Drake

So I will just post my own learning here to help others.

What I have concluded is that for layered models, the initial smoothing operations after creating the 2D supermesh is the way to help form a good mesh. 

I realize after digging that for layered meshes, "Condition Number" in the auxiliary data (Data Panel), is the better item to review when trying to create a reliable mesh for a layered mesh (compared to a completely unstructured mesh).  I did do this smoothing mentioned above earlier, but found that my model was still showing a lot of spread in the condition number (from 1 to the 1000s). 

Then I realized that this was a problem in only one layer.  While creating the 3D mesh from the 2D mesh, I had a layer that was very thin and in some areas would "pinch" to the minimum thickness defined in the "problem settings">"Editor Settings">"Minimum slice distance" which I had set earlier to 0.1m.  So when I inspected some of the larger 3D elements in my 3D mesh, I realized they were "deformed" in the z direction (they were not a nice "slab" like some of the elements in thicker layers).

So I increased my minimum thickness to 1m, re-applied my elevations, and the popup asking me how to resolve slice elevations to adhere to minimum thickness came up.  I chose bottom up, and let Feflow reassign all the elevations but now with the thicker minimum thickness of 1m.  So now most of the red indicating a bad (high) condition number disappeared, and the scale legend showed max values that are in the 100s instead of 1000s.  Still some work to do, but much better.

In my case I have some leeway to change this minimum thickness. That may not be the case for all models.  I am not sure yet if the model would have trouble converging with these high condition numbers, but hopefully I will not need to find out :)

Hello Björn,

I am fairly new to modelling with Feflow and it is a steep learning curve.

So if I understand, improving our layered mesh using tetgen converts the mesh to unstructured, even if we turn on the switch to "preserve layers and element groups"

Is there a way to improve the mesh of a layered mesh other than Tetgen?  I have reviewed the Auxillary data in the Data panel, and I have quite a bit of spread on the histograms.

Would someone please describe how the conflicts are resolved in these two scenarios of elevation assignment?  The manual does not seem to describe what happens in the relevant "verify assignment" section.

Given a three layer model, I assume bottom up means that the bedrock is set at the defined elevations, and then the overlying soil stratigraphy is checked for minimum distance and this distance is set.  Then the next slice in the overlying strat is set at its elevations, for those nodes where there is not conflict with the underlying bedrock which can now be considered fixed. Finally the overlying soil straigraphy stratigraphy slice is compared to the topography slice to ensure a minimum distance between is upheld.  If this is not upheld, the topography is adjusted upwards.

And top down would be the reverse.  So the topo would be set and everything below pushed down to maintain a minimum distance.

Is this correct?

Also, after assigning the new elevations, I have edges from slice 3 showing above the domain of slice 2.  Is this just a result of a very thin slice in these areas, or is there a possible problem with slice 3 rising above slice 2?

Thank you Christian.  There are numerous areas where similar information is provided online and in the manuals it is hard to know which one has the most current or best information.  I've come across the feflow.info page before but tend to go to other sources (pdf manuals, feflow integrated help) first since the page you linked for us tends to have some display issues on my monitor which oriented 90 degrees (portrait) for reading documents.

Phuntz:  regarding your grain size distribution, I would expect anything with that much clay to not transmit much water (just intuitively).  Is it possible the layer is layered/varved and transmitting the fluid horizontally along the bed of the ditch more like baseflow?  I guess that could end up seeping out the edges if the river depending on the base layers geometry.

Also, have you resolved your modeling issue yet or are you still waiting for answers? 

Pardon, I misunderstood that you knew what D was and were trying to apply it to the model.

DHI may yet correct me, but I would expect that your D is the thickness of the river bed material.  This link is for ModFlow, but maybe it will be of some assistance conceptually:

https://www.waterloohydrogeologic.com/help/vmod-flex/index.html?vm_about_boundary_conditions____t.htm

Click on "River" in blue to get down to the pertinent section.

Out of curiosity, how have you done your node selection for the Cauchy/fluid-transfer boundary condition?  Did you select all the nodes on slice 1 (horizontal node selection) that represent your river bed, or did you select a vertical plane of nodes as is shown in the Feflow manual on page 113?

