A method for deriving a machining process of a plate part based on volume decomposition

Abstract

This invention discloses a method for deriving the machining process of plate-like parts based on volume decomposition. The method treats plate-like parts as a series of machining processes obtained by removing material from the original blank through subtractive processing. The steps include: geometric preprocessing and volume difference calculation: chamfer identification and filling based on a three-sided chain restores the digital model to its state before finishing, and material removal generation based on the OpenCASCADE geometric kernel; backfill complexity analysis and volume decomposition: convex hull inspection determines whether the backfill block is convex, and cutting guided by concave edges reduces the "concavity" of the backfill block; feature recognition and machining process mapping: datum surface identification and reachability analysis generate a mapping table, feature rule matching identifies feature types and establishes a feature-machining action correspondence, and the complete machining process sequence is output through reverse derivation of the machining process. This invention solves the problem of feature adhesion identification and achieves a leap from "geometric recognition" to "process derivation".

Description

Technical Field

[0001] This invention relates to the fields of computer vision and computer-aided design, and in particular to a method for deriving the machining process of plate-type parts based on volume decomposition. Background Technology

[0002] Traditional methods for deriving processing steps primarily rely on a series of rule-based and geometric reasoning techniques, including but not limited to: feature recognition methods, feature modeling methods, and expert system-based methods. These methods have the following limitations:

[0003] 1. When multiple processing features intersect and overlap in space (e.g., a groove passes through a hole), their boundaries will merge or disappear, forming a complex mixed region. Predefined feature templates cannot match this non-standard, mixed geometric topology, leading to recognition failure or incorrect results.

[0004] 2. Traditional methods (especially purely geometric feature recognition) can only identify the "shape" of features, but cannot understand their "function" and "processing intent";

[0005] 3. Rule-based and expert system-based methods rely heavily on domain knowledge, and acquiring and formalizing this knowledge is a time-consuming, expensive, and error-prone process (i.e., the "knowledge acquisition bottleneck").

[0006] 4. Even if all features are successfully identified, sequencing them into a reasonable and efficient machining sequence is a huge challenge. Sequencing needs to consider process constraints (such as roughing before finishing), geometric constraints (such as machining the reference surface first), clamping constraints (such as reducing the number of clamping operations), and resource constraints.

[0007] With the digital transformation of industry, there is an urgent need for a part processing derivation method with multi-feature generalization and adaptive capabilities to improve the automation level of CAD systems and intelligent manufacturing platforms. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a method for deriving the machining process of plate-type parts by combining computational geometry and topology with digital manufacturing and processing simulation.

[0009] The objective of this invention is achieved through the following technical solutions.

[0010] A method for deriving the machining process of plate-type parts based on volume decomposition treats plate-type parts as a complete machining process sequence obtained by removing material from the original blank through subtractive processing. Specific steps include:

[0011] 1) Geometric preprocessing and volume difference calculation: Based on the chamfer recognition and filling of the three-sided chain, the digital model is restored to the state before finishing, and the material removal generation is based on the OpenCASCADE geometric kernel;

[0012] 2) Backfill material complexity analysis and volume decomposition: Convex hull inspection is used to determine whether the backfill block is convex, and the "concavity" of the backfill block is reduced by guiding the cutting through concave edges;

[0013] 3) Feature recognition and processing procedure mapping: Reference surface recognition and accessibility analysis generate a mapping relationship table, feature rule matching identifies feature types and establishes feature-processing action correspondence, and the complete processing procedure sequence is output by reverse derivation through processing procedures.

[0014] The chamfer identification and filling based on the three-sided chain is as follows: by traversing the three-sided chain in the digital model, the chamfer surface is selected based on the chamfer topology, and the chamfer surface is filled by stretching to restore the digital model to the state before finishing.

[0015] The material removal generation based on the OpenCASCADE geometric kernel is specifically as follows: input a 3D model of a plate-type part in STP format, use the OpenCASCADE geometric kernel to calculate the axial bounding box of the part as the initial blank for processing, and perform a Boolean difference operation: blank entity - part entity = backfill material volume set.

[0016] The convex hull test is as follows: for each backfill block, construct the maximum convex hull, calculate the volume difference between the maximum convex hull and the body, and determine whether the backfill block is convex based on this.

[0017] The concave edge guided cutting is as follows: for non-convex backfill blocks, they need to undergo longitudinal cutting based on the reference surface under the guidance of the concave edge and transverse cutting based on themselves. The two cuttings are used to reduce the "concavity" of the backfill block.

