Structural or Shrinkage Crack in CMU?

[dylansdad]

While visiting a commercial structure I was asked to look at a crack in a CMU wall. The wall is located in an interior wall in the elevator penthouse above the roof. The wall is approximately 20 feet long and was constructed after the original building, mid 1960 to early 1970s. The crack is located in the general vicinity of the hoist beam. It appears to me to be a shrinkage crack but I would welcome others opinions and thanks in advance.

 

[JedClampett]

I'd vote settlement. The beam (it looks pretty big, probably a W12) is loading the CMU and the stair step crack is the extra settlement under the beam resolving itself.
I'd ask how often the hoist is used, how often the load gets near the wall and if they've ever overloaded it.

 

[masonrygeek]

It's hard to say by looking at just one photo, but it should probably be fixed in either case. When replacing block for the repair you may find an explanation.

 

[phamENG]

I vote a combination but mostly shrinkage. If you zoom in on the cracks, there's a whole lot of lateral movement but almost no vertical movement. The crack under the beam also originates well off center of the reaction, which suggests that the reaction of the beam is restraining the wall below and the section of wall to the right contracted and moved away.

Counter points to this are the presence of two such cracks in very close proximity, and some vertical movement. I suspect the wall developed a shrinkage crack and the wall segments were then able to move a little independently of one another and some settlement (not necessarily foundation if the rest of the building below isn't damaged) has occurred.

I agree with masonrygeek - should be repaired in any case.

 

[KootK]

Put me down for load/settlement:

1) the cracking basically emanates from the steel beam flanges.

2) The cracking actually crosses through solid block in one or tow spots. I feel that's more common with load/settlement than shrinkage.

3) I'd like to know more about the supporting structure. If it's beams or slab, settlement of those supports seems quite plausible.

4) A 20' wall seems kinda short to be seeing large shrinkage cracks in my opinion. My guess is that settlement based rotation in the two sides of the wall caused them to pry apart laterally a bit and then the cracks opened up some when the load was removed..

 

[dylansdad]

Thanks for the replies so far. I am not sure of the supporting structure, so more investigation is warranted. This beam was only used once.

 

[KootK]

The crack seems to fade to nothing at the bottom which might be interpreted as the the result of longitudinal restraint provided by a slab or whatever the wall is sitting on and doweled into. That could tilt in favor of settlement or shrinkage depending on how you tell the story.

Do you know what kind of roof deck was used here? It looks fiberglass-ish and, therefore, probably not prone to much expansion in the heat.

 

[dylansdad]

Roof material is Tectum decking, roof membrane, ballast.

 

[phamENG]

Didn't see the gap - I've seen several older buildings where the hoist beam was incorporated into the roof structure (intentionally or not).

I'm even less overloading or settlement now and almost exclusively lateral shrinkage.

I disagree with KootK on the split block being more likely settlement. I've seen quite a few CMU foundation walls on houses with blocks broken clean through with perfectly level floors and no distress in the foundation. Same thing in the back/side walls of large commercial spaces. It all comes down to which is stronger in that spot - the mortar, the block, or the bond between the mortar and the block. The bond is usually the weak point, but sometimes it's the block. (And restraint, and how much stress you've got, etc.)

Unless there's evidence of differential movement in the structure below, I come back to the lack of differential vertical movement in this wall. Also, if you look at where the second crack is to the right, it coincides to much smaller cracks in the main crack on the left. If you were to add the total width of the vertical cracks in each course, I bet they'll be really close. The horizontal cracks look to be well within the margin of a shadow line for slightly displaced mortar or a small shift in the wall alignment as it contracts.

If there was settlement, I'd expect to see larger gaps in the horizontal cracks and/or a variation in the vertical crack width as either the left side of the wall drops relative to the right or the right side pivots as the far right side of that 'panel' drops away off of the screen. I just don't see it.

 

[phamENG]

The crack seems to fade to nothing at the bottom

I'm not so sure. I think it may be a combination of perspective and the conduit hiding things from us. OP?

 

[dylansdad]

The crack stops at the location of the conduit coupling, one course above the lowest block in the picture. The pic was taken pretty much "head on". I wish I had taken more pics while I was there, but it was a "while you are here" things that I am sure we all get a lot of. Thanks for the replies. I need to do some more investigating.

 

[KootK]

Unless there's evidence of differential movement in the structure below. I come back to the lack of differential vertical movement in this wall.

