wall intersections in 3d

[adamsb]

Hi all,

I have noticed that, even when two walls intersect properly at an angle in plan, one wall always has priority over the other in 3d. This causes the materials to not line up very well. (I know that individually setting the 3d alignment manually can overcome this, but that's not the point.) This also ends up affecting the way the sections work through cuts near this intersection.

This isn't a wall priority problem as both walls are the same building material, surface material, fill, layer, wall priority, and everything. They also meet properly on plan. I would like the wall "mass" in 3d to create a mitered affect, instead of one wall passing all the way through the other wall.

Can somebody help me with this one? Thanks.

[David Maudlin]

adamsb:

Are the Reference Lines on opposite sides of the Walls? If they are, try changing the Reference Line for one wall so both are on the same side.

David

[Mats_Knutsson]

I'm also interested in how/why the walls don't join mitered. I want to use BM.

[adamsb]

Agreed. The program should split the angle to create a simple mitre. This would allow one surface to be overridden to a different material, with the corner remaining the change point.

[Mats_Knutsson]

adamsb wrote:
Agreed. The program should split the angle to create a simple mitre. This would allow one surface to be overridden to a different material, with the corner remaining the change point.

However I agree with previos posters that this is not a normal situation...but let's say I want a brick wall in my garden..I'm curoius about the quantities I'll get with this kind of wall joins. Need to try that in the week-end.

[Steve Jepson]

Mats_Knutsson wrote:

adamsb wrote:
Agreed. The program should split the angle to create a simple mitre. This would allow one surface to be overridden to a different material, with the corner remaining the change point.

However I agree with previos posters that this is not a normal situation...but let's say I want a brick wall in my garden..I'm curoius about the quantities I'll get with this kind of wall joins. Need to try that in the week-end.

I think this is a graphics issue that will not make any difference in the schedules, which you would probably want to be based on the reference line.

But I think it is an important graphics issue that should be fixed.

I would prefer that the clean up not be automatically mitered, but that it would clean up just like the walls skins will be constructed. Wall skins are rarely mitered. I would like to set the priority for which skins overlap or clean up as a miter in the Composite settings.

[Laszlo Nagy]

I am afraid the way skins of joined Walls are constructed in L- and T- and other types of joins is something that occurred as a consequence of Priority Based Connections (PBC) in ARCHICAD 17.
Before PBC we had mitered joins. (This was not perfect either because if could give us incorrect results - the mitering was not performed for the whole wall, and not on a skin level.) With PBC the strength of each skin is evaluated by the program. The stronger skin cuts the weaker skin (a Solid Element Operation is performed internally based on the Intersection Priority Number values of the Building Materials (BMs) of the connecting skins). If the Intersection Priority Numbers of the two connecting skins are the same (e.g. because they are the same skin of the same Composite), then ARCHICAD randomly decided which skin cuts the other skin.
This means that there is no mitered join on a skin-by-skin basis. On skin always cuts the other. And this is what you guys see in those attached screenshots: at each joint one skin is always extended as far as possible and it cuts the other skin with the same BM.

My guess is that technologically it is a totally different ball-game to be able to handle each skin of each structure separately and create mitered joins on a skin level - that is totally different from creating a mitered join for the whole wall.

[Steve Jepson]

Yes. Someday controlling the walls skins independently might be possible. We need that especially for independent vertical control of the skins. When we can do that, we will also be able to control the horizontal skin connections.

My idea/wish for a long time has been for composites skins to behave as one, or temporarily ungrouped so they can be controlled independently, then regrouped. When the skins are ungrouped would be the time to configure the ends of the skins to mitered or lapped.

I believe this will be possible in ArchiCAD someday because the demands for more precise details are increasing exponentially with the implementation of LOD's that are becoming closer and closer to 100% virtual modeling of the entire building.

I like the clean up as it is in this image. No miter.