Y'all are a bunch of wankers!

### engineering question - bridge collapse

So this afternoon some gas tanker fell over and caught fire and ended up wiping out a gigantic section of freeway interchange in Oakland, which will create traffic jams for the next year or so, crippling the entire bay area.

How the hell can these bridges be designed that one gas truck can wipe out the whole thing? I can't imagine these things are really earthquake safe.
Practical Economist
April 29th, 2007 11:08pm
They said the heat from the fire melted the steel structure holding it up.
sharkfish
April 29th, 2007 11:10pm
If a meteor hit would you ask how could it be designed not to survive a meteor hit? There are always extremes that aren't designed for. And that's extreme. It's basically a giant bomb.
son of parnas
April 29th, 2007 11:12pm
I used to ask how that kind of thing wasn't part of the 9/11 attacks - blow up tanker trucks on every major highway interchange.
Philo
April 29th, 2007 11:22pm
Yeah, Philo, exactly. If the bridges are that weak. Would be hard to time it though.

But even so, gas trucks fall over and blow up from time to time, I don't see why it's a given that bridges have to have complete failure in those conditions.

With the meteor example, so what? What size meteor? You're talking a big one right, that leaves a half-mile wide crater? One that comes every 250 million years? Can't protect against that, but I see gas trucks all the time.
Practical Economist
April 29th, 2007 11:33pm
> but I see gas trucks all the time.

Explode? I've never seen it. Now calculate the probability in one particular 1/2 mile. Pretty low.
son of parnas
April 29th, 2007 11:34pm
Compared to the chance of a meteor hitting it?
Practical Economist
April 29th, 2007 11:36pm
BTW, chance of gas truck exploding in that particular 1/2 mile was 1 in 1 today.
Practical Economist
April 29th, 2007 11:37pm
> particular 1/2 mile was 1 in 1 today.

The probability that highly improbable things will happen everyday is 1.

> Compared to the chance of a meteor hitting it?

Probably both probabilities are outside the cost parameters. Much like a skyscraper handling a 747.
son of parnas
April 29th, 2007 11:42pm
A truck that was just a bit too tall tried to go under a bridge here last year.  Bunged it enough to screw up its structural integrity to the point where the entire thing had to be ripped out and rebuilt, taking months in the process.

That's what I heard, at least.  It could have also just smacked the center column, I suppose, but I think that would have resulted in a more spectacular catastrophe and stuck out more strongly in my memory.

April 29th, 2007 11:50pm
"Much like a skyscraper handling a 747."

Actually the WTC towers WERE designed to handle being hit directly by a 747, part of the design specs.

But not a 747 flying with the wings vertically with all wing tanks filled.

If a 747 had hit a WTC tower flying with wings horizontal, the tower would not have collapsed.
Practical Economist
April 29th, 2007 11:53pm
> But not a 747 flying with the wings vertically with all wing tanks filled.

Then I guess it wasn't designed for a hit by a 747.
son of parnas
April 29th, 2007 11:55pm
It's tough to protect against both fire and severe earthquake - steel box girders can withstand earthquake far better than prestressed concrete and vice-versa for fire. Looks like the designers were less concerned about fire and chose steel on hazard probability grounds.

I base this on seeing steel metioned in the news reports - I am not familiar with the bridges concerned.

Good reason *not* to have large numbers of hydrogen tankers running around.
trollop
April 30th, 2007 12:18am
OK, now THAT's a good answer. Thanks trollop.
Practical Economist
April 30th, 2007 12:57am
> Actually the WTC towers WERE designed to handle being hit directly by a 747, part of the design specs.

That would be a pretty amazing feat considering the the WTC was designed in 1966.

The WTC was designed to handle a 707 (which was the largest airliner in service at the time), at low speed:

"After the 2001 attacks, Leslie Robertson, who had participated in the structural design of the towers, said that the towers had in fact been designed to withstand the impact of the largest airliner of the day, the Boeing 707-320, in the event one was lost in fog while looking to land. According to Robertson, the modeled aircraft weighed 263,000 lb (119 metric tons) with a flight speed of 180 mph (290 km/h), as in approach and landing, which would have been much slower than the actual impacts of 9/11. He also said that they lacked a good understanding of the effects of such large fires on the structures."

Source: Robertson, Leslie E. (2002). Reflections on the World Trade Center. The Bridge Volume 32, Number 1. National Academy of Engineering. Retrieved on 2006-07-28.

April 30th, 2007 5:24am
Every fact mentioned by Practical Economist needs to take into account the Practical Economist Distortion Variable (PEDV). This variable depends on the P.E. Bullshit Quotient which is a compicated forumla highly dependent on the situation and how much P.E. has had to drink.

When talking about native americans, be sure to multiply the highest estimated number by at least 2.5. If a repuable source claims there were 100 million native americans when Columbus sailed the ocean blue, the PEDV makes the actual number of native americans at least 250 million.

Likewise, if there is a legitimate source claiming that the WTC towers could withstand the impact of a 707...combining this fact with the PEDV will let you know that the real number is a 747, so long as the wings are not vertically oriented.

