Tilt-up concrete panel cracks at connections 1

[todh]

A client asked me to look at a building that another engineer designed for him. The building is a pre-engineered metal building with a concrete tilt-up wall panel facade. The typical concrete panels are 25' wide (one bay width) and approximately 25' tall. The panels have embedded (8"x8"x3/8") steel plates with 4 studs embedded in the concrete. These plates are spaced vertically about 6' from the bottom of the panel and 6' from the top of the panel and are used to connect the panels to the pre-engineered metal building rigid frame.

The issue is a crack has appeared during construction at embedded plates welded to the pre-engineered metal building columns. The contractors initial thought was the cold concrete (30-40 degrees when erected) cracked due to the heat caused by welding on the embedded plate.

I do not think this is the cause, but would like a second, third and fourth opinion on the matter to see what others think. The locations of the cracked panels are random throughout the building. They occur in panels with openings and without openings. The attached photo of the worst crack is at an opening for an overhead door. I will attach a close-up in a follow-on post.

At first, it almost appears that the concrete panels shrunk compared to the metal building columns and cracked due to the restrained connection at the columns. The worst case crack happens to be on the first panels poured, so they cured the longest. All panels were erected within 2 days.

Any thoughts? Are repairs required or are the embedded studded anchor rods sufficiently bonded and the crack superficial? The connection just prevents hurricane force (120-mph) winds from pulling the concrete panels off the building. The majority of the cracks appear too small to epoxy inject, however this crack is wide enough to inject.

Thanks in advance for your insight.

 

[todh]

Close-up of cracked connection...

 

[hokie66]

The connection is much too rigid for the concrete to tolerate. Concrete shrinks, and steel buildings, particularly the skinny PEMB ones, move all over the place. Bolted/clamped connections should be used.

 

[JAE]

I would agree that concrete shrinkage is a good possibilibly. But that is one heck of a lot of welding and the crack occurred on an END panel where there was no restraint from the opposite end. There could be restraint from the bottom of the panel.

 

[TXStructural]

The welds need to be on flexible elements, like the opposing edges of angles. That is a lot of weld and since there is probably another on the opposite edge of the flange, they are acting against each other. Because of the size and weight of the panels, I doubt there has to be any other significant restraint.

We see this all the time where two tilt wall panels are welded rigidly together. If the cracks are tight and there is reinforcement spanning the cracks, you can probably inject epoxy into the cracks and move along.

You say that these panels have to resist hurricane winds. I wonder if the columns were designed to take that kind of lateral point loading.

 

[Ron]

Agree with all the comments thus far....combination of issues, most prominently drying shrinkage, restraint and strain incompatibility between the structural components.

That's a lot of weld for the connection. Looks like a large gap between the panel embed and the column. Are the panels bowed?

Considering that this crack is vertical, there is no restraint beyond the crack, I would lean toward drying shrinkage of the panel section, pulling in toward the center point between the two columns (assuming your columns are spaced at 20'), and restrained at the columns (obviously) by the welds.

 

[csd72]

Definately a restraint issue this.

This is a classic example of a rigid connection where a flexible one is required. Not sure why they couldnt use a simple panel clip.

 

[todh]

@ JAE - This is not an end panel. The free end you see is the edge of an overhead door opening. The photo was taken in the middle of a 200' long building. The bottom of the panel is restrained. There are 3-4 embedded angles in the bottom of the wall panel which are welded to embedded angles in the foundation.

@ TXStructural & Ron - The panels are not bowed. They were designed to have a 1/2" gap between the column flange and the wall panel. There is a 1" diameter rod behind all that weld so that the panels is welded to the rod and the rod is welded to the column flange.

With the cracks being so tight in most cases, I am assuming that the connection is still good with the headed studs and additional edge reinforcement in the panels and that epoxy injection would not be required. Perhaps they could install a panel clip after the fact in addition to the rigid connections as extra insurance. Thoughts???

 

[dik]

What sort of deformation allowances are there for the PEMB? I once had a significant extra for a horizontal girt that was supporting the top of a masonry wall... The PEMB people had designed it for a deflection of L/90...

