Benefits of splitting project into Core and Shell

[rtaube]

I'm working for a structural engineering office and we received an architects model that had been split into two models (Core and Shell) each file is only 80 Mb apiece, so i was wondering what the benefit is to splitting it up, since the file size isn't that big. Wouldn't it cause more problems with the joining?

I also noticed that all the sheets are laid out in the Shell file, with the Core drawing linked in, FWIW.

Thanks for any insight!

[greg.mcdowell]

They certainly could have used Worksets for this instead of separating the files. Probably just an internal workflow decision.

[Revitaoist]

They probably just have their standards set up that way.

[MikeJarosz]

There could be lots of reasons, some good, some not so good:

They may have more than one office, or they may have a separate interiors group, or they outsourced it to India, or they don't have a good interoffice network (WAFS), or they may be trying to replicate Acad xrefs or they contracted the interiors or the interiors were a type they are uncomfortable with (think hospitals).

I have actually seen all of the above and more.

[DaveP]

We set up almost all of our projects this way.
They do tend to be larger than 80 Meg, though. More like 100-120.
Partly we do this out of habit. We started doing it back in the days of bad Worksharing, and 2 Gig XP machines.
We found that even with decent machines now, using a Shell and and Interior model is still an efficient workflow. Typically our projects have a separate Bid Pack for the Shell anyway, and our project teams usually have one Shell person and one "inside" person.

You say the Arch models are around 80 Meg at the moment. How far along are you in the project? We've had a couple of smaller projects that we thought we would not need to split up. Worked fine early on, but by the time we got to the end of CDs, those "small" models had grown to over 150 Meg and started getting slow and unwieldy. Then we got into PRs and really wished we had split it.

Especially with the ability in the recent versions to use Visibilty per Link View, and Room Bounding and Tagging through Links, there's not a whole lot of downside to using 2 models.

NOW. All that said, I'm talking Architecture here. Structurally, you'd probably be OK, but it does have more impact in MEP, since a lot of systems tend to cross the entire building. Them you'd probably want to split by function.