Learning Forum

[Krasa]

I am working conjugate heat transfer of channel which consist of two fins and I applied vibration to channel to investigate the effect of vibration on convective cooling. I used structured mesh for channel and unstructured mesh fluid domain ( due to the use of dynamic mesh).n I am getting an error in my simulation. after some iterations I am getting AMG solver divergence - temperature error. nn I have tried with reducing time step and well as reducing relaxation factors, but error is still coming. I think problem is my mesh. because I have higher skewness in my mesh. nnHere I have attached my geometry as well as my existing mesh. could you please suggest me a method to obtain a fine mesh and run my simulation without any error?n

[System_delete]

This discussion was created from comments split from: Divergence detected in AMG solver and floating point exception in ANSYS Fluent.

[Rob]

Looking at the mesh, is the fluid in the lighter grey tet volume? If it is, you need to do the tutorials on meshing and the Fluent solver, paying special attention to mesh resolution: 1-2 cells across a gap is too coarse and also look at leaving space before & after the obstruction. n

[Krasa]

fluid is outer domain and heat sink is inside the fluid domain. I have changed my mesh like bellow. nn

[Rob]

And how does that compare with this

?n

[Krasa]

actually they used cutcell meshing method but I want unstructured mesh for fluid domain due to the smoothing and remeshing settings which I used in dynamic mesh.n

[Rob]

And the domain size? n

[Krasa]

my fluid domain is 112 x 31 x 37 mm and inside that domain I have heat sink with dimension of 100 x 15 x 32 mmn

[Rob]

And compare the amount of space around the heat sink in the tutorial. In a flow experiment how close would you put the item of interest to the inlet/outlet? n