Beam supporting reinforced masonry deflection limit 1

[milkshakelake]

What's the deflection limit of a steel beam supporting reinforced masonry? I use L/600, but my beams are getting too massive, so I'm wondering if L/240 is appropriate.

ACI 530-13 5.2.1.4.1 uses L/600 for unfactored dead plus live loads. However, the commentary has something interesting.

5.2.1.4.1. The deflection limits apply to beams and lintels of all materials that support unreinforced masonry. The deflection requirements may also be applicable to supported reinforced masonry that has vertical reinforcement only.

The commentary says that the limit "might" apply to reinforced masonry that has vertical reinforcement only. All my reinforced masonry walls have ladder trusses at 16" on center vertically, which might count as horizontal reinforcement. All this reinforcement gives the masonry some crack control and ductility. Is that enough of an argument to definitively allow L/240?

Bonus question: Separately from the previous question, let's say I have to use L/600, like if it was theoretically unreinforced masonry. Would the deflection limit apply to the long beams in the picture below? The distance between them is about 6' average. My thinking is that it does apply to long beams. Although I usually consider local deflection of the beam directly under the masonry, in this case they're pretty rigid and most of the deflection would come from the long beams. On the other hand, they'd all deflect pretty much the same amount (they're all controlled by deflection), so I don't see a lot of potential for differential deflection.

I did see an old thread about this, but it's maybe referring to some outdated code commentary:

https://www.eng-tips.com/viewthread.cfm?qid=125144

 

[Lomarandil]

I used L/600 for unreinforced masonry, and I'd be reluctant to reduce that very much based on reinforcement. Masonry cracks were the number one complaint against structural engineers in my environ.

I presume you're already taking advantage of arching when appropriate?

 

[milkshakelake]

I'm not sure how arching would apply. It's kind of a long CMU wall that I apply as a linear load.

Maybe I can reduce to L/480 because it's reinforced? I know a big company around here that does that, but I've never found a code reference for it. Not sure how well that would hold up to an audit.

 

[WesternJeb]

I concur with Lomarandil.

I don't see how anything except l/600 would handle in an audit.

 

[milkshakelake]

Understood, thanks!

 

[JAE]

I think your "bonus question" is the key here.

The L/600 limit is for beams directly supporting masonry walls and is an attempt to minimize the curvature under the wall in a deflected state.
To design the long beams, you'd want to keep the deflection under the wall, along its length, under control.

In this case, the concern would be some type of equivalent L/600 control along the line of the wall, not the beams.

The parallel beams might be loaded exactly the same or they might be loaded in some various pattern, etc., where some beams deflect downward more than others and the concern then is the differential deflections along the wall length. I might be inclined to assume various spans (two beam spaces, then three, then four) and determine an L/600 deflection for each and then pattern the live loads to see what sort of maximum differential deflection occurs.

The other thing that you can do is add expansion joints in the walls at a much closer spacing that typically required.
This essentially breaks up the walls into individual segments that are much more adept at performing well under deflections than a single, long brittle wall.

 

[KootK]

I mostly see this as JAE does: the concern is curvature you generate in the walls. No differential movement between beams, no problem. Moreover, it's likely only the curvature that is induced after the mortar in the walls cure. I'd be wanting fairly regularly spaced control joints along the length of the walls which would, I suspect, mitigate most of the problem.

I also feel no compunction to respect any published deflection limits for something like this other than to observe them as being intelligent recommendations for most common situations. My only concern for something like this is whether or not I truly feel that it's a serviceability problem that the owner would object to.

 

[milkshakelake]

Thanks for the tips. I'll liberally use expansion joints. Seems a lot cheaper than enforcing L/600. I'm getting beams like W18x100+ with just L/240 due to the long span. It seems like I can justify this pretty well, as opposed to relying on some code commentary that's very open to interpretation.

I understand the reasoning, that the concern is mostly about the curvature. I'm afraid that the burden of proof is quite high when using patterned live loads. The person reviewing it might not be very experienced or nuanced, or might pick at small things. So I'll take the easy road every time rather than be a fancy engineer.

@KootK I wish the auditors around here thought like you. The code is more often gospel than intelligent recommendations. It's something I'm trying to navigate daily. When I practiced in a different state, my mentor told me that the code was mostly suggestions, and he was correct for that market. The stringency here is probably due to the sheer volume of job sites and lawsuits around here, which causes the building department to clam up and go by the book on everything.

