Edited triangles and LandXML

InRoads version 615

Triangles in a DTM were edited (flipping and deleting), the surface saved and a LandXML file generated from the surface. The DTM was then opened in Trimble Busines Center. The problem is the original triangles, NOT the edited triangles are displayed in Trimble. Is InRoads retirangulating as it exports to LandXML or do the triangles simply need to be locked before generating a LandXML file? 

Parents
  • The artical addresses triangles deleted from around the perimeter of the surface. I don't see how it will resolve the issue with triangles that are 'flipped' in the interior of the surface.

  • For PowerGEOPAK, here's what I've put together as a workflow recommendation for our users, based partially on work in this and related forums.  I don't know if something similar would make sense for InROADS, as I don't use it. 


    1. Finish editing your source data for your surface (don't edit triangles yet)
    2. Right-click on survey project in Project Explorer “survey” tab, select “deactivate survey processing rules”.  Otherwise, it will show up as “not an editable terrain”.  (Note: either the data is editable, or the terrain is editable.  Not both.)  
    3. Triangles can now be edited.  Survey data is NOT live; points cannot be added or moved and have triangles follow the changes.  
    4. It makes no difference if the terrain model (in the civil model tab of Project Explorer) is locked (rules deactivated) or unlocked (rules activated).  
    5. IF RULES ARE REACTIVATED in the Project Explorer “Survey” tab, ALL TRIANGLE EDITS WILL INSTANTLY REVERT.  
    6. To save a clip boundary to protect the triangle editing along the boundary follow this process.
      1. Exit MicroStation and make a copy of the file containing the terrain.
      2. Open that copy
      3. Select the boundary
      4. Activate the “drop element” tool
      5. In the “Drop Element” dialog, uncheck all boxes except “Application Elements”.  Click a data point to accept.
      6. Export this boundary line, using FileFence or something similar
      7. Exit this file, re-open your original terrain file
      8. Import the new boundary line (ref merge)
      9. Select this line
      10. Open the survey tab in Project Explorer, right-click on “ALL Linear Features” under the appropriate field book, and select “Add graphic linear feature”.  
      11. Left-click on the line and hold to get to the Properties menu.
      12. Change the feature definition to CLIP and the Terrain Model Attribute to “Determine by feature definition”

    RECOMMENDATIONS

    Note that saving a clip boundary using the method outlined here will ONLY preserve your triangle edits along the boundary of the project.  No internal edits will be preserved.  Our suggestion is to NEVER edit the internal triangle structure; rather, you should edit the source data to fix any issues, so that the triangles reflect your intent.  

    Don't forget that one of the best advances in SS3 terrain modeling is that ALL functions can be undone.  If you do something and your triangle edits revert, stop, take a breath, and UNDO.  Now take time to think about what you need to do, and prevent the problem if possible.  

  • Frank,

    Thanks for the suggested workflow.  While it sounds tedious, it will certainly save a lot of headaches preserving deleted edge triangles.

    Much of the internal terrain editing I do involves flipping a lot of triangles so that the resulting contours look better (smoother ground, probably more representative of reality).  I haven't found a way to edit the source data to reproduce this in a terrain.  Have you experienced this?

  • Unknown said:


    RECOMMENDATIONS

    Note that saving a clip boundary using the method outlined here will ONLY preserve your triangle edits along the boundary of the project.  No internal edits will be preserved.  Our suggestion is to NEVER edit the internal triangle structure; rather, you should edit the source data to fix any issues, so that the triangles reflect your intent.  

    Don't forget that one of the best advances in SS3 terrain modeling is that ALL functions can be undone.  If you do something and your triangle edits revert, stop, take a breath, and UNDO.  Now take time to think about what you need to do, and prevent the problem if possible.  

    We are just now being forced to use Inroads for certain DOT work and from what I'm reading here this scares the hell out of me. How could you possibly ever create a surface through source data that was 100% correct when you create your surface?  I have worked on thousands of surfaces and never once have I ever created the surface with 100% accuracy the first time through.  Sure, some are really easy and probably would be okay without my edits but we're talking about miles of highway topo with thousands and thousands of points.  The editing can take quite some time, and now we find out under these conditions they are preserved!  I'm amazed.

  • Mauritzj - Yes, that workflow is VERY tedious.  That's why I say Bentley needs to address this.  Paul Brandl at PennDOT suggested importing the triangles into the DGN so that they would be fixed in place (see above).  This is probably a simpler solution, although it basically makes the terrain less "live", if you know what I mean.  Maybe that's a good thing.  I think the workflow I've adopted still keeps the terrain model "live", for future work. 

    As far as flipping triangles, I guess we don't worry about that as much here.  In the "old days" I worked for an aerial survey firm, where the "look" of the output was of utmost importance.  That dated back to drafting days, and the fact that we were still trying to meet those visual standards.  Here at MDOT, the attitude is more like, "so what if it's ugly, as long as it's accurate".  Obviously this is an oversimplification, but you get the idea.  While I might want to find and fix spikes and holes that shouldn't be there, I don't worry about the look of individual contours.  As a photogrammetrist, my viewpoint has always been that if there's a problem with the triangles, then there's a problem with the underlying data.  (Not counting edge triangles.)  To answer your original question, no, it probably isn't practical to try editing your source data to get triangles to flip.  Paul Brandl's method would preserve your triangle editing work in this regard. 

  • Obviously, we are all still learning the best way to approach this problem.  Please don't take anything I've suggested as gospel - it will likely be a work in progress for some time.  As better ideas evolve, and are put forth, we may well adopt them.  

    Bentley, I hope you're listening!  

Reply Children
No Data