CMU wall base connection to slab

[Delchi]

I have a large pre-fab metal building on a post-tensioned slab, and I am designing masonry walls inside to divide up the space and to supprt a wood-framed 2nd floor. (In other words, there are a combination of CMU bearing walls and CMU non-strucutral partitions on this large PT slab.) All of the walls are away from the edge of the slab because the slab extends out away from the metal building below a covered patio. The project is in a southern location where frost should not be an issue.

What is standard practice for anchoring the base of the wall to the slab? An embed plate every so often with a few headed anchor studs sticking up into a grouted cell? The lateral loads are not high, and wind controls.

Also, since I do not have extensive experience with CMU design, how are CMU walls typically anchored to a slab, or do they rely on the bond between the lowest mortar bed and the slab?

Thanks!

 

[hokie66]

1) Starter bars set in the slab into grouted cores. Using this method, the bars are difficult to locate accurately before the floor is cast. 2) Post-installed starters into grouted cores set in epoxy. Be careful to miss the strands.

 

[Delchi]

Thanks for the reply. I would also be concerned about hitting the strands, which is why I figured embed plates. BUT, either way, is there a typical minimum rule of thumb, like a #4 @ 4'-0, regardless of the load on the wall?

Thanks!

 

[jike]

Interior walls, whether bearing or not, need to be designed for a lateral internal pressure of 5 psf or seismic loads, whichever controls (per IBC). These loads (in combination with wall thickness that you choose) will determine whether the CMU walls need to be unreinforced or reinforced. You may design them as empirical or engineered.

 

[MiketheEngineer]

Don't forget to check seismic - if need be.

 

[Delchi]

So, it sounds like it is common practaice to place CMU walls on slabs with no dowels to the concrete, relying on bond and friction alone to resist lateral movement. Is that true? It was surprising to me, because we put dowels in cast-in-place concrete at the slab-to-wall connection as a matter of course, even with no calculated load.

Would I treat the CMU-to-slab bed joint the same way as unreinforced masonry, and just check stress across the joint (per ACI 530-05 section 3.2.4)? the interface between concrete and masonry does not seem to be addressed per se in the MSJC codes.

 

[KBVT]

My firm always provides dowels or embedded plates with studs where CMU bears on slabs as a matter of good practice.

 

[concretemasonry]

The first course of masonry should be laid using a full mortar bed, although the remainder of the CMU wall is laid with face shell bedding. You are just talking about a little bit of shear at the wall/floor intersection. - Of course you might be able to come up with some method that shows more or different loading.

How high are the interior walls and is there any forklift (or similar) traffic inside? If you have high walls, you may need bracing or restraint at the top because of the potential to pick up some load (due to the 5 psi or casual loading).

Dick

 

[Delchi]

I figured out what I was missing in the MSJC code. For any wall taking shear, which is most walls, the base connectoin must be designed for at least 200 plf of in-plane or transverse shear, and the connectors cannot be spaced farther than 4'-0. (ACI 530-05 1.14.2.2.2.2)

Jike, as you pointed out, there are also additional requirements for any seismic category, so they must always be considered. (1.14.3 through 1.14.7)

Thank you all for the help. Pretty basic stuff, but I wanted a better feel for what is standard construction practice in addition to the engineering requirements. I believe the designs I've seen where there were no mechanical connections to the slab were erroneous.