Why the Bay Bridge broke (again)
David Nebenzahl
saying that the failure was due to Caltrans rushing the recent repair of
the broken eyebar. (Astaneh, you'll recall, is one of the severest
critics of the new "signature" span being built. Channel 7 introduced
him as a leading expert on steel structures. He teaches engineering at
UC Berkeley.)
He also dinged Caltrans for having a PR person, rather than an engineer,
addressing the media on this latest fiasco.
--
Who needs a junta or a dictatorship when you have a Congress
blowing Wall Street, using the media as a condom?
- harvested from Usenet
David Kaye
>Just now listening to my main man Abolhassan Astaneh on the teevee
>saying that the failure was due to Caltrans rushing the recent repair of
>the broken eyebar.
I'm wondering what's next. All the eyebars are the same age, so if this one
was fatigued and had to be fixed, what's to say that the other are not about
to fail, too.
I take back everything I've ever said about the rebuild of the cantilever span
being a monument to Jerry Brown (though I still think the new design is crap
because of instability). Yeah, I guess it *was* needed after all.
>He also dinged Caltrans for having a PR person, rather than an engineer,
>addressing the media on this latest fiasco.
Caltrans seemed to have way more to say in thanks to the people who worked on
it than they did actually talking about what was done, why, etc. It wasn't
one of their better moments.
--
"You're in probably the wickedest, most corrupt city, most
Godless city in America." -- Fr Mullen, "San Francisco"
spamtrap1888
> Just now listening to my main man Abolhassan Astaneh on the teevee
> saying that the failure was due to Caltrans rushing the recent repair of
> the broken eyebar. (Astaneh, you'll recall, is one of the severest
> critics of the new "signature" span being built. Channel 7 introduced
> him as a leading expert on steel structures. He teaches engineering at
> UC Berkeley.)
I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
of stress and strain in structures like bridges? And then don't they
check their work by, oh I dunno, SLAPPING STRAIN GAUGES on important
things like beams? They might even simulate wind loads and vehicle
loads.
>
> He also dinged Caltrans for having a PR person, rather than an engineer,
> addressing the media on this latest fiasco.
Good idea -- maybe Caltrans can find an engineer who's been through
Toastmasters.
Keith Keller
>
> I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
> of stress and strain in structures like bridges?
Do you really think that Caltrans can create an accurate computer model
based on 80 year old construction blueprints, 80 years of maintenance,
and 80 years of wear?
--keith
--
kkeller...@wombat.san-francisco.ca.us
(try just my userid to email me)
AOLSFAQ=http://www.therockgarden.ca/aolsfaq.txt
see X- headers for PGP signature information
spamtrap1888
francisco.ca.us> wrote:
> On 2009-10-28, spamtrap1888 <spamtrap1...@gmail.com> wrote:
>
>
>
> > I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
> > of stress and strain in structures like bridges?
>
> Do you really think that Caltrans can create an accurate computer model
> based on 80 year old construction blueprints, 80 years of maintenance,
> and 80 years of wear?
>
Why not? They've had eighty years in which to get it done. Or twenty
years, when the failure caused by the Loma Prieta quake should have
pushed bridge analysis to the top of their To-Do list. But if they had
no model, how did they come up with the eyebar fix? The SWAG process?
Guesswork because they didn't know what made the eyebar fail? From the
Chron after Labor Day:
"The fractured component is part of a network of eight eyebars. When
the single eyebar failed, the load shifted to the other seven. Crews
fashioned a system of steel braces and high-tension rods to
redistribute the load away from the damaged area.
"It is the first time such a fix has been made on an eyebar, said Mike
Forner, principal bridge engineer for Caltrans. "But it's a standard
system used throughout a lot of construction work," he said."
Read more: http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2009/09/09/MN3219K9RO.DTL#ixzz0VG4mKXRA
David Nebenzahl
> On 2009-10-28, spamtrap1888 <spamtr...@gmail.com> wrote:
>
>> I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
>> of stress and strain in structures like bridges?
