Modelling blast furnace layered burden in ANSYS Fluent, convergence issue?
 

I am modelling layered burden in blast furnace for producing pig iron. I would like to know the effect of layer thickness in the heat transfer between ascending hot gas and descending alternating burden layers. According to User's and Theory Guide, I am using Eulerian Multiphase with k-epsilon tubulence model and enabled energy equation for the heat transfer calculation. I create the model as 2D axis-symmetry in steady state condition. After several trials like redefining the mesh, calculating all input values based on theory, I still find AMG solver divergence results mainly in mp x-momentum and pressure correction. When I change the pressure setting into selective and using CG stabilizer in AMG solver window, I will find AMG divergence in k calculation. All divergences happen in the first iteration.

Do I set the model wrong or is just the problem couldn't suit in Fluent? What should I do to get the convergent result? Thank you for reading this question and I am really looking forward to hearing your answers since I am literally confused what to do next.*