Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Thursday June 13 2019, @02:27AM   Printer-friendly
from the a-bridge-too-far dept.

Florida International University ("FIU") needed a foot bridge to cross a canal and busy street. An FIU committee selected a design without redundant structural support because they were wanting a dramatic landmark (the bridge looked like it was a cable stayed design, but it wasn't -- the faux cables were almost entirely aesthetic). The original specs had called for structural redundancy so that the failure of one structural member would not cause a collapse -- the committee ignored this requirement in favor of visual appeal.

The engineering firm selected for the bridge (FIGG) made an error in calculation for a critical member at the end of the span. The engineering firm providing peer review of the design (Louis Berger) has refused to turn over to OSHA, certain documentation regarding what it was supposed to evaluate and what it actually did. When the bridge section which had been built on the side of the road was moved into place, experienced workers became extremely worried about cracks that started appearing and made their worries known to those up the chain. A FIGG engineer examined the cracks but determined they "did not present a safety concern even though its engineers did not know what caused them — and despite clear evidence that they were growing daily." Apparently, the guidelines are that cracks deeper than a half inch are to be taken seriously and these were much deeper and growing daily -- one photo shows a crack 4" deep.

The final bridge would have two sections -- the long section over the roadway and a short section over the canal. The canal section was to be built in place and tied into the long section. Had the canal section been built first, the risk of collapse for the section over the roadway would have been reduced because it would have shored up the longer road section.

The bridge collapsed killing six and permanently disabling another.

Article regarding the OSHA report: https://www.miamiherald.com/news/local/community/miami-dade/article231428938.html
Article regarding independent engineering review with some good explanations which I, as a non-engineer, found informative: https://www.miamiherald.com/news/local/community/miami-dade/article212571434.html
Time Lapse Video of Bridge Collapse (released by FIU): https://www.youtube.com/watch?v=vrBOF2jugFM
Original Soylent item: https://soylentnews.org/article.pl?sid=18/03/19/1746219
Link to the OSHA report itself: https://www.osha.gov/doc/engineering/pdf/2019_r_03.pdf


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 1, Redundant) by RS3 on Thursday June 13 2019, @02:50AM (11 children)

    by RS3 (6367) on Thursday June 13 2019, @02:50AM (#854977)

    The engineering firm selected for the bridge (FIGG) made an error in calculation for a critical member at the end of the span.

    Was this a brand new type of bridge design?

    Don't they model this stuff in CAD before building it?

    Starting Score:    1  point
    Moderation   -1  
       Redundant=1, Total=1
    Extra 'Redundant' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   1  
  • (Score: 5, Informative) by Anonymous Coward on Thursday June 13 2019, @03:09AM (6 children)

    by Anonymous Coward on Thursday June 13 2019, @03:09AM (#854985)
    They say they did. But this was a multi-stage construction, and the first crack appeared after the supports were removed from the span after it had been made on the side of the road. The span did not want to stay in one piece in the stages 2 and 3. Transportation made cracks worse; placing it across the road made it worse; tightening the bolts made it fall. My opinion is that the firm (Figg) just couldn't handle this technologically challenging project and was not smart enough to realize it until the bridge fell.
    • (Score: 2) by RS3 on Thursday June 13 2019, @03:26AM (4 children)

      by RS3 (6367) on Thursday June 13 2019, @03:26AM (#854990)

      Great summary, thank you. I'm just stunned that there were not more inspectors involved- 3rd, 4th-party, etc. I know several civil engineers. I'll ask their take on the whole thing. I'd love to know the stress margins used. Again, I would think a CAD model would have shown how weak the thing was. Of course, sometimes materials aren't up to spec.

      • (Score: 4, Informative) by Anonymous Coward on Thursday June 13 2019, @03:57AM (2 children)

        by Anonymous Coward on Thursday June 13 2019, @03:57AM (#855001)

        Think what you want about CAD, if it isn't used by someone with basic engineering training _and_ engineering common sense, it is just as easy to get wrong answers as right answers. It's a tool, you have to know how to use it.

        Besides, what you probably mean is called CAE, computer aided engineering, with stress analysis, probably using finite element technology. This is also far from fool proof, many, many ways to build an incorrect FE model. CAD is computer aided drafting, replacing what was formerly done with pencils & ink on large pieces of paper/vellum--and there are still some advantages to working on a large drawing, much larger than ordinary computer screens.

        From the written descriptions, I think my elementary stress analysis course (taken from an architecture department) would have been good enough to determine that this structure was inadequate -- by hand or spreadsheet approximate calculations. With my limited background, I would never attempt to design something that could fail and hurt people, but the basic principles are not that hard to understand. Yet, it appears that the design was signed off without anyone having done a simple sanity-check type of calculation.

        • (Score: 2) by RS3 on Thursday June 13 2019, @04:28AM

          by RS3 (6367) on Thursday June 13 2019, @04:28AM (#855011)

          I agree with everything you wrote.

