Combined ramp and squeeze
Procedural overview of squeeze
Observations
- For a combined ramp and squeeze, the final betas of the squeeze in the ramp (IR1 and IR5, say) must be pre-established.
- Single quadrant power converters drive the insertion quads. At constant energy the current in some insertion quads decrease during the squeeze. This decrease must take place within the natural time constants of the circuits concerned. Tests show that this shouldn't be a problem for power converter regulation. These considerations limit the speed at which the squeeze can be performed.
- Will be starting at 7 TeV unsqueezed - natural starting point for progress through squeeze.
Squeeze in ramp - arguments for
- The main advantage is that all insertion quadrupoles currents will be increasing, negating the need to deal with decreasing currents in magnets with single quadrant power converters.This in itself will mean that the squeeze can be performed more quickly.
- By performing a simultaneous ramp and squeeze we clearly save time by performing two operations at once. Depending on the final betas and preparation time etc. this saving could be of the order of 20 minutes or so.
- Does is make sense to do what is possibly the trickiest operational phase at top energy where the protons that we are bound to lose are at their most damaging? Possible argument for stopping in the ramp, squeezing and then continuing the ramp - of course, again, one has to pre-establish where one wants to go to in terms of beta*s. Bear in mind also, we will be performing the squeeze commissioning with single pilot++
Squeeze in ramp - arguments against
- During commissioning, it is planned to commissioning the squeeze IR by IR, and by stepping through from beta = 11m to beta = 2 m. Clearly to deal with the required flexibility a combined ramp and squeeze is precluded. Clearly related to rule number one of commissioning rule: don't try and do two things at once.
- Instrumentation: critical here is the PLL, again simple argument of trying to do two beams and the ramp and the squeeze suggests factoring out.
- Collimators: the emittance is shrinking during the ramp - this has to be tracked by the collimators. Obviously the collimators also have to track the decreasing aperture in the insertion region(s) in the squeeze. To attempt both before completely mastery of collimator adjustment during the squeeze would be foolhardy.
- TCDQ: similar arguments to collimators.
- Orbit feedback: the orbit feedback system needs to track any change in optics and energy. Although not impossible, it is clearly more complicated to track both at the same time and initially at least it makes sense to factor out the ramp and squeeze.
- Machine Protection: Essential establish good control of collimators, TCDQ, energy tracking of beam dump in ramp, BLM thresholds, etc. - trying to squeeze at same time - potential source of confusion.
- Crossing/Separation bumps: the correctors generating the bumps are bi-polar and there should not be any regulation issues. However, the separation bumps come down in the ramp (limited corrector strength). The crossing angle bumps and separation correctors move with the squeeze. Although in principle orthogonal it would probably be wise to deal with one issue at a time.
Recommendations
- Software functionality should be in place to support the combined ramp and squeeze. This is includes standard settings generation (power converters, collimators) and the use of feedback systems. Further the ability to ramp to a given energy, perform a partial squeeze at that energy and then continue the ramp should be provided.
- Phase A: given the variability in the path to a given beta* and the variability of the final beta* and the consideration of dealing with one challenge at a time, commissioning of the ramp and the squeeze in stage A should be kept separate.
- Phase B: keep ramp and squeeze seperate. Final betas probably still to be establsihed. Check crossing and separation bump plus collimators/aperture in squeeze.
- Phase C: machine development on combined ramp & squeeze.