We all know about bridges that float directly on the surface of the water, and hence only work well on relatively calm surfaces, like that lake east of Seattle. But I was thinking about submarines, which when submerged at relatively small distances below the surface don't feel the waves above, even in powerful storm. So build a bridge which rests upon a series of air-filled tanks, spaced out on both sides of the bridge and tethered to the bottom by long chains to keep them from drifting away. Balance the weight so that the tanks have just enough lift to keep them a hundred feet or more below the surface, enough that they won't be affected by waves. The bridge level itself will be far enough above the surface that the tallest waves won't hit it either, and ships can pass underneath between the floatation tanks and their support struts.

Advantages: these bridges can cross *any* ocean strait no matter how deep the water for almost the same cost/mile, no deep piers need be dug, no super-long suspended sections. Possible sites: Japan-Korea, Spain-Morocco, Britain-Ireland.

Disadvantages: despite chains, this structure would wobble a little bit; it would need lots of little expansion joints. A ship could hit the lower structure and snap a chain, destabilizing it; it would need to be monitored continuously, and closed instantly should any problem arise.
-- scottinmn, Oct 11 2002

Submerged bridge http://www.halfbake...ns-Oceanic_20Tunnel
as well as pontoons. [DrCurry, Oct 11 2002, last modified Oct 04 2004]

the waves crashing into the support struts have no effect?
-- Mr Burns, Oct 11 2002


I think your structure would still suffer serious damage in mid-ocean storms (I guess you didn't pay much attention to Perfect Storm).

Someone else has proposed submerging the roadway itself at a sufficient depth to avoid surface waves, which seems a much likelier solution.
-- DrCurry, Oct 11 2002


// ...submarines, which when submerged at relatively small distances below the surface don't feel the waves above, even in powerful storm. //

Yes, but your bridge wouldn't be submerged. The tanks would be, yes, but the supports will rise up and out of the water, and feel the wave effects.
-- waugsqueke, Oct 11 2002


I see a trend in the criticism--but let me point out that the support struts would have a relatively small surface area compared to say a ship or a floatation tank. Yes, they would get some pressure from the waves, but then that's what the chains holding the tanks down are for (and I wouldn't make these absolutely rigid chains or cables, which would be more likely to snap eventually, but equipped with large coils that would absorb the forces after contracting/expanding a few centimeters, reducing the movement down to something the expansion joints can handle).

Waves crash against ordinary bridges too, and they don't fall down if the base is strong enough. So you put more cables down to the bottom until you get something that's semi-rigid enough to take the waves.

I also don't think this would work mid-ocean; I'm not crazy like the people talking about bridging the Atlantic. :-) For that kind of situation, a wholly subsurface tunnel makes more sense, although I for one don't look forward to driving or taking a train 1000's of km all underwater. But the floating bridge might work for shorter distances near large landmasses (limiting storm powers), but which are currently unfeasible due to water depths.
-- scottinmn, Oct 12 2002


[scottmmmm] Do try again, the title was a nice descriptive piece of bait. Gave me pause thinking of "A Moveable Feast", more than once.
-- blissmiss, Oct 13 2002


Quite an odd one this, I can not see any advantage in building these crazy bridges. Maybe I just don't have the intelect to understand what the heck your going on about.
-- big_boy_gun_bell, Oct 14 2002


The armies of the world have been building pontoon bridges for years, but these are considered temporary, and indeed if you've ever seen films of the storms that hit Normandy shortly after the invasion you'll understand why. The force of moving water is easily underestimated.

While submarines aren't subject to surface waves they're moved about quite a lot by currents nontheless. Try anchoring one and it would probably rip the chains from the hull.

Even if we get past all that, the tanks are going to need an active system for maintaining bouyancy under different conditions of load, water density (yes, it changes), temperture and so forth. That means compressors, motors, and power. Radar beacons to signal their presence to ships. Active monitoring to make sure a support isn't sinking. And regular maintenance.
-- mje, Oct 28 2003


At this point it's probably "water under the bridge" (sorry), but as a naval architect I just have to point out that the concept of balancing the tanks so they have "just enough lift to keep them a few hundred feet or more below the surface" is not possible. The tanks can have either positive, neutral or negative buoyancy. Positive, and they will float, negative they will sink. However, neutral buoyancy means they will drift aimlessly according to minute changes in temperature, density, etc., certainly offering no support for a roadway. (A submarine that is ballasted can controll depth dynamically when underway, and only through minute adjustments of the ballast when stopped.)

