> but it would surprise me if you could scale them up to a kilometer in size.

It's like the intuitive argument that heavier objects fall at the same rate as lighter objects: throw two shoes off the roof, if you tie their shoelaces together will they fall faster? You start with N kites and connect them, each kite retains its airworthiness and connecting them doesn't change that.

> nearly [constant]

The ratio falls off much slower than lifts add as you get bigger.

> you wouldn’t need stronger beams for huge kites

I don't think so, because you're just connecting small kites together, but you need to be flexible, or maybe modulate the airfoils' area (maybe open/close like butterfly wings.)

> a huge number of kites connected to each other would provide less lifting weight than the sum of the lifting weights of the individual kites.

If you just make a ball or cube, sure, but that's optional, eh? Most of my designs come out looking like modified 3D Sierpiński gaskets.

> You start with N kites and connect them, each kite retains its airworthiness and connecting them doesn't change that.

If you do that, you have N wires between the kites and the ground.

Keeping them untangled may be a problem. Your best bet probably is tying them together and having only one thicker wire towards the ground.

Making sure each of those wires takes 1/Nth of the load from the wind definitely will be a problem, even in a perfectly stable uniform wind. If you can’t guarantee that, you’ll have to make the wires a bit stronger than for the individual kites.

If you think “we won’t need 1 wire for each small kite”, you’ll need to make the connections between the kites stronger. To see why, think of the similar problem of a plank over a ditch. If a 1m plank over a 80cm ditch just holds your weight, do you think a similar 25m plank over a 20m ditch will hold you, standing in the center of the plank? Do you think it will hold 25 persons along its length?

One possibility would have a large membrane with wires connected to a pendant load. The wires would be fractal: the lines from each small kite in a unit would combine to make that unit's single line, and those units' lines would do the same, and so on.

In terms of keeping the load balanced, I figure the individual kites can modulate their surface area somehow. Maybe they fold like butterfly wings, or dilate, or have accordion pleats, whatever. Some of them would have motors and propellers for active control.

You would also want to allow for flexibility at large scales, by making the connectors elastic or even having motorized spindles.

Last but not least, the failure mode of the large structure is to break up into smaller structures. If you allow this to happen in a controlled fashion at the connectors, then the large structure breaks up into functional autonomous substructures. They can eject any damaged parts and reform.

- - - -

When I say "kite" I mean the form of the airfoil (as contrasted with planes or blimps, etc.), not that they are tethered to the ground.

- - - -

I don't think the "plank over a ditch" is similar. If anything it's more like a suspension bridge?

> I don't think the "plank over a ditch" is similar. If anything it's more like a suspension bridge?

It’s a suspension bridge if you keep “one wire per kite”, but I mentioned that example to help make clear that the “we won’t need 1 wire for each small kite” scenario would scale the weight of the kite at a speed that’s faster than linear, and that’s not the suspension bridge scenario.

I think your suggestions to modulate the surface area/adding motors or propellers/etc. also mean the larger, composite, kite would be heavier than the set of smaller kites. The structure certainly would be more complex than “just tie a thousand kites together”.