Workers have installed the first of 137 strands of the nearly mile-long main cable. Each strand is comprised of 127 individual wires.
The SAS's cable is anchored into the east end of the roadway, traveling up and over the single tower to wrap around the west end before traveling back up and over the tower to anchor back into the east end. The cable features 118 miles of 2 1/2-inch steel strands and more than 17,000 5mm wires. The cable weighs 5,291 tons or nearly 10.6 million pounds.
Bay Bridge SAS Main Cable
Construction update, Jan 26, 2012:
10 Responses:
So uh, someone explain to me the non-obvious reason why they used one cable, rather than two?
The two cable method seems to have worked very well for other bridges in the area, and the total failure of one side won't impact the other side because the anchors are independent of one another.
With a single cable, it seems that if one side fails, it will weaken the anchor for the other side immediately, and make replacing it in the future much more difficult.
As far as I can tell, the reason is "because we can."
It seems like a completely retarded design from a structural point of view.
I searched for an authoritative answer, but couldn't find one. (Apparently the lead engineer is out of Cal Berkeley, maybe they do presentations or answer email...)
At first, I thought that the cables actually crossed at the top of the tower; that would actually equalize cable length as the top swayed. But it looks like they go over straight.
My next best guess is that the saddle on the western end of the span is cheaper / lighter / easier than a full anchorage (which is at the eastern end). It also makes the western end continuous, instead of cable-deck-cable (with anchorage bolts between).
I'm curious why you think it's "completely retarded from a structural point of view".
Perhaps, with two cables, if one of them failed then the other one wouldn't be enough to hold up the loaded roadway by itself and they'd be fucked anyway? What would happen if one of the two cables on the Golden Gate Bridge or on the Bay Bridge west span failed?
Yes, one could say "But with this single mast design they could use two cables where each one could normally bear the unloaded weight of the bridge, so they'd have a safety margin. If one failed they could then stop new traffic from going onto the bridge and the remaining cable could hold it long enough to get the existing traffic off the bridge, though it'd be overly stressed so it'd have to be replaced." The cost of doing something like this would probably be prohibitive, though, and the costs of this bridge have been ballooning over time like 90%+ of similar big engineering projects, anyway.
My dad used to be a reliability engineer for Lockheed. In the 70's and 80's this meant 3x redundancy, and sometimes up to 5x for really critical systems. They also used to always build two+ of each major satellite or space probe (Pioneer, Mariner, Voyager, Viking...). Since NASA's return to manned spaceflight with Shuttle in 1981, though, they tended to only be able to budget one craft for each mission (Hubble, Kepler, Chandra, Mars Observer [failed], Mars Polar Lander [failed], Pathfinder, Galileo, Magellan, Cassini–Huygens... We were lucky to get both Spirit and Opportunity rovers as a double Mars mission). So instead of redundant vehicles they had to build as much reliability as they could into each single vehicle, and count on it being enough. This is probably the kind of thing they have to deal with regarding this bridge cable, even though by its design it appears that they could have redundancy by building in two fully-supportive cables.
I expect that the deck would still fail, if it's only held up on one side. I also wonder if the towers on typical two-main-cable designs (Golden Gate, West Bay Bridge) are designed to handle a lopsided load like that, and likewise the anchorages.
As above, I'd expect that even a single-side failure on a typical suspension bridge would result in a write-off of almost all the bridge components, possibly including the towers and the anchorages.
(I'm not saying that the design discussed in the article is any better -- but it might be that they're economizing where it doesn't matter: if the failure of one main cable on a two-main-cable bridge still results in the collapse of the bridge, then why bother using two?)
Finally, replacing main cables on suspension bridges seems exceedingly non-trivial; this is a good rundown of a typical situation: http://www.forthroadbridge.org/maincablereplacement
Even if a bridge advertises a "safety factor of 4" (e.g., the Brooklyn Bridge), I believe that is referring to calculated / expected loads: static load, dynamic load, wind load, etc. The asymmetry produced by losing one of two main cables seems likely to be a "beyond design basis" scenario.
It might be that a suspension bridge made with four main cables, two on either side, could successfully have one cable at a time replaced. This implies that you're building twice as much cable as you originally need, though, which has many different cost impacts, possibly even enough to keep the project from being started in the first place. Similar to how small towns don't build superhighways when they're first starting out, then find themselves spending lots of money to upgrade the system as the town grows.
The old Tacoma Narrows bridge had a complete structural failure of the road, but both the towers, cables, anchors and tower pedestals were completely reusable. They ended up reusing the tower pedestals for the new bridge, but scrapped the old towers in favor of a wider bridge due to increased traffic (and with them the cables & anchors).
We're moving deep in to theory crafting territory here, but I suspect that in the event of a cable failure, the road surface would fail but the rest of the bridge would remain intact. That's quite a bit of a shock load for the remaining cable, but as another poster pointed out, they're grossly overbuilt in the first place.
I totally agree that we're hypothesizing about things which (I, at least) have zero formal training...
The fact that both cables survived the Tacoma Narrows collapse means that this isn't the same incident at all. Whether the oscillations provided enough differential load to exercise the same failure mechanics, I don't know.
And I further recall reading that the Tacoma Narrows towers were, in fact, damaged. Let me see if I can find the cite... Ok, not much of one, but:
Ah, here we go:
The upshot of all this is that suspension bridges — both traditional and self-anchored — are structures in tensile and compressive equilibrium; removing any substantial portion of that structure risks destroying that equilibrium. Suspension bridges are economical for long spans for exactly this reason: they use less material to provide a given load-bearing capacity. That same economy makes for less redundancy in the equilibrium.
The cable anchors are big and difficult and have to be really geologically sound. Using one cable like this they get rid of the anchors on one side. Instead of having to pull the cable tension away from the tower on that end, the tension is made into compression on the deck. That makes the deck significantly stiffer.
Losing a cable doesn't enter into it, no modern suspension bridge is designed to survive losing a cable. There aren't really any realistic failure modes for breaking a cable anyway. Overloading the bridge will take out suspender cables first; a big earthquake will damage the deck, pier or tower first.
I recommend "Structures: Or Why Things Don't Fall Down", it's quite readable and talks a bit about bridges.
In this case, it's a *self-anchored* suspension bridge, so geology doesn't enter into the anchor design at all. (Except that the self-anchored design was prompted by bad geology on the western end in the first place.)
i agree with the first sentence, but the second one is ambiguous to me. There are definitely realistic scenarios where a main cable can fail: see my previous link to the Forth Road Bridge studies. On the other paw, I agree that suspension bridges are not built to tolerate a main cable failure, so there's "no realistic *survival* mode for breaking a cable"...
In mostly unrelated news, I'm getting OpenID errors, but it's still managing to find my LJ userpic. Not sure what's going on there...