| Summary: | Flowfield Dependent Variation (FDV) method is a mixed explicit-implicit numerical
scheme that was originally developed to solve complex flow problems through the use
of so-called implicitness parameters. These parameters determine the implicitness of
FDV method by evaluating local gradients of physical flow parameters, hence vary
across the computational domain. The method has been used successfully in solving
wide range of flow problems. However it has only been applied to problems where the
objects or obstacles are static relative to the flow. Since FDV method has been proved
to be able to solve many complex flow problems, there is a need to extend FDV
method into the application of moving boundary problems where an object
experiences motion and deformation in the flow. With the main objective to develop a
robust numerical scheme that is applicable for wide range of flow problems involving
moving boundaries, in this study, FDV method was combined with a body
interpolation technique called Arbitrary Lagrangian-Eulerian (ALE) method. The
ALE method is a technique that combines Lagrangian and Eulerian descriptions of a
continuum in one numerical scheme, which then enables a computational mesh to
follow the moving structures in an arbitrary movement while the fluid is still seen in a
Eulerian manner. The new scheme, which is named as ALE-FDV method, is
formulated using finite volume method in order to give flexibility in dealing with
complicated geometries and freedom of choice of either structured or unstructured
mesh. The method is found to be conditionally stable because its stability is dependent
on the FDV parameters. The formulation yields a sparse matrix that can be solved by
using any iterative algorithm. Several benchmark stationary and moving body
problems in one, two and three-dimensional inviscid and viscous flows have been
selected to validate the method. Good agreement with available experimental and
numerical results from the published literature has been obtained. This shows that the
ALE-FDV has great potential for solving a wide range of complex flow problems
involving moving bodies.
|
|---|