Woven Composite Microstructure Reconstruction via Topological Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for reconstructing the microstructure of woven composite materials are either damaging to the samples, overly complex, or result in models that do not accurately represent the real structure due to neglecting deformation and interaction between fiber bundles, leading to inaccuracies in predicting mechanical properties.
Innovation Solution
A method involving μCT scanning, parametric modeling, and image processing using software like ImageJ to extract and modify characteristic parameters of fiber bundles, creating a highly consistent microscopic model without damaging the material, reducing calculation costs and errors, and applicable to various woven structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detailed microstructure modeling is carried out completely according to real materials, then the accuracy of mechanical property prediction is improved, but the calculation cost increases greatly
Solution Approach 1:
The patent creates a simplified copy of the microstructure that preserves essential topological features (fiber bundle connections, weaving patterns) while omitting unnecessary geometric details. This copying approach maintains predictive accuracy for mechanical properties while significantly reducing calculation cost by using simplified geometric representations instead of complete detailed models.
Solution Approach 2:
The patent extracts only the critical topological features from the complete microstructure (such as fiber bundle connectivity, weaving sequence, and structural topology) while discarding less important geometric details. This extraction enables accurate mechanical property prediction with reduced computational resources by focusing only on the essential structural characteristics.
2Measurement precision
If detailed microstructure modeling is carried out completely according to real materials, then the accuracy of mechanical property prediction is improved, but the numerical calculation method may be difficult to converge
Solution Approach 1:
The patent creates a simplified copy of the microstructure that preserves essential topological features while removing small-scale geometric irregularities that cause numerical convergence problems. This copying approach maintains the structural topology necessary for accurate mechanical property prediction while eliminating features that would destabilize numerical calculations.
Solution Approach 2:
The patent segments the microstructure modeling into two levels: topological features (fiber bundle connectivity and weaving patterns) that are preserved for accuracy, and geometric details (small defects and irregularities) that are simplified for numerical stability. This segmentation allows the model to maintain predictive accuracy while ensuring reliable numerical convergence.
3Manufacturing precision
If cutting method is used to obtain actual fiber bundle cross-sectional profile, then the geometric model can be reconstructed, but damage is inevitably caused to the woven body
Solution Approach 1:
The patent replaces the mechanical cutting method with a non-contact μCT scanning method to obtain cross-sectional profiles of fiber bundles. This substitution eliminates physical damage to the woven body while still providing the geometric information needed for accurate model reconstruction, as μCT scanning uses X-rays to capture internal structures without physical contact.
4Ease of manufacture
If idealized model is used without considering bending and cross-section deformation, then the modeling process is simplified, but the model does not reflect actual woven structure
Solution Approach 1:
The patent creates a simplified copy of the fiber bundle cross-sections that captures the essential deformation characteristics (elliptical shapes, size variations) without requiring complete detailed modeling. This copying approach maintains modeling simplicity while improving consistency with the actual woven structure by incorporating observed deformation features.
Solution Approach 2:
The patent changes the geometric parameters of fiber bundle cross-sections from idealized circular shapes to elliptical shapes with variable dimensions that reflect actual deformation. This parameter modification allows the model to account for bending and cross-section deformation while maintaining relative modeling simplicity through parametric definitions rather than complex geometry.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method provides a fast, accurate, and cost-effective reconstruction of woven composite material microstructures, ensuring high consistency with real structures and broad applicability across different woven configurations, reducing errors and calculation costs compared to existing methods.
Implementation Method 1
scanning using a μCT system: scanning the composite material sample with a high-power microscopic μCT system to obtain the point cloud data of the composite material sample
Data Source
AI summary
A method for rapid reconstruction of a woven composite material microstructure based on topological features is provided. First, geometric modeling is performed according to the topological features of fiber bundles and matrix in a woven composite structure, then the topological features are identified on the basis of a μ-computed tomography (CT) sequence image, the change laws of the features are extracted, and finally, a geometric model is reconstructed to complete the modeling of a microstructure. Compared with the structure modeling method based on weaving process parameters, the consistency between the reconstructed model and a real woven structure is improved, facilitating the improvement of the accuracy of the subsequent calculation of the mechanical properties of the material; and compared with the method for structure reconstruction directly on the basis of a μCT image, a tedious point cloud data process is simplified and calculation costs are greatly reduced.