Solid by extrution from 2D cell in GC

Question

Hej ppl, 
 having another question about GC again. 
 We have our stand steel cross section as 2d cell libary, which can be placed in mircostation or with a GC node type 'cell'. 
 But if I want to extrude 2d cell into a solid, I couldnt find the correct method/option in solid node type. 
 please look at the attached pic here: 
 
 those steel cells are placed as 2d cell. I can extrude them manually by extrudtion tool, but GC node seems not really work..... 
 any ideas?? thanks 
 Leo

Thomas Mahon · Accepted Answer

You need to extract ('promote') the curves from the cell first, then perform the solid extrusion. You cant extrude a cell object in GC. Also, if you have object types that are not promotable using the method I'm showing below, you'll need to do an exhaustive search for them using other element types then join them back together again:

Thomas Mahon · Answer

It depends on the normals of the profile. The simplest solution if they are ALL extruding in the wrong direction is to simply reverse your offset values. 
 However, you might find that some profiles extrude 'updwards' whereas some extrude 'downwards' meaning the simple fix is not possible. In this example, the offset for both curves is positive 10m but they extrude in different directions as their normals are different. 
 
 In these cases use a one line if statement and evaluate the dot product of the normals against world Z (if all you columns are vertical). Return statements simply invert the offsets:

Heres the code you need to enter into: 
 OffetAboveCurve 
 DotProduct(baseCS.ZDirection, new Direction().FromDRay3d(baseCS, curve01.DRay3d)) > 0.9 ? 10.0 : 0.0 
 
 OffetBelowCurve 
 DotProduct(baseCS.ZDirection, new Direction().FromDRay3d(baseCS, curve01.DRay3d)) < 0.0 ? 10.0 : 0.0

Mueller · Answer

With well-constructed cells the results should be predictable and the potential complications indicated by Tom are more likely to be worst-case scenarios. If there is a concrete case that remains unresolved, please provide that case so that we may help you make progress. 
 
It is important to remember, that a cell element can contain a collection of a variety of element types, or nested cells for that matter. GC looks at cells from that perspective. Then, wholesale-extrusion of a cell does not make much sense. Therefore, from the collection of elements in the cell, those need to be identified and selected that make sense to be extruded. Usually, that's one additional step which at the same time allows some curation of the contents. So far that had not been identified as an issue, yet; however, now it has. 
 
Therefore, we will look into the possibility of implementing some simplification for cells that contain a profile-type arrangement of 2D elements. Perhaps there is a way to draw on MicroStation's methods how it handles this situation.

Mueller · Answer

Whoa, what an interesting discussion. My apologies to all those who try to find out about extruding profiles from cells... 
 
I think there are good points being made on both sides of the discussion. I spare everyone lengthening this thread by responding to details. It will take some time for me to filter out what this may mean for GenerativeComponents, some of which won't be trivial at first glance and most likely neither at second. Also, for some of the references to GH I will have to consult the GH manual. However, perhaps even fortunately, there are fundamental differences in concepts and technology that makes it difficult for both to mimic each other --which has not and will not prevent everyone to learn from everyone else's strengths and weaknesses. 
 
In any case, the GC team is very aware of the fact that there is room for improvement in both, capabilities GC already has and capabilities GC has not acquired, yet.