If anyone finds some usefull info on this tragedy beyond the normal news agency fluff, please provide a link. Thanks Jed for the Denver Post link, much better than the crappy news network.
This failure looks very much to me like a lateral buckling failure under self weight. This check is often overlooked for composite beams, particularly by those using software to design the beam. Enercalc in particular does not check stability under self-weight during erection. I do indeed hope I am incorrect, and that some freak accident occured. Though, why it would occur several days after it was erected is unclear.
You raise an interesting and important point that I don't believe I've seen discussed here. Could you elaborate on the stability during steel erection issue or provide a few links or references?
Is the reason some engineers don't reveiw this because they feel it is a steel erection issue or because they don't see the problem at all?
Well, you also touch on a "touchy" subject. Is erection stability solely the erector's responsibility, or does the SEOR share in it?
My personal feeling is that the SEOR has to play some part in erection stability. I would break it down into components and systems. I feel the SEOR should be responsible to ensure that individual components (beams, columns, etc..) are individually stable under typical construction loads. The erector should be responsible to ensure that the system (masonry or steel framework) is stable as a whole.
By this I mean that the EOR should make sure that a beam will not buckle under its self-weight, and some nominal construction load, say 20 psf, prior to bracing by other beams or deck. However, the erector should be responsible to ensure that the partially erected structure is stable by providing temporary erection bracing. I would not expect an erector to have to brace an individual beam prior to fastening of the deck.
However, in the case of the girder collapse, the divider line becomes a little less clear. I think the engineer should check the stability of the girder. If a stability problem does exist, either he should adjust the girder shape to preclude instability, or make it explicitly clear on the contract drawings that temporary bracing will be required during erection. If the EOR takes the latter route, I think he is responsible for dictating where to brace the girder, and what capacity the bracing requires.
Now, as for whether this sort of check is required, well, I have found that this problem is most prevalent for lightly loaded (LL=40-80 psf)longer span composite beams. During erection, the beam is initially unbraced over the entire length. As the erector installs the metal deck, or builds up formwork for a concrete slab, the beam could see significant loading while unbraced.
I don't know of any links or references which talk about this issue. I just know I would not want to be the EOR in a case like this girder collapse, if they should discover that the girder had a potential for buckling under self-weight, and the engineer did not alert them to that fact.
I would love to hear input from others on the topic, but maybe we should move the discussion to another thread.
One last question. You mention the fact that the software you use does not check for this stability condition, should it? Or, should the software at least make note of the fact that it doesn't check for certain conditions of use? Afterall, the software developer can't know all the uses of the members their software may design.
I think software like RAM will check a pre and post composite bracing and loading scenario for the same beam.
1)For shored composite beam designs, we would have a note that states the contractor is required to provide the necessary shoring. Since we do not design shoring, it would be his responsibility to hire an engineer that does. I would specify a max shored length requirement.
2)I think beams definitely should be able to resist buckling under their own weight. How can any bracing be attached if the beam can't even sit there by itself?
3)For unshored beam construction, I think what needs to happen is they need to incrementally weld down the deck first. Then workers can start walking on it and bringing up additional materials. Instead of laying down deck across any beam and people walking on it without it being welded down. This is easier said than done though.
However, if I am safe and put in add'l cross bridging on an unshored composite or non composite floor system (to resist construction loads before deck is welded), I am going to get a call "why do we have to do this we've never done it before". The owner would be notified and I would get a call from him about the additional labor cost.
I am sad for the loss of life in this painful accident.
Like other failures and when the smoke is settled, we will find out that many things went wrong in this disaster.
Construction safety and stability is a huge matter. It is so complex, there is no way that we can pin the responsibly on one entity alone.
I am currently working on a project where I have to demolish an existing load bearing wall to allow for any expansion of a building. The situation is very tricky, the schedule is ambitions to say the least and the fees are not adequate for the SEOR to perform the design. We could not find as-built plans nor were we allowed to do extensive field investigation because it would have required some destructive effort to ceilings and work place.
We developed a logical sequence of construction, detailed the final configuration. We have delegated the shoring design to the contractor. We are asking for signed and sealed calculations and plans to be reviewed by our office before they can proceed.
I've seen the photo in this morning papers in the Midwest. I clipped the photo to post on the work bulletin board. I can tell you immediately without a second thought that is failure is due to erection problems.
To qualify my statement I've more than 20 years on construction and design of heavy highway infrastructure. I've worked on many projects just like this one.
This is not an error on the part of a designer who didn't check girder stresses before the concrete slab was placed.
Where, pray tell, is the lateral bracing?! AASHTO requires designers, like it or not, to place cross-frames on 25' centers, needed or not. When I was a construction inspector (many, many years ago), the contractor was either required to devise a bracing scheme to tie off a girder or use the final cross-frames to tie it off. Moreover, most contractors I've worked with would not unhook the crane even though the its the end of the day and the pick was completed. If the girder wasn't braced, no one was going home until it was.
In the photo that I have, the splice is visible and the different sizes in top flange are also noticible. I would expect the section with the larger top flange to be placed over some intermediate support, in which case, would have smaller cross-frame spacing due to bracing requirements. Many state DOTs do not allow for a full composite section in the area over the pier (negative moment region) thereby not allowing for a positive mechanical anchorage for stability calcs. Moreover, the compression flange is on the bottom where it is not braced by the slab.
In my opinion, this was the work of a unknowledgeble contractor and, if the specifications were written like many others, also the fault of the inspector.
I hope a valuable lesson was learned...it certainly cost someone dearly...
