Unified structural optimization method using topology optimization and genetic algorithms
This paper presents a new structural design framework that incorporates the concept of topology optimization and genetic algorithms to improve the manufacturability and structural robustness of the optimal structure. The level set function is employed as a topological design variable to obtain clear...
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | article |
| Lenguaje: | EN |
| Publicado: |
The Japan Society of Mechanical Engineers
2021
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/269ce59711954b6e957711f9e5cbfa6f |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: | This paper presents a new structural design framework that incorporates the concept of topology optimization and genetic algorithms to improve the manufacturability and structural robustness of the optimal structure. The level set function is employed as a topological design variable to obtain clear structural boundaries, and the manufacturability of the structure is mathematically defined based on the manufacturing directions and the fictitious heat fluxes. To gain the manufacturable structure design, the optimization problem is formulated to find both the optimal shape of the structure and the optimal directions of the adjustable manufacturing tools. A level set-based optimization regarding manufacturability has been studied in previous papers, however, due to the influence of the manufacturing directions, the objective value tends to be captured in local optima as the structure becomes more complex. To cope with this issue, we decided to adapt a heuristic based approach, genetic algorithm, to the optimization method. Simultaneously, to reduce the computation time, we applied Design of Experiments for the initial population of the genetic algorithm. The initial population of the manufacturing directions is installed using Latin Hypercube sampling for both a good representation and computational efficiency. To demonstrate the effectiveness of the proposed method, several design examples are provided, and the differences from the optimal solution, derived by a previous gradient-based optimization scheme, are mentioned. |
|---|