[WSc]

Hallo everybody,

I want to do an open-channel simulation with shallowWaterFoam.
How can I define the bed-roughness? Or do I have to model it somehow?

Regards,
Wolfgang

[dai_bach]

Hi,

This reply may come far too late for the original post but I hope it may help other openfoam users. I've been trying to model certain classes of free-surface gravity currents for which the horizontal length scale (x and/or y) is much larger than the vertical scale (z). In this framework, vertical accelerations are small compared to gravity, and the conservation equations can be depth averaged which leads to what are often called the shallow water equations. For this reason, I started to investigate the shallowWaterFoam solver. However, as is described in

http://www.tfd.chalmers.se/~hani/kurser/OS_CFD_2010/johanPilqvist/johanPilqvistReport.pdf

this solver has been written for atmospheric/oceanic flows where the depth (z) is large compared to the Earth's radius. Furthermore, the top and bottom stresses are assumed to be zero. Hence, as is, this solver cannot be used to treat the kind of problem that I've described above. However, with modifications to the equation set in shallowWaterFoam.C, the same files can be used to tackle these classes of problems via a finite volume method.

[yhaomin2007]

I think you can just use a rough wall function. You may look it up in turbulence model wall function source file.

[dai_bach]

That would mean that the "rough wall" would be one of the "frontAndBack faces". I don't think that would work as the domain is only one-cell thick in the depthwise direction, but I could be wrong.

[haze_1986]

you are right. it would not work since it is only one cell thick. In depth averaged models the wind and bed stresses act as souce and sink to the momentum equations. I am currently working on this.

[andyru]

hi,
for bottom roughness you just have to impose an additional source. I did that already and that works brilliantly. It is easy:

...=lambda*hU

Best,

andy

[haze_1986]

Originally Posted by andyru

hi,
for bottom roughness you just have to impose an additional source. I did that already and that works brilliantly. It is easy:

...=lambda*hU

Best,

andy

Hi Andy, did you include this in the hUEqn, momentumPredictor or inside the piso loop?

[andyru]

According to Jasak PhD http://powerlab.fsb.hr/ped/kturbo/Op...jeJasakPhD.pdf
I added it to momentum predictor as well as in piso-loop.

[haze_1986]

Originally Posted by andyru

According to Jasak PhD http://powerlab.fsb.hr/ped/kturbo/Op...jeJasakPhD.pdf
I added it to momentum predictor as well as in piso-loop.

Can you show me how it can be implemented in both momentum predictor and pisoloop? I have a couple of term that I need to add in shallowWaterFoam.

[haze_1986]

Fixed, thanks

[Kdelavy]

Hi everyone,

I am trying to simulate a river with openfoam, using the shallowWaterFoam solver. Is the solver still not considering bottom roughness? How may I add the manning roughness number in the simulation then? Thanks a lot in advance.