Alternate steel diaphragm design

[stickmotion]

I have a project where the contractor did not puddle weld the steel roof deck to the bar joists with 5/8" puddle welds. The deck was just spot welded to the joists with a MIG welder from below. It was not noted by the inspector until after the roof had been completely constructed. I have talked to the deck supplier and SDI and they have never heard of anyone fastening the deck from the bottom and basically said the design tables would not apply to fastening from the bottom of deck. Has anyone ever ran into this situation before?

I looked at some type of screw/rivet system that would be equivalent to a #12 screw and provide compression from the top of the deck similar to a screw from the top. The contractor is concerned with the amount of predrilling that would be needed.

Could you neglect the deck and transfer the load with structural steel between the bar joists? What would be the most efficient and cost effective way to transfer the load? I would appreciate any input.

 

[structSU10]

As long as you can verify the load path still works well, you could add horizontal x-bracing in the plane of the roof.

 

[steellion]

I don't know how or why you would try to weld through the member from the otherside of the deck. Were sidelaps installed? You can create a diaphragm from X-bracing, but is the roof deck OK for wind uplift if it was not properly attached? I wouldn't bet my seal on such an improper construction method with no SDI backing.

 

[KootK]

Man, you have my sympathy on this one.

I'd also be worried that the deck welds would be insufficient for uplift and the smaller amount of shear capacity required to brace the joist top chords and transfer seismic inertial forces out of the roof and into the joists. QC on those overhead welds might be a problem.

As StructSU10 noted, you can use a discrete bracing system so long as a complete load path is provided. That's no small thing however. Among other difficulties, some of your joists will likely become axially loaded members. Vulcraft has a manual that deals with this although not with your specific situation: Link

If you install bracing across the entire roof, I imagine that you'll be stuck having to do something below your joists which is tricky. Depending on what your framing looks like, you may be able to scale your fix back to just smaller horizontal trusses around the perimeter of the building.

A colleague of mine did something related, albeit on a much smaller scale. She installed new drag struts into an existing roof. Hilti helped us work out a system for fastening below the deck. The issue seemed to be whether or not you could install the fasteners from the underside without simply pushing the deck away from the drag strut rather than piercing the deck and engaging it properly.

We ended up using hot rolled channels for the drag struts. Overhanging cold formed sheets were fastened to the channels on the ground. Once the channels were lifted into place, the last step was to screw the cold formed sheets into the deck. The thinking was that a CFM to CFM connection would be more successful than a hot rolled steel to CFM connection would be. I'm not sure if that's true but one advantage of this system was that it was pretty easy to verify the results.

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

 

[JAE]

Contractor screw up - they should pay for the extra effort to respond.
You have the diaphragm shear and uplift loadings so you might be the best to come up with a solution -
or you could put the monkey on their back and ask them to hire an engineer to solve - with your cooperation in providing required loads, etc. and demanding a PE produce calculations backing up any solution.

I've attached a sketch similar to what KootK suggested - only using all light gage and fabricating a large number of bent clips with screw fasteners.


Check out Eng-Tips Forum's Policies here:
faq731-376

 

[stickmotion]

Thanks for all your feedback. The roof is a ballast roof so that will help for the uplift. I was thinking about the steel X-Bracing alternate as well. In order to adequately design for uplift, bracing of the bar joists and diaphragm forces I need to provide some type of attachment between the deck and joists.

I was thinking about using Fab Lok fasteners. See the link below. The diameter would be larger than a #12 screw. Is it reasonable to consider it as an equivalent to a #12 screw and use the diaphragm values for #12 screws? Maybe it makes the most sense to use a combination of X-bracing and fasteners to achieve all of the design requirements. Thoughts?

http://www.elcoconstruction.com/Products/Metal-Fastening/Rivets/Fab-Lok-R

 

[KootK]

Structurally, those fasteners would be great. However, given the large diameter of the sleeves, I'd be worried that you'd pop the roofing. Why don't you like the sheet metal screw solution?

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

 

[stickmotion]

I don't feel that screws from the bottom would be equivalent to screws from the top. I was actually told this by the deck supplier and SDI that it would not be considered equivalent. When you fasten from the top the head of the screw will provide some compression between the deck and the joists. Also I would be concerned with the connection between the deck and screw with such a thin material.

There is insulation between the metal deck and roof membrane so I don't see any issues with penetrating the roof membrane.

Is it reasonable to use the Fab Lok fasteners as an equivalent to #12 screws?

 

[hokie66]

Installing all those Fab Lok fasteners will be a nightmare, as that does require predrilling. With self drilling screws, it will still be a tough job, but doable.

 

[XR250]

Run away or have him pull the roofing and do it right.
Unlikely it will ever get done correctly from below.
Why would they have done overhead welds from below when doing it from above is so much easier?

 

[KootK]

It might also be worth investigating whether or not the finished product will be able to be warrantied. That's important to the owner and you don't want to be the at fault party on that one -- not even partially.

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

 

[UcfSE]

Did the SDI say the tables would not apply (as in calculate the capacity yourself), or that the fasteners used (welds from below) were incorrect or not permitted? Have you considered calculating what would be needed if welded from below and directing the contractor to provide that?

 

[stickmotion]

I am looking at doing x-bracing and just using the deck for vertical load. I was looking at the x-bracing layout. I have two separate options. The first option has continuous bracing throughout. The second option has x-bracing staggered every other joist space. The second option appears to adequately transfer the load path. The first option seems like a lot of x-bracing. Any feedback would be appreciated.

 

[KootK]

Either option would get the job done. I'd try to justify something even simpler though. See the attached sketch. This would deal with all of your lateral loads save the seismic loads originating from the inertia of the roof deck itself. And those seismic loads should be cut down to a level that you can feel pretty comfortable with them based on the overhead welding.

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

 

[KootK]

For interest's sake, check out the attached photo. I took it at the Tokyo airport. It's the sexy, Cadillac version of what you're trying to do.

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

 

[spats]

I would think the diaphragm/shear values would be the same screwing from below. What wouldn't be the same would be the uplift values. It's a matter of pull-out strength (screw shaft) vs. pull-over strength (screw head). ICC ESR-1976 (2013) is what I use to evaluate ITW Buildex TEKS screws. It provides values for both. For 22 gauge deck, the allowable pull-out with a #12 screw is 83 lbs., and the allowable pull-over is 280 lbs. with a HWH.