Dik

 

[ron9876]

Looks like a fairly wide crack to me. What is the reinforcement in the panel at the joints? How far is the edge of the panel from the column?

 

[a2mfk]

Agree with Ron and Texas et al, textbook restraint crack at a tilt panel connection. Saved to my "DON'T DO THIS" library, appreciate the contribution. I'll send you a nickel everytime I look at it

Grind the weld and set this connection free!

 

[todh]

@ron9876 - The reinforcement at the panel edge consists of 2 - #5 bars. The embedded plates are 8"x8" and are aligned on one edge with the panel edge. The centerline of the column is the edge of the panel.

 

[hokie66]

The cracks will just get worse with time, and the ones that haven't cracked yet, will. Why are the panels connected to the columns? Do the panels span horizontally or vertically? If they span vertically to an eaves strut, then the connections are not required for the panels, but may be required to brace the steel.

I think you need to cut out the welds and provide a cee shaped fabrication which is then welded to the plate and slips on the column flange.

 

[todh]

The panels span horizontally. They are connected to the foundation also, but not to the roof via an eave strut or any other way. They were not supposed to be used to brace the steel, but duing my inspection I did observe that there was no wind bracing or moment frames for lateral loads.

hokie66 - I don't follow how your cee shaped fabrication would work. Could you give me a quick sketch to explain it? It would need a double lip to prevent movement inwards and outwards. Also, since it appears there is no other wind bracing except for the panels, would that still work?

 

[hokie66]

You've got the idea about the cee. Yes, plates in front of and behind the flange to allow the panel to shrink, but still provide support normal to the wall. But with no wind bracing, you are in trouble. The whole design needs a rethink.

 

[BAretired]

As a professional courtesy, if you have not already done so, you should advise the original engineer of your involvement. It has the trappings of a potential lawsuit.

I agree with those who have said that the connection is far too rigid. Shrinkage may have been the primary cause, but temperature fluctuations are likely to make matters worse over time.

Your latest observations about the lack of wind bracing adds a new dimension to the problem. As hokie says, the whole design needs a rethink.

BA

 

[todh]

@ BARetired - The real problem is that there is no true Engineer Of Record for this building. The PEMB engineer and the foundation engineer (who also designed the tilt-up panels) most likely never talked or coordinated anything, but left the coordination up to the contractor which hired each of them separately.

I plan to ask for a copy of the PEMB drawings and the foundation/tilt-up wall panel drawings so I can check the question of wind bracing and see if anyone accounted for it or not. Once I check that out, I will contact both engineers and the contractor if indeed there was no bracing provided in the PEMB design and the tilt-up panel engineer did not design the panels for bracing.

Thanks for all the insight.

 

[Ron]

If there is no separate structural engineering firm, the PEMB engineer is the engineer of record. That's not an issue. Correction of the deficient design is an issue.

 

[hokie66]

Ron, why would you pick the PEMB engineer as the EOR rather than the footing and wall designer? Whichever is the EOR, I think the one who designed the wall panels and their connections has the case to answer. The PEMB designer will slither away.

 

[a2mfk]

I've never heard of a PEMB not being designed to independently resist lateral loads, no matter what the cladding (metal panels, CMU dressing, tilt...)

However, depending on the size and dimensions of the building, I would think unless you have horizontal slip connections the tilt panels are so stiff that they will be attracting the majority of the lateral load as shear walls...

Agree with others that this thing is sounding messy with nobody being the boss.

Ron, in the PEMB projects I have been involved with I am usually the EOR. I design the foundations, the connections to the foundations, and many times many other miscellaneous portions of the building outside the scope of the PEMB. I also dictate the loading and if necessary the drift criteria. But on this one I am not sure how the PEMB engineer could say they weren't responsible for laterally bracing their own building unless they made the huge assumption that the engineer designing the foundations and tilt panels was "taking care of that".

Todh- I think I would do a preliminary report of your findings and instruct your client that the foundation and tilt engineer and the PEMB engineer need to explain their design assumptions in writing so you can review all of the information. If they are local a face to face meeting between you and them may help a lot. I am not sure if getting everyone together in the same room would be a good idea or not...