CERN Accelerating science

Squeeze

Essentially driven by pre-loaded functions to power converters.

• Predefine squeeze table (see below) which defines temporal evolution of squeeze in terms of optics. This would respect pre-established current change limits in the quadrupoles concerned. To be established. See Oliver's Cham 2005 paper for initial estimates.
• Establish top energy un-squeezed conditions. For completeness during power converter tests, using nominal LHC ramp could be illustrative. Although it would take 25 minutes. Bear in mind I end ramp has to match I squeeze initial to a good degree of accuracy.
• Input IP, beta* final
• Generate squeeze functions for relevent elements with appropriate parabolic/exponential round in, parabolic round out.
• Load functions to power converters. For tests need not be all converters in a given IR.
• Initiate squeeze with timing event (or FGC ramp now capabilities).

Max & Min beta

Squeeze Table - illustration

LSA settings generation module, as implied above, has to generate functions from/to any given beta for any given IP or IPs - settings labeled by context (beam process). BP name generated on fly. Beam process type would probably exist already. Details to be established. Besides current you also get Kn(t) and beta(t).

• Mad optics files pre-generated and loaded to LSA database. Includes, of course, nominal strengths for all magnetic elements. Note unless we get really clever we take the whole ring even for a squeeze in one point only.
• Squeezing more than one point at a time (generally IR 1 & 5) will presumably enforce equal durations.

Current limits with negative currents steps

• For each optics step, for each circuit, for each magnet (b1,b2) we calculate the initial current, final current and delta current.
• If the delta is negative
  • get nominal max di/dt from database
  • evaluate max di/dt due to coupling
  • evaluate natural decay time and di/dt at end of natural decay time

Use the minimum maxiumum di/dt to evaluate step time (linear current change). Note that we put in a parabolic round in and round off at each optics step.

Issues

What changes? What elements are involved? Everything else stays put.

Speed; consideration of the limits imposed by the quadrupoles power converter di/dt

Another issue: we are running out of volts on the RQX Inner Triplet if we use the max di_dt that has been requested. This happens during the first squeeze transition. Ideally we want to keep below 8V. This would impose a di/dt of ~4 A/s

Number of steps? During commissioning we step through...
Combined ramp and squeeze - forget it for the moment
One IP at a time (in which case we will have to worry about the effect of the X-ing angle and the effect on long range interactions)

Orbit

Going to have to have to maintain:

• horizontal and vertical crossing angles.
• separation bumps

Feedback:

1. Prepare matrices before and as close to squeeze as possible. -> controller  [take into account BPM dropouts, have to recalculate]
2. FB -> OK
3. Start squeeze. Optics changes to GOFB.
4. End Squeeze

Tune feedback

Corrections

• Higher order multipoles in inner triplets

Collimators

As elucidated, primary and secondary collimators to follow squeeze. Really going to have to worry about beta beating. Perhaps to 8 sigma at first, stop in squeeze, place at 7 sigma, check before continuing.

Hysteresis effects in the quadrupoles.

Squeezing more during collisions in LHCb

Worry about: x-angle change with beta function, will need to be monitored and adjusted during the squeeze to keep the beam in collision.

RF

Beam Dump

TCDQ has to track. When does it move?

Measurements

Any of the below dynamic?

• coupling
• dispersion
• beta beating