Don't fret if none of this makes sense; the PEDV is still not completely understood even by the most experienced and well trained mathematicians and statisticians.
hello.
April 30th, 2007 5:35am
OK, designed to take the impact of the largest commercial airliner.
Practical Economist
April 30th, 2007 5:37am
No you guys just don't look at the evidence. 100 million is the conservative estimate of the expert in the field and in later years he has revised elements of it upwards as he obtains more hard data. You can look at how much it has gone up to see how much the whole figure is going up - at least by a factor of 2.5. Just stop and think about it. If you have an actual enumerated census done before 1500 of adults on half the island of Hispaniola and it's 1.1 million, that's just one very small island. How could you possibly claim, with that hard number, that there were only 100 times that many on the whole hemisphere? It doesn't even pass the straight face test.
Practical Economist
April 30th, 2007 5:40am
hello, I knew that it was the largest airliner, and that it was for a normal approach of the airplane, not a weird vertical one.
Practical Economist
April 30th, 2007 5:41am
P.E. that is rad; I have no idea what the WTC towers were meant to withstand. Apparently it wasn't the planes that actually hit it. In regards to Native Americans, I have not seriously looked at the evidence, but I'm totally going to multiply everything by 2.5 in the future just because 2.5 times anything is more impressive than reality. (my penis is now at least 18 inches long)
hello.
April 30th, 2007 5:45am
Jesus, even this guy's corrections are wrong.

> not a weird vertical one.

It's not the angle, it's the speed (these weren't planes coming for landing, lost in the fog), and weight.

And by the way, it was 767-200 which hit the WTC.

April 30th, 2007 5:45am
The thing that did the towers in was the wings and hence the fuel tanks going into the building. If it was flying horizontally that wouldn't have happened. The speed was totally irrelevant - the impact damage from the collision was not a factor in the collapse at all.

Comparison for you all of 707 vs 767 - they are almost exactly the same size and weight and fuel capacity:

http://www.whatreallyhappened.com/boeing_707_767.html
Practical Economist
April 30th, 2007 6:03am
The other thing that might have been an important factor in the Oakland freeway collapse is the wind.  It's possible that it fed the fire, making it much hotter than normal.
xampl
April 30th, 2007 9:15am
FWIW - have you actually been at a tanker fire? 3500 gallons+ raging in a fire? I have 3 times - and I haven't seen a bridge survive it.

The problem is that either the concrete cracks, or the steel reinforcements expand due to the heat, pushing on the outer supports, then contract as they cool, pulling themselves right out of the supports and falling. Or, they melt, depending on if the fire is underneath them.
Cory Foy
April 30th, 2007 9:18am
LOL @ muppet's awesome link - explains EVERYTHING.
Practical Economist
April 30th, 2007 1:35pm
A few years ago, I was living and working in South Florida. There was a road construction project (Cypress Creek Rd) near the office that I considered a death trap, so I'd take the offramps north (Atlantic) or south (Commercial) of that one. One day, when the construction company fucked it up more than usual, a gas tanker truck got caught on the rail tracks and was sliced in half by a train going 45 mph. It incinerated the 4 lanes of traffic stopped (many people got out of their cars and ran when they saw the crossing arm dropping on top of the tanker) on the other side of the road and the structure that held up the lights and stuff got so hot from the flames it sagged. Some people didn't run in time, and when people are alive before being burnt to death, their arms curl up into a pose that looks more like a preying mantis, or maybe a boxing pose.

The structure looked somewhat like this one:http://www.cdc.gov/niosh/fire/images/200019p1.jpg
Imagine one of those sagging so that part of it almost touches the ground.

http://www.ntsb.gov/publictn/1994/HAR9401.htm

Heat from the flames of a gasoline fire is enough to weaken most structures.

April 30th, 2007 1:57pm
FWIW here's a (partisan) overview of why concrete is far preferable (PDF):

http://www.mace.manchester.ac.uk/project/research/structures/strucfire/DataBase/References/Concrete%20&%20Fire%203557%20lo%20res.pdf

Looking at the NY Times' photo here:

http://www.nytimes.com/2007/04/29/us/29cnd-collapse.html?hp

It's apparent that two spans have fallen following the failure of the common supporting crossbeam that would have run between the two pillars seen on either side of the lower roadway.

I'm guessing that beam was unprotected steel holding up a steel framed, concrete surfaced deck and that it failed due to strength loss during incineration. Had the beam been reinforced concrete properly designed for fire resistance the fire would have burned and died before the core temperature climbed above that needed to weaken the steel reinforcement. Fibre admixtures can be employed to reduce surface spalling during a fire.

I'd guess that that had the tanker burned anywhere else except in that particular spot the roadway would be back in service once safety checks were complete.
trollop
April 30th, 2007 7:55pm
It's amazing that after the truck exploded and plummeted a hundred feet in a pile of melting concrete, the driver crawled out of the flaming wreckage, through the wall of fire, climbed up into another freeway, waved down a taxi, and asked the driver to take him to the nearest hospital, where he was treated for second degree burns.

Man? Or robot from the future? I seem to remember this scene from Terminator 3.
Practical Economist
May 1st, 2007 1:01am