When to dowel into slab-on-grade

[HeavyCivil]

I am the sole structural designer at my firm. The majority of our work is small buildings and large tanks for WWTP's or municipal water supply plants. I often come across the issue of whether or not to dowel slabs into frost walls, especially when we use 8" cmu walls and 8" foundation walls. With a stud wall I would 'shelf' the foundation wall to provide direct bearing for the slab on the wall with either dowels/ties or isolation joints along the perimeter. I am hesitant to allow the cmu wall to bear on two surfaces without a ridgid connection, but we try to avoid cracking at all costs and I sometimes worry about providign too much restraint to slabs, even when using appropriate reinforcing and control joint spacing.

When and where is it appropraite or standard practice to provide ridgid connection between slabs and frost walls and what does a typical connection consist of? Would the slab dowel into walls on all four sides, just two or one?

Thanks,
VTeit

 

[HeavyCivil]

Also-

I apologize for the double post-

In laying out control joints I understand it is desirable to place them under walls and partitions. In other locations where there may be substantial loads on the slab but are more visible (like storage racks or machinery) is it still desirable to locate joints beneath objects that induce large quanties of shear?

On that note, in pouring a topping over a slab (for slope or to accomidate drains, or for a retrofit) I understand that matching existing control joints is desirable. Is this necessary where toppings are thin (4" or so) and contain appropraite reinforcing?

 

[Teguci]

Q1 - If you are relying on the slab to restrain the wall laterally then you should tie them together. A concern is that a SOG will settle more than a footing in which case some rotation will occur at the joint. YOu can allow for some vertical deflection at the rebar by sheathing the rebar in pipe insulation for 6" of its length.
Q2 - I've never placed control joints under walls. It is probably better to have the control joint along one edge of the wall instead.
Q3 - If there is sizable shear at a joint, it may be best to go with a thickened slab and a large key. I would have to look at the loads to make a final verdict.
Q4 - You can match the joints below the new topping or it will do that for you. I prefer to have a nice caulked joint as opposed to a crack.

 

[HeavyCivil]

1 - Walls are desiged to resist any lateral earth pressure (cantilver). I have always thought it would be best not to rely on a thin slab as a compression member.

2- Okay, good.

3- Shear is bellow capacity of 8" slab - It is due to 96 kip load distributed over two 4" chanels 108" apart, 144" long. So there some shear there (Tao=350 psi) but it is a fraction of the slabs capacity. That said, forcign a crack at this point (a relative maximum) seams like it could be troublesome - or maybe logical if one will form there anyway.

4- Okay, thanks.