I understand the often debated bridge between the boot and Sicily will be built (u$ 15bn, ETA 2033) …
which brought me to the thought of what are the length maximums for bridges (7 miles bridge makes me think there is none) … What about the span between 2 pillars … there must be some “laws of xyz” at work …
So, fine ingeneers of SD, step up to the plate and educate this unwashed mass of one.
The length between supports depends on the construction method, materials, and the expected forces (‘live’ and ‘dead’ loads of traffic, maximum wind and seismic loads, thermal stresses, current/tidal and buoyancy loads on the pilings, and loads experienced by incomplete sections during erection). Part of how the supports are placed depends on depth and condition of the riverbed or ocean/sea/lake floor and what kind of marine traffic may transit below it. There is no general rule of thumb or specific physical limitation that applies to all bridges; just the basic mechanics of structural engineering and empirical knowledge of why types of construction works best for a particular application.
I believe @Snarky_Kong is an experienced civil engineer (and there are at least a couple of others with experience with bridge and dam construction although I don’t recall having seen them post recently) and may be able to provide more authoritative and informed answers to your question.
Stranger
I read that both of the main towers will be anchored on dry land, which is why the main span length will end up being the longest in the world for a suspension bridge.
I’m not licensed anymore (moved fields to AI), and while I have structural engineering experience my expertise is more in soil mechanics and geotechnical engineering. And to separate myself from the question even more, I’ve never worked on bridge design or supports.
For overall maximum length, I think that’ll depend on what you consider a bridge. If the Lake Pontrchartrain Causeway fits the bill, then the question is basically just “where is the longest stretch of shallow water” in the world. And if you are pedantic enough, consider the Millau Viaduct. You could construct a bridge of arbitrary length (well, continent spanning at least) if you wanted to.
If you restrict the question to the longest span, I’m not sure what the limiting factor would be. My initial guess is that the diameter of the cable becomes unwieldly. Anchoring the cables or compression of the towers I think would be lower concerns. In any case I’d be shocked if economical reasons weren’t the actual limiting factor. After that, constructability. Then after that, actually resisting the loads.
This video may be of interest (10 minutes long):
understanding the “unwieldly” aspect perfectly …
wouldn’t you run also into the systemic problem of “Weight of cable” scaling to the cube (^3; volume), whereas the cable’s “resistance” scales to square (^2; diameter) …,
so, double the cable, you get 4 times the strength but 8 times the weight
just like the problems the largest dinosaurs had (replace cable with leg-bone)
Which bridge are you thinking of by that name?
the one in your backyard … ;o)
…. (and yes, I am aware there are lots of pylons in very shallow water) … and didn’t it feature prominently in a Schwarzenegger movie?
Yes.
If you answered “how large of a span could a catenary cable span without tearing itself” you’d have a decent back of the envelope estimate of the longest possible real span.
I suspect that, with suitably-tapered cables, any length is theoretically possible, but that the total mass (and other things that increase cost) required scales very poorly with length, such that there’s a practical but slightly squishy upper limit. Sort of like how, for any given desired final delta-v for a rocket, there’s some ratio of fuel to payload that’ll accomplish that, but it grows exponentially quickly… but if you had sufficient incentive and resources, you could always break whatever the current record is.
That’s what I thought. The 7 mile bridge is in effect 7 miles’ worth of bog-standard freeway overpass structures tacked end to end. As @Snarky_Kong said, if you’re just talking longest total length of adjacent individually trivial spans, you could build a bridge the breadth of Asia with a support post every 100-ish feet.
That doesn’t really get to the spirit of your question, which is what is the longest single span one can build? Or at least that’s the question relevant to the Strait of Messina bridge.
For sure the answer to longest bridge type is “cable suspension”. Then, as the others said, it’s mostly a matter of how hard the builders are willing to try. As measured in jillions of dollars.
Which also reminds me of this important work on rocket equation scaling: Model Rockets - What If? xkcd. Which also encounters certain engineering scaling issues as it gets bigger. 
Bridges tend to be made very conservatively, but we don’t have to do that.
The cables of the Golden Gate Bridge are carbon steel with a tensile strength of 1500 MPa and a specific strength of 193 kN-m/kg.
But a good carbon fiber has a tensile strength of 7000 MPa and a specific strength of 3900 kN-m/kg.
