The case setup of Sandia Flame D

[banyangtju]

The mean field i calculated cannont fit the experimental data.

During my simulations, the temperature field in 0/T has been updated to the experimaental setup. The term USE_TURBULENCE_GENERATOR in 0/U is also changed to 1 to open the turbulent inlet. However, the mean filed i get cannot agree with the experiment.

[moreff]

Did the profile converge? It might be so that you need to run the simulation for a longer time.
Also, make sure that you start averaging after the initial phase, i.e. when the flow is statistically steady

[banyangtju]

Thanks for your reply. The profile was conerge and no divergence was reported during the running of OpenFOAM. The time starting averaging is 0.1 s, consistent with the setup in your paper. In fact, expect for the files 0/T and 0/U, other setup was consistent with the case you provided here.

[moreff]

I mean not the convergence of an OpenFOAM matrix solver, I am talking about the convergence of the statistics that you are gathering. You need to gather it long enough so that the profiles you are getting do not change anymore. If I now remember correctly, I had to run this simulation until 0.27s or even longer to see that the statistics were not changing significantly anymore

[banyangtju]

Thanks for your patient reply. A longer time i wiil try.