Choice of the turbulence model and mesh near walls
Posted: Mon Aug 15, 2016 6:04 pm
Hello,
I start to modelize an after heater furnace in 3D. It's a test case with simplified geometry and arbitrary input parameters.
The furnace is composed of a gas burner with high velocity (inlet), electrical heaters (modelize in syrthes with volumic heat source) and two outlets (atmospheric pressure). The hot fluid flux heat directly a solid charge in front of the inlet (See attached schematic 2D of the furnace)
In a first approximation, I neglect the natural convection, assuming that the forced convection is larger. I neglect also the gas radiation assuming that the gas enter in the domain at T_hot and v_in=100 m/s. My goal is to compute a steady stream in the furnace and calculation the temperature distribution using a frozen velocity and pressures fields (CS and syrthes coupled).
My question is about the choice of the turbulence model because I have, in the same model, high Reynolds (where the gas velocity is high in the inlet canal and on the solid charge) but I have also some recirculating fluid stream where the Reynolds number can be low (in the corners of the furnace).
I think that I have to use a k-w model (low Re number) where the mesh first cell close to the walls must be < y+=1. I have to adapt manually the boundary layer mesh in function of Reynolds number.
Have I another alternative less sensitive with the mesh first cell distance in the choice of the turbulence model ?
Best regards
Julien
I start to modelize an after heater furnace in 3D. It's a test case with simplified geometry and arbitrary input parameters.
The furnace is composed of a gas burner with high velocity (inlet), electrical heaters (modelize in syrthes with volumic heat source) and two outlets (atmospheric pressure). The hot fluid flux heat directly a solid charge in front of the inlet (See attached schematic 2D of the furnace)
In a first approximation, I neglect the natural convection, assuming that the forced convection is larger. I neglect also the gas radiation assuming that the gas enter in the domain at T_hot and v_in=100 m/s. My goal is to compute a steady stream in the furnace and calculation the temperature distribution using a frozen velocity and pressures fields (CS and syrthes coupled).
My question is about the choice of the turbulence model because I have, in the same model, high Reynolds (where the gas velocity is high in the inlet canal and on the solid charge) but I have also some recirculating fluid stream where the Reynolds number can be low (in the corners of the furnace).
I think that I have to use a k-w model (low Re number) where the mesh first cell close to the walls must be < y+=1. I have to adapt manually the boundary layer mesh in function of Reynolds number.
Have I another alternative less sensitive with the mesh first cell distance in the choice of the turbulence model ?
Best regards
Julien