PROJECT:   LANL/TA-54 Unstructured 3D Mesh Vers01 (EPLM)

Mesh Information README.txt

Gallery of Selected Images images/gallery_txt.html

Gallery of Mesh Issues: Ximages/gallery_txt.html

Gallery of All Images: images/gallery.html

Vers02 Mesh (on hold): images_workv02/gallery.html

Mesh Type: 3D Unstructured Tetrahedral
Number of Nodes:   452,618
Number of Elements:   2,582,936
Features: Regional Watertable, Tb4 (Basalt), and 83 wells with top and bottom screens
Data Source: Watertable Surface from Weston EV draft WC09e May 25 2011
Data Source: Bottom Cerros del Rio (Tb4) from Weston EV draft WC09e May 25 2011
Data Source: 83 Wells and their screens from MYSQL extracted Jul 12 2011
XYZ min max length: (490000.00 510800.00 20800.00) (526697.87 543000.00 16302.13) (1100.00 2227.64 1127.64) meters
XYZ min max length: (1607611.55 1675853.02 68241.47) (1728011.38 1781496.06 53484.68) (3608.92 7308.53 3699.60) feet


This mesh was built with LaGriT strategies that capture the resolution at 83 wells and the basalt and watertable interfaces, but also keep the total number of nodes below 500,000. The result captures resolution in important areas along the basalt and well locations - but only coarsely captures features such as the thin basalt arm on the west. This mesh is considered a first draft and new strategies can be used to improve the resolution and the mesh quality.

Mesh issues to address in next versions:

Using old version water table surface wt_surface_ev_WC09.inp 
update to Aug 12 2011 From directory /scratch/er/monty/lanl/water table/maps/2011-02
cp wl*feb*811*ascii.grd /scratch/er/tam/er_monitoring_2011/watertables/wl_2011_feb_s_tw_20

Fix extreme elevation at intersectin of polygon and TB4 surface that creates a false spike
and bad ccoefs at elements at that edge. This is caused by TB4 "cliffs" where values past
TB4 boundary are very high or very low to represent the null data for the material.
The TB4 surface and water table surface should be clean and be coincident in the null area.

Negative coupling ccoefs along steeper west slope of water table surface (410 m edges).
- Reduce edge length of triangles to 230 - 300 meters to avoid problems in slope area (10
meter from next layer)
- Consider adding a layer of nodes above the water table

Negative coupling ccoefs at some wells along steep slope because the well object
is horizontal and does not form good connections where slopes are steep.
Use stack method to add layers to keep well columns (do not use offset)

Negative coupling ccoefs along TB4/water table interface caused by wedge shapes in pinchout. 
   - Form a 10 meter high pinchout, blunt edge to TB4.
   - push TB4 zic down to 10m buffer layer ( wtr - 10 )
   - color tets by coarse hex and polygon fence

Fix well x,y locations that differ from mysql data
- R-53 has 2 different x,y locations 7 meters apart
- R-41 and possible others suffer from smoothing that should not have happened.
  8 and 24 m radial well pts should have true radial distance
- Need automatic way to deal with well objects that overlap such that one
  well will "win" over the previous and remove points that are too close to winning well
- Need automatic or script to buffer vertical well screens 
  and perhaps add 1 meter vertical resolution to pumping well objects

What to do with well points that exist very close but not the same as mesh point.
- Remove mesh well wheel, add real well wheel? Or remove real well wheel, keep mesh wheel?
- What is "too close"? 1 meter? 8 meters?

Buffer the Tb4 polygon and remove interface points. 
- need to create buffer polygons + - about 5 meters
- should have strategy or code to coarsen any given polygon to some edge length h


The mesh domain is bounded on the east and south by the Infiltration boundary. It is cropped at the west edge at X = 490000 meters (1607611.55 feet) and on the north by Y = 543000 meters (1781496.06 feet). The top of the mesh is truncated by 2010? watertable and the bottom is flat at 1100 meters (3608.92 feet).

Horizontal spacing is based on a triangulated template with edge lengths near 400 meters away from Basalts and wells, and near 40 meters where Basalts and wells meet. Vertical spacing is based on the watertable and basalt surfaces. They each have a 10 meter "buffer" layer. Layers are added within the basalt volume and beneath the watertable down to the flat bottom. The vertical spacing is larger as the layer is deeper except along elevations which have the deep well screens.

Wells are represented as cylinders with 8 meter distance from center well point to surrounding cylinder wall. The 8 meter wall is buffered by additional walls at 24 and 56 meters. These outer walls are influenced my smoothing algorithms and may not have these exact distances. Each cylinder center point is located at 83 x,y well locations. Center points are also included at each top and bottom screen on the vertical line of that well.

Local File Locations

FEHMN Modeling files:

Mesh Geometry and Connectivity (AVS ascii format)
 -rw-r--r-- 1 tam pf 395M Jul 29 13:55 tet_fehm.inp

Mesh Geometry and Connectivity (FEHM ascii format)
 -rw-r--r-- 1 tam pf 132M Jul 29 13:53 tet.fehmn

Sparse Matrix Voronoi Coefficients (FEHM ascii format)
 -rw-r--r-- 1 tam pf 130M Jul 29 13:56 tet_zero.stor  

FEHM zone files
 -rw-r--r-- 1 tam pf 4.8M Jul 29 13:53
 -rw-r--r-- 1 tam pf 574K Jul 29 13:53
 -rw-r--r-- 1 tam pf 188K Jul 28 08:32
 -rw-r--r-- 1 tam pf 188K Jul 28 08:32
 -rw-r--r-- 1 tam pf  37K Jul 28 08:32
 -rw-r--r-- 1 tam pf  30K Jul 28 08:32
 -rw-r--r-- 1 tam pf  24K Jul 28 08:32
 -rw-r--r-- 1 tam pf  12K Jul 28 08:32

Zones for each node as top screen and bottom screen
-rw-r--r--  1 tam  pf   7665 Aug  8 13:48
-rw-r--r--  1 tam  pf   6750 Aug  8 13:48

Zones for all nodes in each screen interval between top and bottom
-rw-r--r--  1 tam  pf  11144 Aug  8 13:48

Mesh Screen Information
  -rw-r--r-- 1 tam pf 4.5K Jul 29 13:29 tops_mat_elev_node.dat
  -rw-r--r-- 1 tam pf 5.5K Jul 29 13:29 bots_mat_elev_node.dat
  -rw-r--r-- 1 tam pf  11K Jul 29 13:39 tops_vorvol_coords.dat
  -rw-r--r-- 1 tam pf  13K Jul 29 13:41 bots_vorvol_coords.dat

Input Screen Data
 -rw-r--r-- 1 tam pf  27K Jul 28 14:17 well_screens_bot.dat
 -rw-r--r-- 1 tam pf  22K Jul 28 14:17 well_screens_top.dat

Well data:

Earthvision Surface data: 

Mesh work directory: 


Grid Team Contacts

Carl Gable gable -at- lanl -dot- gov 505-665-3533
Terry Miller tamiller -at- lanl -dot- gov 505-667-8009