[zryan civil]

dear all
how to plot depth of water in the free surface model at 90% air concentration, I used contour with polyline at contour No. 2, is that right??
thanks

[ghorrocks]

You have not described what you are trying to do or how you have done it. So I have no idea if it is correct or not.

[zryan civil]

Originally Posted by ghorrocks

You have not described what you are trying to do or how you have done it. So I have no idea if it is correct or not.

OK, for a channel I create a section along the center (user specified location) to draw water depth with x (distance), then create contour of volume fraction of water for that location, and then add polyline with intersect with contour at (contour number =2), then add chart draw y versus x, for polyline. is that y (water depth) is at 90% air concentration.

[ghorrocks]

OK, and so what are you asking if it correct or not?

[zryan civil]

Originally Posted by ghorrocks

OK, and so what are you asking if it correct or not?

Iam asking about is this depth of water that obtained from (contour no. =2) is exactly coresponding to the water depth at 90% air concentration. similarly contour no. 5 produce depth of water at 50 air concentration. is that right or not

[ghorrocks]

In a free surface model with the homogenous model, the width of the transition from 0% to 100% volume fraction is controlled by mesh size and the accuracy of the numerical model. The mathematical model it is trying to use has the surface infinitesimally thin. So the smearing you see of the volume fraction is just numerical error and nothing real.

[zryan civil]

Originally Posted by ghorrocks

In a free surface model with the homogenous model, the width of the transition from 0% to 100% volume fraction is controlled by mesh size and the accuracy of the numerical model. The mathematical model it is trying to use has the surface infinitesimally thin. So the smearing you see of the volume fraction is just numerical error and nothing real.

thanks, as I understood from your reply as transition zone increase it may be related to mesh size, so two question arise:
1- mesh adaption at free surface can treat this?
2- at some location like high air entrainment (jump) even with small size mesh high transition zone was observed, so in this case using DNS or LES can treat this, rather than using RANS
thanks

[ghorrocks]

1) Yes, a finer mesh will result in a better resolved surface
2) I am guessing here as I have not seen details of your application, but in my estimation if you have a feature which is entraining lots of air such as a hydraulic jump, and you are using a homogenous free surface model - then the most important thing is to have a mesh fine enough to resolve all the air bubbles being entrained. This is going to be much more important than the turbulence model.