2D lines and smartlines for DTM

Hey all:

I have a .dgn file that contains a mixture of 2D lines and 2D smartlines that represent the proposed contours. What is the best procedure to convert these for using in creating a DTM?

Using Geopak Eng Suite V8i.

Thanks!

Parents
  • Hi,

    2D linear elements can be used to build a DTM within GEOPAK using the DTM>Extract Graphics (Mode = Interpolation).

    To accomplish this, the software needs two (2) pieces of information. One (1) the linear elements AND two (2) text values that represent the elevations at which the linear elements are expected to be set at.

    Do you have text values on (or near) the linear elements that designate the expected element elevations?

    If yes; using the Interpolation Mode, you set the symbology of the linear elements and, in a separate panel on the dialog, set the symbology for the elevation text and then "extract" to create the DAT file. When the DAT file is complete (you can append as much data as necessary), you will triangulate the DAT (DTM>Build DTM) to build the TIN file that is the basis for design within GEOPAK.

    You will note that there is an "extract tolerance" at the bottom of the dialog. This is a scan distance that tells the software how far to scan from the vertices of the linear elements in an effort to find a text value that designates the vertice elevation. The scan is measure to the justification point of the text.

    HTHs,

    Michael

  • Hi Michael:

    1. should the feature be set to contours?
    2. should stroking be used?
    3. the help menu states "for contours within a 2D file, the associated elevation must be within the Con Z-range"... what does this mean?
    4. is only one text per string of connected lines required?
    5. does every line need a text, or will it interpolate if say every other line has text?
    6. under spots criteria does the types field need to be checked? and what does types mean? and should content be selected?
    7. the help menu states for content - "the x,y coords are derived from the text origin"..."however the z value is not the z value within a 3D file, but the actual value of the text, whether a 2D or 3D file"
    8. if extract is set to View 1, it seems the software only finds the elements visible on the screen and a fit view should be done before apply
    9. is this the best method to use or is it better to assign z's to all the lines and use extraction?

    thanks for your help!

     

    greg

  • Hi,

    1) If the extract elements represent "contours" then you should set them as a contour feature.

    2) Stroking: Depends on how many data points you have on your graphical elements. Typically, if the contour elements were digitized, they will have more vertices than you would ever need so you would not use stroking (and may even consider "filtering" the points once the DAT has been created to reduce the amount of data). If the data is b-splines or mostly curves/arcs, you may want to add "curve" stroking to densify the data along the cureved elements so the eventual triangulations closly follows the curves. Its a tough call to respond on without seeing the data you have.

    3) Absolutely no idea :-)... I'd have to find the context of that line in the Help before I could explain that one.

    4) Yes. The single value will be applied to all connected elements.

    5) Every line needs text. It will not interpolate between them. No text; and the lines will be ignored.

    6) Yes. You will need to set it to "text" and then set it to "content" (drop down option just below the types field). This tells to software to look for an element type of text and then to read the text content as the elevation to be applied.

    7) This is true IF you set the "content" option. If you leave the option as Origin, in a 2D file you will get zero as the elevation (as zero is the Z value of the origin of any element in a 2D file).

    8) Correct!

    9) This is the only method... if the elements are 2D (as in your case) there is a certain amount of data required for any software to assign an elevation automatically. There are manual methods where you could pick elements and assign elevations on an "element by element" basis; but I'm reasonably confidenet you don't want to go down that path...

    HTHs.

     

  • 3. The interpolation mode is only valid for Feature Types: Breaks, Contours, and Break Voids. For contours within a 2D file, the associated elevation must be within the Con Z-range, more commonly known as "tagged." Elevations for spot elevations may be determined by textual value (if the origin of the text is the origin of the spot) or "tagged" elevations for other element types, i.e., zero length lines or cells.

     that's the full quote from...

    Bentley GEOPAK Site V8i/command reference/DTM tools/Ditigal Terrain Modeling/DTM Tools/Extract Tools/Extraction of Graphic Elements...around the 3rd paragraph

  • Hi again,

    I see that explanation in the Help and the only assumption I can make is that the tool has changed since that Help was written. It's certainly not valid now.

    If you'd like to send your data too me I'd happy to take a look at what you have and let you know where you stand (and what dialog settings will/will not work.

    ZIP the DGN and send it to michael.gilham@bentley.com

    HTHs

  • When I do the proposed contours for parking lots, I prefere to use lines for the proposed contours since the parking lots vary so much and for the many changes that will come. After I've created the proposed DTM is there anyway to make the parking lot an object in a model so that I can do the side slopes, ponds and roads by objects in the same model?

    Greg

  • Hi Greg,

    I'll tentaively answer YES.

    Once you build the TIN from your graphical elements (as discussed earlier in this thread), you can import that as an Object into Modeler.

    Then:

    1) Create a new Object (type Parking Lot would make sense).

    2) Drape the graphical elements that make up your Parking Lot onto that Object that was created by importing the TIN.

    3) Now your PL Object has the graphical elements and it should merge in your Site Project so you can do the other Objects (ponds, buildings, roads etc).

     

    This does seem a more difficult way to do things. I'd typically suggest you design the Parking Lot inside Modeler as well as the other objects. While I've always found grading linear elements using the Site Modeler tools the most efficient way to design Parking Lots, you can still use contours if thats the way you want to design... Just add the contours within Modeler to the PL object.

    Does this help?

Reply
  • Hi Greg,

    I'll tentaively answer YES.

    Once you build the TIN from your graphical elements (as discussed earlier in this thread), you can import that as an Object into Modeler.

    Then:

    1) Create a new Object (type Parking Lot would make sense).

    2) Drape the graphical elements that make up your Parking Lot onto that Object that was created by importing the TIN.

    3) Now your PL Object has the graphical elements and it should merge in your Site Project so you can do the other Objects (ponds, buildings, roads etc).

     

    This does seem a more difficult way to do things. I'd typically suggest you design the Parking Lot inside Modeler as well as the other objects. While I've always found grading linear elements using the Site Modeler tools the most efficient way to design Parking Lots, you can still use contours if thats the way you want to design... Just add the contours within Modeler to the PL object.

    Does this help?

Children