- This topic has 28 replies, 4,507 voices, and was last updated 9 years, 5 months ago by .
-
Hi everybody.
I’m trying to run a simple geometry dam break case with HEC 5. The case refers to a laboratory experience that I have already run in HEC 1D environment.
The lab set up is represented by a reservoir connected by a pneumatic gate to an horizontal flume where is located a bridge, that would be hit by the steep wave produced by the quasi instantaneous gate opening.
I’d like to riproduce it using 2D flow areas.As first step I have created the Terrain model in Ras Mapper environment using a .txt file built in a DEM format, as follows:
ncols
[these are the number of columns]
nrows[these are the number of rows]
xllcorner[this is the x coordinate of the center or lower-left corner of the lower-left cell] yllcorner [this is the y coordinate of the center or lower-left corner of the lower-left cell]
cellsize[this is the resolution of your data]
nodata_value [use -99]
5280 5283 5271 etc [these values are the data; the elevations]
4986 4991 4994 etc [these values are the data; the elevations]I haven’t set a projection for project because being a laboratory experience I don’t need it.
Once done it I have exported the terrain as an image (jpeg format) into the geometry editor, in order to draw properly the element shapes.
I have followed the manual instruction to design the 2D flow areas, then I have set the computation mesh fields as desired.
At this point by clicking “Generate point in 2D flow area”, comes an “error in generating points” and I cannot create any mesh. Moreover onto the “2D flow areas” editor, I cannot see any computation points showed under the icon, but the string “current mesh contains no computation points”, as reported in the image here below.From reading the manual I couldn’t discovered where I’m wrong, it seems to me I followed step by step the instructions, but of course something left.
I hope to have explained it clearly.Greetings
Simone M.
I’ve tried to run the project, the same I had been concerning in the previous talk, and I’ve experienced bugs both with the geometry preprocessor and so with the unsteady flow simulation.
I know that geometry preprocess could be done from Ras Mapper, by compute property tables, or as always has been from the unsteady flow analysis editor. I’ve tried both and the problem seems to be same, about the association between Terrain model and geometry, as shown here:Everything seems to have been done properly, actually I have set the right terrain association from Ras Mapper. I’m using the version markwood sent me, that’s the April one.
Could you please run it as well and see what gonna happen if you experience the same or not
Here you are the project .zip filesI have tried to reperform the project with a new DTM, more detailed than the previous one. Now the grid has a 1mm resolution, the previous one had 1cm. By doing this I can deal better with the error introduced in the representation of the domain vertical borders, that are linearly interpolated between two follow cell center points.
I have also remaked: meshes (with a spacing of 1cm) and 2D connectors (the dam and the bridge).
Every file stands in the same folder, in order to avoid the rise of any problem like you told me.
I have then tried to run the project. It stops at the geometry preprocessor step, reporting the follow errors: 1)connector invert elevations exceed the terrain; 2)the lateral structure width exceed the 2D flow area edge.
Firstly it seemed strange to me, because nor the invert elevation nor the width of the structure seemed to exceed any domain edge.
Thus I have tried to find similar cases reported on the blog. There’s one related to the first error (invert elevation). The solution seems freaky. Infact just by representing the problem in US unit system should work. I would directly check it it but I cannot because I cannot export the DTM (built with the US system) image and using it as a background image in the geometric editor in order to draw the new file. The problem is that I’m allowed in exporting the image but it comes without georeferentiation. The strange is that in order to have a georeferenced image I need an existing geometry file (georeferenced) already uploaded in the geometry editor, that is a nonsense if it had to be built right based on the background image. Fortunately I have available a georeferenced geometry file (in SI unit system) earlier built with the old 5.0 version (where instead it was unproblematic doing it) and I’m using it to georeference the background image and than use it to redraw the geometry for all the case (in SI) I need.Omitting this bug and carrying on the matter, I have tried to decompose the problem in order to see where the problems were.
By erasing the bridge and adjusting the dam (I have drawn it exactly in the middle of the area between the two 2D flows ones, with an exact width as well) I have got a successful preprocessor run. So I have tried to run the Unsteady simulation but the follow error message appears:I couldn’t get its meaning. Do you have any experience with this?
I would fix it before doing any try to introduce the bridge.
I would like to attach the zip project but I have seen that by doing this many datas disappear as well as different program versions read it differently, so I cannot see the sense of doing it.I hope to have been clear enough
I was able to add the bridge so right now the geometry file is completed.
The only lack I would to highlight is about the hydraulic structures representation, that not match exactly the flume edge. If I do it properly, as it’s explained in the old HEC manual (the 1D) i.e. by letting the hydraulic structure be deeper and wider than the terrain in order to ensure the structure to right follow the river edges, the program reads the structure out of the mesh edges and that unabling the simulation.
So I’m try to work in this way.
Now the geometry preprocessor works properly but not the same for the unsteady running, that shows error. I’ve tried both the SI system and US one (I convert it as you suggest).
The error is different than the last I’ve reported here and seems reflect unstability of the model but I’m not sure of it.
- You must be logged in to reply to this topic.