strange problems 2

[gyroman10]

This problem has occurred in a six story concrete building, 42ft x 100ft. simple plan, regular structure. The strange thing is that all of the sudden the floor marble tiles simply popped out completely.
please see the photograph and plan as attached for illustration
what could be the possible causes.
i personally think that the floor beam below this specific strip has deflected, as there are bending cracks visible in the beam immediately below.

 

[sdz]

You are possibly correct. Can you take levels to check deflection? Why has this beam deflected more than the others to cause this problem?

 

[msquared48]

Don't overlook concrete slab shrinkage too.

Mike McCann
MMC Engineering

 

[gyroman10]

slab shrinkage ?
the temperature change is not too much. it is a controlled environment. secondly the slab is 5" thick and not large spans involved.

 

[pba]

You don't say how old the building is or how long that flooring had been in place.

Also where are the joints in the finishes? There is a possibility that this is simply due to movement stresses within the finishes being released due to one tile de-bonding and then propogating across the floor. - Lets hope so.

I concur with sdz - Please get a level survey undertaken

 

[Ron]

I agree with Mike..I'd look at shrinkage.

We see this a lot in slabs on grade, less so in elevated slabs, mostly because they are not usually covered with rigid tile. While it usually occurs early on in the life of a structure, I've seen it take several years to develop.

Many times you do not see it simply because the tiles are not bonded well to the slab, but when good bond occurs and remains, this can happen with drying shrinkage in the concrete. This is likely to manifest itself over larger masses of concrete, because of the larger volumetric shrinkage. Further, rigid tile somewhat "protects" the concrete from drying faster due to ambient dehumidification of HVAC systems, thus these areas will be slower to realized comparable drying shrinkage to those exposed or carpet covered.

While some deflection and creep can exacerbate the issue, it is more likely that the drying shrinkage over time has resulted in pent up compressive stresses in the tile and that ultimately cause the tile to shear, either at the tile plane or the thinset/grout plane. Most often these failures occur quickly (though the onset takes a long time)and they can be quite loud, sounding like a gunshot.

If this were a post-tensioned structure or even prestressed, I'd look at strand relaxation or slippage; however, you stated it is a "regular structure", implying conventional reinforcement.

The cracking you see might well be shrinkage cracks, not flexural cracks. As others have suggested, check the deflections against other beams to see if this location is different.

 

[jike]

Could be a combination of shrinkage and long term movements (deflection and shortening). This is a common problem with tile bonded to slabs over large areas and inadequate expansion joints. The tile moves with the structure and tries to take on compressive load until it cracks or buckles.

 

[dik]

Clay tiles generally expand over time (like brick) and the concrete shrinks and creeps. If the tiles are located in a compression area of the beam, there might be some compression load due to creep.

Dik

 

[archeng59]

How old is the building? Possible effects of creep plus shrinkage and deflection.

 

[nuche1973]

Ron,
nice answer very informative...star for you.

 

[ali07]

Check also for any visible cracks above or below the concrete slab surface. These type of secondary beams are assumed normally as pinned in design and i think that beam release for becoming pin at support.

 

[Ron]

CCB1...thanks.

 

[Philip86]

I think what ali07 has said makes an interesting point.

Due to the location over the secondary beam it would be interesting to know if you can see cracks near the main beam support. In becoming a pin there would be significant increase in compressive force at the top layer of the secondary beam. It would also explain why this has not maybe occoured elsewhere such as midway in the slab.

As a question, normally when you design a secondary beam as a pin, what level of negative moment reinforcement would you provide? Even though you are designing as a pin, there will be some negative moment until the concrete cracks. Is this cracking acceptable?

 

[beton1]

Is there a possibility of loading change, such as a new
tenant or renovation? Redistribution of heavy filing cabinets?

 

[pba]

I must confess that when I first posted, I'd only looked at the first photo. I now see that there are two and a diagramme showing where the problem occurs - Well done Gyroman10.

If I'm now viewing things correctly the floor tiles have lifted along the length of a first interior beam position, with a span of 20 ft. The 20 ft. beam spans between main beams of 40 ft. span. No floor bays are shown in the diagramme to either side but the photos do show floor continuing to at least one side.

There is still no comment on the age of this building or the positions of joints within the floor finishes. Both could be very significant to the cause of this problem.

It is probable that the 20 ft beams are designed as simply supported. If this is the case, then cracking of any significant size to the underside of the beam may be due to some form of overstress. Looking at the furniture in the photos - it's difficult to see what that overloading might be.

If however the beam was designed as continuous, the cracking the the underside may be due to re-distribution of moments as the building "relaxes" into its steady state condition.

Now the floor slab would most likely be designed as continuous. Thus the top surface where the floor tiles have lifted would be in tension. The span of the floors would seem to be 8 ft. and again, this is the first interior support - where the bending moment will be at its highest.

Now the original post talks about cracks in the beam but does not indicate a size. I'm guessing that if these cracks were significant then more would have been said about them...

So if the building is recent, there is a possible mechanism for failure as a result of the frame settling somewhat under load.

If the building is older then the additional mechanism of shrinkage of the concrete and possible expansion of the floor tiles needs to be considered. Now the bending stresses described above are still likely to be present in an older building but are of a lesser magnitude such that they do not cause failure. The failure then occurs when the shrinkage stresses are added. As these build up slowly they can lead to a failure at any time.

I hope that helps.

 

[jeffandmike]

I would guess concrete shrinkage combined with a lack of control joints in the floor finish.

We commonly see this in concrete slabs on grade with a floor tile covering in residential construction. This is rarely seen in commercial construction as those contractors (and building designers) are typically more skilled and aware of the need for movement joints in rigid floor finishes (i.e. tile) when installed over a surface where shrinkage occurs (i.e. concrete).

One could include improper installation of the tiles/prep of the floor surface (prior to placing the tiles; i.e. allowing the thinset to skin over, using thinset that is past its shelf life, installing on a floor surface that wasn't properly cleaned, etc.) as another possible contributing factor. That said, the primary problem is probably a lack of control joints (in the floor finish).

See the link below for additional information:

http://www.tileusa.com/movementjoint_faq.htm

 

[ali07]

Location of control joints in floor finishes are important over reinforced concrete structure also. I don't think their locations are commonly set keeping in view the possible bending of the beams below in reality.

 

[hokie66]

Coming in a bit late on this, but I suggest that the tile popping is due to a combination of the reasons proposed above. Shrinkage of the concrete would be a contributory factor, but I believe deflection of both the slab and beam would also be involved. As pba said, the slab over the beam would be in tension transverse to the beam and in compression parallel to the beam, therefore the compressive force in the bonded tiles was unrestrained. As the failure occurred over the first internal beam, both the slab moment and the beam moment are at their maximum in the system.

We have talked about the tile failure purely from a structural point of view, but marble tiles are a strange material which can behave unexpectedly. Have you investigated the method of installation of the tiles? Was the appropriate adhesive used? There should be some flexibility in the adhesive if not in the grout.

 

[rcd445]

It sounds like the deflection is the biggest culprit. The bonding agent could be part of the problem also. Tile is sometimes set in epoxy and epoxy is too rigid if there is going to be any deflection or other movement. Tile normally requires expansion joints that aren't always installed. The problem may not be entirely with the structure.