Library Floor Deflections

[marinaman]

I am designing a library for one of my clients. The building will be steel framed with a composite floor system.

Given this building is a library.......how do you guys feel about what an allowable floor deflection might be? If I set my live load deflection to L/360....and my overall deflection to about L/320...that seems about right....but I feel like I may need to be slightly more conservative.

For example, the code requires I design the floor for 150 psf uniform live load within stack rooms. My architect is saying design the whole floor for the 150 psf, as we don't know where stacks might end up. That said, as an example, for a 32' long span, 10' beam spacing.....my live load deflection is 5/8" and my overall deflection is 1.15"....for a W24 x 55. I can't determine if I feel like that's ok or not.

Have any of you guys designed library floors before out of steel? What deflections did you feel comfortable with?

 

[XR250]

I personally never use L/360 unless I am trying to get some existing condition to meet code. Maybe L/600 for this? I imagine stacks are going to be sensitive to deflection as it will be magnified by their height. Even if they ain't gonna fall over, it may look that way to the customer.

 

[JoshPlumSE]

L/360 for Live Load / storage load seems about right to me. Maybe 480 if you're really concerned. I've only used L/600 for when really brittle material is connected to my members.... i.e. brick cladding or such.

Be careful with what member stiffness you use. Remember the code suggests taking something like a 25% stiffness reduction of the "Equivalent" moment of inertia. Alternatively, you could use the I_lower_bound instead.

There is also a technique (in the AISC commentary) for calculating long term composite beam deflection due to creep / shrinkage. If you went with L/360 then it might be a good idea to consider this.

 

[phamENG]

I generally set a hard limit on floor or flat ceiling deflections at 0.75" wherever possible. On really long spans even the curvature limitations (like L/360) can become noticeable - especially if there's a nearby 'straight' line to compare it to. I could imagine stacks with adjustable legs forming a nice, even line not too far below the ceiling.

 

[kwinner]

Not sure what your spans are, but a vibration calculation might be helpful here too.

 

[dik]

Concur with Josh... L/360 is OK.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[marinaman]

Thank you guys for all of your input. Very much appreciated.

I've studied these members pretty hard, here in the last couple days. My spans aren't terrible.....but aren't short either. My beam spans are from 28' to about 32' in length, with beam spacings in the 8' to 10' range.

I've tried to consider the system movement as well. This building is nearly squared-up, in terms of the building grid.....in other words, my girders are about as long as my beams. That said, I've got their movement to consider as well.

Additionally, I've spent a considerable amount of time looking at my deck (5 1/4" lightweight.....3 1/4 over 2" 18 gage deck). Of course there's movement there as well.

Here's the deal: When I restrict the live load deflection....we end up with a member that's stiff enough that no camber can be used when loaded with wet concrete...as its too stiff. In my dead load, I'm accounting for a 1" overpour, as the GC will likely overpour slightly to level the floors when poured. All that said, adding my girder, beam, and deck deflections....there could be some significant combined movement here.

I'm thinking the L/600 live is the correct move for beam designs. Here's why: A beam member at 32', at L/600, deflects 5/8" under live load. If I had allowed L/360, we'd be talking 1.1" of movement....not including any additional system movement....and I'm not liking those magnitudes. Loaded up with stacks, I don't think those magnitudes are acceptable.....at least not to me.

I think I'm going with the conservative numbers. That said, to all, I appreciate your time in responding and your inputs.

 

[JLNJ]

For such a long span between beams I'd also look closely at the floor load itself. The 150 psf averages out the extremely heavy stack load with the normally unoccupied aisles. More like 300 psf or more in the middle of the span and nothing on the ends (in the aisles).

 

[Enable]

I found this article on the Indiana University library shelf failure in 2010. It doesn't really outline any guidelines for deflection limitations, but it does provide some cool takeaways to think about

1) Book asymmetry. The failure occurred while books were being taken off one side of the shelf. This would make the entire thing less stable if the side removed was the opposite side of tilt
2) Perhaps the better way to go instead of by L / criteria would be to back-calculate how close these shelves are to tipping based on a series of absolute deflections. Find something comfortable.
3) Dont forget incidental loads in addition to the tilt (someone running into the stack or a seismic event)
4) If the stacks are to be bolted and/or braced you can bet that the bolting / bracing will be removed at some point for maintenance / moving as it was in the Indiana case

This makes me concur with phamENG that an absolute deflection criteria should be looked at rather than L / .

CWB (W47.1) Div 1 Fabricator
Temporary Works Design

https://www.enable-inc.com/

 

[KootK]

1) My main concern would be for rolling stacks. The last thing you need is some poor librarian struggling to push a rack up hill. I'd try to scare up some rack supplier recommendations even if you don't yet know the particular system to be used.

2) All my libraries have been plain, CIP slabs. Not that I'm suggesting that composite steel can't work.

3) The libraries that I've looked at all had raised floor systems. I believe that those permitted some vertical adjustment capability if needed.

4) This is one of those situations where a high percentage of your design load may actually come to pass. That always makes me a little nervous.

5) I'm guessing that you're low seismic on this but, if not, watch your hold down forces for overturning. I've struggled with that even with a CIP slab as you can get some substantial anchorage force requirements.

 

[SlideRuleEra]

Our office was built with high density shelving similar to this:

Floor loading can increase to 300 PSF:

I doubt a public library would use high density shelving for the everyday stacks, but they might for reference or archive areas.

http://www.SlideRuleEra.net

 

[JLNJ]

Hey SRE - we did one like that where the architect had high density shelving on their slab on grade (filled with tile and brick samples, no less) and they wanted to move it up on a 60 psf office elevated slab. I took one look and said, "Nope".

 

[jayrod12]

I've had a few retrofit type jobs where they wanted to use High density filing like the ones shown there. And then they had the gall to complain when I gave them the required reinforcing to get it to work out. No amount of warning at the beginning about how much work it would be made them re-think the plan. Once they saw the amount of work required to accommodate it, magically they felt they could do without the high density part. So back to 100 psf storage loading.

 

[SlideRuleEra]

JLNJ - Tile and brick! Surprising the shelves held up to that.

http://www.SlideRuleEra.net

 

[marinaman]

Thanks for all the input.

I went back and reviewed the composite deck itself. Turns out, to get deflection results that were acceptable to me for library loading, of just the deck, I had to thicken the slab to 6 1/4" (4 1/4" lightweight over 2" composite metal deck) and, I closed my beam spacing down to no more than 8'-0" o.c.

Now I'm looking at beams.

I have several beams with 32' spans.

At L/600 for post composite loadings, that's about 0.65". That's tipping a 8' tall shelf about 3/8" at the top.

I've been thinking about setting a hard post composite deflection of 1/2".....or L/600, whichever is smaller, and letting that be my deflection design criteria.

That generates pretty good sized beams. That said, the initial load causes an initial deflection that is too small to be cambered out.....so what will likely happen is the GC will likely overpour by roughly 0.4" to get the floor level when poured (but I am including an extra 1" of dead load for likely overpour in my design loads).

I'm thinking this does what I need it to do.

 

[bones206]

I've never designed a library, but I did work on a facility with massive heavy battery racks in an arrangement similar to bookshelves. The creep deflection was quite significant.

Your absolute midspan deflection of 1/2" feels reasonable to me. If you consider the long term deflection ending up around say 1", that's closer to L/360, which is what Josh was getting at earlier. So I'd give creep deflection a check (using 100% live load since it's a library) just to make sure you still end up under L/360 in the long term.