Το έργο με τίτλο Aerodynamic shape optimization of diffuser augmented wind turbine shrouds using asynchronous differential evolution από τον/τους δημιουργό/ούς Leloudas Stavros, Lygidakis Georgios, Strofylas Giorgos, Nikolos Ioannis διατίθεται με την άδεια Creative Commons Αναφορά Δημιουργού 4.0 Διεθνές
Βιβλιογραφική Αναφορά
S.N. Leloudas, G.N. Lygidakis, G.A. Strofylas and I.K. Nikolos, "Aerodynamic shape optimization of diffuser augmented wind turbine shrouds using asynchronous differential evolution," in ASME 2018 International Mechanical Engineering Congress and Exposition, 2018. doi: 10.1115/IMECE201886820
https://doi.org/10.1115/IMECE201886820
An adaptable numerical scheme for the aerodynamic shape optimization of axisymmetric diffuser-augmented wind turbine shrouds is demonstrated in this work, using an asynchronous and parallel version of a Differential Evolution (DE) algorithm. The simulation of the incompressible flow field about each candidate geometry is succeeded by means of an in-house Computational Fluid Dynamics (CFD) solver, that has been developed based on the specially modified, by the artificial compressibility approach, Navier-Stokes equations, expressed in non-dimensional form, for 2D-axisymmetric frames of reference. The discretization of the computational domain is made with 2D hybrid unstructured meshes, composed by both triangular and quadrilateral elements, combined with a node-centered finite-volume scheme, while the Free-Form Deformation (FFD) technique is applied, for both the parameterization of the design geometry and the morphing of the computational mesh. The required data transfer between the DE algorithm and the CFD solver is accomplished with appropriate text files, while the parallel implementation is achieved utilizing the Message Passing Interface (MPI) library functions. Further acceleration of the optimization procedure is succeeded by the combination of the DE with surrogate models, in order to replace the costly CFD-based evaluations of the candidate solutions with fast, but approximate estimations of their cost function.