Octree Shrink Wrap Generation for Complex CAE Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mesh-fitting methods in CAE systems fail to generate accurate shrink wraps around complex, unsymmetrical objects, resulting in severe wrinkles and extreme stretching due to differences in surface metrics between the finite element model and the shrink-wrap, leading to inadequate approximations for simulations.
Innovation Solution
The method involves detecting non-manifold edges in an octree-generated shrink wrap, reconnecting or cloning edges, and generating projections using a seed-based closest point method to improve the wrapper's fit, ensuring it adheres closely to the model while avoiding excessive stretching and creases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a mesh-fitting scheme is used to generate a wrapper around a finite element model, then the process is simple and works for symmetrical surfaces, but it fails to generate good wrappers around unsymmetrical objects resulting in severe wrinkles and extreme stretching
Solution Approach 1:
The patent segments the wrapper generation process into multiple stages: initial octree-based wrapper generation, projection stage with angle threshold detection, and secondary projection for problematic vertices. This segmentation allows the system to handle different regions of the model with appropriate methods, avoiding wrinkles and stretching in unsymmetrical areas while maintaining simplicity for symmetrical regions.
Solution Approach 2:
The patent performs preliminary actions by first generating an initial wrapper using the simple octree method, then identifying vertices with excessive projection angles or rotational angles before performing secondary projection. This preliminary detection and classification allows the system to apply corrective actions only where needed, improving accuracy without sacrificing the efficiency of the initial generation process.
2Manufacturing precision
If the wrapper is shrunk towards the model to improve fit, then the approximation quality improves, but non-manifold edges and topological defects are created
Solution Approach 1:
The patent implements feedback mechanisms by detecting non-manifold edges after shrinking and automatically reconnecting them. The system counts faces adjacent to each edge, identifies non-manifold conditions, and performs corrective reconnection operations. This feedback loop ensures topological correctness is restored after the shrinking process, maintaining both fit quality and reliability.
Solution Approach 2:
The patent changes parameters by introducing angle thresholds (maximum projection angle, maximum rotational angle) to control the shrinking process. Vertices exceeding these thresholds are identified and subjected to secondary projection with modified parameters, allowing the system to maintain topological correctness while achieving good fit quality in critical regions.
3Manufacturing precision
If projection angles and rotational angles are increased to improve wrapper adherence, then the wrapper fits the model more closely, but extreme stretching and distortion occur
Solution Approach 1:
The patent applies local quality by treating different vertices differently based on their projection characteristics. Vertices with projection angles or rotational angles exceeding thresholds are identified and subjected to secondary projection, while other vertices use the standard projection method. This localized approach ensures good adherence where needed without causing extreme stretching in sensitive regions.
Data Source
AI summary
A system and method for generating a shrink wrap around a model. The method includes detecting non-manifold edges in an octree generated shrink wrap by counting a number of faces adjacent to each edge, removing the non-manifold edges by cloning the edges or vertices shared by the non-manifold edge, and generating a first projection for the wrapper by moving each wrapper vertex towards a nearest location on the model. The method includes determining a set of wrapper vertices for reprojection based on the computation of a projection angle and a rotational angle and generating a second projection for the set the wrapper vertices using a seed-based closest point method or the center of the adjacent wrapper vertices.