Adjacent PEMB Modeling Tips

[jplay2519]

I have a building that has a PEMB adjacent to my building. It's all new construction and the PEMB is next to my CMU fire separation wall. The CMU wall will support steel joists on the west side and there will be steel frames in the storefront the east side. I don't want to model it as just the CMU/steel building because the load generation won't be accurate. If I try to model this structure in a 3D finite element analysis should I make the frames some sort of infinitely rigid members adjacent to the buildings or model something tapered that's similar to the PEMB and just ignore it's design?

 

[JoshPlumSE]

It's not that clear to me what your question is.... or what really is happening with your structure.

Your structure is CMU. But, the wall between the buildings supports framing from both your building and the adjacent PEMB. Right? So, the gravity systems are not completely independent. Are the lateral systems independent?

Sometimes it can be convenient to support add columns / frames adjacent to your CMU walls, just to keep the structures truly independent for both gravity and lateral.

 

[jplay2519]

You can see in the file the areas. The clouded area is where the PEMB is going and the right side is the bar joist roof and steel frame going into that storefront. The heavy red line is the CMU separation. After reading your question I think I can just make a joint to make them independent. My question is if both buildings used that CMU wall how would I model the building because I can't put in the PEMB i don't have their model and I wouldn't model a building half the size because it would just throw off the load calculation. I think I'd just place in columns where the PEMB grid is and try to get some info from the manufacturer to determine stiffness's.

 

[jplay2519]

Sorry I left that out.

 

[KootK]

Looking at your plan, you might model both buildings and just assume that the PEMB has no independent lateral system. There appears to be lots of block wall on the non-PEMB side so this may well be a fairly accurate representation in many respects. If this is a single development, I have a hard time seeing an owner wanting to pay for:

- redundant vertical elements at the fire wall and or:

- an expansion joint on a 150' long building.

That said, the particular nature of your firewall may push things towards separation. As it stands, one side has to be able to burn down without bringing the firewall down with it, right? That can be a tall order sometimes.

 

[JoshPlumSE]

Thanks for the image, that makes things a lot clearer. Some comments:

- If the PEMB system uses your CMU wall for gravity load, then you should will have to get information from them about what load they're putting into your wall. This can be reasonably estimated by you in the early stages, probably.

- If the PEMB is relying on your CMU wall for lateral resistance, then I don't see how the two structures can be be designed separately. That leans towards modeling both buildings. The PEMB guys would probably have to use a reduced R value (for seismic) due to the two different lateral force resisting systems.