>
> Do you really think that Caltrans can create an accurate computer model
> based on 80 year old construction blueprints, 80 years of maintenance,
> and 80 years of wear?
Seems like they ought to be able to.
Teevee news this evening said the failure of the repair was due to
vibration, possibly caused by wind gales. It was explained that while
the rest of the bridge structure was designed to flex with vibration,
it's possible the repair wasn't, which eventually loosened it.
All of which would seem to cry out for some kind of modeling,
simulation, *something*.
Keith Keller
> On 10/28/2009 10:45 AM Keith Keller spake thus:
>
>> On 2009-10-28, spamtrap1888 <spamtr...@gmail.com> wrote:
>>
>>> I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
>>> of stress and strain in structures like bridges?
>>
>> Do you really think that Caltrans can create an accurate computer model
>> based on 80 year old construction blueprints, 80 years of maintenance,
>> and 80 years of wear?
>
>
> All of which would seem to cry out for some kind of modeling,
> simulation, *something*.
On the repair itself? Sure, I agree. But other people are asking for
simulations of stresses on the entire bridge span. In order for a
simulation to be effective, it needs to know the starting point. Does
anyone know with a decent enough degree of accuracy what state all of
the bridge components are in right at this moment? I think it'd be hard
enough to get the initial simulation of the bridge 80 years ago correct,
much less determine its current condition. (And since its current
condition would have a significant effect on the repair, it wouldn't
surprise me if they did model the repair, and the model just wasn't
accurate enough.)
Disclaimer: I don't do computer modeling, but many people I work with do.
van...@vsta.org
>> All of which would seem to cry out for some kind of modeling,
>> simulation, *something*.
> On the repair itself? Sure, I agree. But other people are asking for
> simulations of stresses on the entire bridge span.
Well, the basic requirement is that bridges not fail. If the USA has
gotten to the point where we lack the skills needed to have that sort
of bridge, what's left? Start decommissioning them? Or--as seems to
be a popular trend in Federal Government--grant unconditional immunity
and ignore the problem?
Andy
David Nebenzahl
> On 2009-10-29, David Nebenzahl <nob...@but.us.chickens> wrote:
>
>> On 10/28/2009 10:45 AM Keith Keller spake thus:
>>
>>> On 2009-10-28, spamtrap1888 <spamtr...@gmail.com> wrote:
>>>
>>>> I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
>>>> of stress and strain in structures like bridges?
>>>
>>> Do you really think that Caltrans can create an accurate computer model
>>> based on 80 year old construction blueprints, 80 years of maintenance,
>>> and 80 years of wear?
>>
> [snip]
>
>> All of which would seem to cry out for some kind of modeling,
>> simulation, *something*.
>
> On the repair itself? Sure, I agree. But other people are asking for
> simulations of stresses on the entire bridge span. In order for a
> simulation to be effective, it needs to know the starting point. Does
> anyone know with a decent enough degree of accuracy what state all of
> the bridge components are in right at this moment? I think it'd be hard
> enough to get the initial simulation of the bridge 80 years ago correct,
> much less determine its current condition. (And since its current
> condition would have a significant effect on the repair, it wouldn't
> surprise me if they did model the repair, and the model just wasn't
> accurate enough.)
Well, anyone who's really interested in this could (and should) ask Dr.
Astaneh-Asl what he thinks of all this. I'm sure he'd be able to tell
you what current best practice is in this area. (He was, after all, the
expert that Caltrans itself has turned to in the past on matters
concerning steel bridge construction.)
John David Galt
> I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
> of stress and strain in structures like bridges? And then don't they
> check their work by, oh I dunno, SLAPPING STRAIN GAUGES on important
> things like beams? They might even simulate wind loads and vehicle
> loads.
Strain gauges can tell you how much stress is on a structure or part, but
not how much it can take.
