geostatnoob

My partial fix was to get a Nvidia (6800) OpenGL video card with 512mb of memory and up my system memory to 2gigs. Well it fixes it a little. I revert back to old Rockplot3D's from about 1 year ago when this is not enough video memory and I don't have this problem but it is alot slower. High resolution is the reason I switched to using arcscene to view my RockPlot3D models. Don't bother with an ATI video card as their opengl port sucks. You can also trick the 6800 to think it is a Quadro FX 4000+.

Basically my question has to deal with the contour intervals. Right now in the plume caps the contours are drawn with hundreads of triangles when it could be just simply a line/polygon per contour interval. I wonder if Autocad 2006 has dealt with the problem of not being able to import surface (isosurface etc.) with more than around 50,000 polygons. Seems to be about the max with autocad. Many of my plumes have on the order of 100,000~200,000 polygon/triangles and really tax the system. My research involves the direct sensing of data at 1" intervals or less for ultra high resolution contaminant mapping.

This is not an option. As direct sensing technology logs in higher resolution and computing power increases software will need to keep up with this.

I have a couple of questions. Is it possible to make a p-data ESRI shapefile cross section (3D fenceplot) not have so many triangles? The ESRI exported Stratigraphic cross sections do not have any triangles in them. It would be nice if the exported cross section also did not because it takes alot of resources in Arcscene and it would look nicer too because some of these triangles are not situated along the cross section plane. Also the contour caps on P-data or I-Data plume models could be made so there are not any triangles if possible. It would help the resources in ArcScene. I export plan maps from Rockworks in .dxf format then convert to a shapefile. IN arcgis you can do a merge and it removes all the triangles and in doing so reduces the size of the shapefile greatly. This is difficult or impossible to do in the z direction. Check my attachments to see what I mean.

Okay I will try this.

Thanks this should work.

Rockworks seems to have to reinterpolate the 3D stratigraphic model everytime you make a cross section or any 3D model. Why is there not a way to use a already produced .mod or .grd. For instance I like to use a different kriging software to prooduce 2D grid models. It would be nice if I could import this into a rockworks grid and then use it for my strat model. Rockworks does import this and convert this format into a rockworks .grd file.

I have both a geologic and plume model. I need to overlay them in a cross section diagram. Does rockworks do this?

If you go to this page and look you can see a 2D grid. If you go into the literature about this the 2D grid is basically interpolated from collapsed boreholes to max at that given xy. http://www.ctech.com/industry_solutions/an..._assessment.htm

That is too bad about the graphics. This is why I export everything to arcscene. How ever lately it seems they changed things and they made RockPlot 3D hardware accelerated. I have 512mb professional video cards in my machines running 2GB+ of ram. The biggest benefit to me is the ability of importing gslib or kriged 3D grids. I don't use the xyzc import function. I simply make sure both RockWorks settings are the same as the other kriging programs and then make a dummy .mod file and paste the c or variable data into the ascii .mod file. Long live the ascii 3D grid.

One function that would be nice to have would be a max of borehole data and then grid that to a contour map. So xyc contour with c = max concentration at a non specified level z of the borehole log. Basically colapsing borehole data to a max at that xy. This is nice for work in the field and since it can be kriged in rockworks then geostatistics can be done to help optimize placement of additional boreholes.

I am importing an ascii file that is about 300,000 rows long and 4 columns wide ascii xyzg file. It has taken about 10 minutes and ran my processor at 100% the whole time. Surfer 8 and Minitab 14 can do the import in less than 30 seconds. I was wondering if this could be tweaked or maybe improved in newer versions. I am importing 3D grid data from other kriging software and want to do a warp to grid. Rockworks has the best 3D shapefile exporter I can find and I like the new benefits of it being ported to a true database. Maybe this would work better if I imported from a dbf? I have not tried the new rockworks much yet.

Look at my attachment. The 2D grid is larger than the 3D grid. As you can see by the picture the middle is being pushed down too much and the edges are not being pushed down enough. It appears to be warping the whole grid but not together as a solid piece. Also wanted to add.... You should add the ability to warp a already interpolated grid without going through the hassle of converting to a xyzg file and then using closest point. I am not sure if that even works this way. I tried it last night and the grid was not even warped at the end. Also It takes too long. It takes about 10 times longer to try and warp a xyzg ascii 3D grid than it does to use the warp to grid function when interpolating otherwise.