Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Proceedings
ISSN 1759-3433 CCP: 89
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY Edited by: M. Papadrakakis and B.H.V. Topping
Paper 7
Fluid-Structure Coupling for Large Simulations Using Radial Basis Functions T.C.S. Rendall and C.B. Allen
Department of Aerospace Engineering, University of Bristol, United Kingdom T.C.S. Rendall, C.B. Allen, "Fluid-Structure Coupling for Large Simulations Using Radial Basis Functions", in M. Papadrakakis, B.H.V. Topping, (Editors), "Proceedings of the Sixth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 7, 2008. doi:10.4203/ccp.89.7
Keywords: radial basis function coupling, fluid-structure interpolation, information transfer, compact functions, aeroelasticity.
Summary
Meshless fluid-structure coupling with radial basis functions is performed for a large case using a partition of unity for each fluid surface point. This partitioning keeps the bandwidth of the coupling matrix (and therefore storage cost and multiplication time) to a minimum but retains a smooth fluid surface mesh suitable for a computational analysis. The method is used to couple an eight million cell finite volume mesh to a simple structural model to find the static aeroelastic deflection of a typical transport aircraft wing, showing that the partitioned method achieves an acceptable result with a maximum matrix bandwidth that is less than 5% of the full method. Since the partitioned method is inherently local, a local force transfer is also guaranteed which is physically more logical than the global influence of the full method.
For the eight million cell mesh studied a reduction in matrix size of 58x was achieved. Since the only expense that scales with the total size of the structural grid varies as N ln(N) the method may also be used on much larger structural grids at little additional cost. purchase the full-text of this paper (price £20)
go to the previous paper |
|