Need help from Rhino experts.

Discussion in 'Software' started by Alexanov, Oct 29, 2020.

  1. Alexanov
    Joined: Feb 2003
    Posts: 215
    Likes: 27, Points: 38, Legacy Rep: 10
    Location: Norway Sola

    Alexanov Senior Member

    upload_2020-10-29_17-49-40.png
    I have two trimmed surfaces. First surface has 4x4 control points, second one 5x5 control points.
    In second case Rhino can't calculate Area. What can be wrong? Iges file attached.
     

    Attached Files:

    pafurijaz likes this.
  2. pafurijaz
    Joined: Dec 2016
    Posts: 177
    Likes: 22, Points: 28, Legacy Rep: 18
    Location: Italy

    pafurijaz Senior Member

    I tested that with moi3D the little brother, but the file seem corrupted after remke it I can get the surface area.

    here how looks your sample in Moi3D
    Screenshot.png

    and that after remake.
    Screenshot(1).png
    left area
    2153790201,14 mm²
    right area
    2153790200,35 mm²

    here the result from your original file with FreeCAD and moi3d

    1815496126,75 mm²
    1815496126,15 mm²

    and with FreeCAD I have the same result of moi3d but edited one
    2153790201,14 mm²
    2153790200,35 mm²

    But that is a bit strange result and I think the result of original file is wrong.
    and below how looks the file exported as e mesh.
    Screenshot(2).png
    more test and I thinks this file has many issues..
    inspection of original
    Screenshot(3).png
    edited inspection
    Screenshot(4).png
     

    Attached Files:

    Last edited: Oct 29, 2020
  3. Dr34m3r
    Joined: Mar 2012
    Posts: 146
    Likes: 3, Points: 18, Legacy Rep: 10
    Location: Europe

    Dr34m3r Senior Member

    upload_2020-10-29_21-35-27.png

    Its looks like tolerance issue . rebuild the edge with increase tolerance . also note the free naked edge point between these two surface.
     
  4. Alexanov
    Joined: Feb 2003
    Posts: 215
    Likes: 27, Points: 38, Legacy Rep: 10
    Location: Norway Sola

    Alexanov Senior Member

    Thank you! I will check it. What tolerance do you use for it?
     
  5. Dr34m3r
    Joined: Mar 2012
    Posts: 146
    Likes: 3, Points: 18, Legacy Rep: 10
    Location: Europe

    Dr34m3r Senior Member

    i was playing with 0 , probably i used 0.000001
     
  6. pafurijaz
    Joined: Dec 2016
    Posts: 177
    Likes: 22, Points: 28, Legacy Rep: 18
    Location: Italy

    pafurijaz Senior Member

    Hi, I have made further checks on the file and I am convinced that the file has problems. the correct result is the following.

    01 surface area =2492084275,52 mm²
    01 trim area =338294074,39 mm²
    01 trimmed surface area =2153790201,13 mm²

    02 surface area =2492084274,85 mm²
    02 trim area =338294074,20 mm²
    02 trimmed surface area =2153790200,65 mm²
     
    Last edited: Oct 30, 2020
  7. DCockey
    Joined: Oct 2009
    Posts: 4,730
    Likes: 325, Points: 83, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    Alexanov, the triangular hole is the source of the problem. Each surface has an extra set of edges around the triangular hole. The Split command also fails if the split line goes through the triangular hole. The What command in Rhino reports 11 edges when there should be 8 edges in total - 5 around the outside and 3 around the hole. If the hole is eliminated by Untrim then the area computes properly. I don't know why only one surface has the problems. It may be related to the surface with the problem having two spans with the hole located across the knot line.

    I could submit the file McNeel to see if they can identify the problem via the Rhino forum.

    What command results for the surface with the problems:

    surface

    ID: d0dd9b83-4997-4f7c-9a5b-aee13cadaffb (2747)
    Object name: (not named)
    Layer name: IGES level 0
    Render Material:
    source = from layer
    index = -1

    Geometry:
    Valid surface.
    trimmed surface with 2 hole(s).
    NURBS Surface (SubD friendly)
    "U": degree =3 CV count = 5 (0.000 <= U <= 2.000)
    "V": degree =3 CV count = 5 (0.000 <= V <= 2.000)
    Edge Tally:
    11 boundary edges

    Edge Tolerances: 0.000 to 0.000
    median = 0.000 average = 0.000
    Vertex Tolerances: all 0.000
    Render mesh: none present
    Analysis mesh: none present​
     
    pafurijaz likes this.
  8. Alexanov
    Joined: Feb 2003
    Posts: 215
    Likes: 27, Points: 38, Legacy Rep: 10
    Location: Norway Sola

    Alexanov Senior Member

    The Iges file has been generated by Opencascade library. Intersting. To definition of trimmed surface I have to make outer contour an inner contour. May be this is create problem in Rhino. But why it is works different???
     
  9. DCockey
    Joined: Oct 2009
    Posts: 4,730
    Likes: 325, Points: 83, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    The only reason for it to work differently that I can think of is the 4 point surface is single span and the 5 point surface has two spans. Should I ask McNeel (Rhino developers) to look into it?
     
  10. bhnautika
    Joined: Feb 2006
    Posts: 848
    Likes: 55, Points: 38, Legacy Rep: 571
    Location: australia

    bhnautika Senior Member

    David you're on the right track with the points but I think the degree has any influence also.I did a rebuild (see pic) and was able to get a surface area.
     

    Attached Files:

  11. DCockey
    Joined: Oct 2009
    Posts: 4,730
    Likes: 325, Points: 83, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    Both surfaces are degree 3. The difference is the number of control points and corresponding number of knots and spans.
     
  12. bhnautika
    Joined: Feb 2006
    Posts: 848
    Likes: 55, Points: 38, Legacy Rep: 571
    Location: australia

    bhnautika Senior Member

    I did find I could leave the points at 5 but change the degree’s to 4 I could get the area to work
     
  13. DCockey
    Joined: Oct 2009
    Posts: 4,730
    Likes: 325, Points: 83, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    That is useful information. Changing the degree of the 5 point surface from 3 to 4 makes it a single span surface. That is consistent with the problem being associated with the problem surface being multispan.
    (Number of spans = number of control points - degree)
     
  14. pafurijaz
    Joined: Dec 2016
    Posts: 177
    Likes: 22, Points: 28, Legacy Rep: 18
    Location: Italy

    pafurijaz Senior Member

    an answer from a developer...

    It looks like the interior triangle hole boundary is malformed. There should be 3 edges there, in both MoI and Rhino I see there are actually 6 edges. It looks like it's 2 duplicate trim boundaries stacked up on top of each other.

    It also isn't great that it has unusually large numeric coordinate values. It's not good to have coordinates like x = 153093.0 , values that are far from the origin like that will accumulate more rounding errors in floating point arithmetic. But that's minor compared to the malformed inner trim boundary.
     

  15. DCockey
    Joined: Oct 2009
    Posts: 4,730
    Likes: 325, Points: 83, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    You or someone else? A developer of what?
    Same as I reported above. Both surfaces have duplicate trim curves around the holes. But only one surface causes a problem computing area.
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.