Machine Tolerances in Cleaning Insertions

1
Machine Tolerances in Cleaning Insertions
J. Wenninger AB-OP SPS Beam Operation

• Tolerances
• Orbit stabilization
• Beam optics

2
A world of collimators

• Operation without collimators (All OUT) is only
  possible at the LHC with very low intensity and
  around injection
• One pilot-ish bunch (5 109 p) ? no quench
  expected
• One nominal bunch (1011 p) ? no damage, but
  risk of quench
• At high energy use of collimators will be
  mandatory, but coarse settings are acceptable for
  low intensity and during the initial phases.
• This is of course independent of the collimator
  design issue.

3
Getting started at injection

• The available machine aperture at injection is
  8.5s with a margin of
• ? 4 mm closed orbit 20 b-beat mom. offset
  mech. tolerances
• Protection devices for injection will be set ? 7s
• ? primary collimators will be set to 5.5-6s
• Both absolute orbit excursions and b-beat must be
  under control, or else the collimator settings of
  5-6 s must be tightened even more !

4
Constraints on machine changes

• The following machines changes (wrt a reference
  situation) lead to a 50 degradation of the
  nominal betatron cleaning efficiency
• 8 b-beating
• 0.6 s orbit shift
• 50 mrad angle change

Collimation inefficiency versus position error

• Tolerances are cleary tigher if there is a
  combined change of b-beat, orbit as is often
  the case !
• ? may have to take a factor 2 off from those
  numbers !

Note the simulations were made for an older
version (2002) of the cleaning system !
5
Orbit tolerances for the LHC

• With time demands for orbit stabilization have
  poped up everywhere around the LHC. A more or
  less exhaustive list
• Cleaning section IR3/7 lt ? 0.3 s 70
  mm
• TCDQ absorber in IR6 lt ? 0.5 s 200 mm
• Q-meter and transverse damper in IR4 ?
  200 mm
• Injection points IR2/IR8 200 mm
• Injection protection devices IR2/IR8 lt ? 0.5
  s ? 500 mm (?)
• Stabilization for collisions
• TOTEM experiment IR5 20 mm (!!!)
• Protection global orbit 500 mm
  rms
• e-cloud() global orbit lt1000 mm
  rms ?

7 TeV
Note the expected BPM systematic errors
intensity (pilot ? nominal bunch) ?100
mm bunch length changes (injection flat top)
?100 mm ?
Presently studied at the SPS
() not formally expressed but to be expected
from SPS experience
6
Expected orbit drifts
Phase Total drift / rms Time
scale Comment Injection
?2 mm 15 min Decay Start
ramp ?2 mm 20 sec Snapback Ramp
few mm 20 min SPS/LEP
experience Squeeze 2-20 mm few
min Depends on orbit
quality in insertions Collisions few
mm hours LEP orbit

• Some drifts (ramp, squeeze) are probably
  sufficiently reproducible to use feed-forward
  from one fill to the next for the bulk part.
• Most drifts become critical on time scales gt 1-10
  seconds.

7
Ground motion at LEP

• f ? 0.1 Hz no problem expected

Average 1s
LEP rms orbit drifts in 1998 for 390 fills,
normalized to b1 m

• f gt 0.1 Hz
• Amplitudes ? O(few mm)
• ? should be OK

? If the LHC moves like LEP we are safe.
Note The large time constants of the orbit
corrector power converter (10-200 s) and the
available voltage limit useful orbit corrections
to f ? 1 Hz (at 7 TeV) !
8
Orbit feedback overview
A global real-time orbit stabilization local
refinements is considered to satisfy all the
demands

• Sampling rate 5 25 Hz (design is 10 Hz) for
  corrections at up to 0.2-1 Hz.
• Upper limit is 50 Hz due to power converter
  controls.
• Data transmission from ? 70 front-end computers
  (1000 readings/plane) to central feedback over
  switched Gigabit Ethernet (LHC technical
  network).
• Central processing on Linux systems
  (multi-processor) with (almost) hard real-time
  capabilities. A total processing delay lt 40 ms is
  feasible.
• Fan-out of corrections to PC front-end systems
  (500 correctors/plane).
• If there are performance problems ? local loops
  in cleaning insertions !

9
b-beating

• With only 8 b-beating change tolerated, a good
  correction of the optics is required at all
  stages
• Ramp (decay snapback seem OK).
• Squeeze
• Dynamic squeeze in collisions for LHCb.
• Fancy knobies (arrrgh !!).
• There are lots of distributed sources of b-beat
• Spool-piece corrector alignement.
• Orbit in sextupoles.
• Quadrupole calibrations (nominal accuracy 2 x
  10-4 ? ?5 b-beat change during squeeze, Ok but
  near the limit).
• ? work ahead !

10
b-beating measurement

• Careful optics adjustments will have to be made
• K-modulation available (as far as I know !) in
  cleaning IRs
• - control of hysteresis effects ? At 7 TeV the
  quadrupoles will not be far from saturation.
• - good measurements require continuous / PLL Q
  measurements. Not expected to be available
  before some months after startup.
• Kicks / AC dipoles combined with multi-turn BPM
  data
• - beware of oscillation amplitude limits !
• This issue clearly deserves a closer look.

11
On the road to a nominal LHC

• Beam cleaning offers new challenges for retired
  LEP operation cowboys
• lots of tuning ahead at least as far as the
  tight tolerance allow it !
• The relatively tight orbit control in the
  cleaning sections is manageable.
• Tight optics control is much more tricky.
  Deserves further studies and lots of work on
  the beam !