[jadams]

HEC-RAS does not perform hydrology calculations; it will simply perform hydraulic calculations based on the flows you input. However, to answer your question the size limitation is based on your methodology of calculating run-off i.e. rational, SCS, Regression, ect. Each method does have a minimum applicable drainage are size.

[jadams]

When a model jumps from steady state flow to critical flow or vice versa it creates a dramatic change in Energy grade. Often these changes will exceed the tolerances of the model and thus cause it to be “unstable.” One common step for “unsteady” flow models to counter this is to first run a draw down model and store the information for the initial conditions of the actual model. To do this you create a boundary condition at the lower point of the model that will have a stage that will completely flood all the reaches. You then lower the stage over a model period. Should the model crash the section where the water stage is at is usually the culprit. This allows you to troubleshoot several instabilities prior to an actual run. Once you have achieved a draw-down model that runs for the entire period you can tell the software to store the condition as a start-up file. For flow you can usually use 2-10% of the reach flow. This will be an iterative process because you are looking for the flow that will yield a stable model. Chris used to have a nice “How-to” on draw-down models. If he reads this and it’s still accessible I am sure he will provide a link.

[jadams]

In a quick look at your model I noted a couple things. You have a few sections that are near vertical drops. Is this intentional? If so that could definitely cause some instability. Also if you haven’t already I would run a draw-down model for your reaches to get your initial conditions. The flows you have in the reaches seem low for how steep the reaches are.

[jadams]

You could try adding culvert blockage that is equivalent the intersecting pipe as depth blocked.

[jadams]

The hydrograph output interval has several things you should consider to determine. First is that the increase in output interval and computation interval will be more resource intensive for the PC and therefore will increase the overall model run time. Second, typically an output interval is not less than the estimated time of concentration for the contributing drainage shed. Third, as in all modeling programs that include the dimension of time you need to ensure you have captured what you want to see. If you are looking for the peak discharge you need to ensure that your timestep is sufficient to capture this data and not skip it on either side and giving you a falsely lower value. All of these factors need to weighed by the modeler with the specific goal of the model in mind and a decision made on a case by case basis.

[jadams]

I am currently using the 10.1 beta so it may be slightly different than the 10.0 I can’t remember. But I don’t think you need to manually create the layers. By choosing the create river CL, XSEC, flowlines, etc the program should create the layer and prompt you to digitize the features, this should solve the projection issue. Also ensure that you are digitizing your XSEC from Left to Right as facing downstream.

[jadams]

Have you checked out Chris’s post about bridges. Even though it seems you have already re-evaluated your geometry I would still start with the post.
http://hecrasmodel.blogspot.com/2009/08/problems-with-bridges-and-culverts.html

[jadams]

In order to import the file you need to open a blank “Geometry.” Then import the .sdf to the geometry.

[jadams]

A way that isn’t exactly the most accurate to make the software update the lengths is to create a new dummy cross section. Just the section will do no need to populate it. Then delete it. RAS should prompt you and ask if you want it to automatically adjust lengths.

[Author]

Posts