joints in buildings 3

[wildehond]

We are designing a 7 storye building that is approx 70m long. the floor plates will be post-tensioned prestress concrete. to avoid shrinkage and other problems, we have called for a joint in the building. most rule of thumb guides recommend that the joint be 40mm wide, but I can't make out the logic of this. the PT slabs will certainly make the buildings shrink away from one another, so 40mm is more than enough. And during a possible seismic event, a smaller joint would simply mean that the slab edges would knock against each other. defintely causing some damage, but surely not catestrophically? Any thoughts?

Thanks

 

[msquared48]

40 mm seems small for a 7 story building to me. That's less than two inches. I would not be so concerned about shrinkage for the gap as I would pounding from a seismic event. I would make the gap at least equal to double the deflection expected at the top from the required seismic event in the area of the project.

Some of these gaps can get to 6" (150 mm)depending on the height of the structure and seismic event modeled.

However, if you are not in a high seismic area, maybe shrinkage will control.

Mike McCann
MMC Engineering

 

[slickdeals]

I don't think 70m is long enough to warrant a joint. Maybe a shrinkage strip to allow the concrete to shrink and also to negate the post-tensioning effects depending on the center of zero movement.

 

[hokie66]

Without knowing any further details about your building, I tend to agree with slickdeals that a permanent movement joint is more trouble than it is worth. Leave temporary movement joints as long as you can before tying the parts together, but make it all work together in the final structure.

 

[wildehond]

thanks to all for your input.

your suggestion to connect the two "halves" once the initial pre-stress shortening has been allowed to occur is an idea i have been toying with and is the one i will actively pursue now. that way i avoid all the hassle that the joint will bring.

fantastic to feel complete clarity so that I can "get on with it"

 

[hokie66]

Be aware that the elastic and creep shortening due to the post-tensioning only accounts for a small fraction of the total shortening, in the order of 20%. About 2/3 of the shortening is normal concrete shrinkage. Maybe 15% can be temperature variation.

 

[wildehond]

hokie
thanks for your additional comments. The bulding is in Cape Town, South Africa. Not as big a variation in temp as you might have experience with. But I get your point about the total only been made up of a part of the post tensioning effects.

Current thinking here is that we will construct the slabs in 2 parts (near half) but with "slots" in the ends of the slabs to allow later reinforcing bars to be grouted in. For all intents, the break will be treated as a joint as far as masonry panels etc goes. But before the finishes go in (probably about 5 months after the slabs were cast), we will have the contractor grout in reinforcing into the slots to combine the 2 "halves". By then we expect approx 70% of the shrinkage to have taken place already.

The bonded bars across the "joint" will keep the building together seismically

 

[rapt]

hokie66,

I will disagree on your numbers on this Hokie66, but the end result is still the same,

About 10% from PT and equal portions of the remainder to shrinkage and temperature. So about 45% of the total is not time dependant.

Wildebond
You will find that the temperature difference does not vary that wildly in different p[arts of the world. If they pour the concrte on a really hot day and it is open to the atmosphere, the temperature difference in Cape Town could still be 30C or more.

 

[wildehond]

Rapt & Hokie

Thanks for your on-going input. With the Atlantic sea a mere 100m away from this project keeping things moderate I'd say a 30degree C range is probably what we're looking at. For a 70m building this implies around 30mm change in length.

It could be that it would be better to spend the money on some extra reinforcing, ignore the joint idea completely and detail all the masonry with goods joints.

We'll get down to calculating the overall change in length of the building to check this out.

Something else that is a pain is that the shrinkage shortening & PT work against the shear wall elements. This too, will require appropriate detailing

 

[hokie66]

wildebond,

You can use lapped bars across the delayed pour strip rather than having to install tie bars later.

And yes, one of the main reasons for using delayed strips is to lessen the shrinkage effects on the restraining elements. The other is to reduce the direct tension forces, thus reducing slab cracking issues.

rapt,

My numbers came from the memory bank and may not be based on up to date data, but which 45% are you saying is not time dependant? Shrinkage is time dependant, PT creep is time dependant, temperature goes both ways.

 

[dik]

for 70m I would likely pour the floor in two parts with a permanent pour strip between them to be filled in prior to stressing. Also for 70m you will likely find that you can stress the tendons from one end if you are using the polypropylene/ethelene sheaths. Spec it as from both ends, if you like, and check the elongation when stressed from the opposite end. You will likely find little added elongation. Also, stressing a 35m building will give you greater losses.

Dik

 

[slickdeals]

Dik:
The numbers I remember are:
125 ft or less : Stress from one end
125 - 250 ft : Stress from both ends
250' or more : Use a pour strip and stress from both ends.

 

[slickdeals]

This particular research paper by Ken Bondy suggests 100 ft and 200 ft as cut-off values.

http://www.concrete.org/tempComDocs/107394-34393/ci1401bondy.pdf

 

[dik]

That's why I suggested stressing from both ends and measure the elongation. I've found that at 200' the losses are very small. The plastic sheathing provides excellent 'lubrication' and the curvature losses are small.

Dik

 

[hokie66]

dik,

In using a pour strip, you have to stress both parts separately and join with nonprestressed reinforcing. If you leave the two sections unstressed for a period of time, there goes your crack control.

 

[dik]

The strand goes through the pour strip and the strip is cast later, prior to stressing to help alleviate some shrinkage. The stressing takes care of crack control.

Dik

 

[rapt]

wildehond

If a joint is a problem (as it usually is) then, yes, reinforce for the restraint due to shrinkage + temperature. It will be a lot of reinforcement. Possibly .6% or more. As I mentioned below, the PT tendons will not help much for this. This is assuming there are stiff concrete cores at both ends of the building.

Even without the expansion joint, I would still provide a pour strip so that the P/A gets into the slab. And stress in both directions from the pour strip, so there is no external stressing and stressing length is only 35m.!

dik,

Stressing takes care of crack control until the concrete cracks. Then, especially unbonded tendons do nothing and bonded tendons do very little as they are too widely spaced. The stress from full restraint of shrinkage if there are cores at each end is equivalent to about 6MPa, so it is cracked and significant reinforcement is needed for crack control. Add another 6 for temperature and you need reinforcement for crack control.

Also, I assume you mean that the pour strip is filled "after stressing", not prior to!

hokie66,
45% that is not time dependant is the temperature one. Shrinkage stress will increase with time and can be reduced by leaving the pour strip open longer. It is also reduced by creep.

The temoerature 45% could happen at any time over the life of the structure and is not affected by pour strips and very little by creep.

 

[hokie66]

dik said the same thing twice, that he would fill in the pour strip prior to stressing. He needs to rethink if that is what he means.

 

[dik]

For a longer structure, the pour joint is just to take up a little shrinkage and is then filled with concrete to provide the continuity of the compressive material prior to stressing. By stressing each half, without the continuity of the strand, you are likely looking at "creating a joint."

If you measure the elongation of the strand when you stress it from the opposite end, with the initial overstress in tensioning, you might be surprised in how little the elongation is when stressing the second end.

Dik

 

[hokie66]

Doing it your way, you lose the potential of the prestressing to limit early age shrinkage cracking due to restraint. If the whole thing is unrestrained, fine, but not may structures are designed that way.

The pour strip is just two construction joints, and deformed bars across the strip can tie the two together.

I know you are talking about unbonded construction, and granted you will have a lot less friction loss in greased strands than in hollow ducts.