[0018] The reference surface identification and accessibility analysis involves simulating a multi-axis machine tool to find the machining reference surface, determining the reachable area of ​​the tool through grid point cloud sampling and collision detection, and generating a mapping table between the reference surface and the machinable point cloud.

[0019] The feature rule matching is as follows: establish a standard processing feature library, match the geometric shape of simple material blocks with the feature library, identify feature types, and establish a feature-processing action correspondence.

[0020] The reverse derivation of the processing steps: the actual processing sequence is deduced in reverse according to the material backfilling sequence, taking into account process constraints, geometric constraints and clamping constraints, a structured dictionary is established, and the complete processing sequence is output.

[0021] The features in the standard machining feature library include cavities, holes, steps, grooves, chamfers, and fillets.

[0022] The process constraints are roughing first and refining later, and primary first and secondary later. The geometric constraints are based on the benchmark. The structured dictionary includes process steps, backfill material, blank state and production characteristics.

[0023] Compared to existing technologies, the advantages of this invention are: 1. It fundamentally solves the problem of feature adhesion identification: Traditional methods attempt to "identify" features from complex, blurred-boundary final geometries, while this invention takes a different approach, using inverse Boolean operations and volume decomposition to reduce mixed adhesion features to multiple independent, simple material removal units. This method analyzes the feature adhesion problem from a physical perspective, significantly improving the accuracy and robustness of feature identification for complex plate-like parts;

[0024] 2. This invention achieves a leap from "geometric recognition" to "process derivation," with output results directly applicable to process planning: Most existing technologies stop at recognizing feature geometric parameters. This invention not only identifies features, but its core value lies in automatically deducing a reasonable sequence of processing steps. Its output is a structured processing procedure (including process steps, corresponding material removal, changes in blank state, etc.), which can directly provide input for Computer-Aided Process Planning (CAPP) systems or Manufacturing Execution Systems (MES), greatly improving the automation level and efficiency of process design.

[0025] 3. Possesses strong generalization ability and physical interpretability: Because it does not rely on deep learning models trained on specific feature libraries or manually compiled rule templates, this invention has a natural adaptability to unseen part structures and novel feature combinations. Furthermore, the entire method is based on the physical essence of "subtractive manufacturing," with each step of reasoning having a clear physical meaning (e.g., material removal, tool reachability). The results are easy for process engineers to understand and verify, avoiding "black box" decision-making and enhancing credibility in practical industrial applications.

[0026] 4. High computational efficiency and strong engineering applicability: Developed based on mature geometric kernels such as OpenCASCADE, it directly processes industry standard formats such as STEP, making it easy to integrate with existing CAD / CAM / PLM systems. Through efficient convex hull detection and intelligent decomposition algorithms, it avoids the complex graph matching or large-scale neural network calculations of traditional methods, significantly improving processing speed while ensuring high accuracy, meeting the timeliness requirements of actual production. Attached Figure Description

[0027] Figure 1 This is a flowchart of the present invention.

[0028] Figure 2 This is a screenshot of the STP file of part A in an embodiment of the present invention.

[0029] Figure 3This is a schematic diagram of the material removal process for part A in an embodiment of the present invention.

[0030] Figure 4 This is a demonstration diagram of the longitudinal cutting and transverse self-cutting of the complex geometric backfill block in part A of this invention.

[0031] Figure 5 This is the result of merging the maximum convex hull of the atomized backfill block of part A in this embodiment of the invention.

[0032] Figure 6 This refers to the minimum machining reference surface and the corresponding number of material removal point clouds for part A in this embodiment of the invention.

[0033] Figure 7 This is the accessible point cloud of material removal in the reference plane direction [0 0 -1] of this embodiment of the invention.

[0034] Figure 8 This is the accessible point cloud for material removal in the reference plane direction [-1 0 0] of this embodiment of the invention.

[0035] Figure 9 This is the accessible point cloud of the reference plane direction [0 0 1] in the embodiment of the present invention.

[0036] Figure 10 This is a schematic diagram of the processable material removal of the reference plane [0 0 -1] in an embodiment of the present invention.

[0037] Figure 11 This is a digital model of the punching process in an embodiment of the present invention.

[0038] Figure 12 This is a digital model of the grooving process in an embodiment of the present invention. Detailed Implementation

[0039] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0040] like Figure 1 As shown, this invention employs a three-step method of "reverse material backfilling - volume decomposition - process mapping." The core idea is to treat plate-type parts as the result of removing material from the original blank through subtractive processing. The core includes the following steps:

[0041] I. Geometric Preprocessing and Volume Difference Calculation

[0042] 1. Chamfer identification and filling based on three-sided chains: By traversing the three-sided chains in the digital model, the chamfer surfaces are selected based on the unique topology of the chamfer. The chamfer surfaces are then filled by stretching, restoring the digital model to its state before finishing.

