open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone!

So my problems with my current study are getting slowly fixed, one by one.

But I now face a whole new issue: I have a dam in the case I'm studying, but when water goes over it, it gets eroded by the liquid flow.

Hence my question: how could I define that particular area (the dam) to be non-erodable?

Thanks in advance!
lpilavoine

Hi
I think it's quite well explain in the Sisyphe manual and probably also in some other topic...
You could do that by modifying noerof.f

it's also possible to fix the bottom thickness with a previous computation file or directly in the geometry file

regards

Hi lpilavoine

As c.coulet says, you can modify the file noerof.f, for that I leave the following file modified by me :).

What you should do is make a file, where you have the point x, y and node of your mesh that you want that is not erocionable, example file 0.25rec.4.txt.

Then in the file joaco.f modified row 63 and place the total number of non-erocionable points, row 80 where you can put if you want a maximum erocion in all your mesh (in my case is 0.05 meters), row 95 where you place the address Where is your txt file, row 96,107 and 116, where you put again the total number of non-erocionable points you have.

Remember that in the .cas file, you must put "FORTRAN FILE = 'joaco.f'"

regards

Joaquin

Thank you both for your answers.

Since I had around 190 nodes of non-erodable domain, I opted for a polygon enclosing my whole dam, and I wanted to set the ZR variable at those points using the "inpoly" subroutine.

It worked, but I am now facing a new problem: the thing is that I've set ZR to ZF, meaning a maximum erosion of 0 on those nodes.

When I launch a simulation without suspension, it starts without any issue.
But when the suspension is enabled, I have an error from the get-go:
If I understand correctly, it has something to do with the layer thickness (ELAY), but I don't know what this EST value is. Something estimated by Sisyphe?

This message doesn't show up if I set ZR to ZF-0.1 in my NOEROD subroutine though (but then this isn't a non-erodable dam anymore).
Do we need to have something like ELAY<EST with the suspension activated?

OK, so I have tried several things, and yet my dam is still getting eroded during the simulation.

I have tried to use the inpoly function, the nodes numbers and, right now, to set all points with an X value lower than 55 to have ZR = ZF-0.1

And yet, my dam is getting eroded way more than this.
I have attached an image of the phenomenon as well as the noerod subroutine I am using.

It is worth mentionning that I am currently trying to run this case with ZR = ZF-0.1 all over my domain and it looks like my current simulation is on the same path than the one I'm showing in my image.

Is the fact the non-erodable area is connected to a free exit boundary condition for the water can explain my issue?
Thanks in advance.

Hi,

I do not understand what you want to model, but I did a model of a channel, where the output was not erocionable and I had no problem.
Question, why do not you do it as I explain? I'm new to this, but the way I use it works great, I point out the points that non-erod and ready.

It's a bit slow but it works.

Joaquin

I tried your method as well, but it doesn't change this behavior: my dam is getting eroded on the boundary and it eventually makes the simulation fail (it creates empty layers)

So I tried a couple of things: add an area where the water will exit the domain (so that the erosion on the boundary doesn't cause my dam to slowly disappear).

Then, two cases :
1) The first one, with ZR=ZF-1 everywhere expect around the dam where ZR=ZF-0.1;
2) The second one, with ZR=ZF-0.1 everywhere.

The two cases end up with the same results... Am I doing something wrong here? It doesn't look like my noerod subroutine really works.

It will make it much easier to help if you upload the steering files for the simulation. It looks like the .f file is not being used. If you are running TELEMAC and SISYPHE coupled, the .f file should be in the TELEMAC steering file. Any link to a .f file in SISYPHE will be ignored.

Michiel

Sorry, I was sure I had uploaded them on an earlier post.

Here they are: the fortran file is called in the TELEMAC steering file (and the subroutine is called, as using write(LU,*) to print some stuff in the shell works as well)...