OpenCrossProfileDef used for IfcSectionedSurface can also reference breaklines captured as IfcBuiltElement

[civilx64]

Section 10.2 states

IFC-Alignment-Geometry-Implementation-Guide/10_Sectioned_Surfaces_and_Solids.md

Line 23 in 1504404

In IFC, stringlines are represented as `IfcOffsetCurveByDistances` instances. The `Tag` attribute on `IfcOffsetCurveByDistances` connects a guide curve to the cross-section points that carry the matching tag value. A tagged point in a cross-section profile is constrained to follow the guide curve carrying the same tag. Throughout this chapter, the industry term *stringline* and the IFC term *guide curve* refer to the same construct — a tagged `IfcOffsetCurveByDistances` — and are used interchangeably.

IfcBuiltElement [1] also includes a Tag attribute. When represented by IfcPolyline or IfcIndexedPolyCurve it is suitable for use as a "string line".

[1] https://ifc43-docs.standards.buildingsmart.org/IFC/RELEASE/IFC4x3/HTML/lexical/IfcBuiltElement.htm

[RickBrice]

Hmmm, interesting thought. From this conversation

https://community.osarch.org/discussion/comment/25421#Comment_25421

it was my understanding that that IfcElement.Tag and IfcProductType.Tag are an entirely different kind of tag than the tags for IfcSectionedSurface and IfcSectionedSolidHorizontal (and the related sweeping curves).

[civilx64]

Here's where I'm coming from: current practice for digital delivery on the roadway side involves 3D breaklines and surfaces that are used for automated machine guidance. We can capture geometry for the former via IfcPolyline or IndexedPolyCurve and the latter with IfcPolyfaceMesh. But what semantic entity should be used? We have guidance and concept templates for IfcAnnotation for existing condition survey data, both points and list of points. But does that extend to design information? And what about surfaces, both survey and design?

That is how I arrived at IfcBuiltElement as the semantic entity, and recently happened to notice that it can carry a tag.

With IfcOffsetCurveByDistances you are carrying redundant information in that the widths are part of the OpenCrossProfileDef but also required to define the curve by offsets.

Also FWIW the workflow that I see in most authoring tools is to define the offset curves in 2D first (offset from IfcCompositeCurve) and then define Z values by cross slope or elevation differences. But maybe that's a discussion for a different chapter.

[RickBrice]

I had a meeting this morning and this exact topic came up in the context of AMG - I'm sure it is related to your work.

Would CT 4.1.7.1.13.3 Surface Tessellation Geometry be suited for this? The break lines AMG needs could be modeled with IfcTriangulatedFaceSet.Coordinates with IfcCartesianPointList3D.TagList. In this case, the TagList would not contain unique tags, but rather like a game of connect the dots.... A break line along a surface would connect all points tagged as "A" and a second line with all points tagged as "B".

TBH, I really know nothing about AMG and ground surface modeling, so maybe I'm way off base.