Another Sewage Treatment Wall Collapsing 7

[JedClampett]

I wondered about that myself. But that isn't even the worst mistake. Since they didn't have the calculations, they should of studied every possible mechanism that might of been assumed. I think they could easily show that the design was inadequate no matter what load carrying pattern (cantilever or two way) is used. Plus they assumed the water was up to the top of the wall in their design calculations. I agree it should of been designed that way, but in the pictures, the liquid is a good foot or so below the top of wall. If the wall failed at a lower water level, the stresses at that level need to be checked.
They dwell on reinforcing lap lengths. But to prove that they had something to do with a failure needs a study of the stresses in the reinforcing at the lap location. After all, development lengths are allowed to be reduced if the analysis shows the bars aren't fully loaded.
This sloppy report, with its multiple offenders, will result in a muddying of liability, and no one being held accountable.

 

[hokie66]

I just skimmed the report, and agree with Jed. One thing which caught my eye was the statement that lapped bars are required to be spaced rather than touching, which is obviously a misuse of the wrong section of ACI318.

All those leaking cracks prior to the collapse, and in walls that haven't collapsed, shows the lack of adequate horizontal reinforcement. And the little stubs of bars sticking out of the interior walls answers the question we had from the original photos.

 

[Ron]

Whole lotta unbonded rebar there! Elite7 noted as well.

 

[hokie66]

I don't think that bond was the issue. The unbonded bars in the wall got to be that way when the wall collapsed.

The wall was clearly deficient in both vertical and horizontal reinforcement (only about .0038 Ag horizontal in a water retaining wall).

The wall wasn't cantilevered off the base slab, but according to the drawings was supposed to be pinned to the base slab as a propped cantilever. The photos show very short lengths of bar protruding from the base slab, rather than the hooked bars which should have been there, similar to the internal wall. The whole cantilever was dependent on passive resistance of the earth, which would have been fill material. When the ground finally moved enough, down she came.

The failure of the report to investigate the actual mechanism of collapse is inexplicable.

 

[hokie66]

Another report has been issued, by the well respected firm Simpson Gumpertz & Heger. I don't think it is online, but if anyone comes across it, please post.

 

[MJB315]

A news article references the report findings, and can be found at the link below. (This is not a link to the SGH report itself.)

http://www.democratandchronicle.com...bad-news-Binghamton-Johnson-City-sewage-plant

"We shape our buildings; thereafter they shape us." -WSC

 

[Gumpmaster]

The first report had a ton of problems. Why no mention of the lack of 90deg hooks at the ends of walls and the base slab (they're shown in the details)? That, to me, is the primary failure mechanism. It looks like the tension at the bottom of wall and the intersecting walls caused the failure of the 6 inch dowels embeded therein.

It is a very poor report. It seems that the investigators had never designed a water holding structure.

It will be interesting to see the SGH report and its findings.

 

[JedClampett]

Gumpmaster, the lack of hooks probably was the last straw. But if the original design assumed a cantilever or a propped cantilever mechanism and it was designed correctly, the wall should have worked, hooked bars or not. The horizontal reinforcing was so light (#5's @ 9") that it doesn't seem likely that any horizontal bending was assumed.
I think it's also very possible that one-way (vertical) bending was the planned design, it was done poorly, the wall went into two way action, and the inadequate horizontal steel and lack of support from the side walls (planned or unplanned) was the end of the story.
The calculations would be a big help (hopefully they haven't "disappeared"), but more complete drawing information would go a long way.

 

[hokie66]

Jed,
Gumpmaster has identified the same major deficiency I saw, but which not mentioned in the report by the insurance company's engineer. Granted, the wall vertical reinforcement was inadequate, but that is not why it fell. Neither was the woefully inadequate horizontal reinforcement. It collapsed because of the connection between the wall and the tank base slab.

 

[graybeach]

Here is the SGH Report:

http://images.bimedia.net/documents/BING-JC-STRUCTURAL-SIGNED-8-15-2011.pdf

 

[hokie66]

Disappointing in that SGH did not discuss the actual failure. They obviously were not asked to do that.

 

[strguy11]

This failure has me thinking.....

A contractor is currently asking our firm on a water treatment plant to use mechanical splices for the intersecting walls in a water tank. Currently, we have designed and detailed corner bars with lap splices in both directions.

I have used these mechanical splices before, but never in a liquid retaining structure. In light of this, has anyone used these in the past on tank structure, and would you continue to use them in light of the failure? I realize that mechanical bar splices may not have been the cause of the failure, but it has me questioning their use.

Thoughts??

 

[JedClampett]

If they're threaded couplers, like Lenton Formsavers,

http://www.erico.com/products/FormSaver.asp

and they're properly installed, I wouldn't have a problem with them. They market their couplers as developing 125% of the bar strength. They look beefy, too. There's other systems I'm not as familiar with, but I suspect they're OK, too.
When you think about how little of the ultimate capacity is utilized if these bars are designed correctly, you appreciate how only a cascading series of errors can cause a catastrophe like this.

 

[MJB315]

Speaking of rebar couplers, let me share this:

We had a job were we specified the couplers shown in the link below. During the inspection, we noticed that not one of the coupler screws were tightened properly (the tops torque off when properly tightened).

The Contractor commented, "Huh, I never saw those twist off before!" Which is a little unsettling.

So keep an eye out for this. It's clear in the manufacturer installation instructions that the screwheads torque when properly installed, but sometimes that information doesn't always make it to the site.

http://www.daytonsuperior.com/Lists...07b1472&ID=113&NodeId=420&FilterType=&Filter=

"We shape our buildings, thereafter they shape us." -WSC

 

[ARLORD]

Failures like these are good topics for CEU seminars.

 

[hokie66]

How do you use threaded couplers or ferrules at an intersection? Looks to me like you would not be able to turn the corner bars into the ferrules.

 

[strguy11]

Hokie -

See the attached link for more information:

http://www.barsplice.com/Barsplicer_Brochure_RevD_pg4.pdf

 

[CivilPipe]

Anyone notice that the engineering report lists the lap splice length as the primary cause of the failure? The report says the actual lap splice length was 5'-4" and should have been 8'-8". From the equation listed at Section 12.2.3 of ACI 318-08, I calculate 46" as the minimum length for a Class B Splice vertical bar and 60" for a Class B Splice horizontal bar. This also agrees with a CRSI design table that I have.

I would usually avoid a lap splice at this location but by my interpretation of ACI 318, the lap splice length meets the code.

 

[JedClampett]

I agree with CivilPipe that blaming lap splice lengths is bogus. I'm sure that there have been failures due to lap splice inadequacy, but to just state that the lap splice lengths were inadequate, and throw that out there as a cause of failure is negligent. As I stated above, development lengths can be reduced based on bar stresses, and they're the basis of lap splices. And that is assuming that they were correct in the first place.
This kind of report has lawyers licking their chops.

 

[hokie66]

Yeah, I think we came to that conclusion above. The wall fell because it came loose from its supports, not because it failed in bending. The report was amateurish.