Hole Seeding in Level Set Topology Optimization via Density Fields

Jorge L. Barrera,Markus J. Geiss,Kurt Maute

DOI: https://doi.org/10.48550/arXiv.1909.10703

2019-09-24

Abstract:Two approaches that use a density field for seeding holes in level set topology optimization are proposed. In these approaches, the level set field describes the material-void interface while the density field describes the material distribution within the material phase. Both fields are optimized simultaneously by coupling them through either a single abstract design variable field or a penalty term introduced into the objective function. These approaches eliminate drawbacks of level set topology optimization methods that rely on seeding the initial design domain with a large number of holes. Instead, the proposed approaches insert holes during the optimization process where beneficial. The dependency of the optimization results on the initial hole pattern is reduced, and the computational costs are lowered by keeping the number of elements intersected by the material interface at a minimum. In comparison to level set methods that use topological derivatives to seed small holes at distinct steps in the optimization process, the proposed approaches introduce holes continuously during the optimization process, with the hole size and shape being optimized for the particular design problem. The proposed approaches are studied using the extended finite element method for spatial discretization, and the solid isotropic material with penalization for material interpolation using fictitious densities. Their robustness with respect to algorithmic parameters, dependency on the density penalization, and performance are examined through 2D and 3D benchmark linear elastic numerical examples, and a geometrically complex mass minimization with stress constraint design problem.

Optimization and Control,Computational Engineering, Finance, and Science

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to effectively generate holes dynamically during the design process in topological optimization. Specifically, the traditional level - set method has difficulty in automatically generating holes during the optimization process. Usually, a large number of holes need to be pre - arranged in the design domain, which not only increases the computational cost but also may lead to excessive dependence of the design result on the initial hole pattern. In addition, if the initial hole pattern is inappropriate, it may lead to sub - optimal design or instability problems during the optimization process.

To solve these problems, this paper proposes two density - field - based methods to dynamically generate holes during the level - set topological optimization process. These two methods couple the level - set field and the density field, so that holes can be generated at appropriate positions and shapes as needed during the optimization process without pre - arranging a large number of holes. This can not only reduce the computational cost but also improve the robustness and performance of the optimization results.

### Main Contributions
1. **Dynamically Generate Holes**: The proposed methods can dynamically generate holes during the optimization process instead of relying on preset hole patterns.
2. **Reduce Computational Cost**: By reducing the number of initial holes, the computational complexity and cost can be reduced.
3. **Improve Robustness**: The dependence of the optimization result on the initial hole pattern is reduced, and the stability of the optimization process and the quality of the final design are improved.

### Method Overview
- **Single - Field Coupling (SFC)**: Use a single design variable field to control the level - set field and the density field simultaneously, and connect the two through a linear relationship.
- **Two - Field Coupling (TFC)**: Use two independent design variable fields to control the level - set field and the density field respectively, and use a penalty term to promote the generation of negative level - set values in low - density regions, thereby generating holes.

### Mathematical Expressions
- **Definition of Level - Set Field and Density Field**
\[
\phi(X)=\begin{cases}
> 0, & \forall X\in\Omega_I\\
< 0, & \forall X\in\Omega_{II}\\
= 0, & \forall X\in\Gamma_{I,II}
\end{cases}
\]
where $\Omega_I$ and $\Omega_{II}$ represent the material and void regions respectively, and $\Gamma_{I,II}$ represents the material - void interface.

- **Definition of Density Field**
\[
\rho(X)=\begin{cases}
\frac{\hat{S}(X)-\phi_{sh}}{1 - \phi_{sh}}, & \forall X:\phi(X)\geq0\\
\text{undefined}, & \forall X:\phi(X)< 0
\end{cases}
\]
where $\phi_{sh}$ is a threshold used to distinguish between materials and voids.

- **Penalty Term for Two - Field Coupling**
\[
\bar{p}_{\rho\phi}(X)=\begin{cases}
0, & \forall\rho\geq\rho_{th}\\
\max\left(0,\frac{\phi-\phi_{th}}{\phi_{up}-\phi_{th}}\right), & \forall\rho<\rho_{th}
\end{cases}
\]
where $\rho_{th}$ and $\phi_{th}$ are the density and level - set thresholds respectively.

### Conclusion
The two methods proposed in this paper can effectively solve the shortcomings of the traditional level - set topological optimization method in generating holes. By dynamically generating holes, the robustness and performance of the optimization results are improved. These methods have been verified in 2D and 3D linear elastic problems and have shown their application potential in geometrically complex design problems.