Why did the Control Joints not work?
Why did the Control Joints not work?
(OP)
I inspected a PEMB that had large (8'x8'x2') footings at each column. Columns were spaced 25' apart on gridlines that are 50' apart. The contractor poured the footings monolithically with the 5" slab on grade (#4 @18" ocew). Control joints were 15' square grids. I'm told that cracks began about 9' away from the columns in a circular pattern and than propogated throughout the building. Cracks go right through the control joints into all sections (a section being a 15' x 15' square area of the slab as outlined by the control joints). The slab was exposed to rain and snow for 3 months prior to the roof being placed. The cracks began (apparantly) 3 weeks after the initial pour and slowly progressed over months. Weather was not too cold or hot during these first three weeks. The roof is now on and its been 6 months and the cracks seem to have stopped. I would appreciate any thoughts on why this happened and how to prevent it. Why did they first crack and why didn't the control joints stop them from propogating? Inspections document that the contractor did things correctly but who really knows. I realize more info is needed and I thank you all for your input.






RE: Why did the Control Joints not work?
The foundations carry all the load and will want to settle - the slab on grade is very lightly loaded in comparison and does not want to settle.
My first thought is that the foundations are pulling the slab on grade downwards, leading the the circular cracks around each column.
RE: Why did the Control Joints not work?
RE: Why did the Control Joints not work?
The slow progression of the cracks indicates drying shrinkage. That is also consistent with the weather you described.
The only odd part is the crack pattern. Usually drying shrinkage cracks are somewhat linear, not radial; although the size of the column footings and the stress pattern set up from shrinking toward the center could do this.
RE: Why did the Control Joints not work?
I would typically expect the cracks to occur at the thinnest section, with a 8' square footing that is 2' deep and sloped side walls the thinnest section occurs 6' from the columns, the cracks occur around 9', there is an additional bar that is bent 24" into the slab. That would mean the rebar is transferring the cracks to the slab section that has only one bar in it. Does that happen? Also, what is a typical time line for shrinkage cracks? It thought they began and ended within days of the pour.
RE: Why did the Control Joints not work?
RE: Why did the Control Joints not work?
Have you had heavy snow load during this time frame. That could cause the footings to settle and crack the slab around the columns.
I am in South Florida and we design for 20 psf live load but in reality the roofs never receive real gravity loading so I have never seen the cracking pattern that you describe.
RE: Why did the Control Joints not work?
Depending on the variation in thickness of the slab, as ron9876 noted, the joints might not have been cut deeply enough. This is a common problem.
As for shrinkage cracks and their timeline....they start quickly but can take days or weeks to show up, particularly if you are using fiber enhancement. They become wider with age and some new cracks can develop over time, depending on the restraint conditions and the amount of shrinkage expected from that particular mix design.
RE: Why did the Control Joints not work?
The radial crack pattern could be a failure of the tension tie. Again I'm saying this from the perspective of someone who has no idea of the history of the project. You might want to ask the builder if he substituted fibermesh for the WWF. The radial cracking would be occuring around the extent of the hairpins. The cracks might have stopped because the snow load is now gone. Still a 25 foot span shouldn't have must thrust. Just consider it. Later tonight I'll try to cut and paste and excerpt from the metal building book I have but it is probably copyrighted. But I think this is important if this might be the problem.
John Southard, M.S., P.E.
http://www.pdhlibrary.com/
RE: Why did the Control Joints not work?
RE: Why did the Control Joints not work?
These cracks are far more than is typically seen and I assume it has more to do with just standard shrinkage - how about the fact that the slab was exposed to snow/rain for months prior to the roof being placed - how much part do you guys think that played in these cracks?
thanks for all the input.
RE: Why did the Control Joints not work?
Proper control joints would have encouraged cracks to propagate along them, but the slab will crack at the point of lower resistance (lighter reinforcement, thinner concrete) and where there are higher tensile forces (tension induced by flexure due to settlement or curling due to drying during curing.)
Precisely why foundations should be isolated from slabs where cracks are of any concern. Fortunately, the reinforcement across cracks should help maintain aggregate interlock, and reduce vertical displacement at cracks.
