Parent/Child Hierarchy with Building elements

[Anonymous]

Imagine if you could create a parent/child hierarchy with building elements.

Consider the following work flow.
- Create the footprint of the building; it could be a slab, zone, plan or fill.
- Define it as a “Parent” object.
- Using the Magic Wand, create exterior walls whose reference line is “linked” to the outline of the footprint. (An operation similar to a boolean operation would link the child/children to the parent.)
- Multiple exterior building components can be a child of the parent; footing, foundation walls, exterior walls, floor slabs, etc.
- Even profiled cornices and trim elements that are built as walls could be a child element.
- When a change or revision is required, as it always is, a simple change to the parent element will update the children.

All of the linking is connected to the reference line of the wall(s) and roof(s). There are other bells and whistles that can be incorporated; add child dependency to parent object; remove child dependency from parent object; set the children to auto/manual update.

Imagine changing one reference line in a multi-story or even high-rise building and every linked element is updated! (Maybe this is a connection that can be done similar to Rhinoceros?)

[Barry Kelly]

jfa_ra wrote:
Imagine changing one reference line in a multi-story or even high-rise building and every linked element is updated!

This can already be done since version 18.
So long as the elements share the same reference line (walls and roof pivot lines) and perimeter position (slabs and fills), then all can be selected in plan or 3D for elements on multiple storeys and they will all adjust together.

The is no permanent link - only when selected.

See the new features guide in the 18 reference manual if you have it or I am sure it will be in the current manual as something like 'Multi-element editing'.

Barry.

[Laszlo Nagy]

Something like this exists in Revit and its merit is also its weakness: when you create many of these relationships, it starts to take its toll on the performance of the application.
If you check it you will see that in recent years Autodesk is not really emphasizing these constraints as a distinguishing factor. Also, in various technical papers Autodesk warns against using many constraints as it adversely affects performance so much that on large/complex projects it is practically unfeasible.

I think this constraint thing would be a great thing if we had practically unlimited processing power, otherwise it is not feasible under realistic project circumstances.
Hopefully, we will get there someday.

[DGSketcher]

I could see the OP suggestion as a worthy evolution. If you consider that we currently link walls to storeys and trim to roof to create the envelope then taking the next step of linking the (external) walls to a floor slab would be quite useful. The same slab could also operate in the same way as the "trim to roof" to adjust the length of internal walls that exceed the slab boundary. Not sure whether you would link your roof plan to the slab changes, as some of the results may be unwanted.

As has been mentioned it would need to be implemented in such a way that there wasn't a performance hit.

With regards to the adjustment of sub components complex profiles offer an easy way to incorporate these items that can adapt to a changing envelope.

[Laszlo Nagy]

Yes, I agree, this would be useful.
But implementing constraints in general is a can of worms so one should proceed with caution.

Let me just give you an example: in AC19 we got unlimited associative labels, so each element can have multiple Labels that can display all kinds of data about elements. But then we started getting reports from people placing hundreds of such labels and consequently complaining about a performance hit.
So yes, one needs to proceed carefully in implementing these kind of features that create a host of associativity among elements that constantly need to be kept track of by the application. So a cautious approach is indicated.

[DGSketcher]

I found a GS video for the multi editing Barry referred to...

http://helpcenter.graphisoft.com/videos/archicad/3d-modeling/archicad-18-new-features-editing-multip...

I really should keep up with the new features training as this pretty much does what I was suggesting without creating a hard link between elements

[Erwin Edel]

If you work with element IDs it becomes quite quick to setup find and select criteria (which you can save in your project) to select related elements. You just need to include something in the element ID that is common among those elements you want to relate to eachother.

It might become more work to maintain them though.

Of course a simple group of elements works too.

[Anonymous]

There are certainly ways to achieve the result in a semi-automatic way as suggested. It requires a well organized ID naming convention and object grouping to work successfully. Although surprising results can occur when objects are grouped over multiple stories. Also, to edit the reference line of elements in a group, the grouping needs to be temporarily suspended. (A bit of Option <G> toggling...)

Maybe it would be better if nested groups could be created. (But that is another tangled wish altogether...)