Slab control joint on expansive site - greased or not greased 1

[Euler07]

Hey all. We have a building that is being built on an expansive clay site. It is built using shotcrete concrete domes connected by a domed hallway. The slab is polished concrete so is highly susceptible to shrinkage cracking at the hallway. As a result, I am wanting to place control joints in the hallway slab at the locations shown in the image (dashed lines are walls of the concrete domes above).

I realise that dowels through slab control joints are best using a non-bonded dowel (such as greased or sleeved etc). However, since this is a reactive site and the concrete dome shell is above, I are trying to avoid any hinge action at the joint and would want full moment transfer. Therefore, I was thinking of using dowels that have full development on each side (eg. non-bonded with 500mm lap each side to slab reo). Would this at all reduce the plastic shrinkage cracking in the slab, or do you think that non-bonded dowels are the only way to go?

Thanks

 

[JAE]

In San Antonio, TX, where we used this, there was really no other option for commercial structures.
They typically used left over concrete from the trucks to start making these from day 1 of concrete operations. Could then place these after concrete construction had moved away or upward to floors above.

The only other option we used (for more minor structures) was a stiffened slab system (similar to what Euler07 suggests) but with true deeper beams (vs in-slab beams) - sort of like a waffle. But with these (even with 36" deep beams @ 10 ft. o.c. each way) we'd get the cracking from clay movements.

 

[KootK]

They typically used left over concrete from the trucks to start making these from day 1 of concrete operations.

Ahhh... so precast but not plant precast. Slick.

 

[canwesteng]

Not sure how a grid of beams would stop differential settlement. Really you should be designing for differential settlement and some uneven support below the concrete beam grillage - which is not feasible for something of this scale

 

[KootK]

Not sure how a grid of beams would stop differential settlement.

It wouldn't stop it end to end overall but the presence of stiff beams will tend to encourage a uniform gradient which improves matters for finishes, transitions, etc. You may wind up with a building tilted 5 degrees but still be able to ride a shopping cart down the hallways smoothly. But, yeah, piles would be better.

 

[Euler07]

Thanks guys. I guess with the crawlspace and cardboard void former there would be reduced uplift on the piles, although the slab would need to be designed as a suspended slab with removable propping (access to the crawlspace between the internal beams, in order to remove props, would also need to be considered carefully). If we were to go with that system my prediction is that the client would go to another engineer, which I guess happens from time to time and shouldn't be worried about.

What are peoples thoughts on using a plastic membrane just below the surface of the soil around the perimeter of the building, let's say minimum 3m wide, and graded 1:100 away from the building to allow water runoff. The purpose being to stabilise moisture levels in the soil and reduce shrinkage movement in the summer. There is plenty of room around the building to allow for this. Thanks.