Hi, everyone,

How to deal with boundary conditions in the Smoothed particle hydrodynamics ?

Thank you

Hi THere,

For open boundaries there is no problem, you can just assume that once a 'particle' is far from you computational domain then it has to be removed. In regions where the number of particles becomes fairly small, the flow is not represented accurately and the results are much less accurate (regions of low density), that's another problem you have to take into account.

For rigid boundaries it is more tricky and I have heard of several ways of implementing the rigid boundary conditions. The simplest thing is just to 'fix' particles at the location of the boundary and not allowing them to move Another method is to put a potential barrier that the particles would not be able to 'cross'.

I am sure if you make a litterature search for Joe J. Monaghan, you should find enough papers to answer all you questions. Since your question is quite general it is difficult to go into more details.

I hope this helps. Patrick.

Hi, I found one: Monaghan, JJ (1994), "Simulating Free Surface Flows with SPH", J of Computational Physics, V110, pp.399-406.