Abstract:Trimming is a core technique in geometric modeling. Unfortunately, the resulting objects do not take the requirements of numerical simulations into account and yield various problems. This paper outlines principal issues of trimmed models and highlights different analysis-suitable strategies to address them. It is discussed that these concepts not only provide important computational tools for isogeometric analysis, but can also improve the treatment of trimmed models in a design context.
What problem does this paper attempt to address?
The main problems that this paper attempts to solve are the compatibility and robustness issues with **trimmed CAD models in numerical simulation**. Specifically, these problems stem from the fact that traditional geometric modeling techniques do not fully consider the requirements of numerical simulation, resulting in various challenges when dealing with trimmed models. The following are the specific problems and solutions mentioned in the paper:
### Main Problems
1. **The Gap between Geometric Modeling and Numerical Simulation**:
- Traditional Computer - Aided Geometric Design (CAGD) techniques are mainly used to create geometric models, and these models are not always suitable for direct use in numerical simulation.
- Trimming is a core technique in geometric modeling, but the trimmed models often have inaccuracies and inconsistencies, which affect their application in numerical simulation.
2. **The Complexity and Uncertainty of Surface - to - Surface Intersection (SSI)**:
- Computing the intersection line of two parameterized surfaces \( S_1(u, v) \) and \( S_2(s, t) \) involves solving a system of nonlinear equations:
\[
S_1(u, v)=(x_1(u, v), y_1(u, v), z_1(u, v))
\]
\[
S_2(s, t)=(x_2(s, t), y_2(s, t), z_2(s, t))
\]
- The results of these intersection line calculations are usually approximate, and different SSI algorithms may produce different approximate results, leading to model inconsistencies.
3. **Robustness and Interoperability Issues of Trimmed Models**:
- Trimmed models are prone to problems during the data exchange process between different software tools because different systems may use different data structures and robustness checking methods.
- The internal representation of trimmed models is not uniform, resulting in possible misunderstandings or errors during transmission and conversion.
### Solutions
The paper proposes two main strategies to address the above problems:
1. **Local Approaches**:
- Directly improve on the existing trimmed models so that they can better meet the requirements of numerical simulation.
- For example, handle the elements cut by the trimmed curves through special integration techniques, or enhance the stability of the model by weakly coupling adjacent surfaces.
2. **Global Approaches**:
- Convert the trimmed models into more regular representation forms, such as subdivision surfaces or T - splines, to eliminate the inconsistencies caused by trimming.
- This method aims to fundamentally solve the core problems of trimmed models, making the models more suitable for numerical simulation and other downstream applications.
### Summary
The paper emphasizes that these advances are not only of great significance for Isogeometric Analysis (IGA), but also bring new insights and technological progress to the CAGD field. Through these methods, the robustness and interoperability of trimmed models can be improved, thereby better meeting the needs of engineering design and numerical simulation.