Dissertação
Desenvolvimento de turbinas eólicas com geometria bio-inspirada e uso de perfis aerodinâmicos gerados por métodos heurísticos
Fecha
2022-03-25Autor
Christiano Pagung Neto
Institución
Resumen
The use of renewable energy has gained importance in recent years. One of these energy sources
that is also growing is wind energy. The use of this source is extremely important for society
both in the environmental and in the economic sphere. There are two approaches to improving
gains for wind turbines: location and turbine design. This work proposes to expand on the second
approach. For this, three geometry improvement techniques will be used in order to obtain an
optimal geometry. Initially, a commercial wind turbine is chosen to define the flow parameters.
Then, using the blade element momentum method (BEMM) a primary geometry will be created
using the best NACAs for each blade section. Soon after, the sweep from selected seeds is
introduced in each geometry, and a second geometry is calculated. The use of this sweep will be
analyzed along the length of the blade, because, despite reducing some forms of drag, it also
reduces lift. Finally, using the genetic algorithm method, profile geometries for each section will
be created. With these geometries in hand, it is possible to create final turbine geometries based
on each studied seed. The final results demonstrate a significant power gain with the use of this
methodology. For the best case, the gain is almost double the power for the same flow conditions
as the commercial turbine. After the theoretical results, validation takes place using commercial
computational fluid dynamics (CFD) software. The CFD results obtained help to corroborate the
results, even considering a certain final difference in potency. This difference can be attributed
to the method employed of sweep compensation within BEMM or other three-dimensional
aerodynamic phenomena not considered. In the end, it can be concluded that the final geometries
actually present a significant power gain. Even so, further validation studies should be carried
out as some dimensions of analysis were not considered in this method.