[MichaelG]

There is no reqirement to include a storage area in the system. Have modeled a few surges with RAS comparing RAS results with laboratory data.

[MichaelG]

The problem that you report sounds like it is related to the time window. Be sure that your simulation time window (set in the unsteady flow manager) is encompassed by the data in the DSS file. Another consideration might be a connectivity error. The reach connections can be found on the Geometric Data Editor under Tools – reach connectivity. For example; two cross sections are required upstream of the dam – one to connect to the storage area and one for the unsteady flow equations to work. Two boundary conditions, internal or external must always be separated by at least one computational cross section.
A tutorial for developing the input data for modeling a dam breach with RAS using a storage area to depict the reservoir was published by the State of Colorado a few years ago. Their document can be found at http://water.state.co.us/DWRIPub/Documents/GuidelinesForDamBreachAnalysis.pdf
That document describes how you enter the data for using a storage area to describe the reservoir. An example of the data structure for describing the reservoir with cross sections is included in the example data sets with RAS 4.1 (Bald Eagle Cr.).
Michael

[MichaelG]

Hi: I looked at your data and noted several problem areas; one being overlapping of a storage area with cross sections and another being cross sections spaced too closely for an unsteady flow simulation. You can find my notes at: https://files.secureserver.net/0st9XnfLcyi4BJ

Regards,
Michael

[MichaelG]

When looking at the time cycles of RAS computations and subsequent output, we should separate the computational aspects (solving the equations) from the output aspects (viewing the results).
The RAS computational time step (Computation Interval) considerations that do indeed affect the simulation results are governed by:
1. The stability of the computations.
2. The numerical accuracy of the computations.
3. The resolution of the input hydrographs.
Computational stability governs the computational time step for many applications. There are some text book situations where stability may be satisfied, but the numerical accuracy of the solution requires a smaller time step. Also, there are situations where numerical stability and accuracy may be satisfied by a relatively large time step, but the time scale of the inflow hydrograph (perhaps a dam break), may require a smaller computational time step so that the peaks will be captured by the computations. On the other side of applications, a tidal forcing function has a period of 12 to 24 hours. Such an input stage hydrograph can be adequately resolved at a one hour time step; indeed the stable computational time step for tidal situations often can be on the order of 10 to 30 min. However, resolution of inflow hydrogaphs and details of velocity and stages may require smaller computational time steps so that the output can be obtained at more frequent intervals.
RAS Output time steps controls. The concept here is that an unsteady flow model such as HEC-RAS can generate voluminous output, so the user gets to control what gets saved – where and how often. The output selections do not affect the unsteady flow computations.
As a first step in testing the operation of the unsteady flow application of HEC-RAS, one should just look at the hydrograph plots. The information sampled for these plots (Hydrograph Output Interval) should be equal to an integer multiple of the computational time step. The post-processing computations (Detailed Output Interval) may be done at a different (usually longer) time step than the hydrograph output. Again, integer multiples of the hydrograph output interval are necessary. Post-processing of a large unsteady flow simulation can take some time; that is why the hydrographs are available after the unsteady flow run. These hydrographs can be used for simulation diagnostics immediately after the run is finished.

[Author]

Posts