Viv

Hi,

My goal is to simulate and observe hydraulic heads and out flows in a cubic tank discharge model with a double porosity continuum.
In order to illustrate the model, the problem would be a tank filled with 4 "towers" of rocky material (matrix) separated from 1) each other and 2) the tank border by an aperture of size X (fracture).
In this model, the discharge would take place through a hole located at the bottom of the tank, in the middle of one side.

I have some questions regarding with my FEFLOW (5.3) models 1) the boundary condition I should have at the outlet, 2) the free surface parameter, and 3) discrete feature elements and a 3D approach.

I tried to work step by step from a simple discharge of a homogeneous material through a node 2D model to a 3D model with higher conductivities areas.

Here are my questions :
- In a 2D horizontal model (*.fem file attached), considering a discharge in the case of 2 conductivities media. Top elevation of the aquifer is 0.30 [m], bottom is 0 [m], and a square mesh of 0.30x0.30 [m]. I change the "dry" parameter (residual water depth) to 0.000101 [m] in order to set a fixed head boundary (1st kind BC) on an outer border node at 0.001[m], because 0 isn't working right?

- I tried to use a discrete feature element to represent the fracture medium but it gets messy, is this changing a lot from a 2 conductivities approach?

- Would it improve a lot the first 2D results to switch to a 3D problem? I tried to, but I am not sure of how many slices I should use (3 or more?) and 1) which [i]"free surface"[/i] condition I should use in this case [i]"constrained"[/i] with a [i]"seepage face"[/i] ?, 2)Is [i]"phreatic"[/i] recommended for the top slice ?


Thanks for your answers!


Vivien

Hello everyone,

I tried to run Feflow with a 3D model on a linux based installation in command line mode but I get this error :

[quote]ERROR: Element parameter #134 (Anisotropy angle phi (3D)) was used out of scope![/quote]

I also ran this model on a win based installation and I never came across this error with the same model.

I talked with another  Feflow user and the same error occurs when Feflow is launched with a command line in a linux environnement.

We use different linux distributions, but the same Feflow version (5.4), and of course we are both working on different models.


Has anyone had and fixed the same kind of problem?

Thank you.
Vivien.