Convergence
 

Hello,

I am running a multiphase VoF simulation in a very large domain, I have subtracted the shape of my body from this domain and am interested in the drag forces experienced by the subtracted body, it has been split using symmetry. I am using the turbulence model and open channel flow with velocity-inlet and pressure outlet.

I am running the transient simulations however my residuals are not converging to a particularly high degree and the drag results are continuing to fluctuate... this is using a time step of 0.005s and after 6.5seconds and over 3500 iterations... is this expected or a sign of something wrong?

Does anyone have any suggestions as how to improve this? It is for a deadline and i am now becoming very short on time as i have a total of 3 simulations to run.

Thanks in advance!

VoF
 

For a transient, VOF simulation, drag value will always oscillate depending upon the oscillations in the free-surface. Furthermore, you also need to decide about the reference values. If you are comparing against some available data, then you need to use similar reference values. However, if that is not the case, I'd suggest determining the ratio between drag force (not coefficient) on the submerged portion to that on the portion in the air. If the ratio is >> 1, then you can neglect the region above the water line, use properties of water and freestream velocity of water to determine the coefficient. Use only submerged area. And while reporting, report these values along with the coefficient for clarity. Since values are always oscillating, you have to use a time-averaged value.

Thanks, should the residuals be decreasing though? as when i run it through they tend to either remain similar all the way or even increase sometimes..

Is there a standard acceptable value for how close the results have to become? For example the percentage difference between this? I have tried to find some literature on this but no luck so far

Residuals
 

The values for the residuals should be much smaller than the absolute values for the fields, e.g., if velocity is of the order of 1 m/s, then residuals should be of the order of 0.001. Do note that the residuals are usually normalized and, even scaled. So, don't look at the residuals alone. If those are stalling, you can try change the gradient limiter from default to cell-to-cell under Solution Controls > Advanced.

Hi,
could you advise me on how to seperate air and water drag forces?
thanks