Comment of the Day: A Structural Engineer Explains Parking Garage Collapses

COMMENT OF THE DAY: A STRUCTURAL ENGINEER EXPLAINS PARKING GARAGE COLLAPSES Drawing of Broken Concrete with Rebar“. . . Structural design is generally done by computing an anticipated load (how much will the stuff in the building weigh?), multiplying that by a safety factor, and then designing a structure with enough strength to support that “factored” load. Modern building codes also estimate the anticipated strength as less than it actually is too (for errors in materials/construction). Structures generally end up with at least 3 times as much strength as they need. That said, parking structures typically have the lowest factor of safety built into their design. They fail much more often than other building types because a) they’re so cheaply constructed, b) the loading is so much lighter than other types of structures that factoring doesn’t increase the loading by as many tons, and c) failure rarely results in loss of life. @TL: You mentioned that there was a loud creek and then it failed? Any guess how long that creak lasted? 5 seconds, 3 minutes, an hour? I ask because concrete structures like this (ESPECIALLY pre-stressed concrete structures) are designed so that IF they fail, the rebar in them is the last thing to go, which will stretch and stretch and stretch gradually so that people have a chance to GTFO. The alternative is what’s called a brittle failure, where there’s just one loud pop and then bam; no warning at all. . . . Engineers always talk about this stuff in terms of ‘strain’ and ‘yield.’ Strain is how much a material can deform (stretch or compress). When the strain gets too much, the structure ‘yields’ or permanently deforms. For a concrete structure, deform === collapse. The last thing to go before a concrete structure collapses is the reinforcing steel, which has a maximum strain of about 0.02 (2%) before yielding. That means if the clear span (beam-to-beam distance) is 30 feet, you can have a sag of 30 ft / 2 * 0.02 = 3.6 inches before it actually damages the structure. Parking garage widths are typically 64 feet, which can have ~7.5 inches of sag in the middle.” [Ornlu, commenting on A Top-Down View of Last Night’s Parking Garage Collapse at One Riverway] Illustration: Lulu

14 Comment

  • “Safety Factor” = The biggest inside joke in engineering.

  • i thought steel was designed to fail first so you don’t’ have a beam crumbling and collapsing if the concrete fails first.

  • ““Safety Factor” = The biggest inside joke in engineering.”

    I’m an engineer. Please let me in on the joke. Thanks.

  • Speaking strictly as a Jane Q. Public, you first guys aren’t helping build “public trust” in parking garages.

  • This was a great comment of the day. Highly informative, made very good sense.

    Can I ask for a follow-on? If you were going to go out and inspect a parking garage to evaluate its structural condition, what would you look for as being red flags?

  • @Mathew,

    Youre correct (mostly), assuming if the failure was in the center of a beam or slab. But this appears to be a pre-cast panel. Movement of the supports as speculated could allow movement of the entire panel and thus allowing it to lodge loose.

    Please elaborate on the joke, I don’t get it.

  • @Niche: When I inspect parking garages (concrete) I have to oblige by ACI 207.3r.

    A basic visual survey, without testing I’d hammer sounding all elements of the structure, listening for delaminations. Checking connections, Corbels, pre-cast elements, checking for any released strands (if post/pre tensioned). The big check that I always look for is for bearing surfaces between pre-cast panels and their respective corbels. Point load failures, cracking are extensively mapped and documented. I also check for evidence of weathering, Improper loading or uses of the structure, improper water drainage, visible bending moment cracks. If were spending more time , we’ll do Semi-destructive and non-destructive testing on the structure…etc…

    Overall, the answer is: It depends on the structure.

  • If the maximum strain of steel is 2% along its length (is that what you meant?), that does not translate to 3.6 inches of vertical sag in a 30 ft horizontal beam. Rather the beam would lengthen to 30 ft, 3.6 inches, and the vertical sag would be somewhat shorter than 3.6 inches, right?

  • purdue – would that movement of supports be most likely be caused by something shifting in the ground/foundation failure?

  • Sure, But I’ve personally never seen lateral movement of a support column from changing Geo-technical conditions in a parking structure. I think the speculation came from another thread, but that’s not my first intuition for failure mechanism. Lets just say, I’m not convinced in regards to foundation failure is what caused this. Of the parking garages I’ve inspected: Many times – its a corbel failure. (for reinforced concrete) Ive seen some load failures before as well, but normally its when the client was plowing all their snow into the corner of the garage, or it was improperly loaded in some other factor. I would expect an out of plum column to result in a significantly more catastrophic collapse.

    Note: This is speculation based off of the pictures I’m seeing… not actual design calcs or examination. It could be geotech— but call me skeptic. Different Project: But read this report, its got some good info on how many garages are put together and how they can collapse.

    This project had an lateral impact load, that resulted in an out out plum column, coupled with corbel failures: Catastrophic collapse.

  • In engineering school, my “Strength of Materials” professor used to call the creaking, groaning, sagging period before material failure “run time” (as in the time allowed for people to run away before the building fell on them). 15 years later and I’m still totally paranoid about every strange sound I hear in my high-rise office suite.

  • I wish that there were a way for me to ‘like’ this thread about a hundred times over.

  • Oops, step away for a day and I get a COTD. Cool!
    @Matthew – No, steel gives you warning before it collapses. Concrete just collapses w/out warning. This is a concrete collapse: while this is steel failure: . Granted no one wants to be in a collapsing structure, but wouldn’t you prefer door #2?
    @Karma – Let’s approach it this way: There’s about 40,000 parking garages in the US. ( ). Last year, 4 of them collapsed. So far this year, 1 has. You are MUCH less likely to be injured in a parking garage collapse than you are in the car on the way to the garage…
    @TheNiche – My strength’s not in inspections so I defer to Purduenginerd. Nearly all the failure’s I’ve seen related to bridges, and they’re overwhelmingly bearing failures (where the beams meet the piers). I’d start there and then look for excessively long/wide cracks in the bottoms of the beams next. Finally, rust staining on the concrete shows through a lot when the rebar’s being corroded.
    @Mike – Yes, I oversimplified. The ~2% elongation’s along the bottom of the beam, and that results in an arc length of 30.3′ instead of 30′. That translates to 1.9″ of vertical deflection in the center. Sorry for the misleading math; I was just trying to do it in a way that others could follow along.

  • Steel structure depends on the structural planning with new frame so it is very effective way to plan before start working with strongly erectly.