Jun 21 2011

The Big Assumption

When I graduated from Colorado State University and entered the working world I had very limited exposure to hydraulic modeling.  I had taken fluid mechanics and hydraulic structures where I was first introduced to a hydraulic model.  I also had some exposure to HEC-RAS, but didn’t even know that 2-D modeling existed nor how it was different than 1-D modeling.  Although I new how to setup a basic HEC-RAS model I did not fully understand how everything worked, or all of the assumptions that I was making when creating and running the model.

2D Model_velocity

Over the course of my career I have had the opportunity to learn from some very knowledgeable hydraulic engineers and have worked on several 2-D models for a wide variety of applications.  Having this exposure to 2-D modeling has really helped me become a better 1-D modeler because I am more aware of the assumptions I am making in the 1-D environment.

While there are several assumptions to be aware of when setting up a 1-D model, one of the biggest ones is what you are doing when drawing cross sections.  Cross sections should be cut so as to be perpendicular to the anticipated flow lines.  The more 1-D the flow actually is the easier this is to do, like in a canal.  When you cut a cross section you are essentially stating that at this location the water surface is equal across the entire section.  When confronted with a more complex hydraulic condition the anticipated flow lines are a lot more difficult to assume, which can lead to errors in your model.

Of course the best way to understand this is visually, so lets look at an example.  Below is an aerial image and USGS Topo Map of the Little River at the Highway 122 crossing east of Barney Georgia.  The floodplain is a little over a mile wide.  Flow is from north to south The skew between the floodplain and the roadway alignment is around 45°.  There is also a relief bridge east of the main span bridge.  Model limits are shown in blue.

Little River aerial

Little River topo

So, how would you position cross sections for a HEC-RAS model?  Remember cross sections represent equal water surface elevation.

In the next post I will show the 2-D model results that display the water surface elevation contours across the floodplain.


3 pings

Skip to comment form

  1. Matt Anderson

    Dusty – The cross-sections will likely different depending on the extent and depth of flooding. As this river stretch begins to flood, the mass of water likely isn’t going to flow along the meanders, but over top of them.

    1. Dusty Robinson

      That’s correct, but the larger hydraulic control for this reach is the roadway crossing the floodplain. The flooding for the 100 and 500 year events fills the extents that I have shown. So yes, the channel is of little significance during a flood event at this location.

  2. Bill McDavitt

    Good post and good question. Many thanks Anthony and Dusty.

    Combining your question about xsec orientation with the mention of the roadway alignment angle would cause some additional thought to go into what value should be entered into the Bridge/Culvert Skew editor.

    Not being familiar with the site, at a Q100 flow, does Highway 122 create a backwater?
    This would influence how I think about xsec orientation upstream and downstream of the road.

    Does Meetinghouse (‘Creek’ I assume) essentially get backwatered by the Little River at Q100. In other words, does the modeler assume that Meetinghouse isn’t increasing WSE to Little River in any significant way? This would influence how I think about whether to add Meedinghouse as a reach and add a junction.

    Finally, can the 2D results help you fine tune your expansion/contraction coefficients?

    1. Dusty Robinson

      Highway 122 does create a backwater, and the roadway gets over topped in the 100 and 500 year event. Meetinghouse creek is a very small drainage and does not affect the hydraulics for the Little River. A 2-D model would help inform the expansion/contraction coefficients as well as the exact location of ineffective flow ares. I will include a graphic of the velocity vectors to show this in my next post.

  3. Anthony henry

    Having done some 1D and 2D models myself, I agree with the ideas expressed here. However, I don’t think it is just 1D or 2D. This scenario does look like it would be a good candidate for a quasi-2D approach utilising storage areas with lateral and in-line links. Basically a 1D approach approximating 2D conditions by defining variable 1D flow paths.

    1. Brian Varrella

      Not a bad suggestion, but when you are talking about inline and offline storage you are diverging from a steady flow analysis to an unsteady flow situation. Dusty’s suggestions and description of the importance of cross selction location are still valid for unsteady flow — is that correct Dusty?

      1. Dusty Robinson

        That is correct Brian. As I noted in the next post, the modeling approach is dependent on the needs of the analysis. Since this specific model was used for a bridge redesign, the 2D approach was a good choice.

    2. Dusty Robinson

      Thanks for the input Anthony. Are you referring to the Mike and Tuflow models?

      While I am aware of some of their capabilities, I have not personally used either.

      What benefits would you say they have over fully 2D models?

  4. Marcio Nobrega

    I just want to prais your blog for the visual richness. Congratulations! A bliss to a Hydraulic Engineer.

    1. Dusty Robinson

      Thank you. I am glad you find it helpful.

  1. » Assume No More Hydraulically Inclined

    […] No More In my last post I asked how to best lay out cross sections for a HEC-RAS model for a section of the Little River in […]

  2. » Which Way Does the Water Go? Hydraulically Inclined

    […] trying to figure out which way the water is flowing and layout your cross sections accordingly, as I discussed earlier.  This is not that difficult when the flow for the reach of river you are modeling is primarily […]

  3. » Our Top Posts of 2011: Year of the Flood Hydraulically Inclined

    […] The Big Assumption (June 21): In this article, Dusty begins the discussion of 1D versus 2D models and when 2D modeling […]

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>