Written by Chris Goodell, P.E., D. WRE | WEST Consultants
Copyright © RASModel.com. 2009. All rights reserved.
I was recently emailed this question. Here are a few of my thoughts:
You could model the barbs themselves as cross sections with adjusted station elevation points, or as blocked obstructions within cross sections. The blocked obstructions make it easy to compare with and without barbs, however, you are limited to a rectangular (or series of rectangular) shape(s). Make sure you account for ineffective flow between the barbs. The barbs should have their own n values to ensure that the conveyance over the barb is treated independently.
You technically should place the cross sections perpendicular to the anticipated flow paths. However, the “angling” of flow over the barbs will be a very localized effect and you could probably get by with placing the cross section perpendicular across the stream through the middle of the barb. Though it is easy enough to angle the cross sections. Just make sure you understand what you are modeling. If trying to model an extreme flood event, you probably should cut your cross sections straight across the stream.
I would not try to model the barbs as inline weirs in HEC-RAS. Being a 1-D model, defining weir flow on one side of a channel and conveyance on the other doesn't really make sense. However, this does make me wonder if you could try this using the multiple opening analysis in the bridge/culvert editor.
Keep in mind that you are modeling a very 3-dimensional flow pattern (over and around barbs). You should only use HEC-RAS to get a feel for the overall rise in backwater due to the barbs. Do not use RAS to analyze local velocities over and around the barbs, or for sizing of the material (i.e. riprap sizing). If you do want to design the barbs using RAS, make sure you use appropriate safety factors.
Any local analysis around barbs would require a multi-dimension model (at least 2-d, preferably 3-d).