          Yes, I know the term CAE but I haven't heard it used in a long time. Many CAD and CAE tools get lumped together, or the CAD tool calls up a CAE tool, and the lines are blurred, and most in my world (EE) just call it CAD. Actually the big ones like Altium are being called EDA- Electronic Design Automation, but even some free ones called CAD are getting very powerful, like KICAD.

          My gut feel is that the bridge is a very bad design. I wonder how many of the tension cables could loosen or break before the bridge would collapse under its own weight.

        • (Score: 1, Troll) by schad on Thursday June 13 2019, @12:18PM

          by schad (2398) on Thursday June 13 2019, @12:18PM (#855088)

          From the written descriptions, I think my elementary stress analysis course (taken from an architecture department) would have been good enough to determine that this structure was inadequate -- by hand or spreadsheet approximate calculations.

          Amateurs screw up and cost you money. It takes a professional to screw up at such a scale that people die.

      • (Score: 2) by fraxinus-tree on Thursday June 13 2019, @09:49AM

        by fraxinus-tree (5590) on Thursday June 13 2019, @09:49AM (#855060)

        I read TFReport. The design was peer-reviwed and considered bad, but workable. The lack of redundancy was mentioned in the review. The intermediate stages of construction were not peer-reviewed (they had to be).

        The bridge could be built, but not the way they tried - the pass was clearly not stable without the suspension and they had to put the cables before removing the support under the span (and close the road for much longer).

    • (Score: 1) by Coward, Anonymous on Thursday June 13 2019, @03:26PM

      by Coward, Anonymous (7017) on Thursday June 13 2019, @03:26PM (#855164) Journal

      But the design was formally reviewed by a different engineering company. If there was a basic design flaw, the reviewers should also be held responsible.

      Overall, this fiasco would make me stay away from FIU as an engineering student or professor, and avoid hiring FIU graduates as an employer. It might be an attractive place for failure analysts.

  • (Score: 5, Informative) by Anonymous Coward on Thursday June 13 2019, @07:06AM

    by Anonymous Coward on Thursday June 13 2019, @07:06AM (#855026)

    The idea was a hybrid of a truss and cable stay. The truss was done in concrete and was expected to handle most of the load, including being self-supporting during installation. The cable stay portion was needed to handle resonance caused by the motion of humans and wind. At the time of the failure the cable stays were not in place, but it wasn't particularly windy and there were only 3 humans on the bridge. The materials have been tested; they were as specified.

    Modern engineering doesn't leave much of a safety margin and often uses inappropriate materials. Since the time of ancient Rome, we've known that concrete sucks for tension and shear forces. Romans used nice solid arches, and later people used steel trusses. Nobody sane uses concrete trusses. Well, somebody thought a thin concrete truss would be pretty and cheap, so they tried it. That was a mistake that should be obvious if you aren't blinded by a determination to make it for cost and aesthetic reasons.

    Maybe do concrete suspension cables next, eh?

    A big problem comes at the joints that make up the truss. They are fixed joints, like the sutures of the human skull. They really are joints, with the concrete poured separately. Here you're asking thin pieces of concrete to resist bending and shearing, and you have a joint right at that stress point. That is nuts. Of course it will crack.

    For the transverse direction, nothing was done to deal with that. In the other directions, they hoped to solve the problems with post-tensioned metal bars. On the day of the failure, these bars were tightened to excess in a futile attempt to stop the rapidly growing cracks. In some places the bars snapped. In some places the concrete got crushed, with blow-out events that involve chunks blasting out sideways due to the pressure.

    CAD was used of course, including structural analysis. The designer modeled it with two different kinds of software. An independent reviewer again modeled it, with a third distinct piece of software, and raised some concerns that were mostly dismissed.

  • (Score: 2) by hemocyanin on Thursday June 13 2019, @01:14PM

    by hemocyanin (186) on Thursday June 13 2019, @01:14PM (#855114) Journal

    The article about the engineers who did an independent analysis explains that these things are always modeled in a computer program but that it is important to do some manual sanity checks in case the inputs are in error. It was a garbage in, garbage out situation.

  • (Score: 4, Interesting) by Rupert Pupnick on Thursday June 13 2019, @01:25PM (1 child)

    by Rupert Pupnick (7277) on Thursday June 13 2019, @01:25PM (#855119) Journal

    As has already been pointed out, CAD systems only work well when you set up the model correctly, know how to interpret the results, and then compare them to the physical system they are modeling. When you have early evidence that the CAD result isn't matching the physical result (cracks in the concrete??) you stop what you're doing, schedule and aesthetics be damned (especially in the case of something having an impact on public safety), and find out what the hell is going on. Errors in calculations are supposed by caught in design reviews and tests. You can only have a failure of this magnitude when management is deliberately bypassing these crucial phases of the development process.

    As we used to say in my line of work, "You can't bullshit the circuit."

    • (Score: 0) by Anonymous Coward on Thursday June 13 2019, @05:47PM

      by Anonymous Coward on Thursday June 13 2019, @05:47PM (#855216)

      "As we used to say in my line of work, "You can't bullshit the circuit.""

      that's a good one. thanks