Extend PartOf facet with IfcRelSpaceBoundary

[NickNisbet]

I would like to specify
a. That every significant space should have space boundaries

This may be an applicability (anything with a space boundary should have a temperature.)
Or a requirement (any space bigger than 3m3 should have space boundaries. )

[atomczak]

So, in IDS terms, are you, @NickNisbet, proposing adding IfcRelSpaceBoundary to the list of allowed relations of the PartOf facet?

I'm not an architect. What is the practical use of such a requirement? Why not just saying "any space needs a temperature"?

[NickNisbet]

The benefit of being able to require IfcRelSpaceBoundary are that this is necessary for: 1. Thermal and energy analysis (not just target temperature). 2. Routing, accessibility and escape analysis for fire-safety, security, usability. 3. Appropriate/ inappropriate space adjacencies including Sound transmission acoustics analysis. 4. Quantity reporting for building fabric elements (for costing and sustainability). NN

[Yoshiyuki-Miyauchi]

@atomczak
IfcRelSpaceBoundary is what I need too.
Specifically, it would be great if it was a PartOf facet and could be selected.

For example, it would be useful when there is a condition that a wall belonging to a space called A should have a property called B.

[mdjska]

I also support this. It would be very useful for defining requirements for building elements that delimit a fire compartment.

[Yoshiyuki-Miyauchi]

@mdjska @atomczak
I would love to see IfcRelSpaceBoundary available in IDS, it would be great.
It is clear that it will be useful at work.

[Moult]

I'm against the idea of PartOf being used for this, since PartOf (strange name, but that's another issue) currently addresses the spatial hierarchy. Rel space boundaries is a relationship that is a separate concept. If it is supported, either PartOf should be more generic (i.e. turns into a "Relationship" facet) or a separate facet should be created.

[Yoshiyuki-Miyauchi]

Yes, you're right.
Maybe we should have a different facet or a more generic facet name for the PartOf facet.

[mdjska]

@Moult I've seen this stated before, but the definition of 'spatial hierarchy' confuses me a bit here. Under IfcRelContainedInSpatialStructure (and other places in the docs) it's said that a hierarchical relationship = an element can only be contained within a single spatial structure element. But if you look at IfcRelAssignsToGroup, which is also part of the PartOf, it says that 'Grouping relationships are not hierarchical'. I don't really get how IfcRelSpaceBoundary, which has a physical definition and representation, is different from IfcRelAssignsToGroup or IfcRelNests. Is there a place in the docs that specifies those 6 relationships specifically as defining the spatial hierarchy?

[Moult]

Yeah you're right groups arent part of the hierarchy either. Dion Moult Sent from Proton Mail Android

…

-------- Original Message --------

On 10/3/25 9:32 pm, Martina Jakubowska wrote: ***@***.***(https://github.com/Moult) I've seen this stated before, but the definition of 'spatial hierarchy' confuses me a bit here. Under IfcRelContainedInSpatialStructure (and other places in the docs) it's said that a hierarchical relationship = an element can only be contained within a single spatial structure element. But if you look at IfcRelAssignsToGroup, which is also part of the PartOf, it says that 'Grouping relationships are not hierarchical'. I don't really get how IfcRelSpaceBoundary, which has a physical definition and representation, is different from IfcRelAssignsToGroup or IfcRelNests. Is there a place in the docs that specifies those 6 relationships specifically as defining the spatial hierarchy? — Reply to this email directly, [view it on GitHub](#247 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AAAVR3XQ32YMNA4LK367POL2TVS2ZAVCNFSM6AAAAABDOCJU3CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMJQGEYTGOBWG4). You are receiving this because you were mentioned.Message ID: ***@***.***> [mdjska]mdjska left a comment [(buildingSMART/IDS#247)](#247 (comment)) ***@***.***(https://github.com/Moult) I've seen this stated before, but the definition of 'spatial hierarchy' confuses me a bit here. Under IfcRelContainedInSpatialStructure (and other places in the docs) it's said that a hierarchical relationship = an element can only be contained within a single spatial structure element. But if you look at IfcRelAssignsToGroup, which is also part of the PartOf, it says that 'Grouping relationships are not hierarchical'. I don't really get how IfcRelSpaceBoundary, which has a physical definition and representation, is different from IfcRelAssignsToGroup or IfcRelNests. Is there a place in the docs that specifies those 6 relationships specifically as defining the spatial hierarchy? — Reply to this email directly, [view it on GitHub](#247 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AAAVR3XQ32YMNA4LK367POL2TVS2ZAVCNFSM6AAAAABDOCJU3CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMJQGEYTGOBWG4). You are receiving this because you were mentioned.Message ID: ***@***.***>