Christopherus Braun

Gemma,

If you introduce recharge to your modell it is allways freshwater (0 mg/l). This would be the same for injection wells. I would not use injection wells because your modell may become confusing.

There are several ways to define time dependent recharge. You could look up the discussion about "aquifer recharge" in this forum.

zebra

Hector,

did you look through the FEFLOW-Tools? This is a collection of several scripts for data conversion. I don't think you will find exactly what you are looking for, but maybe you could adapt one of them to your workflow.
If you have a lot of timesteps in your dar-file you could maybe get some problems with file width.

Zebra

Hello,

has anybody experiences of programming FEFLOW filters or IFM modules with Python?

Zebra

Hector,

I once thought of developping such an IFM module. But things are a little bit more complicate. The IFM functions are dealing with interpolated velocities at the nodes. This interpolation will give you a more smooth distribution of velocity. Espacially at interfaces where conductivity has an aprubt change (and talking about critical gradients this could be the interesting parts of the modell) this could introduce some errors.
So what you have to do is a calculation of the gradients from the head distribution in each element. In the general case (3D modell maybe with a moving surface) this could be quiet complicate...

So what we do now, we export the FEFLOW results to tecplot. Tecplot360  offers some data evaluation functions which also include computation of gradients on nodes or elements.

Zebra

Hi,

Look in the main menu
--> File
--> Export Filter
--> Superelement mesh

Here you can save the superelement mesh as Shape- or DXF-File.

Hello,

if you use time varying material properties FEFLOW stores this in a temporary file.
By default this temporary file is saved in the "import+export" directory of your standard feflow directory.
I had once the same problem described by you since this directory was a read only directory.

What you should do is:
- Use a standard FEFLOW directory tree (i.e. save the *.fem file in a subdirectory "femdata" and assure that there does exist a directory named "import+export")
- check whether you have the permission to write files in that directory
(Sometimes it is not clear which is the actual standard directory of feflow. To check that go in the main menu  --> tools --> disk space, there you can see the actual path)

If you want to be independent from the standard feflow paths, you have to do some editing in your windows registry:
look for "HKEY_CURRENT_USER/Software/Wasy/FEFLOW/x.x/Global" and edit the „ProjectPath“ to some existing directory.


Zebra

Hello,

the budget analyser will give you a volumetric flow value. In a vertival 2D model FEFLOW assumes that your aquifer is 1 m thick. So you can scale these results to some other thickness easily.

To your question about Genuchten or linear model:
a linear model would not produce realistic results in the unsaturated zone, you should use a Van Genuchten type model. I have quiet good experience with the modified Van Genuchten model. If you have numerical problems during the computation you can adjust here the parameter for capillary pressure and relative permeability seperatly (low values of A and Delta should help to "smooth" your calculations).

Hello,

there are 3 different ways to define a time varying rechege in FEFLOW.

1. Flow Materials --> "T-LIST":
Here you can define different time stages and assign different spatial distributions of recharge. You can use the "cyclic" option, so that you have only to edit one period of recharge and this is reused for later teime periods.
Be careful in the visualization of the recharge, you have to select the proper time step!

2. Definition via power function:
You can difine power functions (in the boundary condition menu) for recharge and join them to the material property data. Also in the power functions menu you could use the "cyclic" option.

3. "formula editor"
For some material properties you can define mathematical functions in the formula editor. Here you can also define a time dependency of recharge (e.g. a sinuos function of time).
There is no way to visualize the distributio nat a given time step in FEFLOW.
Be careful: you have to adjust the timestepping to this function! In opposite to the above described method time stepping is not influenced by this definition.

Zebra

Michael,

instead of using an upwinding scheme, you can also enlarge your dispersion. This should also work and maybe you have more controll over your computation.

The theory of numerical computations would tell you that your simulation should be stable if you use mesh Peclet numbers smaller or equal to 1. (The mesh Peclet number is defined: Pe = ds / alpha, where ds is the length of your element edges and alpha your logitudinal dispersivity).

To my experience you can accept Peclet numbers up to 2, but beyond that numerical problems may begin.

So there are 2 things you can do: assume a higher longitudinal dispersivity (if you do this, you should also enlarge the transversal dispersivity; ratios od longitudinal/tranversal 5:1 - 10:1 should be ok) or refine the mesh.

Nomally the most critical parts of the simulation are arround your transport boundaries where you start the simulation with high concentration gradients. So maybe a mesh refinement arround your wells could help.

Zebra

Hello,

check the FEFLOW Help Menu (Info --> Help --> Batch Mode) you will find a description and an example.