to perform dynamic fracture process simulation, and generating an initial fracture contour surface including randomly distributed protrusions / pits;

[0276] S204, performing finite element contact analysis on the initial fracture profile surface, using the bolt preload as a boundary condition, iteratively adjusting the geometric parameters of the protrusions / pits until the contact stress uniformity reaches a preset threshold, and outputting an optimized fracture profile surface.

[0277] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings. The above is only a preferred embodiment of the present invention, but the scope of implementation of the present invention is not limited to what is shown in the drawings. Any changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which do not exceed the spirit covered by the description and drawings, should be within the scope of protection of the present invention.

Claims

1. A method for generating a connecting rod expansion section, characterized in that: The method comprises: Historically broken connecting rods were classified based on the inner diameter of the connecting rod's big end hole, connecting rod thickness, and expanded cross-sectional area. 3D scanning was used to obtain actual contour point cloud data of the expanded cross-sections of various connecting rods, and a classification contour database containing material grades and heat treatment parameters was established. Performing fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set; The fracture characteristic parameter set is called, and the dynamic fracture process simulation is performed in combination with the material fracture toughness value to generate an initial expansion fracture contour surface containing randomly distributed protrusions / pits; wherein, the gamma distribution parameters in the fracture characteristic parameter set are analyzed to generate a random concave-convex radius sequence that conforms to the actual statistical law, and an initial concave-convex size sample pool is output; according to the initial concave-convex size sample pool, random concave-convex is spread on the theoretical expansion fracture surface based on the coordinate association rules of the spatial topological graph to ensure that the spatial distribution of concave-convex conforms to the historical statistical characteristics, and an initial geometric model with random concave-convex is output; the material fracture toughness value is loaded into the explicit dynamics solver, the concave-convex generation timing in the initial geometric model is controlled by the crack propagation rate, a high strain rate fracture process simulation is performed, and a dynamic crack path is output; the cross-sectional morphology in the stable crack propagation stage is extracted to generate an initial expansion fracture contour surface containing randomly distributed protrusions / pits; Finite element contact analysis is performed on the initial fracture profile surface. With the bolt preload as the boundary condition, the geometric parameters of the protrusions / pits are iteratively adjusted until the contact stress uniformity reaches a preset threshold, and the optimized fracture profile surface is output.

2. The method according to claim 1, characterized in that The method of classifying historical broken connecting rods according to the inner diameter of the connecting rod big end hole, the connecting rod thickness and the expanded cross-sectional area, obtaining the actual contour point cloud data of the expanded cross-sections of various connecting rods through 3D scanning, and establishing a classification contour database containing material grades and heat treatment parameters includes: A laser interferometer 3D scanner is used to scan the cross section of the historically broken connecting rod with sub-micron accuracy, and outputs raw point cloud data containing surface topography details. Based on the process parameter combination of the inner diameter, thickness and expanded cross-sectional area of ​​the connecting rod's big end hole, a three-dimensional parameter space grid is constructed, the original point cloud data is mapped to the corresponding grid cells, and a parameterized classified point cloud cluster is output; Add material grade and heat treatment process labels to each parameterized point cloud cluster, align them to a unified coordinate system through a point cloud registration algorithm, and output a standardized cross-section contour point cloud set; The standardized cross-section contour point cloud set is imported into the NoSQL database, and a classified contour database is established with process parameters as index keys and point cloud data and material properties as values.

3. The method according to claim 2, characterized in that The fracture mechanics analysis is performed on the classification profile database. The theoretical expansion section is used as the reference plane. The actual profile is fitted using a sphere envelope algorithm. The number of protrusions / pits, center coordinates, and sphere radius distribution are counted to output a set of fracture characteristic parameters, including: Read the cross-section point cloud corresponding to the target process parameters from the classification profile database, fit the theoretical expansion reference plane using the least squares method, and output the normal deviation field from the point cloud to the reference plane; Perform multi-scale curvature analysis on the normal deviation field, identify local extreme points as concave-convex feature seed points, and output a set of candidate concave-convex region coordinates; Taking the candidate concave-convex area as the center, the adaptive radius sphere envelope algorithm is used for surface fitting, and the envelope sphere radius and the spatial coordinates of the sphere center of each concave-convex area are output; The gamma distribution parameters of the envelope sphere radius are statistically analyzed, and combined with the spatial autocorrelation characteristics of the sphere center coordinates, a spatial distribution topology map of the concave and convex features is generated; The gamma distribution parameters are integrated with the spatial topology map to construct a fracture feature parameter set including the number, location, and size statistics of asperities.

