MEP models need more data than your usual architectural model. The reason is that you need to make a lot of calculations with them, and then the sequence and relationship between the MEP elements really matter.
Here another important class comes into play: the IfcDistributionPort. Ports are those parts of your elements that connect to other elements via a distribution system. This works for electrical systems, ventilation, piping etc.
Again, ports are derived from the IfcRoot class and therefore have their own identity and life-cycle. That means that you actually don’t need a hosting element, theoretically you can work with them directly and add the host later (according to the IFC schema only; this way of working is not really supported by BIM applications).
Why is it not enough to just add properties that tell you which system they are connected to? As it is a one to many relationships (one element can host many ports) you would need to provide many properties for the same kind of behavior, like Port1, Port2. It would be nearly impossible to filter the right one. You need metadata, and that’s why it needs to be an object on its own.
Ports have a few specific attributes that define them. You can select the flow direction that is allowed for this connection:
• SINK means that only incoming substances (air, water, etc.) is allowed
• SOURCE means that substances can only flow out
• SOURCEANDSINK allows incoming and outgoing substances.
Ports can also be nested to indicate subports (we will look at nested in a future post).
As for all objects you can also add property sets and properties to your ports which can contain details about size, pressure, voltage etc.