How to reduce velocity change at 1D/2D lateral structure connection

[Anonymous]

I have a 1D model that is attached to a 2D flow area by means of a lateral structure (weir) that basically captures “leakages” out of the main river through a bunch of tiny connected channels that were too many in number to just do 1D junctions.

Right now I am using the “Normal 2D Equation Domain” computation method and also using the “Use Velocity” option for the 2D Boundary.

You can see from the results that there is a sharp change in velocity at the interface of the 1D/2D model. There is also a sharp directional change, but that is expected I guess as the direction is perpendicular to the 1D XSs and the “leakages” are essentially perpendicular to the main river or parallel to the 1D XSs.

My question is, how can I make the change in velocity at the boundary smoother? Maybe it is not possible but to me it just “looks” wrong and feel there must be something I can do.

Any thoughts?

Thanks

Derek

[Anonymous]

Two things you need to consider:

1: The interchange location and the mannings value associated with the cross section AND the cell it is attached to. The manual suggests that the mannings values at the exact location should be the same

2: The weir coefficient, the RAS manual suggests values based on the criteria of submergence. Double check these and make sure you are using the correct value

[Anonymous]

Thanks, I have taken all this into account before writing this post but still seem to be getting these issues.

When I use the full 2D equation at the connection I get a sharp velocity gradient

When I use the weir equation (with appropriately chosen coefficient) I get a sharp WSE gradient.

>< any other thoughts would be appreciated

[Anonymous]

The final two suggestions is to look at the interface. The manual recommends the elevation and geographic location of the 2D and 1D interface be at the same spatial location/elevation. If the cross section is dropping into a well on the other side this could cause problems.

Another potential solution is to play with cell sizes at the interface. By utilizing breaklines and increasing the minimum and maximum cell sizes along the interface it would allow you to vary and test different ranges of connectivity

[Anonymous]

thanks, ya this is what I’ve been trying for the last couple days (to no avail yet).

Location/elevation is the exact same, everything is geo-referenced and comes from the same terrain file. Manning coefficients are the same at the interface. The connection is at the top of the banks and flows over into relatively flat terrain (maybe a couple ft. slope downwards after the bank).

If I am able to find a solution/improvement I will definitely share it.

[Anonymous]

Theres also a hidden weir setting too that can be altered. Its in Calculations options and tolerances. Bottom right-ish there is a lateral structure flow stability factor coefficient

[Anonymous]

Thanks..will try this too.

[Lonnie A]

Derek,
I find the 2D equation to be unstable a lot of times. Particularly if the lateral is cutting across tributaries. If you really want to use the 2D equation I’d suggest breaking the lateral at tributaries and for the short piece across the tributary use the normal weir equation with a weir coefficient of 1 to 2.
Most of my models I use the normal weir equation with a Cd of 0.5 when attached at grade to the floodplain and a 1 to 2 Cd across tributaries. I also bump the stability factor up to 3. You want to try and keep the laterals to a nominal length. I try to keep them to less than 3000′ typically.
I then review the hydrographs across them. If I see some stability issues either tweak the Cd or try the 2D equation. If you start with the weir equation and get it running you can go back and try turning 1 or 2 laterals at a time into a 2D equation and narrow down which ones you have to keep with the weir equation. My models are largely coastal so with the flat terrain the difference between weir equation and 2D equation is usually minimal like 0.2’+/-

[Author]

Posts