ML-Based CAE Mesh Generation for Complex Geometric Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The time-consuming process of meshing parts with complex features in computer-aided engineering (CAE) modeling, as existing methods require extensive hours or days to identify and mesh features accurately, hindering efficiency and compliance with customer requirements.
Innovation Solution
A system utilizing a machine learning model to recognize and classify features into families, applying feature-specific mesh parameters to generate meshes efficiently, thereby reducing the time and improving accuracy in mesh generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional meshing methods are used to identify and mesh features accurately, then manufacturing precision is improved, but productivity deteriorates due to extensive hours or days required
Solution Approach 1:
The patent replaces traditional manual or rule-based meshing algorithms with a machine learning model that automatically identifies and classifies geometric features. The ML model learns from training data to recognize feature families (such as holes, protrusions, recesses) and applies appropriate mesh parameters, substituting the mechanical step-by-step meshing process with an intelligent system that achieves both high accuracy and rapid processing.
Solution Approach 2:
The system changes the approach by using machine learning to dynamically determine meshing parameters based on recognized feature families. Instead of using fixed or manually adjusted parameters, the ML model selects and applies feature-specific mesh parameters automatically, adapting the meshing strategy to the specific geometric features present in the part, thereby maintaining precision while reducing processing time.
2Measurement precision
If manual feature identification and meshing is performed, then measurement precision is improved, but loss of time increases significantly
Solution Approach 1:
The system performs preliminary action by pre-training the machine learning model on extensive datasets of geometric features and their corresponding mesh parameters. This training phase prepares the model in advance to quickly and accurately identify features during actual meshing operations, eliminating the need for time-consuming manual feature identification while maintaining high measurement precision.
Solution Approach 2:
The patent substitutes manual feature identification and classification with an automated machine learning-based recognition system. The ML model processes geometric representations and identifies feature families rapidly, replacing the slow manual process while maintaining or improving identification accuracy through consistent application of learned patterns.
3Productivity
If automated meshing is implemented without machine learning, then productivity is improved, but manufacturing precision deteriorates due to inability to handle complex features
Solution Approach 1:
The patent enhances automated meshing by integrating a machine learning model that specializes in recognizing complex geometric features. The ML system substitutes basic automated meshing algorithms with an intelligent recognition engine that can distinguish and properly mesh complex features (such as undercuts, variable thickness sections, and intricate geometries) that traditional automated methods cannot handle accurately.
Solution Approach 2:
The system dynamically changes meshing parameters based on the complexity and type of features identified by the machine learning model. For complex features, the ML model selects specialized mesh parameters and strategies, while for simpler features, it applies standard parameters. This adaptive parameter selection maintains high productivity while ensuring accurate meshing of complex geometries.
Data Source
AI summary
A memory stores a representation of geometric features of an object, a machine learning model configured to identify one or more feature families of features of the representation, and feature-specific parameters defining how to mesh the one or more feature families. A processor recognizes and classifies features of the representation into the feature families utilizing the machine learning model, applies feature-specific mesh parameters to the recognized and classified features of the representation, and generates a mesh of the representation in accordance with the feature-specific parameters.