RE: Why did the Control Joints not work?
Control joints don't prevent propagation of cracks perpendicular to the joints.
RE: Why did the Control Joints not work?
Thanks to all.
RE: Why did the Control Joints not work?
RE: Why did the Control Joints not work?
I've designed lots of tilt-up and basically you should never connect more than 3 panels together at a time which would around 75 feet max. And of course this is a bit different anyway since the bars don't go through joints. Concrete retaining walls usually have joints around 25 feet apart. Both types of examples are reinforced of course pretty heavily. Post tensioned slabs obviously don't have problems. ICF walls they really don't worry about it because the cracks won't be exposed. I'm not sure I agree with it but the industry had no standards yet.
The good news is that with all that reinforcement the crack size should be limited. #4's at 18"o.c. is way more than WWF. The only other idea I have at this time is to make sure expansive soils don't exist in your area. I'm thinking of the pockets we have in Florida and Texas. Up North though I really don't know.
I think right now though the best ideas going are the one mentioned above with regard to continuous steel and slab restraint.
If you get to a point where your are sure this is a slab shrinkage issue and not a structural collaspe issue than the question becomes what do you do with this situation. If Thrust isn't an issue and a tension tie for the frames is not required you might consider having the foundations isolated with full thickness saw cuts. Now again I don't have all the project info so I'm just throwing ideas out there. Its possible that relieving the restraint will allow the cracks to close up a bit. I'm not good at repairs. Hopefully this a building where the owner won't mind the cracks. I'll say this. I had never though about monolithic slabs restraining a slab like that. I've done lots of metal building monolithic slabs and luckily haven't had this issue. WWF probably yields more. I've always been a bit uncomfortable with metal building foundations. Just so much more that can go wrong. So far no cracking on m allow each square section to shrink individually instead of as a big giant unit. ine at least that I've heard of. Lots of my metal buildings have flooring on them since I do lots of educational buildings.
I've always prefers columns be centered on the slabs anyway. Plus doing two pours allows for easier uplift solutions. And I can only think that overall the contractors are only saving a little bit off the total construction cost of the building and really making the liablity for us a lot higher. But here in Florida they expect all metal buildings to be constructed like this. At times the column on edge of slab excludes this. OK now I'm rambling. And I hate the whole anchor bolt situation. I wonder if these issues are something that us engineers as a group should be trying to standardize or at least have official limitations published in a building code or something. That way we would have something to fall back on when explaining to contractors why we can't do something. We've all dealt with the classic complaint from contractors that they've done it so many times before why can't you do it. Uggh! I'm not a fan of metal buildings but they are unavoidable especially now with all the budget cutting.
John Southard, M.S., P.E.
http://www.pdhlibrary.com/
RE: Why did the Control Joints not work?
I don't think a control joint can work through a footing because of the reinforcement. It seems logical to me that it would generate some type of random slab cracking. If the cracks are wider than normal then you have to start to wonder about the amount of water ,etc. I
I haven't used a slab on grade with that amount of reinforcement before but it still seems to me that all the steel continuous thru the joints could be a problem.
RE: Why did the Control Joints not work?
RE: Why did the Control Joints not work?
maybe the conc in the footing continued to shrink while the slab exposed to rain and snow had delayed schrinkage.....
relative pull down of the slab at the col footing would explain the circular crack pattern....
or shrinkage in the slab occured first while conc in the footing continued to shrink due to it's mass and not being exposed to atmosphere....
or who knows!!!
RE: Why did the Control Joints not work?
If you don't cut any bars across the joint, then it seems to me that you haven't actually made a joint that can shrink more than any old shrinkage crack can. We want to allow our joints to be able to soak up as much of the shrinkage as possible.
Additionally, as discussed by others, if your slab is also restrained by large integrated footings, then you are even more likely to generate internal tension in the slab. From your column layout and joint dimensions, it seems like there is a mismatch, which would make it hard to locate joints in regular locations.
For example a good location would be like having a column in the centre of a panel to minimize its effect on restraint.
RE: Why did the Control Joints not work?