Denim Umeshkumar Anajwala

While the fluid-flux boundary condition is technically set on nodes, it has to be set to nodes enclosing an entire element face in 3D models to actually work. So to specify an inflow into the aquifer, you'd set it to the nodes in slice 1 and 2 to specify and inflow over the entire vertical element face at the boundary.

The typical workflow would be as follows:
- Export the current mesh to a file (e.g., shp)
- Create a new mesh for the additional part by using a superelement that a) has exactly the number of nodes as the existing mesh at the common border. Use Triangle with the option to not insert Steiner points at the outer boundary to obtain exactly this number of boundary nodes.
- Export the new mesh to a file.
- Import old and new mesh at once.
- Set all parameters again.

This is just a temporary workaround, providing some more flexibility than version 6.0. For 6.2, it is planned to support direct adding of new supermesh polygons to an existing model.

Hi Mingo,

The total simulation time will be identical according to your settings. However, in many cases it is highly recommendable to use the automatic time stepping scheme: Just imagine that any temporal changes of model properties - like well pumping schemes f.e., induce a change in the matrix system. The solver has to deal with this, so "The greater the change, the smaller the time step" is more or less a general rule for solving. When using constant steps, the solution may not reflect such changes and in some cases, the solution will be less accurate or even not converge properly.
As the name indicates, the automatic time stepping scheme does this on its own. Of course, in some cases further adaption may be needed.

A second good reason for choosing automatic time stepping is the budget: If there is a change of pumping between day 40 and 50 f.e., but your saved .dac file only has step 30 and step 60, you will not see this change in the budgeting process properly.

When thinking about time-steps, you may also refer to this matter as temporal resolution of a model. This is quite as important as a sufficient spatial resolution is.

We currently do not have plans to extend the current options here. You may find a solution by modifying either solid or fluid conductivities (or both) accordingly via a plug-in.

Indeed my answer was wrong in the other thread... I hadn't seen that it was referring to thermal conductivity. FEFLOW does not consider the dependency of thermal conductivity on temperature at the moment. The only way to implement it would be the development of a specific plug-in.

My answer above referred to hydraulic conductivity while the questions was for thermal conductivity - sorry for this. FEFLOW does not consider the dependency of thermal conductivity on temperature. The only way at the moment to consider this would be the development of a specific plug-in for FEFLOW.

There's 'Streamlines' and 'Pathlines' now, hereby using the more correct terminology (see for example http://en.wikipedia.org/wiki/Streamlines,_streaklines,_and_pathlines).

When using Richards' equation FEFLOW applies porosity and specific storage (compressibility). Under fully saturated conditions, only specific storage is applied.

Currently there are no functions for accessing these data. You'd have to calculate the thickness within the plug-in from raw z data.

Actually FEFLOW 6.1 should support XYZ points. Feel free to email to support@dhi-wasy.de if it doesn't work. I'm not at the office at the moment so I can't check myself.