Composite Metal Floor Deck Deflection With Removed Support

[JSA]

I'm trying to compute the deflection of concrete-filled metal floor deck, two-span condition after the center support is removed. The one-way slab is 3" deck with 6" concrete fill over the flutes for 9" total thickness. There are two 10-foot spans and the deck is continuous over both spans, but interior support is only temporary. There is a #7 in the bottom of each flute (12" spacing} so that the floor can span the 20-feet after removal of the center support. My questions are qualitative and quantitative.

1. When the concrete is placed, the deck section at the interior support has compression at bottom and tension at top (negative moment in metal deck at support). After the concrete cures and the center support is removed, the deck now acts compositely and the slab spans 20-feet. So at the mid-span, the top of the metal deck gets an added tension, but the bottom tension serves to relieve the "precompression." Qualitatively, how does this "pre" stress affect the final deck deflection? Increase it, decrease it or have no effect?

2. How would you compute this, so as to be able to specify appropriate camber?

JSA

 

[HochwaltPE]

In theory the pre-existing stress in the deck would effect the deflections because instead of the entire section of the deck reaching yield strain simultaneously (as would happen with a reinforcing bar), the plasticity will propagate across the depth of the section. Before the deck starts to yield the deflection will be unchanged, and after the section has fully yielded, it will also be unchanged. There would be a difference in deflections while the deck is transitioning from fully elastic to fully plastic.

Given that you're estimating the deflection of a reinforced concrete section, I would suggest this effect is not worth worrying about. I've heard it said (and have seen it in borne out in practice) that you're doing quite well is you can predict the deflection of a reinforced concrete section within 50%. With that level of accuracy, it's not worth the effort to explicitly consider the gradual yielding of the deck.

 

[JSA]

I'm trying to decide if I should camber the system.

Neglecting any effects from the precompression noted above, my calculations show deflections for the 20-foot span to be anywhere from 0.25" slab-only dead to 2" long-term dead plus live. My gut tells me to camber this 20-foot, 9" simple span slab 1", but the area of steel of #7 at 12" combined with 18 gauge metal deck is quite a bit of steel and my sustained dead load moments are not too far above the cracking moment.

Any thoughts?

 

[HochwaltPE]

How real is your live load?

In general, I prefer not to camber for more than the dead load deflection because most of the time the real live load is much less than the design live load. If the structure is cambered based on the design live load, most of the time the structure will have a "hump" to it which may be no less annoying than the sag. The other thing I would consider is what the consequences are of hogging and sagging for the structure's ultimate function. (For example, if this were a roof, the drainage pattern might tell us whether we'd rather risk hogging or sagging.)

Based on the span/depth, this looks like a pretty aggressive design. If you haven't already (and depending on the application), you might want to look into how dynamically lively this slab will be.

 

[JSA]

Thanks for your input. I think I will camber it about an inch. I'll let you know what happens.

 

[JSA]

HochwaltPE,

As promised, I have some actual results.

The shoring (middle support) was pulled after 30 days and the deflection was only 1/16 to 1/8". This surprised me; I thought it would be more. It seems that the enormous cross-sectional area of steel more than compensated for the temporary reverse curvature of the steel deck (bottom fiber prestress).

 

[LPPE]

My $.02 - I agree with HochwaltPE on the dead load deflection camber only. A hump is just as annoying as a sag. And the concrete is placed level, so if you have a LL camber, there will be less concrete section at the middle of the span.

 

[mjohan]

pylco,

I agree with your statement regarding placing concrete level. Say a suspended concrete floor was poured utilizing an open web steel, composite floor joist system. The joists were cambered 1" for dead load deflection. Do you think that their would be a 1" deflection on the top of the slab, and coincidently, a flat ceiling below?

Also, has anyone heard of cutting control joints in a concrete deck, like described above? The reinforcement would be wire mesh. The deck dimensions are 180'x80' with minor jogs.

Thanks,
mark

 

[JSA]

Mjohan,

I have never heard of cutting joints and have never required it for structural [elevated] slabs. I believe this is because most elevated slabs have plenty of reinforcing and pours are limited typically to 20,000 sf and 500 yards, as a function of what can be done in a day.

Regarding floor deflections and reduced concrete section, I had the contractor screed to the hump, which left a constant 9" slab and a hump.