Re: Heat transfer coefficient
Posted: 04 Oct 2010, 08:08
Hi Bill,
yes well, there are two domains 'left' and 'right' with a shared boundary in the middle. Now the
internal middle boundary is defined doubled in .grd file (the fourth number in the boundary definition
is 2). Both boundaries are written with index 1 to the mesh.boundary file.
The sif file defintion 'HeatGap=True' in the BC 1 instructs ElmerSolver to look for this double
boundary and setup heat transfer between the boundaries using the "Heat Transfer Coefficient",
i.e.
Otherwise the two domains are independent due to the doubled boundary.
Hope it helps, juha
P.S. You can (at least try) to make existing mesh have doubled boundaries using
where the 'boundary' is the mesh boundary index you want doubled. The option may
be given several times for different boundaries. Don't know how well it'll work out
in corners crossing different doubled boundaries though, maybe it will?
yes well, there are two domains 'left' and 'right' with a shared boundary in the middle. Now the
internal middle boundary is defined doubled in .grd file (the fourth number in the boundary definition
is 2). Both boundaries are written with index 1 to the mesh.boundary file.
The sif file defintion 'HeatGap=True' in the BC 1 instructs ElmerSolver to look for this double
boundary and setup heat transfer between the boundaries using the "Heat Transfer Coefficient",
i.e.
Code: Select all
flux_L = HTC*(T_L- T_R)
flux_R = HTC*(T_R - T_L)
Hope it helps, juha
P.S. You can (at least try) to make existing mesh have doubled boundaries using
Code: Select all
ElmerGrid 2 2 mesh -discont boundary
be given several times for different boundaries. Don't know how well it'll work out
in corners crossing different doubled boundaries though, maybe it will?