Björn Kaiser

I suggest to define the stages using either Seepage face Boundary Conditions (BC) or WellBC's. You may activate / de-activate / re-activate the BC's of interest by imposing modulation functions. A modulation function is represented by a time-series which is used as a multiplier for the value you imposed. Use a value of 1 in your time-series if the BC is activated and a gap if the BC's is deactivated.

Inactive elements are not available for free & moveable meshes.

I would not set the parameter to zero. However, you can use a small value closed to zero. Please note that this parameter has not only an impact on the physics, but also on the numerical stability. For more details you may also contact the support: https://www.mikepoweredbydhi.com/support

The current implementation allows you to control the field lines by the Property panel (e.g. minimum & maximum travel time) and the Numerical Controls in the Problem Settings (Particle-Tracking Computation). Most likely, this is not exactly what you are looking for.

A workaround which comes into my mind is the following. Export the fieldlines and the water table to GIS formats. ArcGIS has a method called [b] Intersect 3D Line With Multipatch (3D Analyst)[/b]. I never used this method but it seems the method provides what you are looking for: http://desktop.arcgis.com/en/arcmap/10.3/tools/3d-analyst-toolbox/intersect-3d-line-with-multipatch.htm. The fieldlines must be represented as 3D Lines and the water table as the Multipatch. Perhaps you need to manipulate the exported data before you can apply this method.

You can export the K distribution from the fem-file assuming you already saved the fem-file with the optimized K-values returned by PEST. To export the values go to the Data panel and right-click on the K values. Choose [b]Export Data => All Elements => As Center[/b] from the context menu. Manipulate the export file with the external software of interest. Load the manipulated file in FEFLOW and make a Parameter Link as usual.

If you choose General Anisotropy with computed angles you need to enter K_1m, K_2m and K_3m. These conductivities are directed along the axes of a transformed coordinate system. The transformation is done automatically using the inclination of each single element. Using this option, anisotropy can be related to the varying inclination of layers.

More information is provided in the FEFLOW Help: http://www.feflow.info/html/help71/feflow/mainpage.htm#t=08_ProblemSettings%2Fanisotropy_settings.html%3Frhhlterm%3Danisotropy%26rhsyns%3D%2520

The current implementation allows you to add/fill holes by connecting two meshes via nodes. This workflow works for 2D meshes only and Discrete Features, material properties, boundary conditions are lost. 

In general, the workflow is perfectly right. The conversion from "Layer 1" to "Top of Layer 1" + "Bottom of Layer 1" indicates that you re-meshed a 3D layered-based model (3D prismatic elements). The zone you re-meshed is now composed of tetrahedral elements connected to 3D prismatic elements which are located outside your re-meshed zone. The connection between re-meshed tetrahedrons and 3D prismatic elements is done via pyramids. The local unstructured elements (tetrahedrons and pyramids) disappear in the Slice View, because Slices / Layers do not exist anymore. However, you still see the elements outside the re-meshed domain, because 3D prismatic elements are still present. Accordingly, you use hybrid mesh composed of mixed elements.

You can even try to re-mesh the entire model domain. If you do this test, you will see that the Slice View completely disappears. No Slices / Layers are anymore available in fully-unstructured mesh environments. You have to work in the 3D View for fully-unstructured meshes or local re-meshed domains composed of tetrahedrons and pyramids.

If you do not want to work with fully- or partially-unstructured elements, you still can use a layered-based approach. You can refine the mesh locally within the Slice View by using the Mesh toolbar (not the Mesh panel) and drag + snap a mesh node a map point.

If I understand you correctly, you want to allow mass flowing out of the system (model domain) and prohibit mass entering the system. If this is true, you can use a fixed mass Boundary Condition (type Dirichlet) on top of a flow Boundary Condition. Then, you may constraint the mass Boundary Condition with a minimum mass flow of 0 g/d as you already indicated. In this configuration, the Boundary Condition for mass is only active as long as water is entering the system. If water leaves the system, mass is transported out by the advective mass transport component (convective formulation of the transport equation).

Could you please attach the picture to your post?