Composites

[Paul King]

Composite walls, slabs & roofs need each skin to be independently extensible & clean up independently in different ways & with ends that can be shaped & angled in different ways - in plan and section

In almost ALL projects, composites as currently implemented (essentially 2D lines & fills painted over a single dumb object) fail to address the required 3D real world situations they are intended to depict, requiring messy & counter intuitive workarounds

Eg in section outer skins of brick work are always set down lower than inner timber skins, inner wall linings are never as tall as exterior wall linings, skillion roof ceiling linings need to turn down to clean with wall linings, while core / fafter framing passes on over top of wall.

In plan, situations arise regularly in complex 3 & 4 way junctions of different composites where each skin of each wall needs to clean independently in a different direction

Often it is nearly impossible to resolve these elegantly, especially when reference line positions and intersections priorities cannot be adjusted without unravelling other junctions.

In 3D, all layers of a composite should be visible as distinct entities when viewed edge on, with own fills & hatch patterns - including layers that are cavities (these should display as voids in 3D)

In a nutshell, composite skins should behave almost like a series of parallel but independent walls/slabs/roofs, that just happen to share common openings for doors & windows, and move as a group when dragged. A "dumb mode" toggle could leave behaviour as is by default for those few situations where current composite behaviour is ok

Sorry if parts of above are wished for already - can't seem to find exactly this rationalised summary anywhere.

[Anonymous]

Paul,

I believe that most if not all of these items have been wished for in the past but I don't recall them all being put together in this coherent fashion. Thanks.

One minor bit that I'll add is that for any of this to work we need adjoining fills to clean up with separator lines/pens rather than the section outlines. If just this one small improvement were made now we could accomplish much of the other stuff with profiles and SEOs while awaiting the more sophisticated parts.

[Laura Yanoviak]

Matthew wrote:
One minor bit that I'll add is that for any of this to work we need adjoining fills to clean up with separator lines/pens rather than the section outlines. If just this one small improvement were made now we could accomplish much of the other stuff with profiles and SEOs while awaiting the more sophisticated parts.

I very much agree. We are currently able to create very usable sections with little 2D embellishment using these methods.

[Anonymous]

I agree with Paul's summary. I would also add that each skin thickness should be independently variable.

[Geoff Briggs]

This is a huge, longstanding wish of mine as well. The constant
tweaking of offsets, display order, SEOs, separate slabs, etc. is tiresome.

I even took the trouble once of modeling a test building where every part of the envelope was a separate wall, roof or slab. Arguably it was easier than all the workarounds we employ today, with the exception of the window and door openings. If those object had the capability to cut adjacent walls I might just adopt that method right now.

[Anonymous]

Laura wrote:

Matthew wrote:
One minor bit that I'll add is that for any of this to work we need adjoining fills to clean up with separator lines/pens rather than the section outlines. If just this one small improvement were made now we could accomplish much of the other stuff with profiles and SEOs while awaiting the more sophisticated parts.

I very much agree. We are currently able to create very usable sections with little 2D embellishment using these methods.

Yes, this has always been one of the biggest and littlest of the LTWBI (little things with big implications).