Functional Exchange Decomposition across abstraction levels

Hello,
I know that Capella only does functional exchanges between leaf functions, but I’m wondering how realization works when crossing levels of abstraction.
Obviously at system analysis you don’t know all of the leaves and their functional exchanges - you just provide high-level abstractions of what they will be. But there doesn’t seem to be a built in way to split them into more detailed functional exchanges when diving in at lower levels of abstraction. Just assigning exchanges to leaf functions would only work if those exchanges were never composite / broken down into smaller exchanges - which they almost have to be for complex systems.

Example below: FE.1->2 in the SDFB decomposes to FE.1.1->2.1 AND FE.1.2->2.2 in the LDFB. Is there a way to make those exchanges BOTH realize a single SA-level exchange? I can’t seem to find a way to manually apply realization links.


Hi Ryan,

Yes this is a pretty typical situation.

In your example, at the logical level, you can select each functional exchange and double-click (or look at the property view) and here you can setup the “Realized exchange” property for indicating that each of these Logical Exchanges are realizing your System Exchange.
Then you can go back at the SA level, click on your System exchange, and look into the Semantic browser. Now you will see that this System Exchange is realized by 2 Logical Exchanges at the LA level.

Stephane

1 Like

Ah, thank you @StephaneLacrampe. I spent so much time looking through right-click menus that I didn’t fully explore the properties.

1 Like

In general this should be possible to be done in the mass editing view - this would make it much easier to be done if you have a lot of functions, functional exchanges to edit / realisations to be added etc.

Unfortunately this does not seem to work in Capella 6.1.0 - there may be a bug. :frowning_face:
In some cases the multi-value picker does not open, in the other cases the selected items are not taken…

When at the “next layer” (i.e. LA from SA or PA from LA); I tend to choose one functional exchange at random and re-allocate the ports to the appropriate leaf functions (normally in a LDFB or PDFB). I will then create a second, third etc functional exchange as necessary. I will look at what realised exchanges there were from the first chosen exchange and create this link manually for the others too.

If you do it as you go along, its not that much extra work.

Josh

Can you raise a bug maybe?