Cell Center Points in 2D Flow Areas

[robdir]

Hi everybody, I have a question about the criterion of assignment of the Cell Center Points in 2D Flow Area.
On pag.3.4 -3.5 of 2D Modeling User’s Manual there is this notation ” the cell center does not necessarily correspond to the exact cell centroid”, but it doesn’t explain how it is defined.
Do you have some references, documents for this?
Or… can you explain me?
Thanks

[Anonymous]

Did you also look at page 3.13? I think the cell centers are positioned so that a line from one cell center to the next cell center will intersect the face at a 90 degree angle in order to make the computations faster.

[robdir]

I did it jarvus, but I didn’t understand how HECRAS “catched the points of Delaunay triangulation. I supposed that the black points in fig.3.9 represented the Elevation catched from the terrain associated but I don’t know how it works. If you suppose that your terrain has a 10m x 10 m grid size and your 2DArea has a 5mx5m grid size, which would the points be to take to reference? The program do an interpolation between the grid elevations 10x10m?

[Lonnie A]

I think you are confusing how RAS performs its calculations. In simplest terms it treats every cell as a storage area with interconnection to adjacent cells. So there is a stage storage curve for each cell. The center point is just a location for the most part and isn’t reading a elevation from the underlying terrain to create a TIN.

Read item #6 in the 2D user manual introduction for a better description.

[robdir]

???
item#6: “…each computational cell and cell face is based on the details of underlying terrain. This type of model is often referred to in the literature as a “high resolution subgrid model (Casulli,2008) ”
I didn’t understand your observation. I know that the cells are used to create to develop detailed hydraulic property tables, but I would like understand the method used by the program to evaluate cell center points.
If I look at pag.3.13 of 2D Modeling User’s Manual I read that HECRAS makes the computational mesh by following the Delaunay Triangulation technique, etc…
I remember that this method was used to interpolate the height of points, isn’t it?

However, my request was about two “simple” questions:
1) How are located the cell center points?
2) What is the elevation associated to each cell center point?

[Anonymous]

“However, my request was about two “simple” questions:
1) How are located the cell center points?
2) What is the elevation associated to each cell center point?”

1) As I tried to indicate in the previous post:

Once RAS has the cell centers (the points that were created by the grid generation, breaklines, and any editing by hand from the users), it then creates the faces that surround each cell center. In general, it tries to create the faces in such a manner so that if you draw a line from a cell center to an adjacent cell center, that line should intersect the face at 90 degrees. This speeds up the 2D computations. I don’t know if it can always get exactly 90 degrees in all situations, but it does most of the time. I believe this is the point of the Delaunay Triangulation.

2) I don’t think there is anything special about the exact elevation at the cell center point, at least as far as the 2D computations are concerned. RAS creates an elevation/volume lookup table for each cell based on the subgrid. This lookup table starts at the lowest elevation inside of the given cell, which is usually not at the cell center point. This lookup table is based on the detailed subgrid and allows a cell that only has little bit of water in it to be partially wet and partially dry. This is different from most other 2D models where a cell has a flat bottom and the cell is either entirely wet or entirely dry.

The exact location of the cell center point is used in other computations. For instance, in determining the distance between cells in order to compute friction losses and in determining the water surface elevation at the face for determining conveyance.

[robdir]

Thanks jarvus, now is clearer.
1)I noticed that the cell position respect your observations.
It seems that program starts from upper side of 2d “Contour” Mesh, to assign the position of cell points (spaced along horizontal lines if dx=dy and along parallel lines for breaklines).
Later it builds the cells according to Delaunay and Voronoi diagrams.
I’m still not finding anything about point’s position from the boundary sides, but I can be
satisfied, for now

2) I made some attempts with version 5.0.3 and I forgot to set Horizontal Render
mode!?!
When I did it, I checked come cells and I noticed that the water surface values referred
to the max values calculated on the minimun cell elevation.
So, the Cell Center Point doesn’t have an elevation value, it is only used by HECRAS to
assign the water surface (max values based on the min cell elevation).
When you click on any cell in RasMapper, you can notice that is reported only a water
surface cell value… it is referred to the “time variation” of max water surface (WSE,
Depth) calculated in the cell and referred to the min cell elevation.
This value is assigned to the cell center point to solve the hydraulics equations

Can you confirm my observations?
Thanks for your support

[Lonnie A]

The calculation’s assume a level WSE within each cell. The distance between cell centers provide the HGL for driving flow across cell faces. So the cell centers get assigned the cell WSE which is based on the underlying cell data (storage, cell face hydraulics etc).
I’m not understanding your question I guess.

[Author]

Posts