Rhino plug-in for mesh - limitation

[Patrick M]

In recent tests, we (Vincent) was using Rhino+Grasshopper to simplify and unify DWG contours, elevated in Rhino, then run through Grasshopper again to get a mesh into Archicad. 

We are seeing a major limitation to the Grasshopper plug-in: that only points can be plugged into the mesh script; what we really need is "user defined ridges" as an option. 

Anyone have a work-around to get real contours onto a mesh with Rhino? Any GS developers have any influence on expanding this part of the GH plug-in to allow for more usable/versatile/flexible mesh development?

I'll add that Vincent Todosiev (not myself) has been putting in a lot of leg work to figure out this workflow, so I hope he'll be able to chime in here to give more details on what he's done. These are just screen shots I snapped from our last meeting.

   

Operating system used: Mac Apple Silicon Sonoma

[Botonis]

Just curious why you need to create the contour lines by adding so many mesh points. Contour lines are just theoritical after "scanning" the area.

For me making such a complicated mesh may lead to a slower project. There are other simpler solutions like the contour line object in the bimcomponents page... Unless this is a case study with Rhino which looks interesting! 

Greetings. 

[thesleepofreason]

Having contours as direct input would be nice but I guess it comes down to the development of the mesh component. 

But to be sure - is it the Rhino/GH point generation that is an issue? I usually use an approach where points are generated at the intersection of  the rays from a point grid and a mesh generated from the contours - sort of LIDARing the 3D model 🤔. The number of grid segments gives control of the resolution.

[Patrick M]

we do detailed site meshes; at least for points closest to our buildable area. I've done extensive tests on polygon thresholds for various elements, and as long as we keep all mesh elements under 100,000 polygons, we do not see them as a factor in file performance.

the whole point of using Grasshopper/rhino is to take a surveys contours that often have excessive points, as you've mentioned, and reduce them down to something reasonable (a contour node every 5-20 feet instead of every 1/2 inch). I understand the imprecision of surveyors provided data, but it is what we have to work from. This rhino workflow saves a ton of time over manually tracing over excessively complicated contours with splines or directly with the mesh tool.

it's all pointless however, if we can't get the user defined ridges into archicad from the archicad/grasshopper plug-in

[Patrick M]

no, we can take the contours (dwg from surveyor) into rhino, then use grasshopper to unify and simplify those contours, and Vincent even set up a grasshopper command to check for polygon complexity before bringing into AC. the only wall we hit is that the plug-in from grasshopper to archicad only allows for meshes to be defined by points, not ridges/contours. so all the contours are turned to a series of single points and we get no floor plan contours for our site plan  documents.

[thesleepofreason]

Is there any specific reason for wanting the contours as user defined ridges, apart from the simplicity of just one element? I usually just take them as polylines from Rhino/GH.

[Botonis]

Just asking to check if my workflow is wrong. Is using Rhino a matter or 2d presentattion, a better mesh quality or else?

I am taking the survey data in a txt file with a xyz format. This is all the available realistic data you get.Several random points in a plot according to any special characteristics of the area. Those points do not represent contour lines.

After the mesh is created I am using the Mesh contour object to create the contours. 

There are other solutions, not free, like Architera or LAND4 for contours with more settings.

In general I try not to go out Archicad and use other software.

[Patrick M]

separating 2d and 3d is kind of antithetical to BIM-ness. There are definitely work-arounds, and I'm not even beyond doing a site plan 100% 2d on a worksheet; but in an ideal situation (always what we are aiming for), 2d elements are used as little as possible, especially where there is a 3d counterpart, and especially on plans, sections, elevations.
there are obvious exceptions, but wasn't looking to publish a thesis on site plan and site modeling, just looking for some direction on getting contours into a site mesh with user defined ridges through the grasshopper plug-in.

[Patrick M]

using rhino to limit polygons and streamline translation of 2d to 3d. Survey data from txt or xyz has always resulted in a workflow needing excessive formatting and insanely high polygon mesh elements. Rhino and grasshopper make short work of taking a survey with 250,000 contour points and cutting it to 25,000 without losing significant detail or resulting in contours that cross each other, overly straighten curves, etc.
the next step, taking those contours to AC and tracing/magic-wanding with the mesh is pretty straight forward, but the grasshopper/AC connection allows us to take it into AC as a mesh directly; just bummed that it only brings it in as points, which is usesless for anyone who may need to adjust the mesh.

To the point of having the contours as 2d overlay, thats just more coordination and more points for errors and miscoordination, and does not eliminate the high likelihood of needing to adjust contours, which isn't just tedius, but is impractical for a model of mesh points with no user defined ridges.

[thesleepofreason]

Agree - unfortunately GS doesn’t seem too interested in developing a framework for proper BIM. Let’s hope for atleast an improvement to the mesh tool and its GH component.