Pre-Engineered Metal Building - general considerations for openings 2

[ThorenO]

Hello, thank you for any feedback in advance.

Existing PEMB Description
360' wide (6-span frames) x 300' long (12) bays
Eave at 23', Ridge @ 33'
1" dia. X-bracing with hillside anchors. Bracing located at (4) of the (12) bays on each end wall. Probable that (1) or more were removed.
8" "Z-shaped" girts at 5' on outside face of frames.
8" "Z-shaped' purlins at 5' at the roof with wind-bracing.
C8 eave strut with 8" strut-purlin at end walls.

Pretty standard.

What's been proposed is to remove ALL the X-bracing on one side to permit 14' wide x 16' high overhead doors in the center of the 25' bays.

I'd like to get some ideas and general considerations on restoring the building's lateral system and creating the openings.

 

[jayrod12]

Haha, love that request. Remove all the cross bracing.

I recently did a job where we replaced cross bracing with pinned-base moment frames. But it was rather expensive.

 

[CELinOttawa]

I've done this KIND of major modification to a PEMB ONCE. It is a nightmare, pure and simple.

They are so crazy efficient that the least little change, sometimes including typical cyclical code changes, are enough to require major modifications.

In our case I introduced tilt panels as a new lateral load system and reworked quite a few columns and beams.

A great deal of work, and not to be entered into on a quoted rate. I strongly suggest you charge your client for actual effort expended/take this on as an hourly engagement.

 

[msquared48]

Moment frames inserted into each bay where thecx-bracing is removed is the most logical answer in order to maintain the original load path.

Mike McCann, PE, SE (WA)

 

[KootK]

Flying. Buttress. Seriously, bracing external to the building might be worth looking at.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[JedClampett]

Charge him/her out the wazoo. And then remind him/her how much money they saved on the building in the first place.

 

[SteelPE]

I don't suggest being mean to the client. The easiest solution is to see if they can live with openings in only 8 of the 12 bays. When they reject this idea then the best solution is to add the portal frames as suggested by jayrod and msquared. It's not going to be cheap, but it should work.

 

[KootK]

With regard to the moment frame alternative, be sure to give some consideration to diaphragm stiffness. At the aspect ratio of your roof, rigid diaphragm behaviour is a real possibility. If that's the case, the moment frames will need to be pretty stiff in order to shield the remaining braces on the opposite side from attracting too much load. There probably isn't enough reserve capacity in your building for it to start behaving like a three sided structure.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[jayrod12]

KootK,

Not that I disagree with you but let's just say for the sake of argument that the moment frames aren't quite stiff enough. They are still going to have some contribution no? Would you actually go so far as to look at the building as a 3 sided structure?

I think if your moment frames limit your building drift accordingly then they would (should) be stiff enough.

 

[KootK]

Not utterly three sided Jayrod. I'd just put the centre of rigidity where it actually deserves to be. And that would likely be considerably off centre given typical moment frame stiffnesses. So three sided-ish behaviour depending on relative stiffness of the moment frames.

For this issue, code drift limits are not the relevant parameter. It's about the ratio of moment frame stiffness to braced frame stiffness.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[jayrod12]

Fair enough. However in order to limit the building drift (especially one this tall) the moment frames would end up being stiff enough to not have too significant of an impact on the centre of rigidity I would bet.

 

[KootK]

I would bet against you. Height punishes the stiffness of a moment frame much more than a braced frame (cubic vs linear). Regardless, the responsible thing to do is check it out, no?

Perhaps the roof is discretely braced rather than metal deck? That would change the flavour of the analysis.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[ThorenO]

Moment frames are the logical answer with cost & deflection being the two key variables. Would there ever be a consideration for connecting to the adjacent PEMB at the expansion joint? I'd ruled this out completely, but wanted to get your take if that were ever feasible. The two structures are identical with the girts & purlins separated by about a foot. The roof is going to be replaced so, this I was wondering if this could become an option or if it should be dismissed forever.

 

[KootK]

I vote dismiss forever.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[ThorenO]

Thanks! Just to clarify, this would be in conjunction with the moment frames, but the analysis would be too complicated. I'll bench that idea forever!

 

[KootK]

Yeah, with moment frames and braces on the same line, I would think that you'd have even more serious issues with stiffness compatibility. I wouldn't find that option attractive unless I could get all of the capacity that I needed out of the building that would remain braced. One thing that I really like about the moment frames is that you get to reuse foundations intended for the load. That's a boon for sure.

I got lucky the one time that I did this. I had a loading dock retaining wall in line with the braces that I removed. We threw in an external HSS brace over the retaining wall, tidied up some drag struts, and called it a day. It was pretty economical if not particularly attractive.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Archie264]

KootK,

What did you do for anchorage in that situation?

 

[KootK]

There was a long steel strap bolted to the traffic side of the retaining wall. It took a substantial amount of wall dead load to resist uplift. Like I said, it wasn't real pretty. Our only concession to good building design was the installation of a Fabreeka thermal break thing where the strut hit the building proper.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Archie264]

That sounds like a clever approach. Did you give any consideration to drilling in "post-installed" anchor bolts or did the strap make such consideration unnecessary?

 

[ajh1]

Momnet frames are about the only way to go and keep the bracing within the original wall. You might consider only making the frame clear height just enough to clear the door openings and adding a tier of rod bracing about that level. As noted by others, height is a major deterent to frame stiffness. Also consider adding more portals than just the 4 original braced bay locations to boost the overall stiffness. Avoid having rod bracing and portals in the same bracing plane as the relative stiffness will place the large majority of the force into the rod bracing whether you want it there are not. The building is too wide to even remotely think about 3-sided bracing schemes. Our normal practice is to limit something like that to no more than 60' wide and about 16' high. Is there any location within the building that would allow bracing to the floor at an interior column line that might be able to be used to offset the reduced bracing stiffness along the perimeter? That will tend to affect the force distributions in your roof bracing but might help out the problem. Interior bracing in wide buildings while not typically used is one potential way to alleviate the forces.