Also, just so DHI is aware that the question is not answered, phuntz still needs an answer to his/her question in the initial post

Hi phuntz, maybe we can work through some of this together since I am also dealing with 3rd type boundary conditions for a river at the moment and have been doing quite a bit of research into the fairly sparse Feflow documentation.
Hopefully DHI will come back and help us with some of this however.

-Considering your stream/ditch is ephemeral (dries up), and overlies clay, is it necessary to assign boundary conditions to it?  It seem it would not contribute significantly to recharge or discharge and mostly just handle runoff. Especially if the ditch was lined with some form of membrane or geotextile.

-Have you seen that the Fluid transfer BC has limiting conditions available that can be set to avoid to much (unrealistic) loss or gain into the ditch?  Modflow handles this automatically by not letting the head go below the stream bottom by default.  However in Feflow it seems we need to set this manually.  DHI, please correct me if I am wrong.

I will try and provide additional info if I come across anything new.  I know this is not much help and not a definitive answer, but maybe it will spark some thought to help us solve your problem.  I will also keep an eye out for definition of D for you, but suspect it has something to do with thickness since that is what D is frequently used for if I recall.

Cheers,
Drake

Sorry for the explosion of questions.  Result of having to produce a lot in a short period of time.

This one is regarding observation points, and I have several questions related to the topic.

1. When the screen traverses several stratigraphy's, or the hole is open and unscreened, what is everyone's thought of locating the model observation point in the most permeable unit.  In this case, the most permeable unit is sandwiched between a less permeable clay unit above and a low permeability bedrock below. My intuition is that the high permeability until will largely control the head heights (higher pressures when hole is located at low elevations, and lower pressures when the hole is located "uphill"). Maybe my thinking is flawed since I'm a bit rusty these days.
2. [s]When I add observation points to the model, I see the representation of the observation point on top of the original surface topography, but I do not see also represented on the slice I assigned it to (in slice view, or 3D view).[/s] (answered below)
3. Considering an observation point is placed on a slice, which layer controls the pressures observed in the model?
4. When I am converting to observation points, does the snap distance affect anything during the conversion? Just a caution that I am not applying some parameter to the nodes surrounding the wells.  I only bring this up since I am always seeing the observation point information when I move the inspection magnifying glass to various nodes around the well.
5. If I re-assign elevations to a slice, will the observation well adjust its observation point to the newly assigned slice elevation?

P.S: I may have resolved question 2 myself. The zoom distance seems to cause some graphical issues.  When I am zoomed out in the 3D view, I see the crosses (observation points) on top of the terrain.  However, when I zoom in they start to disappear and show up in their proper place on slice 2.

As a helpful fact for myself, and possibly others, I have imported the wells using slice to assign the depth of the point, and then re-exported the observation points which now show up in Feflow's standard table format.  The slice has been complimented with a new column containing  Z elevations which I assumed was the actual elevation of the slice at that coordinate.  This would help, since I may wish to adjust the elevation of the point later on.

Hello again,

I really seam to be having some trouble bringing data from QGIS to Feflow. Feflow often does not like the formats.

My current issue is with DAT files.
I have converted a DEM file to a DAT file inside QGIS, and am constently getting the error "xxxx.dat is invalid" in Feflow.  The operation completed in QGIS without errors.  The Dat file is about 900MB in size.

Any thoughts or solutions?

My work around will be to create a contour map shape file and use that for elevations, however I lose the ability to choose interpolations inside FEFlow when I proceed this way.

Thx

Thanks again Peter,

I've come to the realization that when I import polygons from QGIS, I'm just better off retracing those polygons inside Feflow.  Makes it a bit laborious since it is easier for me to create nested polygons inside QGIS.  But Im getting used to the Feflow split polygon function. I am a little confused why I normally have to create 3 splits however to get the nesting to work.

So far, this has resulted in a clean triangle operation.

Cheers
Drake

Thank you Peter for your reply

I was having issues since I was trying to do these tasks in a fully unstructured mesh. I was also having trouble finding how to reopen the 3d layer config since my 2D mesh got closed and the 3d configure would only become available when I was on that window.

1. Is it possible to select nodes of an element in an unstructured mesh and apply elevations to a single element,  just a portion of an element, or all the nodes below a footprint similar to how this is done for slices.  If so, how is this done?  I'm struggling to find an efficient way to do this like it can be done with slices.

Thanks again
Drake