4. The method according to claim 3, characterized in that The finite element contact analysis is performed on the initial fracture profile surface, and the geometric parameters of the protrusions / pits are iteratively adjusted with the bolt preload as the boundary condition until the contact stress uniformity reaches a preset threshold, and the optimized fracture profile surface is output, including: Apply the bolt preload boundary condition on the initial expansion and fracture contour surface, calculate the local contact pressure field using the Hertz contact theory, and output the initial contact stress distribution cloud map; Calculate the stress uniformity index based on the current contact stress distribution cloud map. If the stress uniformity is lower than the preset threshold, locate the stress concentration area and output the high stress area coordinate set and stress gradient vector. Based on the high stress area coordinate set and stress gradient vector, the concave and convex geometric parameters of the corresponding position are dynamically adjusted, and the expanded fracture contour surface with updated parameters is output; The finite element contact analysis is performed again on the expanded fracture contour surface after the parameter update, a new contact stress contour is generated, and the step of calculating the stress uniformity index based on the current contact stress distribution contour is returned to execute until the stress uniformity reaches the preset threshold, and finally the optimized expanded fracture contour surface is output.

5. A connecting rod expansion section generation system, characterized in that: The system comprises: The acquisition module is used to classify historical broken connecting rods based on the inner diameter of the connecting rod big end hole, the connecting rod thickness, and the expanded cross-sectional area. The actual contour point cloud data of the expanded cross-sections of various connecting rods is obtained through 3D scanning, and a classification contour database containing material grades and heat treatment parameters is established. An analysis module is used to perform fracture mechanics analysis on the classified contour database, using the theoretical expansion fracture surface as a reference plane, using a sphere envelope algorithm to fit the actual contour, counting the number of protrusions / pits, center coordinates, and sphere radius distribution, and outputting a fracture feature parameter set; A generation module is used to call the fracture characteristic parameter set, perform dynamic fracture process simulation in combination with the material fracture toughness value, and generate an initial expansion fracture contour surface containing randomly distributed protrusions / pits; wherein, the gamma distribution parameters in the fracture characteristic parameter set are analyzed to generate a random concave-convex radius sequence that conforms to the actual statistical law, and output an initial concave-convex size sample pool; according to the initial concave-convex size sample pool, based on the coordinate association rules of the spatial topological graph, random concave-convex is spread on the theoretical expansion fracture surface to ensure that the spatial distribution of concave-convex conforms to the historical statistical characteristics, and output an initial geometric model with random concave-convex; load the material fracture toughness value into the explicit dynamics solver, control the concave-convex generation timing in the initial geometric model with the crack propagation rate, perform high strain rate fracture process simulation, and output a dynamic crack path; extract the cross-sectional morphology of the stable crack propagation stage, and generate an initial expansion fracture contour surface containing randomly distributed protrusions / pits; The optimization module is used to perform finite element contact analysis on the initial fracture profile surface, use the bolt preload as the boundary condition, iteratively adjust the geometric parameters of the protrusion / pit until the contact stress uniformity reaches a preset threshold, and output the optimized fracture profile surface.

6. The system according to claim 5, characterized in that The acquisition module is specifically used to: A laser interferometer 3D scanner is used to scan the cross section of the historically broken connecting rod with sub-micron accuracy, and outputs raw point cloud data containing surface topography details. Based on the process parameter combination of the inner diameter, thickness and expanded cross-sectional area of ​​the connecting rod's big end hole, a three-dimensional parameter space grid is constructed, the original point cloud data is mapped to the corresponding grid cells, and a parameterized classified point cloud cluster is output; Add material grade and heat treatment process labels to each parameterized point cloud cluster, align them to a unified coordinate system through a point cloud registration algorithm, and output a standardized cross-section contour point cloud set; The standardized cross-section contour point cloud set is imported into the NoSQL database, and a classified contour database is established with process parameters as index keys and point cloud data and material properties as values.

7. The system according to claim 6, characterized in that The analysis module is specifically used to: Read the cross-section point cloud corresponding to the target process parameters from the classification profile database, fit the theoretical expansion reference plane using the least squares method, and output the normal deviation field from the point cloud to the reference plane; Perform multi-scale curvature analysis on the normal deviation field, identify local extreme points as concave-convex feature seed points, and output a set of candidate concave-convex region coordinates; Taking the candidate concave-convex area as the center, the adaptive radius sphere envelope algorithm is used for surface fitting, and the envelope sphere radius and the spatial coordinates of the sphere center of each concave-convex area are output; The gamma distribution parameters of the envelope sphere radius are statistically analyzed, and combined with the spatial autocorrelation characteristics of the sphere center coordinates, a spatial distribution topology map of the concave and convex features is generated; The gamma distribution parameters are integrated with the spatial topology map to construct a fracture feature parameter set including the number, location, and size statistics of asperities.

8. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 4 when run.

9. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 4.

Images