1D-2D: Lateral weirs at structures/embankments

[Igloo_79]

I’m building a linked 1D-2D model – the 2D flow areas (one on each side of the river) run along the entire length of the 1D network, connected to the 1D XS by a series of lateral weirs.

The river flows through several structures, including bridges, box culverts and railways embankments, so I am wondering whether it’s preferable to end the weir at the start of such structures and start a new one at the outlet? (i.e. flow is not overtopping the ‘weir’ within the actual bridges, rather it is flowing around them in the 2D domain.) Additionally, some structures (e.g. railway embankments) lead to sudden spikes in the lat weir elevation, see below. Feedback appreciated!

[Lonnie A]

No flow is being transferred between bounding XS of structures.

What I do with long culverts in cases like this is model the area over the culvert in 2D…I stich the left and right 2D areas together with a storage area connector along the culvert. You have to set the culvert deck in 1D artificially high so it doesn’t count the weir flow which is already being accounted for in the 2D.

[Igloo_79]

Thanks for this Lonnie. Looks like a good approach!

I notice that your LS/2D Flow Areas do not align to the ends of your 1D XS. The same was true for some parts of my model, leading to ‘dry’ patches where the 1D and 2D domains were not connected. It is quite a pain to realign them to the 1D boundary (i.e. XS Interpolation Surface). Would appreciate your experience of this?

[Lonnie A]

The easiest way to fix the slivers is usually to extend the XS to where it has a small overlap with the 2D. This is easier than pulling the 2D area to overlap with the XS and causing the mesh to change forcing edits to the lateral etc. Usually the slivers are small enough that I know there is no impact on results and is just a graphical thing so I let them slide sometimes.

[Anonymous]

Lonnie, super interesting approach. I will check this out next time i run into this situation

[Author]

Posts