Wednesday, May 20, 2009

Flow/Conveyance Distribution

Written by Chris Goodell, P.E., D. WRE | WEST Consultants
Copyright © RASModel.com. 2009. All rights reserved.

In HEC-RAS conveyance for a given cross section is divided into the main channel, left overbank and right overbank. The distribution of conveyance is controlled by geometric features of the cross section itself, such as the terrain, bank station locations, ineffective flow triggers, and Manning’s n values. The latter three are typically subjective in how they are defined at any given location. It is important for the modeler to define these parameters to accurately define the hydraulics in the cross section, but also to maintain numerical stability for unsteady flow modeling.

A convenient way to check for appropriately defined conveyance distribution is a quick scan of Standard Table 2, in the Profile Output Table. Standard Table 2 displays the flow in the left overbank, main channel and right overbank (Q Left, Q Channel, Q Right). The modeler should view this table and look for cross sections that show a sudden change in distribution. Any sudden change is a good indication of poorly placed bank stations or ineffective flow triggers.
The attached table illustrates how Standard Table 2 can be used to locate areas of sudden changes in flow or conveyance distribution.


Friday, May 8, 2009

Multi-dimensional modeling

Written by Chris Goodell, P.E., D. WRE | WEST Consultants
Copyright © RASModel.com. 2009. All rights reserved.

As much as I am an advocate for HEC-RAS modeling, I recognize that there are some cases where you need to go to a multi-dimensional model. I'm not as quick to jump to that level as others I've worked with, I think mostly because there are a lot of quasi-2-d techniques built into RAS that can be used to simulate flood conditions that are dominated by 2- and 3- dimensional flow patterns. However, there have been cases where it is necessary to go multi-dimensional. My best example is a project I was working on that consisted of a flood down a well confined canyon that opened up onto an urban alluvial fan with no defined channel. I tried to work RAS forward and back to get it to work-tons of lateral structures and storage areas, lots of assumed ineffective flow areas, etc., etc., but it just wouldn't happen. I ended up using FLO2D for the alluvial fan portion of the study area and it worked quite well. I've also used RMA-2 and CCHE2D succesfully in other applications. Does anyone out there have other models that have worked well?