It's hard to reliably computer-model variables like metal fatigue which
are hidden inside huge, 50-year-old steel structures. In theory you
could use radiography (X-ray photos made using large external radioactive
sources, now used for inspecting critical parts in nuclear ships and
submarines), but just imagine trying to safely position packages of
nuclear hazmat under a bridge hundreds of feet in the air for long enough
to take the pictures. I not only wouldn't try it, I wouldn't want to be
anywhere in the neighborhood. I don't think anybody else would, either.
That leaves guesswork.
Hatunen
<j...@diogenes.sacramento.ca.us> wrote:
>spamtrap1888 wrote:
>> I was asking myself: Doesn't Caltrans use, oh I dunno. COMPUTER MODELS
>> of stress and strain in structures like bridges? And then don't they
>> check their work by, oh I dunno, SLAPPING STRAIN GAUGES on important
>> things like beams? They might even simulate wind loads and vehicle
>> loads.
>
>Strain gauges can tell you how much stress is on a structure or part, but
>not how much it can take.
>
>It's hard to reliably computer-model variables like metal fatigue which
>are hidden inside huge, 50-year-old steel structures. In theory you
>could use radiography (X-ray photos made using large external radioactive
>sources, now used for inspecting critical parts in nuclear ships and
>submarines),
And piping in power plants, both fossil and nuclear. I expect
there are other civilian uses, as well.
--
************* DAVE HATUNEN (hat...@cox.net) *************
* Tucson Arizona, out where the cacti grow *
* My typos & mispellings are intentional copyright traps *
PinstripeSniper
I would opin that what is really needed is expertise in bridge
inspection. Aren't there fairly portable crack detection devices
nowadays? Ultrasonic I think? Basically if a metal component appears
not to be severely corroded or otherwise reduced in volume, then the
next thing is to check for and detect fine cracks which become larger
cracks (sometimes very very quickly) and the part fails.
It's my understanding we have many bridges in unknown and suspect
condition and this sort of thing will happen more frequently...
PPS
A fictional account of how to drastically reform the financial world...
More at http://PinstripeSniper.blogspot.com and if that gets banned, check
www.PinstripeSniper.com
Hatunen
Welds can be non-destructively tested a number of ways, but are
usually inspected with UT, ultrasound (which is really better for
butt welds), and RT radiographic testing, which can use either
X-rays or gamma rays. Since gamma ray sources are pretty small
and don't require an whole X-ray machine, radioactive gamma ray
soursces are usually used in the field. I don't expect RT is used
extensively on truss bridges like the eastern spans of the Bay
Bridge because the cost is high; radiographic testing takes time,
and requires celaring the area forsoem distance around.
When I was a quality assurance engineer for nuclear power plant
construction, only certain critcal welds were chosen for RT, and
an entire concrete room where the welds were located had to be
taped off with yellow tape with the radiation symbol.
In fact, though, a good weld inspector can visually examine a
weld and tell if it's good, or at least good enouch for the
particular application. A visual exam looks for cracks in the
welds and for bits of slag in the weld. Also, the steel around
teh weld changes color from the heat and a weld inspector can
tell from the extent and color of the affected area if enough
heat, or too much, was used.
During the Northridge earthquake soem parking garages failed, and
when the published cross sections and photos of the welds it was
perfectly clear why the joints failed.
Undercut, when teh weld material soesn't sufficiently fill the
weld groove leving a "crack between teh weld and the steel is a
real problem since such cracks create stress risers.
The parkign garages were built to commercil standards, and, for
the most part, the welders would have signed off the welds
themselves, perhaps jsut by marking the blueprint next to the
weld
you can see a bad Northridge weld at
http://www.weldreality.com/north-ridge-weld-3.jpg
http://rebar.ecn.purdue.edu/ect/images/civil/posttensionedsteel_06.jpg
http://www.daretoprepare.org/core/public/inline/13560sm.jpg