Extend PartOf facet with IfcRelSpaceBoundary
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. )
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"?
The benefit of being able to require IfcRelSpaceBoundary are that this is necessary for:
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
@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.
I also support this. It would be very useful for defining requirements for building elements that delimit a fire compartment.
@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.
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.
Yes, you're right. Maybe we should have a different facet or a more generic facet name for the PartOf facet.
@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?
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, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>
[mdjska]mdjska left a comment (buildingSMART/IDS#247)
@.***(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, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>