I have a long reach in which we wish to model the channel and overbanks with 2D. However, there are several bridges which do not cross the floodplain so I am trying to figure out the best approach to model the bridges.
My approach has been to model with 2D and then run short 1D reaches through the bridges and then return to 1D. I am running into a mismatch at the 1D/2D interfaces (where flow enters or exits the 1D reach at a cross section). This is especially prevalent at the downstream 1D section. What I am experiencing is a sudden drop in velocity and a sharp rise in WSE.
I am wondering if anyone has experience with a similar situation and if there are any tips and tricks out there for how to handle bridges in the 2D environment.
If you haven't already done so, go into the Options menu and turn on 1D/2D iterations.
If the water doesn't hit the bridge deck, you might try leaving the actual bridge out and modeling with 2D by adding the abutment to the terrain. If you use really small cells, you can also add any piers to the terrain. This has to be done outside of RAS.
You might be able to model the bridge by using a storage area connector and using culvert to represent the bridge opening.
Thanks for the response Jarvus. I have used the 1d/2d iterations and have found no improvement. The image shows a typical example of what I am finding where the velocity at the downstream interface is zero in the 2D section.
Maybe try adding more cells at the 1d/2d boundary. As a rule of thumb you don't want cells to change in size by more than 50% from one to the other, so try small regular shapped cells by the transition and a gradient in size from there. Also could it be from a change in Mannings?
Thanks Toby. I have tried to add more cells and have also adjusted the roughness values to better match, but the problem remains. I am wondering if there might be a problem that is occurring because I am only modeling the channel in 1D and not the entire floodplain (I am modeling 2D over the bridge approaches).