Without "freeboard" (the part of a vessel above the waterline), there is no reserve buoyancy to keep the vessel afloat when additional weight is applied. Perhaps you were thinking that the tanks would have positive buoyancy and be restrained by the cables. In that case the amount of stress in the cables would be comparable to suspending the roadway with cables from above -- There would have to be a lot of them, they would be very difficult to maintain, and you would not want to be on the bridge when they finally corroded through!
-- etsell, Jul 29 2004


[estell] [Perhaps you were thinking that the tanks would have positive buoyancy and be restrained by the cables. In that case the amount of stress in the cables would be comparable to suspending the roadway with cables from above -- There would have to be a lot of them, they would be very difficult to maintain, and you would not want to be on the bridge when they finally corroded through!]

Yes, that's exactly what I meant, I edited the description to make this more clear. While the idea has some problems, I still don't see this as completely crazy, though whether its economically feasible I don't know. But you've correctly identified the basic idea here: the bridge is essentially suspended *from below*, the cables holding the boyancy tanks down and the bridge suspended on the tanks, rather than the bridge being suspended from above, from cables connected to towers. I believe you're correct that the tension on the cables would be essentially the same as for a suspension bridge, which means there must be a lot of cables, but not an inconceivably vast number. Of course they must be longer, since they have to extend all the way to the sea bottom. But since there's no need to drive pilings into the bedrock, a very significant expense for normal bridges, this offsets the cost somewhat.

True, cables can corrode; but so can cables on a suspension bridge. I'd suggest using kevlar cables and aluminum boyancy tanks. And you'd have some regular system of observing them, perhaps electronic monitors, so if one occasionally broke you'd replace it promptly before it created a serious problem. Not sure that this would be incredible difficult to maintain, either--if a cable breaks you just add a new one, perhaps attaching a new cable to the bottom weight with a robotic submersible, and I'm not sure that's any more expensive than, say, replacing a cable on a suspension bridge 100s of meters in the air. Replacing a broken tank would be much more expensive, so these would need more overbuilding (perhaps they should be built with 1000s of isolated compartments--and kept away from icebegs :-)

With sufficient overbuilding the structure could theoretically be quite safe. I'm no engineer but here are some rough numbers. The Golden Gate Bridge weighs 8x10^8 Kg, supported by two piers. Assume that each pier for a similar ocean bridge is actually two columns separated by 50m in order to provide some balance. So each column must support 2x10^8 Kg of bridge, for a force of 2x10^9 N. A steel column has tensile strength 2.5x10^8N/m2; so the cross-sectional area should be at least 10m2. It wouldn't just be a solid column of course; assume it's approximately 5m across. Let's overbuild by a factor of 4 and make it 10m across.

Now, the most devastating ocean waves may be as much as 30m deep with forces of 10^6N/m2. So they would strike each pair of columns with a force of 30m x 2 x 10m x 10^6N/m2 = 6 x 10^8 N, or just over twice the weight normally supported by each column. Half this force would lift one column, the other half would exert a downward force on the other side. We made the tanks 4 times stronger than needed, so each column can hold the bridge, and sustain an extra torque load on the side away from the wave delivered to a boyancy tank 50m below the normal surface/center of the wave. How big do those tanks need to be? Each column must carry 2.5x10^8 N + up to 3x10^8 N during the worst wave, for 5.5x10^8 N total. Each m3 of displaced water provides 10^4 N of force, so the tanks under each column must hold about 5.5 x 10^6m3; 6 spheres 130m in diameter each would work. That's big, but not inconceivable. These should be aluminum, though of course internal structural struts could be steel.

Since their lifting force beyond the weight of the bridge is 3 x 10^8 N for each column's set of tanks, they must be held down by cables in tension by this amount connected to concrete blocks on the seafloor. Here I'd use kevlar, tensile strength 3600MPa; they also have to counteract a lifting force on the side the wave comes from, and we should have a margin of safety...maybe 40 cables with a cross-section of 10x10cm each would do it. Does anyone have any figures for typical ocean current forces? I saw something suggesting that they can flow at a rate of up to 1m/s; that's a force of 10^4 N/m2, possibly as much as 5 x 10^8 N per set of tanks. So double the number of cables and put quite a few of them out at an angle towards the source of the prevailing current, and a few in other directions to handle occasional current changes.

Expensive? Yes. Economically feasible? I have no idea. But possible, I think, unless I have my numbers very wrong.
-- scottinmn, Sep 10 2006



random, halfbakery