[panos_metal]

Hello cfd users

I simulate the air flow in a heater. The wall temperature is approximately 1200 oC. The height is 3 meters and the diameter of the tube is 13 cm. The air flow range from 10 to 220 lit/min. i simulated the procedure in fluent. I would like to use the density of the air in correlation to the air temperature. Can i use the boussinesq approximation? i have read that boussinesq approximation is not appropriate when the temperature differences are large. But this refers to the neighboor cells or to the Toutlet-Tinlet air temperature?

Thank you in advance

[CFDtoy]

Hi there,
Boussinessq goes : rho_modified = rho_ref*(1+beta*deltat). Now, if you have large deltat (Tcell - Treference with which fluid properties are defined)...beta*deltat starts to go close to 1...which means its no longer a small change in rho_ref in a given cell. thats why they indicate u cant use bouss at high Tdiff.

For your case, twall ~ 1200 C - pretty high..yes..but you could definitely try boussinessq approx here. If you see oscillations in the solution, probably you can gain more insight.

/CFDtoy
Visit http://cfdtoy.blogspot.com

Originally Posted by panos_metal

Hello cfd users

I simulate the air flow in a heater. The wall temperature is approximately 1200 oC. The height is 3 meters and the diameter of the tube is 13 cm. The air flow range from 10 to 220 lit/min. i simulated the procedure in fluent. I would like to use the density of the air in correlation to the air temperature. Can i use the boussinesq approximation? i have read that boussinesq approximation is not appropriate when the temperature differences are large. But this refers to the neighboor cells or to the Toutlet-Tinlet air temperature?

Thank you in advance