[0043] 2. Material removal generation based on OpenCASCADE (OCC) geometric kernel: Input a 3D model of a plate-type part in STP format, use the OpenCASCADE (OCC) geometric kernel to calculate the axial bounding box of the part as the initial blank for processing, and perform Boolean difference operation: blank entity - part entity = backfill material volume set.

[0044] II. Complexity Analysis and Volume Decomposition of Backfill Material

[0045] 1. Convex Hull Test: For each backfill block, construct the maximum convex hull, calculate the volume difference between the maximum convex hull and the body, and determine whether the backfill block is convex based on this.

[0046] 2. Concave edge guided cutting: For non-convex backfill blocks, it is necessary to undergo longitudinal cutting based on the reference surface under the guidance of the concave edge and transverse cutting based on itself. The purpose of both cuttings is to reduce the "concavity" of the backfill block.

[0047] III. Feature Recognition and Process Mapping

[0048] 1. Reference Surface Identification and Reachability Analysis: Simulate a multi-axis machine tool to find the machining reference surface, determine the reachable area of ​​the tool through mesh point cloud sampling and collision detection, and generate a mapping relationship table of {reference surface: machinable point cloud};

[0049] 2. Feature rule matching: Establish a standard machining feature library (cavity, hole, step, groove, chamfer, fillet, etc.), match the geometry of simple material blocks with the feature library, identify feature types and establish feature-machining action correspondence;

[0050] 3. Reverse derivation of processing steps: The actual processing sequence is deduced in reverse according to the material backfilling sequence. Considering process constraints (roughing before finishing, main before secondary), geometric constraints (datum priority) and clamping constraints, a structured dictionary is established: {process steps: backfill material: blank state: generation features}, and the complete processing sequence is output. Example

[0051] I. Material Reduction Generation of 3D Digital Models

[0052] 1. For example Figure 2 The STP file for part A shown is used to input the digital model of part A into the subtractive machining generation program. The program will generate a close-fitting blank based on the bounding box size and generate the corresponding machining subtraction, such as... Figure 3 As shown.

[0053] 2. Identify that the material removal portion of part A begins cutting under the guidance of the concave edge, such as... Figure 4 As shown;

[0054] 3. Based on the maximum convex hull, merge the atomized segmentation results, such as... Figure 5 As shown. Figure 5 The merged portion shown can be found in Figure 4 Find the corresponding geometric shape. You can see... Figure 4 The three cuboids in the front center are merged into a single cuboid without a concave edge.

[0055] II. Processability Verification

[0056] By inputting the material removal part of part A into the machinability verification program, the minimum reference surface required to process part A and the machinable part of each reference surface (point cloud output) can be obtained. Then, by comparing it with the convex hull merging result mesh of the previous step, the corresponding reference surface of each backfill block can be obtained.

[0057] 1. Point cloud generation for material removal processing

[0058] Input the material removal information of part A into the manufacturability verification program; the response is as follows: Figure 6 As shown.

[0059] Visualize the material removal point cloud for each reference plane, and the corresponding machinable material removal point cloud for each reference plane is as follows: Figures 7 to 9 .

[0060] 2. Verification of backfill block processing

[0061] During the processing verification of backfill blocks, the system performs meshing on the backfill blocks cut and merged into the largest convex hull, and compares and analyzes them with the processing point clouds of different reference surfaces. Specific verification data can be obtained through coverage statistics.

[0062] Taking the backfill block of one embodiment as an example, the verification results are as follows:

[0063] Mesh and point cloud matching degree: The total number of meshes in the Solid of this backfill block is 24, of which 24 meshes are covered, and the coverage ratio reaches 100.00%, which is far higher than the preset threshold of 90.00%, proving that the backfill block is highly matched with the processing area of ​​the corresponding reference surface.

[0064] Sampling parameters: The grid size used in the verification process was 1.798, and the search radius was 2.698.

[0065] Point cloud sampling changes: After calculation and processing, the number of point cloud samples was optimized and adjusted from the initial 14 to 12.

[0066] Automated processing: The system automatically calculates that the number of internal points is 14, the number of grids is approximately 14, and the automatically adjusted grid size is 1.981.

[0067] Through the above comparison and verification, the minimum processing reference surface corresponding to each backfill block can be accurately determined, thereby ensuring the accuracy and feasibility of the processing derivation. Based on the above, the processable material removal of the reference surface [0 0 -1] can be obtained, such as... Figure 10 As shown.

