Article on a press release about the details of the International Linear Colider. Errr…well sort of linear 😉 :
One unusual twist to the design, said Dr. Barish, is that the tunnels, rather than being laser straight through the ground, would curve with the Earth. “It isn’t obvious and it took us a while to demonstrate that we could actually design a machine that bends” he said, but that feature would allow the digging to stay within the same geologic layers and prevent liquid cryogenics from wanting to flow “downhill” from one part of the tunnel to another.
Prof. Barish is very well qualified to lead this effort IMHO. As head of LIGO, Barish has had plenty of experience dealing with one of the ILC’s central technical challenges, namely the confluence of: (a) the beams must collide head-one to an accuracy of nanometers, (b) the beam is too rigid to steer easily (due to its high energy), and (c) on a scale of nanometers, over distances of kilometers, the earth is far from rigid — it wiggles like jello.
As with a lot of modern engineering (in fact, pretty much all of modern engineering), the ILC is all about model-based design, meaning, validate your design in detailed numerical simulations, before cutting metal. As for the ILC sensing and control, its sensors need to be quantum-limited, and its control algorithms needs to be informatically optimal. Otherwise those beams will miss each other … and that mustn’t happen!
It’s quite a sweet problem, really.