Hollowcore vs. CMU Arching Action 1

[dengebre]

This question pertains to the design of lintels in CMU loadbearing walls that support precast hollowcore panels in multi-story buildings (i.e., “block & plank” buildings). It is common knowledge that lintels can be designed for only the loads within the triangular area below the apex so long as arching action can occur (block in running bond, a minimum of 8” of wall above apex, etc.). But what happens if the apex occurs above the level of the supported hollowcore? In many cases there is only 16”-24” of masonry between the opening and the bottom of the hollowcore above.

Option #1: Assume no arching action and design the lintel to support the full load of all levels above. This can result in a very heavy lintel and seems excessive.
Option #2: Design the lintel to support the full load of the level immediately above. Assume arching action occurs through the hollowcore interface so the levels further up can be ignored.
Option #2: Similar to Option #2 but the base of the triangular area is located at the top of the hollowcore. Arching of the masonry is assumed to occur above the hollowcore plane.

The fundamental question here is how does the hollowcore affect the development of arching action in the CMU? Which of the three design options is most appropriate (the attached PDF may be helpful)?
Many thanks.

 

[jayrod12]

I like option 3 maybe the most. It's a little extra conservative and shouldn't require a significantly different lintel than option 2.

I'm assuming there is no lintel in the walls above at this location?

 

[mike20793]

I always assume no arching action unless I just cannot get the lintel to work. It's conservative, but almost every engineer in my office does the same thing.

 

[dengebre]

mike20793,

That is the problem: The lintel will not work without relying on arching action. We are designing a 5-story and an 8-story block & plank building and the lintel will not support the full load. Intuitively Option #3 seems reasonable but we cannot find anything in the literature to support our gut feel.

 

[BadgerPE]

Personally, I am always hesitant to use arching action in new construction. I usually reserve it for making an opening work on existing construction. I think that the PC would most definitely interrupt the arching mechanism unless there was a mechanism to transfer shear through the PC/CMU interface. Does the PC fully interrupt the CMU?

I would probably install a steel beam below the PC that can carry all loading above it. Then the lintel above the door is carrying minimal loads.

 

[structSU10]

Yea, you could specify that the plank manufacture design and furnish a plank header at all window openings, and then the window header could be designed just for the masonry above. Assuming you have reasonable deflection compatibility... it should work.

 

[WARose]

I don't think I've ever dealt with a hollowcore floor system.....so I'm not sure how they connect together. But in any case, what I would do is sketch out option #2 (i.e. the (assumed) complete arch) and see what parts of the hollowcore fall into this zone and see if the hollowcore can take it as far as joints, compressive stresses, and so forth. Remember: arching action is going to happen whether you want it to or not due to the vertical stiffness of the arch that will form. I've inspected sagging lintels before and seen the outline of the arch in the wall (as the top of the triangle like you've drawn here). So the bottom line is: can the hollowcore participate in the arch successfully?

And the questions do not end there: be aware that arching action generates a lot of thrust that the adjacent walls have to take like a shear wall.

 

[KootK]

I screened down your sketch too much and the scanner didn't pick it up at all. I think you'll get the idea though. My thoughts:

1) Like Mike, I generally don't bother with arching. This is because I'm too lazy to evaluate the thrust capacity either side of the wall.

2) Like Jayrod, I feel that #3 is conservative and entirely defensible.

3) Were I a betting man, I would wager that the precast doesn't mess with the arching action at all. As you've detailed it, horizontal shear transfer through the precast is probably better than it is through the uninterrupted CMU.

4) I propose an option #4 in the sketch below. Basically, you would assume that the compression load turns vertical as it crosses the precast. Effectively, this shifts your load triangle up one block to create a two layer cake. Depending on how you look at things, this might also imply the need to deal with some thrust at the top of the precast but I don't see that as a deal breaker. If thrust works at the level of the lintel, it probably works at the top of the precast too. Also, if this were my baby, there would be a bond beam right above the precast and that would make the thrust moot point.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[hokie66]

Where I am, any 5 or 8 storey block building would have reinforcement in both directions and all cores filled, so the question would be moot.

 

[dengebre]

Thank you all for your input. I too typically ignore arching and design for the full load--not a significant cost driver for low rise buildings. KootK, I like your Option #4 and intuitively I agree that the hollowcore does not significantly impact the arching action (admittedly, I was trying to avoid doing the calculations to prove that we are right). There is a two-course bond beam upon which the hollowcore bears, poured monolithically with the lintel, and I do not know if that is what you are referring to. If not, making the first course that sits on top of the hollowcore a bond beam would certainly create a tension tie and allow Option #3 to develop. Again, many thanks to you all.

 

[jayrod12]

So your lintel won't work as a 3 course lintel? that seems unlikely. I would be using the lintel block and the 2 course bond beam as a 3 course lintel, I would like to think that a 24" deep lintel would work (depending on the span)

 

[KootK]

I think #3 with a bond beam above the PC really is your best bet if you need arching action. With #2 and #4, I see horizontal shear friction between the plank and the block below as the bottleneck. Shear friction stresses me at the best of times. And this isn't the greatest shear friction joint in my opinion.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[dhengr]

Dengebre:
Which option is #3..., #2 or #2?

While the lintels are pretty much confined to carrying the loads up to the immediate fl. above them, when the arching triangle extends above that fl. level; that is, they are all designed essentially the same, without much arching action up to that fl. But don’t forget that the jamb loads are cumulative when the openings are stacked. Of course, these reactions/loads start corbelling out too, immediately below the lintel bearing. But, you might still end up needing more vert. rebar and a wider jamb as you move down, in the bldg. Also, for the arching action to really work, you need a significant run of wall on either side of the arch to take, and distribute, the trust from the arching action. So, your detail at stacked openings in a large expanse of wall will is o.k. But, if that opening occurs near the end of a wall run, within a couple opening widths? of the wall end, then the arching action is (may be) suspect, without some special detailing. And then, the jamb load distrib. comes into question too. And, another issue that can come into play in a taller bldg., with stacked openings, like corridor openings, is that a wall which runs from exterior wall to exterior wall, as a shear wall, really acts like two shorter separate wall elements coupled together at each fl. level by these lintels. Thus, those lintels can see some significant fixed end moments and added shears as they try to couple the two tall wall elements together. And, these moments and shears will reverse depending upon the direction of the lateral loading.

 

[thaidavid40]

The CMU course below the HollowCore is probably a bond beam. And the lintel above the opening is either a precast concrete lintel, or could also be designed as a reinforced CMU bond-beam lintel. Since there are only three CMU courses in action here, why not make the entire three courses into one, reinforced, deep bond beam? Would that carry your load?
Dave

Thaidavid