Regards,
Qshake
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
I have never designed a bridge girder, but it does seem obvious a beam of this size should be braced whether it can sit under it's own weight or not. Any shaking due to vehicles passing on the bridge or even some ground motion could tip it over and off the supports.
Excerpt:
" The girder was temporarily braced to the existing bridge with five metal bars spaced along the 100-foot length. The bracings, fastened to the bridge with 8- to 10-inch bolts, came loose as the girder collapsed. CDOT had no explanation for the girder failure, Stegman said.
Although engineers did not have an immediate explanation for what happened, it appears that the girder "flipped from the bottom out," ripping the bracings out of the concrete as the structure fell onto the freeway, Stegman said.Contractor respected for work"
It appears that the the beam was braced. I think the article is a little confusing when it states that "8 to 10 inch bolts" as the diameter and quantity is more important than the length. Whether or not the bracing was adequate is another question.
I also read in a AP report that some representative blamed vandalism, although the police seemed to dismiss that.
Now I'm assuming things here, but five (single) metal bars shouldn't be bracing a girder of 100' length which is likely to give you a 5'-6' plate girder depending on girder spacing and loading. Not when most DOT standards call for "X" bracing as the norm. Think of 6' deep girders sitting in a windstorm....with a single metal rod.
Most likely those were 8 or 10 1" diameter bolts. But again, that assumes the reporter quoted contractor/engineer/inspector properly. Recall that most reporters are under the delusion that architects are responsible for buildings and bridges!
Once again, my experience is that single girders are picked with cross-frames attached, unless of course its the last one in a row. Some times contractors will elect to pick two girders at a time with a spreader rig.
Regards,
Qshake
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
In response to jheidt2543: I don't necessarily think that the software should check stability during erection, but a lack of this check should be mentioned so the designer is aware. I am a relatively young engineer, and I have seen fellow young engineers blindly following the output from the software. When I ask, "Did you consider this....?", I have gotten responses like " Oh I thought the program checked for that...."
This sort of thing really worries me. In my experience, the senior engineers don't spend enough time passing on learned lessons to the younger engineers. It is a shame to let them learn the hard way. But the young engineers are usually too worried about "having a life" to spend time reading trade journals and learning things not covered at school. So they must take some of the responsibility as well. I am one of the lucky ones to start my career under an exceptional senior engineer with almost 50 years of experience. I learned many things working for him.
"Construction safety and stability is a huge matter. It is so complex, there is no way we can pin the responsibility on one entity alone."
There are many cases where responsibility can be "pinned" on one idividual, let alone an "entity".
Everyone wants to cop out when an accident happens. But the plain facts are, that someone, somewhere is responsible for every construction accident since man first stacked one rock atop another. Sometimes more than one individual is guilty, but very often one lone man makes a decision that cost's the life or lives of others. This is a fact that must be faced daily, and cloaking construction accidents in such a vast, murky shroud as you attempt, is trying to duck the bullet.
There is someone responsible, maybe several, but blame for these deaths will come to rest somewhere.
Lutfi - you used a key word for building engineers - "seqence". We as EOR on structures may not be required to control the erection process, but we do a very good service when we include a sequence of events that are required to maintain stability on a structure during erection. We are the best positioned to identify those elements that are required before other elements are installed.
Bridges are designed somewhat differently than buildings in that most DOT's have very standard requirements for design plans, calculations, and construction. The Colorado DOT, I'm sure, has "systems" in place for most steel bridge structures. In this case, with a long span, perhaps their standard bracing, or the contractor's standard bracing, didn't cut it.
AISC does have some language within the Code of Standard Practice that outlines something about identifying on plans when steel frames depend on other non-steel elements to provide bracing or stability.
The above article states the temporary bracing was attached to the top portion of the girder, which would be useless against LTB for the compression flange over the bent. I hope it wasn't that simple of an oversight.
I fully agree w QSHAKE, that's the failure was a result of the lack of bracing, or rather improper bracing of the single girder. The delay in setting of the second girder has added to the problem, as the braces from the other side of the girder will definitely prevent the catastrophic collapse. The detail of the installed braces, provided in one of the articles, pinpoint the likely source of the failure. (
- scroll to the sketches)
The lateral load capacity of the inclined braces, especially towards existing bridge, is almost zero (until the braces were designed as fixed at both or at list one end).
I have been doing some math on this terrible tragedy. The Weather Channel web site has Denver temperature records on an hourly basis. I have plotted them in Excel from 6 p.m. on May 11 until 10 a.m. on May 15 and looked at the patterns. Result for this entire time period:
Second most rapid 1 hour temperature increase - 9 am to 10 am - May 15
Most rapid 2 hour temperature increase - 8 am to 10 am - May 15
Most rapid 3 hour temperature increase - 7 am to 10 am - May 15
Most rapid 4 hour temperature increase - 6 am to 10 am - May 15
Suggested Area for Investigation: Lateral torsional buckling caused by forces from thermal expansion.
As I live in Evergreen, where the couple and child were from, I travel that road several times a week (Friday afternoon). It is a very high wind area due to I-70 going thru a cutout in a hogback about a mile away. I believe Qshake is on the right track with wind load.
Certainly no disagreement from me that temperature effects can be large and very noteworthy....but a girder erected without restraints? If there were no restraints (constraints) on the thermal expansion, what harm would be done by rigid body movement? Unless, of course, you take the angle bracing as a restraint (though a miserable one) and so not only was the angle skewed in the vertical plane for connection purposes but now is also skewed in the horizontal plane as it moved with the girder's expansion/contraction. That would obviously weaken the bracing.
Regards,
Qshake
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
![[pipe]](/data/assets/smilies/pipe.gif "[pipe] [pipe]")
