scubohuntr

I link to maps a lot, but I can never get the node assignment to work. If you're having trouble with it, try slice instead of node. Not as helpful for point drawdowns, as it averages several nodes, but better than nothing at all.

The .dac files are huge, no doubt (2-15 GB), and unwieldy, and that very likely explains the chart issue, but I ran into the obs point issue in the .fem file (which runs under 50 MB) before I ever ran the model.

I have set the time steps as large as I can (up to 180 days) in the ~90 year simulation, but I need to characterize the drawdown and recovery curves so I need all the time steps. Of course, I only need all the time steps for a relative handful of nodes, but there's no way to selectively limit the .dac file. If there was some way to export fluid rate budgets without going through the chart, this would only be one of a long parade of small annoyances.

My model is a bit over 1.4M elements and runs out in 250-300 time steps. Large model, but not unreasonably so. I'm running it on a dual six-core machine with 32 GB of RAM, so it ought to have the horsepower to do it.

FEFLOW 6.1 is locking up on me quite frequently. Almost every time I try to right click on a chart (apparently the only way to export temporal data) and most of the time when I try to edit observation points, the program freezes. Usually if I hit ctrl-alt-del and cancel the reboot it unfreezes it, until I try to do the same thing again. It's that "almost" that makes this really frustrating. It works often enough to maintain hope, but I am spending a LOT of my time trying to unlock frozen software. I have reinstalled FEFLOW, and I have tried several different machines to no avail. When it is locked up, the mouse moves and obviously the keyboard remains active, since ctrl-alt-del still functions, but I can't click on anything, and alt-tab will bring up the dialog but won't switch between active programs. I haven't seen the problem doing anything else yet, but right clicking charts or editing obs points will lock it up most of the time. Anyone have any suggestions?

I've always exported the slice elevations first (that way you get a .shp file with the node numbers and all the right formats), then do a surface spot on a raster map in 3D Analyst to get the elevations you want. Now import the shapefile and link the "spot" field to elevation. Make sure you have the correct slice and the appropriate nodes are selected, hit the little check mark, and it ought to import elevations. You can use other formats, but shapefiles are by far the easiest way to do it.

Thanks Pete!

I got it to work, after rather severely oversimplifying my supermesh. It won't work with line features of any kind, but as long as I use the lines to split polygons I still get the same general effect. I do wish there was a way to reset the segmentation fault flag without completely exiting the program and restarting everything, though.

I'm trying to put together a presentation on a model I'm developing in the v. 6.1 (Beta 2) interface. When I try to use the "Snapshot of Active View" tool, I get an error message "invalid vector <T> subscript". I know I probably won't get any solutions in time to save tomorrow's presentation, but maybe it can help someone else. Since the only other option to copy output is screenshots, it makes for a much less elegant presentation, and eliminates any possibility of multilayering a slide. This worked last week, before I updated the software.

Hello,

I am trying to generate mesh for a large (~25 000 square km) regional flow model. I have imported all my polygons, lines, and points from GIS and gotten the supermesh beaten into submission. I am trying to use Triangle with an initial guess of 150 000 elements, q-switch=20, i-switch=Divide and Conquer, Y-switch= Allow at all mesh segments, refinement gradations about medium. I get an inscrutable error message: "Error: Segmentation fault during TRIANGLE mesh generation!" Nothing about where this fault may be or how to go about addressing the problem. Gridbuilder just crashes, and Advancing Front makes unacceptable meshes due to not considering points and lines. Anyone have any insights?

Hello,

I am working with a large regional model in support of an Environmental Impact Assessment.  I have had an isue with head recovery after cessation of pumping, which was much improved by tightening the iteration termination criterion from the default of 1E-8 to 1E-10.

Hopefully the attachment worked, in which case the original curve is in blue and the new curve is in red.  If it didn't work, not a problem, the curves themselves are not really essential to the issue.

What I need is a good brief explanation for non-modellers as to how this criterion affects the modelling process.  Does anyone have a reference for a good non-technical overview of this topic?  The FEFLOW manual and white papers aren't helpful.  I can explain it to modellers or mathematicians, but I am at a bit of a loss for how to convince a non-technical reader that changing this setting does not change the model itself, without oversimplifying it to the point that I am providing inaccurate or misleading information.  I don't want to write a textbook, which is what my current efforts are beginning to look like.  A reference to literature would be good CYA insurance, I think.

Thanks!