Institutional Repository
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Development of computational tools, based on radial basis functions anddifferential evolution, for the parametric design of aeroelastic systems

Strofylas Giorgos

Full record


URI: http://purl.tuc.gr/dl/dias/D12E70A2-6C67-4678-A629-07A55AD5CBFD
Year 2021
Type of Item Doctoral Dissertation
License
Details
Bibliographic Citation Giorgos Strofylas, "Development of computational tools, based on radial basis functions and differential evolution, for the parametric design of aeroelastic systems", Doctoral Dissertation, School of Production Engineering and Management, Technical University of Crete, Chania, Greece, 2021 https://doi.org/10.26233/heallink.tuc.89051
Appears in Collections

Summary

In this study the development and validation of methodologies and computational tools, allowing for the effective design, optimization and numerical simulation of aeroelastic systems is reported. Specifically, a software tool named "T4T" (Tools for Turbomachinery) is developed for the parametric design of the external surface and the internal geometry of horizontal-axis wind turbine blades, which is fully parametric and customizable, allowing the user for defining the internal blade structure, including shear webs. Moreover, the software can be used in an automated way (batch mode) to produce several candidate geometries in an optimization cycle, while retaining its topology unchanged. Regarding the use of an optimization methodology in the aforementioned wind turbine design loop, a parallel, synchronous/asynchronous, metamodel-assisted Differential Evolution algorithm is developed. Subsequently, a specialized surface reconstruction methodology is implemented, for the geometry definition of a wind turbine blade as a single Non-Uniform Rational B-Spline (NURBS) surface, from a target set of data points provided in the form of a surface triangular mesh. For the parameterization of the blade surface the dedicated blade geometry modelling software "T4T" is used. The shape reconstruction of the blade surface is formulated as an optimization procedure, which is realized with the aforementioned Differential Evolution algorithm. At next, a methodology for the deformation of computational grids, based on Radial Basis Functions (RBFs), is reported. Additionally, an approach for the acceleration of (RBF)-based mesh deformation procedure via the reduction of the surface points is developed, considering agglomeration of surface nodes’ control volumes. It relies on the strategy followed by the corresponding multigrid methods, aiming to accelerate numerical solutions of fluid flow, radiative heat transfer, etc. Finally, a partitioned Fluid-Structure Interaction (FSI) methodology is developed, based on the RBFs Partition of Unity (PoU) method. With this approach the conservation of energy, momentum and force is ensured over the interface of the flow and structural grids as a result of the radial functions’ properties. The use of the PoU methodology improves the efficiency of the data transfer procedure providing simultaneously a physical formulation of the force distribution.

Available Files

Services

Statistics