Hi Frank,

my original statement was that the BPM centres w.r.t. to the geometric quadrupole centres would be known better than 1 mm. If one would naively steer on these offsets the peak-to-peak orbit would be not larger than that.

The LEP based 4 mm peak-to-peak are good design values and useful for defining the maximum tolerable magnetic field errors but a maybe too "worst case" scenario.

I had another iteration with Rhodri and he confirms that the geometric plus electric errors are even well below 500 um. The geometric error is about 100 um r.m.s. and ultimately limited to be less than 500 um by mechanical constraints of fitting the BPMs onto the cold quadrupoles. To my knowledge, the LEP BPMs were not mounted to the quadrupoles which will be hopefully beneficial for LHC operation. Rhodri further estimates that the electric offset is about 200 um. In addition to that -- though the total global alignment accuracy is targeted to be better than 200 um r.m.s. -- the quadrupole-to-quadrupole misalignment is expected to be less than about 100 um.

Provided that there are not single large alignment errors, it is reasonable to assume that the orbit can be steered (if required) at least better than 1 mm peak-to-peak without rigorous beam-based alignment which is much better than the quoted 4 mm.

Cheers,

Ralph

Email from Ralph Steinhagen from 11/06/2007 with cc to Rhodri Jones