If there was settlement, I'd expect to see larger gaps in the horizontal cracks and/or a variation in the vertical crack width as either the left side of the wall drops relative to the right or the right side pivots as the far right side of that 'panel' drops away off of the screen. I just don't see it.

I feel that's pretty easily explained:

1) The hoist beam is loaded.

2) The structure below the wall deflects.

3) The wall cracks.

5) The hoist beam is unloaded.

6) The structure below the wall returns to its undeflected position.

7) You have cracks but no relative displacement.

Don't confuse what you know about foundation settlement, where the movement is permanent, with elastic support displacement where the movement is temporary. Given the support condition, it's likely that there was not any relative vertical displacement across the crack ever. But that doesn't mean that elastic support settlement didn't still cause the crack.

Do you really think that a 20' wall is long enough to cause shrinkage cracks of the size evident in the photos?

 

[KootK]

I disagree with KootK on the split block being more likely settlement. I've seen quite a few CMU foundation walls on houses with blocks broken clean through with perfectly level floors and no distress in the foundation.

I've seen that too and I didn't mean to suggest that it wasn't possible. However, in a short, 20' wall where the cracks originate under a beam that was temporarily loaded, that tilts things in favor a a load related explanation for me. I see it a as a probability thing, not a certainty thing.

 

[phamENG]

That is certainly possible - I struggle to reconcile the fact that the elevator installer would have to be that oblivious and/or just not care, though. The hoist beam is used to install the car. A process that, while not especially fast, is usually accomplished inside a day on the projects where I've seen it done. They don't like leaving the car hanging from temporary straps if they can avoid it. So, unless it happened on a lunch break, how would they not notice that? But that's the sort of speculation/'hindcasting' we can't really rely on...

And yes, it's possible that you could cracks that size in a 20' wall. Those are certainly on the large end of the spectrum, but not without precedent. The high end of "normal" movement is about 1/8" for 20ft - that looks like it's less than 1/4". So if the blocks were abnormally wet during construction and if it sees a large enough temperature swing (is this in Mississippi? That's quite plausible if it were built in July...), 3/16" cracking wouldn't be a big surprise.

But KootK's theory on elastic deflection in the framing below during elevator installation is compelling.

 

[dylansdad]

Yes, the project is in MS, no idea when this construction was undertaken, but I was told it was after the building was constructed an elevator was installed a few years later. Again thanks for the discussion.

 

[KootK]

@dylansdad: What would you estimate the height of the walls to be? NCMA recommends control joint spacing at 25' but, given the likely height of walls (~10'-12'?) perhaps the 1.5:1.0 aspect ratio limitation is more applicable.

 

[KootK]

I struggle to reconcile the fact that the elevator installer would have to be that oblivious and/or just not care, though.

I could certainly buy into the just not caring hypothesis. In my experience, elevator guys tend to not care about much other than elevators.

On the not noticing front I wonder if, perhaps, the cracks looked quite different when the elevator guys might have been on site than they do now. Consider:

1) Given the nature of the project, the hoist beam might have been loaded fairly shortly after the block walls were constructed. That might have yielded a wall loaded at a relatively early age.

2) An early age wall might be both under strength and still have some shrinking to do.

3) Maybe you load the wall and develop relatively modest cracks that don't freak anybody out too much.

4) As the wall continues to shrink, one would expect all of that shrinkage to aggregate at the pre-existing cracks rather than distribute normally throughout the wall. The cracks expand, even in the absence of an audience to take note.

 

[phamENG]

I think we're zeroing in on maximum plausibility. (And by we I mean KootK with me mostly along for the ride...)

With the corners restrained, a small crack (say 1/16") form that early loading that went unnoticed would become a de facto control joint of sorts. Each half of the wall contracts another 1/16" we're left with 3/16" cracks. I can buy that.

 

[KootK]

Yeah, this has been a fine example of constructive debate. Good diagnostic work takes a village of speculators.

 

[BAretired]

It is common practice to fill the cells of masonry under a concentrated load with concrete or masonry grout. In the pic below, the green shade represents probable fill within the wall. Red shading indicates a possible load path by blocks bearing on each other. The red path would not become loaded until the green path had experienced some strain.

The cracks were likely phased in over time to become the shape we see now, because they are not consistent with either a horizontal tension or pure settlement.

BA