typical moving body problem

Rizwan

i am solving a moving sphere in a pipe flow. i followed different approaches. the physics of the problem is sphere moves along iwth the flow. the size of sphere is more than a computational cell and hence cant be modeled using dpm, the volume occupied by the sphere is not negligible in 1st place. the sphere's velocity is as a result of the shear forces comming on its surface and it doesnot

1. creating a separate geometry for pipe and sphere each of them and meshing the pipe after substracting the volume of sphere from that of pipe. but once the sphere moves at the next tym step, what has to be done with the empty region is first question. dynamic mesh wil not work as its a multi phase flow, need not be only single phase flow of liquid there may be two or more liquid phases present.

2. without substracting the volume, if only mesh is substracted then there it cudnot be done in gambit leading to sequece of errors.

3. using udf's. i tried to begin with udf's using adjust function by assigneing 0 velcity for all components in the cells represented by sphere region (as a begining case i dint consider about parly filled cells with water and solid sphere) but the flow passes through the sphere too irrespective of the usage of interpreted and hooked adjust function.

has anyone experience worked on such a problem? any help in this regard is highly appreciated.

Thanks Rizwan