 

[phamENG]

Look at the wall as part of the beam. If you assume you have no strength in the wall and it's just along for the ride, what kind of stresses are developing when the beam below deflects that much? My guess is compression sufficient to shoot chunks of masonry across the room. Expansion joints may alleviate this.

Now look at the wall as if it has strength. If the masonry wall is effectively a deep beam and reinforced for strength and serviceability, then it will work with the supporting beam and you'll be okay.

EDIT: these are meant to be the edge cases to bracket a final solution...your wall is likely somewhere in between.

Either way, the beam either needs to meet the unreinforced masonry deflection limits for dead load OR it needs to be shored during block laying. Because the wall is effectively unreinforced until the grout sets up and cures.

The L/600 is also accompanied by a hard limit of 0.3 inches. So any beam supporting unreinforced masonry greater than 15ft is ruled by the 0.3in limit, not L/600.

 

[KootK]

The stringency here is probably due to the sheer volume of job sites and lawsuits around here, which causes the building department to clam up and go by the book on everything.

That's too bad. For the most part, I don't feel that it ought to be the purview of codes to dictate serviceability limits. If you've got PE after your name, I feel that it is your privilege, and obligation, to be determining those limits thoughtfully for yourself.

In your case, trying to assure masonry serviceability performance by enforcing L/600 using L_beam is not just needlessly stringent, it's utterly meaningless. A curvature limit needs to be applied over some definition of span that is parallel to the wall itself for it to have any value at all. Otherwise, the limit is little better than saying "it would great if this didn't move a whole lot".

 

[phamENG]

Ohhh....that's what I get for using my laptop in the sun. I thought those were column marks...

Yeah, the long beams don't need to be designed for it so long as the relative deflection between adjacent beams doesn't create a curvature that would damage the masonry.

 

[WinelandV]

That's too bad. For the most part, I don't feel that it ought to be the purview of codes to dictate serviceability limits. If you've got PE after your name, I feel that it is your privilege, and obligation, to be determining those limits thoughtfully for yourself.

Unfortunately, some building departments don't care and refuse to budge, at least until you've started to get the boss' boss involved. But at that point, the process is punishment enough, not to mention the lost time.

Ohhh....that's what I get for using my laptop in the sun.

Look at Mr. FancyPants, engineering out in the sun. Current conditions here: 35 and windy. Yeah, I'm a little jealous.

Please note that is a "v" (as in Violin) not a "y".

 

[phamENG]

Look at Mr. FancyPants

Try Mr. FancyShorts...it's sunny and 85[sup]o[/sup]F here! Cold front comes through tomorrow, though...going to drop back to the 50s...

 

[KootK]

Also jealous. The protracted, overcast Canadian winters are murder on my mood now that I no longer believe in free will nor the long term intellectual supremacy of the human race. I'm going to need to invest in some good scotch and some sweaters woven from vitamin D.

 

[XR250]

Yea, 84 and sunny here today, binches. I'm close enough to Pham that we have similar weather.

 

[phamENG]

KootK - come on down here. XR can drive over and we can spend December drinking Scotch and sailing rather than huddled in your house trying to keep warm.

 

[jayrod12]

KootK - come on down here. XR can drive over and we can spend December drinking Scotch and sailing rather than huddled in your house trying to keep warm.

Would tag along on this trip only to listen.

 

[phamENG]

Sounds good, but we'll have to cap it there. My boat's not that big...

 

[lexpatrie]

It's better than it used to be, there was a 3/4" or 5/8" maximum that got deleted around say 2008. As to beams, are you using a cover plate to land the CMU anyway? That should give you more control of deflection, perhaps, if you include it. I also think the L/600 may apply to the masonry alone if it's applied "wet" after the precast or whatever is also loading the beam is in place. So you may not have a full dead/live load to consider and could reduce it with engineering judgement. Worst case they could shore it? How far into the mattresses do you want to go?

 

[Tomfh]

Designing it to L/600 or thereabouts is standard. In reality you get less because of all the brick arching. A real deflection of L/600 can crack bricks. I recently inspected a cracked brick apartment block. 9m transfer spans. 15mm midspan deflection. So L/600 real deflection, with cracking bad enough to engage an engineer.

I think of L/600 as more of a rule of thumb. It's not necessarily the actual deflection that will occur, but it an established norm that generally results in satisfactory performance.