Shape design optimization of SPH fluid---structure interactions considering geometrically exact interfaces

  • Authors:

  • Youn Doh Ha;Min-Geun Kim;Hyun-Seok Kim;Seonho Cho

  • Affiliations:

  • Department of Naval Architecture, Kunsan National University, Kunsan, South Korea 573-701;National Creative Research Initiatives Center for Isogeometric Optimal Design and Department of Naval Architecture and Ocean Engineering, Seoul National University, Seoul, South Korea 151-744;National Creative Research Initiatives Center for Isogeometric Optimal Design and Department of Naval Architecture and Ocean Engineering, Seoul National University, Seoul, South Korea 151-744;National Creative Research Initiatives Center for Isogeometric Optimal Design and Department of Naval Architecture and Ocean Engineering, Seoul National University, Seoul, South Korea 151-744

  • Venue:
  • Structural and Multidisciplinary Optimization
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Fluid---structure interaction problems are solved by applying a smoothed particle hydrodynamics method to a weakly compressible Navier---Stokes equation as well as an equilibrium equation for geometrically nonlinear structures in updated Lagrangian formulation. The geometrically exact interface, consisting of B-spline basis functions and the corresponding control points, includes the high order geometric information such as tangent, normal, and curvature. The exactness of interface is kept by updating the control points according to the kinematics obtained from response analysis. Under the scheme of explicit time integration and updated Lagrangian formulation, the required shape design velocity should be updated at every single step. The update scheme of design velocity is developed using the sensitivity of physical velocity. The developed sensitivity analysis method is further utilized in gradient-based shape optimization problems and turns out to be very efficient since the interaction pairs of particles determined in the response analysis can be directly utilized.