Monday, September 30, 2019

Using HEC-RAS 2D area breaklines to improve modeling

HEC-RAS 2D area breaklines are a great way to force cell faces to follow a terrain ridge or crest hinge (top of slope) line. Without them, you are never quite sure whether flow is going over or through high points in your terrain.

Divided flow in 1D modeling

Experienced 1D modelers know that HEC-RAS is unable to distinguish between flow on the two sides of any ridge. The HEC-RAS programmers tried to teach us this by providing a warning any time a high point in the terrain splits the flow at a cross section into two regions. "Warning: Divided flow" is what they tell us. This means that the program has no way of knowing whether the flow on the two sides of the high point is divided accurately. If there really is a persistent ridge, the modeler (that's you) has to evaluate how to model the split between the divided areas. Sometimes an ineffective area boundary along the ridge is appropriate to keep the flow calculation to the main area. Other times a lateral flow model and a separation of stream models is appropriate. It can get complex and uncertain, and that's where 2D modeling can help for areas with particular abundance and uncertainty of divided flow.

Divided flow in 2D modeling

But within each 2D flow area cell face, HEC-RAS is no more able to distinguish divided flow than it can at a 1D cross section. Just like in 1D modeling, HEC-RAS will happily allow flow through any and all low points on a cell face regardless of whether they are divided by a high and dry ridge within the face. We could address that situation by using very small cells. But we may not want to because of the issues discussed at the end. So we use 2D area breaklines to force cells and faces to divide at ridges whenever we see or suspect water apparently flowing through a ridge or crest.

What are 2D area breaklines

For the purposes mentioned in this article, "2D area breaklines" are enforced "cell boundaries" or "hydraulic section boundaries" or "hydraulic cell edges".

Where to put 2D area breaklines

To enforce divided flow, we just want the cells to break at high points and sometimes at crest (not sag) hinge points. We do not need 2D area breaklines at flow lines for this purpose. We need them at crowns, weirs, and ridges. 2D area breaklines at other locations such as flow lines could be used to force the orientation of cell faces to follow flow paths instead of being orthogonal.

How to draw/locate 2D area breaklines


There are 2 main ways to create 2D area breaklines. We can draw them freehand or we can import them using shapefiles. For the purposes in this blog post, I recommend each following the reasoning below. For other purposes (2D area connection walls etc) it is often better to import shapefiles.

Terrain imprecisions

Terrain resolution issues may defeat your hydraulic modeling purposes if you mindlessly transfer precise breaklines into HEC-RAS. If a breakline crosses many terrain cells that dip below the ridge or crest, its purpose is defeated. If only one side of a ridge or crest is steep, you may get appropriate results by drawing freehand a breakline slightly back from the ridge or crest to avoid all terrain grid dips. If both sides are steep, you may want to use an imported 2D area connection with a profile to enforce the ridge or crest precisely. If both sides are flat, you can probably import a precise breakline without going to the additional effort of defining a precise profile for it.

Contours for guidance

If you are drawing a 2D area breakline freehand, you should turn on high-resolution contours in RAS Mapper. As of HEC-RAS 5.1, you need to exit HEC-RAS after turning on contours so they will show up in Geometry Editor.

Why not just use a grid of smaller cells?

Reducing the global spacing of your 2D grid gives you questionable value at a cost of increasing a computational time of seconds to a computational time of hours due to the following factors:
  1. Dividing cell spacing by two creates 4 times as many cells.
  2. Reducing cell size reduces the time delay between cell faces, requiring you to decrease computation time step.
  3. Reducing cell size is a coarse approach to solving what usually can be distilled down to a known problem at a few key breakout ridges. It may not solve fake breakouts even after increasing your computation time to hours.


Politics education gate said...

Hello Users, I just need some insights on Breaklines.

I have added breaklines into the 2d mesh area and enforced them as internal connections and entered weir data as top spillway elevation at 700m to the connections assuming they are acting as Embankments/weirs to the reservoir.

But After I ran the Unsteady flow model successfully, in the results section in the RAS Mapper Water still flowing over the embankments at 653.55 only without filling the reservoir up to 700m elevation.

What Should I do?

Thomas Gail Haws said...

I'm grasping here:
1. Is is possible there is a culvert in the connection?
2. Would it make a difference whether you select 2D solution vs. weir solution?
3. Does the length of your weir make a difference?
4. Are you using the full momentum equations vs. diffusion wave?
5. Does this page shed any light?