[davibarreira]

I'm recently trying to validate the twoPhaseEulerFoam solver (OpenFOAM 2.3.x) with the results found in the article "CFD simulation of large-scale bubble plumes: Comparisons against experiments" (Dhotre and Smith, 2007). The simulation is a simple injection of air in an unconfined 2m diameter tank filled up to 1.1m of water.

I've been trying to simulate this experiment for a while now, but the twoPhaseEulerFoam solver has completely failed (which seems odd as the solver has been validated for bubble columns, that are fairly similar to what I'm trying to reproduce).

Basically, it is expected that the void fraction (alpha.air) will follow a Gaussian curve, but this is not happening and the results actually get worst with mesh refinement.

My model is using axial symmetry and I've used many phase properties (such as different lift models and turbulentDispersion), but still no success.

Im posting here my case, some pictures of the results I've been getting and on of the experimental results from the article.



P.S: The results are from a horizontal sample obtained at 75cm above the inlet.The y-axis is the alpha.airMean and the x-axis is the distance from the center of the inlet.