[0068] III. Process Reasoning

[0069] Based on the above verification of material removal, segmentation, merging, and corresponding material removal on the reference surface, the machining process of the part can be obtained by reversing the above process, such as... Figure 11-12 As shown.

[0070] This invention employs a differential volume modeling method based on inverse Boolean operations between the blank and the finished product. It abandons the traditional approach of directly identifying features from the part's geometry and pioneers a reverse modeling approach by subtracting the part's volume from the blank to obtain the volume of the material to be removed. By calculating the precise volume difference between the blank and the finished product, the derivation of the processing process is transformed into an analytical problem of "material removal," laying a mathematically meaningful foundation for the entire process derivation.

[0071] This invention presents an intelligent volume decomposition algorithm for machining accessibility: addressing the problem that "difference volumes" are often complex non-convex bodies and difficult to directly correspond to machining operations, a volume decomposition algorithm integrating geometric characteristics and manufacturing constraints is proposed. This algorithm not only segments based on geometric features such as concave edges, but also integrates machining accessibility analysis (such as tool interference detection and reference surface optimization), ensuring that each simple volume unit after decomposition corresponds to an actually executable machining step.

[0072] This invention is based on an automatic process chain reconstruction technology using a "volume unit-machining step" mapping: establishing a semantic mapping rule base from decomposed basic volume units to standard machining steps. By analyzing the geometric attributes (such as shape, orientation, and size) and technological knowledge (such as machining methods and tool selection) of volume units, it automatically generates a reasonable sequence of machining steps and outputs structured process data that can be integrated into a CAPP system, realizing the automatic conversion from design model to manufacturing instructions.

Claims

1. A method for deriving the machining process of plate-type parts based on volume decomposition, characterized in that... Considering sheet metal parts as the result of removing material from the original blank through subtractive processing, a complete processing sequence can be deduced, including the following steps: 1) Geometric preprocessing and volume difference calculation: The chamfer identification and filling based on three-sided chains restores the digital model to its state before finishing, and the material removal generation based on the OpenCASCADE geometric kernel is performed. Specifically, the chamfer identification and filling based on three-sided chains involves traversing the three-sided chains in the digital model, filtering out chamfer surfaces based on the chamfer topology, and filling the chamfer surfaces by stretching to restore the digital model to its state before finishing. Specifically, the material removal generation based on the OpenCASCADE geometric kernel involves inputting a 3D model of a plate-type part in STP format, using the OpenCASCADE geometric kernel to calculate the part's axial bounding box as the initial blank for processing, and performing a Boolean difference operation: blank entity - part entity = backfill material volume set. 2) Backfill Material Complexity Analysis and Volume Decomposition: Convex hull inspection determines whether the backfill block is convex, and concave edge guided cutting reduces the "concavity" of the backfill block; the convex hull inspection is as follows: for each backfill block, construct the maximum convex hull, calculate the volume difference between the maximum convex hull and the body, and determine whether the backfill block is convex based on this; the concave edge guided cutting is as follows: for non-convex backfill blocks, they need to undergo longitudinal cutting based on the reference plane under the guidance of the concave edge and transverse cutting based on themselves. The two cuttings are used to reduce the "concavity" of the backfill block; 3) Feature Recognition and Machining Process Mapping: Reference surface recognition and accessibility analysis generate a mapping table. Feature rule matching identifies feature types and establishes a feature-machining action correspondence. The complete machining process sequence is output through reverse derivation of machining processes. Reference surface recognition and accessibility analysis involve simulating a multi-axis machine tool to find the machining reference surface, determining the tool's reachable area through mesh point cloud sampling and collision detection, and generating a mapping table between the reference surface and the machinable point cloud. Feature rule matching involves establishing a standard machining feature library, matching the geometry of simple material blocks with the feature library, identifying feature types, and establishing a feature-machining action correspondence. Reverse derivation of machining processes involves deriving the actual machining sequence in reverse according to the material backfilling order, considering process constraints, geometric constraints, and clamping constraints, establishing a structured dictionary, and outputting a complete machining process sequence. The features in the standard machining feature library include cavities, holes, steps, grooves, chamfers, and fillets.

2. The method for deriving the machining process of plate-type parts based on volume decomposition according to claim 1, characterized in that... The process constraints are roughing first and refining later, and primary first and secondary later. The geometric constraints are based on the benchmark. The structured dictionary includes process steps, backfill material, blank state and production characteristics.

Images