psinton@aquageo.us

Hi Didier,  FEFLOW maintains a minimum saturated thickness (0.01 meters?) in all elements because there is no such thing as a no-flow cell like in MODFLOW.  The in/out flow on top is applied the top layer of the model no matter what mode is used.  This can cause problems in the simulation if the layers vary in elevation much more than the observed water table.  If you have to model such elevational changes, such as in a mountainous region, I recommend using 3D variably-saturated mode instead of phreatic.  If you chose to use this mode, you will have to specify unsaturated properties that are compatible with your model setup.

Pete

I see, Perhaps and additional layer or elements or localized zone of elements to help isolate the DE and BC from the shallower aquifer would help.

Francesca,  from your description it seems the flow is never sees the light of day.  Its not really a spring...?  If the flow is internal, your model should have a hydraulic connection between the two aquifers that is perhaps localized.  Perhaps discrete features or 3d elements that have large K. Check the flow between the two aquifers using the flux analyzer rather than the budget analyzer.  If it is a spring, you'll need to separate boundary conditions to simulate the discharge and the subsequent recharge.

Try a 1st-kind BC with constraint set to allow only outflow; should work fine.

Pete

Thanks Peter, moving the feflow window is the one thing I didn't try! :o

I think I am wrong about using discrete features in the vadose zone. I think FEFLOW automatically disables them in the vadose zone due to the assumption of confined conditions (but perhaps this is only for the features that rely on Darcy's law).  If you cannot use DF, you'll have to define channels of hi-k elements.  Probably will also need very fine mesh and time stepping.

In both standard and classic, I cannot see the pressure or saturation diagrams while viewing a dac file.  In the drop-down menu used to switch these on and off (standard gui), these diagrams may be indicated as active but they are nowhere to be seen, even after removing as many windows, panels, etc as possible.  Switch off and back on is no help, neither is a reboot.  But it only occurs on one machine...perhaps the graphics driver?  No, I updated today and still have invisible diagrams. Perhaps it is a setting in a file somewhere?

try removing the free and movable option...run confined to simplify the problem.  its likely that your model set up is ill posed.

Pete

This is a transient model, the negative balance number means there is water removed from the model volume, and amount of inflow is not enough to maintain the original volume of water in the model.  The water levels in your model should be declining.  Eventually, the water levels should stabilize; it is possible that the model will dry out.  If the water levels are not declining (unchanging or oscillating up/down or rising) then your model is probably unstable and has a major problem that needs fixing before it will work properly.  An important factor is the initial water levels you specified.

Pete

You can use the discrete features, which have some channel flow capability (these should work OK if the model is very large compared to the karst channels), or you can use 3d elements with large K values (ok for large or small models).  Or you can program up a fancy ifm module.

Pete