Name: CAIO PATRICK PICOLI DE LIMA
Publication date: 21/02/2025
Examining board:
Name |
Role |
|---|---|
| HERMES VAZZOLER JUNIOR | Examinador Externo |
| JUAN SERGIO ROMERO SAENZ | Presidente |
| MARCUS VINICIUS CANHOTO ALVES | Examinador Externo |
| RAMON SILVA MARTINS | Examinador Interno |
Summary: Optimization is a well-known topic in popular imagination. Regarding engineering, geometry optimization is possibly the most widespread topic to the public, especially optimization of aircraft wings, and, as an example of this topic, there is the work of Muller (2021), that, through optimization, almost halved drag on the wing. Another interesting work was conducted by Zheng et al. (2021) in the optimization of the cooling system of a
photovoltaic panel, in order to maximize its efficiency. Amongst all optimization methods, the one chosen for this work is the Topological Optimization Method, since it does not depend on a previous knowledge of the optimal geometry. Originally, this method was developed using the Finite Elements Method, which has limitations on flow solution. These limitations are overcome by using the Finite Volumes Method, although it has limitations that prevent its use on the Topological Optimization Method. The objective of this work is the development of the tool created by Lima, Aguiar e Romero (2024), which uses the Finite Volume Method and the Finite Element Method together for Topological Optimization, so that it optimizes turbulent flows. The solution of the flow was done using the software OpenFOAM®, the geometry optimization through custom code executed on Octave, and the automatization of this process was done by Bash Scripting. Once the optimization tests of a nozzle, a T bend, and a 90° knee had been conducted, qualitative analyses with the available literature were conducted. The results obtained were qualitatively identical to the reference geometries with smaller meshes, which shows that the tool is functional, although there is still space for improvement. Furthermore, the geometries resulting from the optimization of turbulent flows had a smaller pressure drop when compared to the geometries generated from the optimization of laminar flows.
