Thin-Shell Deformation Simulation Using Curvature-Based Energy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for simulating deformation of thin-shell materials, such as cloth, face issues like instability and altitude collapse due to infinite bending energy derivatives and mesh dependence, particularly in conventional bending models like Cubic Shells and mass-spring systems.
Innovation Solution
A method that determines discretized bending energy using a difference between current and reference curvature vectors, ensuring a finite lower bound, which is calculated using a weighted average of normal vectors and applied to vertices of a mesh comprising polygons and hinges, avoiding the limitations of existing models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional bending models (Cubic Shells, mass-spring systems) are used to simulate thin-shell deformation, then computational performance can be maintained, but the simulation suffers from instability and altitude collapse due to infinite bending energy derivatives
Solution Approach 1:
The patent changes the mathematical formulation of bending energy from conventional models to a discrete mean curvature energy model. By using the difference between current and reference mean curvature vectors raised to an even power (e.g., square), the bending energy achieves a finite lower bound while maintaining computational efficiency. This parameter change in the energy formulation eliminates infinite derivatives and prevents simulation instability.
Solution Approach 2:
The patent replaces the mechanical spring-based bending models with a curvature-based energy model. Instead of using mass-spring systems or cubic shell formulations that rely on mechanical analogies, the invention substitutes a mathematical energy formulation based on mean curvature vectors. This substitution eliminates the pathological behavior of conventional mechanical models while preserving computational performance.
2Device complexity
If conventional bending models are used, then computational simplicity is maintained, but mesh dependence and altitude collapse occur
Solution Approach 1:
The patent formulates bending energy as a function of mean curvature vectors that are computed from the mesh geometry. By using the even power of the difference between current and reference mean curvature vectors, the model achieves mesh independence. The curvature-based formulation naturally adapts to different mesh configurations without requiring mesh-specific parameters or corrections.
3Reliability
If a finite lower bound bending energy is implemented, then simulation stability is improved, but computational complexity increases
Solution Approach 1:
The patent replaces complex conventional bending models with a streamlined curvature-based energy formulation. The discrete mean curvature energy model computes bending energy as the even power of the difference between current and reference mean curvature vectors. This substitution simplifies the computational approach while achieving finite lower bound energy and simulation stability.
Solution Approach 2:
The patent creates a universal bending energy model that works across different mesh types and deformation scenarios. The mean curvature-based formulation is mesh-independent and applicable to various thin-shell materials, eliminating the need for multiple specialized models. This universal approach reduces overall computational complexity by providing a single robust solution for all bending scenarios.
Data Source
AI summary
Methods, systems, and techniques for simulating deformation of a thin-shell material. A processor is used to obtain a mesh of the material, which is made up of polygons and hinges and in which any two of the polygons that are adjacent to each other connect to each other at one of the hinges. The processor then determines forces affecting vertices of the mesh. The forces are determined from a gradient of cumulative potential energy which, for each of at least some of the hinges, includes a discretized bending energy that has a finite lower bound and that is determined using a difference between a current curvature vector and a reference curvature vector. The processor then simulates deformation of the mesh using those forces.