User Forum of Software BASEMENT

rghazanf

User

Registered: 2015-10-20

Posts: 9

asymmetry flow in a symmetric geometry

Hi,

I am modelling an open channel having a fully axisymmetric (around the channel center line) geometry. I generated an unstructured 2D mesh using BASEMESH following all guidelines provided by the tutorials and manual.

Details of BCs are as following:

BOUNDARY {
                name                 = bc_upstream
                string_name          = upstream
                type                 = hydrograph (steady discharge)
                slope                = 1
                file                 = bc_upstream.txt
               
            }
            INITIAL {
                type = dry
            }
            TURBULENCE_MODEL {
                type                 = algebraic
                kinematic_viscosity  = 0.000001307
                boundary_gradient    = zerogradient

            }
            BOUNDARY {
                name            = bc_downstream
                string_name     = downstream
                type            = hqrelation
                slope           = 1
}

The problem is that I am not getting symmetric results. For example, velocity patterns in two corresponding regions are different that is quite strange to me.

I already tried including parameters in the definition of BCs. Also, using zero-gradient BC for the downstream boundary.

But the asymmetry constantly exists and seems to be dependent on something else. However, the water surface profile along the center line of the channel looks correct in both terms of trend and value.

I would appreciate your help to get through this problem,

Reyhaneh

Lukas Vonwiller

Developer

Registered: 2014-09-04

Posts: 127

Re: asymmetry flow in a symmetric geometry

Hi Reyhaneh

Please turn off the turbulence model when starting with dry initial conditions!
You can turn it on again when you defined some well defined initial conditions from a restart file.

Hope this helps!
cheers,
Lukas

rghazanf

User

Registered: 2015-10-20

Posts: 9

Re: asymmetry flow in a symmetric geometry

Hi Matteo,

Here is an example of those geometrically corresponding regions with not matching results

As you see, velocity field is quantitatively different (The center of rotating re-circulation zone is clearly different).

Here my mesh plot,

Hi Lukas,

I did it! Actually, when the turbulence model is working the asymmetry is much more less and apparently it is improving results.

I am looking forward to your help, many thanks.

Reyhaneh

rghazanf

User

Registered: 2015-10-20

Posts: 9

Re: asymmetry flow in a symmetric geometry

Any solutions? I would appreciate.

Reyhaneh

Lukas Vonwiller

Developer

Registered: 2014-09-04

Posts: 127

Re: asymmetry flow in a symmetric geometry

Could you please try to refine your mesh in the channel?
I think that with elments of similar size like in the cavity the results will look better.

Hope this helps!
Best,
Lukas

rghazanf

User

Registered: 2015-10-20

Posts: 9

Re: asymmetry flow in a symmetric geometry

Thank you Lukas. I did refine the internal zone so that the whole domain had more or less the same similar element size.

Except this velocity miss-match between the lower and upper zones, all other results look correct. I don't know what is mistaken.

Reyhaneh

Lukas Vonwiller

Developer

Registered: 2014-09-04

Posts: 127

Re: asymmetry flow in a symmetric geometry

A general concept in numerical modelling with finite volume approach is to refine the mesh until the results don't change significantly and your requirement of accuracy is met.
Therefore, try to refine your mesh further (everywhere) and decide based on the computational time and your requirement of accuracy which mesh to use.

hope this helps!
Best, Lukas

rghazanf

User

Registered: 2015-10-20

Posts: 9

Re: asymmetry flow in a symmetric geometry

Thank you Lukas. It worked finally with a very fine mesh!

Reyhaneh

Lukas Vonwiller

Developer

Registered: 2014-09-04

Posts: 127

Re: asymmetry flow in a symmetric geometry

Dear Reyhaneh, could you please post the results with the finer mesh too?
I think this would be useful for other users!

Thank you and best regards
Lukas