Dear CFX users,

What is the ideal way to Simulate only surface-to-surface HT simulation wihout solving the Convection?

I am trying to benchmark simulated result with theoretically exact calculation for a cube (100 x 100 x 100) mm^3. The two opposite faces are at uniform temperatures 300 [K] and 400 [K]. Remaining 4 walls are black and refractory (re-radiating).

The air was used as dummy fluid (ideal gas at very low pressure 0.1 [bar]) with all 3 components of velocity = 0, thermal conductivity very very low (~ 0.000026 [W/m-K] to prevent any HT to the fluid.

Unfortunately, the run is no way directed towards convergence!!

Any insight would be highly appreciated!

Thanks,

Amod

Hi Amod,

I hope you have a 3D mesh within your box... You can define the air as "solid" and run Monte Carlo s-t-s simulation or, as you've already done, define the domain as dummy "fluid/gas" with low thermal conductivity and zero velocity and use the Discrete Transfer radiation type.

Define the uniform temperatures as boundary condition of the appropriate surface and set initial conditions as automatic. Your mesh doesn't have to be too fine. I suggest a mesh element size of (2x2x2)mm^3. Try first (4x4x4)mm^3 and check your results.

I guess you want to simulate a steady-state situation with thermal equilibrium, right? What is exactly the goal of your simulation?

Hope this works! Post a feedback in any case...

Zoran

Dear Zoran,

Thank you for your reply. ------------------------------------------------------- Actually, I have to ultimately carry out a conjugate heat transfer simulation in automotive Exhaust Manifold where the combustion products (the fluid media) is participating gas.

However, I have not done any radiation problem in CFX earlier. As a Validation and Verification (V&V) test case, I tried to carry out a pure radiation problem to match the simulated result with theoretically exact solution. -------------------------------------------------------- I did follow the method described by you i.e. modeling air as low thermal conductivity medium, zero velocity and Discrete Ordinate Transfer Model. However, "H-Energy" equation is not getting converged. After initial 50 iteration or so, it becomes absolutely flat at RMS residula ~ 0.01. I was wondering whether any setting in "Expert Parameter" would help.

Thanks,

Amod Kumar

Hi Amod,

Set in the default domain under "Fluid Models > Heat Transfer > Option": Thermal Energy With this the "H-Energy" is not solved which is OK since your velocity is zero...

Zoran

Dear Zoran,

The "H-Energy" is solved with "Thermal Energy" option! In case I set "Fluid Models > Heat Transfer > None", "Fluid Models > Thermal Radiation Model" also becomes "none" and it cannot be set to any other value.

In case you share your e-mail ID, I may be able to send you snapshots of the Convergence and Residual Plots.

I thank you again for your time and attention.

Regards,

Amod Kumar amod.kumar@mindid.com

ok