Shape optimization and sensitivity can now be considered a standard to design efficient ships and new unconventional hull forms. There might be very different strategies to achieve such a process, depending on many circumstances, both operational (e.g. available time windows for computations) and methodological (e.g. available numerical methods). As in most of the engineering cases, the best trade-off between accuracy of the solution and the time required to achieve it is searched. When using medium-fidelity methods e.g. Boundary Element Methods (BEMs) for wave resistance and seakeeping, population based optimization algorithm can still be a viable way to achieve a design solution (see e.g. Vernengo et al., 2015 or Vernengo and Brizzolara, 2017). However, when the solution of the quantity of interest becomes more demanding, alternative methods of searching through the design space are needed. In this perspective, a surrogate model based approach for hull form sensitivity analysis is presented. The method relies on the features of a Kriging response surface (see for instance Forrester et al., 2008) to interpolate few computed solutions and to predict the same solution over the whole explored domain. The sensitivity analysis focuses on the effect of hull form variations with respect to calm water resistance at a given forward speed. The total ship resistance is computed by means of a high-fidelity viscous solver based on the openFOAM libraries (Jasak et al., 2007). The shape variations are achieved by a combined approach specifically developed to preserve the fairness of the hull surface based on Subdivision Surface and Free Form Deformation (FFD) (Coppedé et al., 2018). The hydrodynamic solver has been preliminary validated by comparison against available experimental measurements on the KRISO Container Ship (KCS) hull. The variation of the Kriging response surface performance with respect to different sizes of the initial sampling have been studied and possible optimum hull shape have been detected and compared in terms of calm water resistance, wave patterns and pressure on the surface.
Surrogate Model for Ship Resistance : a Sensitivity Analysis of Shape Deformation
Coppedè Antonio;Gaggero Stefano;Vernengo Giuliano;Villa Diego
2018-01-01
Abstract
Shape optimization and sensitivity can now be considered a standard to design efficient ships and new unconventional hull forms. There might be very different strategies to achieve such a process, depending on many circumstances, both operational (e.g. available time windows for computations) and methodological (e.g. available numerical methods). As in most of the engineering cases, the best trade-off between accuracy of the solution and the time required to achieve it is searched. When using medium-fidelity methods e.g. Boundary Element Methods (BEMs) for wave resistance and seakeeping, population based optimization algorithm can still be a viable way to achieve a design solution (see e.g. Vernengo et al., 2015 or Vernengo and Brizzolara, 2017). However, when the solution of the quantity of interest becomes more demanding, alternative methods of searching through the design space are needed. In this perspective, a surrogate model based approach for hull form sensitivity analysis is presented. The method relies on the features of a Kriging response surface (see for instance Forrester et al., 2008) to interpolate few computed solutions and to predict the same solution over the whole explored domain. The sensitivity analysis focuses on the effect of hull form variations with respect to calm water resistance at a given forward speed. The total ship resistance is computed by means of a high-fidelity viscous solver based on the openFOAM libraries (Jasak et al., 2007). The shape variations are achieved by a combined approach specifically developed to preserve the fairness of the hull surface based on Subdivision Surface and Free Form Deformation (FFD) (Coppedé et al., 2018). The hydrodynamic solver has been preliminary validated by comparison against available experimental measurements on the KRISO Container Ship (KCS) hull. The variation of the Kriging response surface performance with respect to different sizes of the initial sampling have been studied and possible optimum hull shape have been detected and compared in terms of calm water resistance, wave patterns and pressure on the surface.File | Dimensione | Formato | |
---|---|---|---|
NuTTS_Coppede.pdf
accesso aperto
Tipologia:
Documento in Post-print
Dimensione
1.65 MB
Formato
Adobe PDF
|
1.65 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.