Learning Forum

[Luigi0]

Hi,I have a problem with discrete phase modeling in a coupled analysis (Two way FSI analisys):Particles lost outside the domain while spanning over partitions. It's strange because before running the coupled analysis, I run an analysis only in fluent with the same settings and particles are tracked in a right way.nThe fluid domain is a parallelepiped, with one inlet and one outlet and the flow is laminar. There is a single injection and the particles are modeled as uncoupled dpm. nAre there special settings that I have to check before modeling dpm in a coupled analysis? nThanksn

[Rob]

How much does the mesh change during the structural part? Given the way particles are handled I can see some being lost if the geometry is no longer present following an update. n

[Luigi0]

Thanks Rob for your interest. nI don't understand when you say how much does the mesh change during the strucutral part, could you clarify it? Thanks. However, I try with a different injection type: I choose release from surface and it works, but for my purpose I need single or group injection.n

[Rob]

You said you were doing FSI so that implies some change in geometry. nFor point injections etc make sure they're a cell or so into the fluid domain, with scattering & numerical errors it's very easy to have an injection that's just outside the fluid domain. n

[Luigi0]

Ok, the change in geometry is very small, the displacements sent from structural to fluent are in the order of nanometers. nSelecting release from surface, it works, but I have 10 cells on the inlet, so I have 10 streams. Is it possible to release particles only from one or two cells from the inlet for example?n