That would enable suspension bridges of a few dozen miles instead of just a few miles, since ultimately strength to weight is what we care about.
Carbon fiber composites are hard to work with and even today are poorly understood, which is why they took so long to show up in passenger aircraft. But we could nevertheless build a bridge out of them. You’d have to spend some of the mass on environmental protection systems and due to the low mass you’d probably need some kind of active or semi-active damping system to eliminate aerodynamic flutter and other things.
That is in fact exactly the case with the Metropolitan Boulevard in Montreal (aka Metropolitan Expressway / Autoroute Métropolitaine). The elevated section of Autoroute A-40 running through the north part of the city is 49.4 kilometers (30.7 miles) long, hence basically a 30.7 mile bridge. Assuming support structures every 100 feet, it would have about 1621 of them.
True Lies.
Yeah, there are a lot of practical issues with engineering a structure that is physically plausible in concept but too expensive or difficult to construct in practice. Someone here once proposed a rail launcher that would have some big truss structure that would go up to 30 kilometers (going from memory; I can’t find the thread), and while there is no fundamental physical reason a structure that size couldn’t be constructed, the practical issues with construction at such height and in near vacuum conditions, making it survive dynamic and seismic loads, et cetera, in addition to the fact that it would take many decades worth of steel and concrete production would make it completely prohibitive.
Carbon or ‘graphite’ fiber is not “poorly understood”; the production of carbon fibers and their precursors have been extensively developed over many decades, mechanical properties and testing methods to evaluate failure mechanics are well characterized and mature; and carbon fiber is used in many applications from commercial items to high performance aircraft and space launch vehicles for its stiffness, high tensile strength-to-weight, and other desirable properties such as conductivity and resistance to thermal environments. It would be a terrible material for a suspension bridge because of both its very high stiffness–creating dynamic load amplification and stress concentrations when flexing under load–poor resilience to both environmental and mechanical damage, and a tendency to fail suddenly due to cross-fiber fracture with no observable warning or inspectable condition. The appropriate alloys of drawn and cable laid steel wire rope, however, are quite resilient and resistant to environmental conditions, provide flexure to absorb shock impulses and protect other more fragile elements of a bridge, and give considerable observable warning from plastic deformation (stretching) before they reach ultimate failure.
Steel alloy wire rope is also a ‘commodity’ item that is relatively inexpensive to purchase in bulk and fabricate into stays versus having to produce and wind carbon fiber specifically for this application, handle it very carefully during transportation and erection to avoid doing any damage, and then somehow manage to construct a suspension bridge that is dozens of miles in length, which would be a massive engineering undertaken even aside from the challenges of designing the finished bridge with foundations, towers, and anchorage that could bear these enormous loads. A structure is only as strong as its weakest element in the load chain, and using a material with a really high tensile strength like carbon fiber doesn’t make the rest of the construction innately stronger.
Stranger
Compared to carbon steel, CF composites are still poorly understood. Especially as there’s no singular thing that is a CF composite. Only very recently have we developed the tools to even inspect it properly. Hence why in high-reliability systems like passenger aircraft, CF composites have made a very late entrance except in limited applications (the 787, first to use it for the majority of the structure, is less than 2 decades old).
Carbon fiber would be an excellent material for suspension bridge cables since the figure of merit is specific tensile strength. You could never build a 30-mile single-span bridge out of carbon steel cables, but you could out of an appropriate CF material.
As with anything, you’d have to engineer the specific properties.
Of course it would. The OP asked what was possible. We don’t build bridges out of CF today, and may not ever. But if you want a long single span, you need something other than steel. Carbon nanotubes are too speculative. Boron nitride fibers might work. But either way, you need materials with properties that you couldn’t get a century ago.
We also build skyscrapers that would be impossible a century ago, due to advancements in concrete, stabilization systems, and so on, so the other elements have improved as well.
Nobody’s going to mention the bridge(s) from Hong Kong to Macao?
Almost 5km (3mi) is tunnel, but it’s a total of 34mi altogether. HOwever, they avid long spans by using the tunnels on the shipping routes.
And, apparently, the single longest span on it is “only” 460 meters.
Hmm, turns out I’m not the only one with these ideas (CF+active stabilization):