Denim Umeshkumar Anajwala

You should try to refine the mesh vertically to get a better spatial resolution.

This is a difficult question. On the one hand, you can decide based on the result (practical approach), i.e., you do the interpolation and check the result for plausibility. On the other hand, you can try do a more theoretically based decision. For this, please refer to related literature about geostatistics.

For recharge with a specified concentration you should use the 2nd kind boundary condition for mass transport all over the top of the model. However, this REQUIRES THE USE OF  THE DIVERGENCE FORM OF THE TRANSPORT EQUATION to obtain the correct boundary fluxes. Please have a look at the White Paper "About the difference between the convective form and the divergence form of the transport equation".

Aurélie,
please see http://www.feflow.info/317.html for a little animation about the usage of the 1D linear interpolation.

Peter

The reason why FEFLOW closes the supermesh polygons with before setting the 4th point is that you have chosen quad meshing in the main menu. Only triangular meshing supports superelements with more than four sides.

I'd typically not force this concentration, but set it as an initial concentration everywhere and fix the concentration also at inflowing boundaries.

The unit is 10-4/d in 3D models, 2D horizontal unconfined models, and cross-sectional models. In 2D confined models, the boundary condition is only set along a line, therefore the transfer rate has to e set as a depth integrated value. For this, the unit is 10-4 m/d.

Wir haben einige Benchmarks auf verschiedenen Maschinen gefahren, um die implementierten Methoden zu testen. Allerdings hängt der Gewinn sehr stark vom Modelltyp ab. Insbesondere ist entscheidend, welcher Teil des Gesamtaufwandes für die unterschiedlichen Schritte während der Simulation anfällt (Matrix-Assembly, Lösung, Geschwindigkeitsberechnung, Bildschirmupdate, ...).
Der Satz heißt in etwa:
"Eine effiziente Methode zur Gebietsaufteilung und OMP-Programmiermethoden wurden eingeführt".
Details über OMP finden Sie z. B. auf http://de.wikipedia.org/wiki/OpenMP. Die Gebietsaufteilung ist bei vielen Prozessoren sinnvoll, um gegenseitige Blockaden der verschiedenen Prozesse durch Warten aufeinander zu vermeiden. Das Zusammenfassen naheliegender Knoten im Modellnetz auf einem Prozesser vermeidet dabei zu häufigen Datenaustausch zwischen den unterschiedlichen Prozessen.

Currently you cannot show both in the same view inside FEFLOW. However, you can 'capture' the layer view, then visualize a single isoline for a pressure 0 in the results and overlay both in FEFLOW's plotting software FEPLOT.

Some time ago, a bug was detected for dac files with (exactly!) 100 or a multiplier of 100 time steps. This bug has been fixed. Please try to install the latest patch version of FEFLOW.