[Luis Partida]

Hi Jeremy, when a model is that large, there are bound to be human error. In my experience, the criss-crossing edge lines are due to random cross sections being drawn in the wrong direction in between cross sections drawn the correct way. This causes left bank and right banks to be in the wrong location hence why the lines cross. Hope this helps

[Luis Partida]

Hi Evan, a couple things here. The first is your model will be extremely unstable and takes a lot of “tricks” if you do not use a 1D storage area to connect the 1D to 2D domain. And now for your error message; The direction in which you draw the connection matters. It sounds like you just need to delete the SA2D connection and redraw it in the opposite direction and youll be good to go.

[Luis Partida]

Hey Daniel, I am glad you figured this out. One way that is best practice to avoid this happening, is to place a dummy breakline along the perimeter of your mesh. This breakline does not even have to triangulate but it forces your cells at the boundary to triangulate perpendicular to your boundary condition line especially when imposing stage. This is also true with lateral structures connecting 1D domains to 2D domains.

[Luis Partida]

What about the computations options and tolerances gate flow submergence decay exponent?

[Luis Partida]

Mannings values are still controlled by the cross-section data editor just as if a Lid is not in place. If you want to vary roughness within the tunnel use vertical variations in mannings values by clicking options

[Luis Partida]

I appreciate you sharing your work flow as so much of this kind of hidden behind unique methods. Thanks!

[Luis Partida]

Hi, so I know I can take Atlas 14 gridded precipitation from their website and process it in HEC-MetVue which is spatially varried for each state. But as far as creating gridded precip from a specific time series….I am not sure, I bet it could be done in MetVue as well. Documentation is not great but I would try it out

[Luis Partida]

In RAS Mapper, in your table of contents, select your imagery and right click and at the bottom select “render beneath terrain (false)”

[Luis Partida]

Oh and of course the computational time step is important as well.

[Luis Partida]

Here are a couple of SA Connection tips:

1.) If a tributary is coming in along the length of the weir, it should be split into 3 separate weirs. 1 for the tributary with a low coefficient, and 2 for the adjacent sides.
2.) In some cases, you can change the equation being used to use the 2D equation instead based on submergence.
3.) In calculations options and tolerances in the General Tab, increase the lateral structure flow stability factor.
4.) Weirs should be filtered down below 100 points, which increases efficiency.

In general the weir coefficient selected should represent the underlying terrain which is why if there is an incoming tributary and or 2 meshes connecting multiple stream networks to the downstream mesh, each stream needs its own weir.

There are also initial depth errors when using spatially varied stage inputs as the DS boundary condition which can be fixed by turning off “Use Initial Stage” in the unsteady state flow editor. Hope this helps

[Luis Partida]

Jeremy, that initial conditions error could potentially be solved by going into your boundary condition data in the unsteady flow editor and at the bottom of the boundary condition data, uncheck the “use initial stage (recommended)”

[Luis Partida]

Yes it should do it automatically. To test it, you can run a simulation, then open RAS Mapper and expand the results “event conditions” and watch the storm generate over your domain.

[Luis Partida]

The first thing I would say is that velocities are determined via a calculated force and since it is not a tabular output currently, we have to use RAS Mapper to determine velocities. If I was modeling this and had this issue, I would adjust the internal XS manning’s n values to ridiculously high (0.5) and see if that stops it. This would then tell us that the velocity calculated is not out of a general range of expectations.

Is the bridge overtopping? In RAS versions 6.4 and earlier pressure flow was not capable of being calculated and that also may be the issue as those versions could only simulate low flow conditions. I would look into that by upgrading the model to V6.4.1.

There are so many things to look at here and hard to debug without the model in front of me. Here are some additional things to check

-Adding near repeats to your SAConn is also recommended by HEC.
-Make sure the “distance to the US XS” in the bridge editor falls within the first cell parallel to the bridge.
-Make sure the bridge is enforced correctly because that will cause odd results, i.e. NO crossing breaklines.
-Make sure the cells on the SAConn for your bridge are not drastically smaller than the global cell spacing

Hope any of this helps!

[Luis Partida]

I bet you have a red cell error that is not being shown in RAS Mapper. I still refuse to build out geometries in RAS Mapper as it is still buggy. Open the geometry data editor with RAS Mapper closed, then check the bottom left for a red message saying there is a cell with more than 8 sides and either fix it with a breakline or manual edit.

The other option I would suggest is to rebuild it again in a fresh environment and do not use RAS Mapper for anything

[Luis Partida]

Hi Kendall, this website isn’t great when I have pasted links in the past. Google “modeling bridges inside 2D flow areas HEC” and a page should pop up where they go into great detail on what is going on with the entire process.

[Author]

Posts