Modelling an instantaneous failure of a Dam

[Najar]

I’m trying to model a Dam break. I have the instantaneous hydrograph and the water level drop scale (which I derived from a physical Model). my question is: how I can set the boundary conditions at the dam location (both the hydrograph and water level). Also, do I need to model the dam in my case (since I’m modelling the breaking/failure, not the breaching, or just define the boundary condition at Dam location?

[Anonymous]

I don’t know of anyway to force both the flow and the water level at a boundary location.

If you try and model it without a dam, you will have to use either a flow boundary or a stage. A flow boundary may not have the correct water surface which means the full momentum solution will not get the correct velocity at the boundary. (I guess you could try a stage boundary but I doubt it would compute anywhere near the same flow as your physical model.)

If you are mainly interested in the results further downstream, the difference in velocity right at the boundary may not matter much. If it is important to get good results immediately downstream of the dam, it would probably be better to include the dam and the pool upstream of the dam and try to do a really fast dam break from inside of RAS.

[Najar]

thanks for your Reply @Jarvus, The idea that I don’t have the topography of the Pool/ reservoir to model it. I used 3D Flow software. it has the condition boundary to inter the hydrograph and the drop in water level for the same entrance condition.

I could convert the drop in water level graph to the drop in volume graph reflecting the how the water will behave/drop when the instantaneous failure happen. Can I use this to model sudden flood in HEC-RAS.

[Anonymous]

If you can create a volume-elevation curve for the pool, you could model the pool as a 1D storage area and then use a storage area connector to connect the 1D storage area to the 2D model.

However, this may have stability problems for an instantaneous failure.

For an instantaneous dam break, it works better if it is all a single 2D area.

[Najar]

HI Jarvus, I do have the volume-elevation curve for the reservoir at the Dam-breaking section- please find the attached .cvs file- can you please explain to me how I can define a 1D storage area for the reservoir on a cross-section. I meant which boundary condition should I use in this case. Thank you in advance

volume-elevation_curve.csv

[Anonymous]

Are you trying to model the dam break flow as 2D or 1D?

If 2D, you want to use a storage area connector. It has nothing to do with a cross section. There are example data sets that came with RAS that you may have installed or should be available under the Help menu. I presume they are also on the HEC-RAS web page.

If you go to the 2D unsteady directory and open up the BaldEagleCrMulti2D data set and then open up the plan that is something like SA to 2D Dam Break. That shows what I am suggesting you might try.

If you are doing this all in 1D unsteady, you connect the 1D reach directly to the 1D storage area. If you go to the 1D unsteady directory and go to the Dam Breaching directory and open up that BaldEagleDamBrk data set, there is an example of a 1D storage area representing the pool connected to a reach with an inline structure representing the dam.

[Author]

Posts