Fermilab Background Issues
- Collisions pbar-p clearly dominate beam gas interactions unlike say HERA or LEP.
- Beam-gas five orders of magnitude lower. So vacuum is not an issue for detector trigger rates, current draws, aging, etc.
- Locally bad vacuum can be an issue. (Bad is 10^-7 Torr).
- Beam losses (some causes outlined below) shower into detectors causing electronics/power supply drop out, and in the worst cases damage to silicon.
Can divide concerns up:
- acute radiation
- chronic radiation - 99% from p-pbar
- background
MACHINE
First observation is that machine's main concern is beam loss rather than background per se.
Abort Gap
This systematically fills every shot. Why is still not clear: RF noise is a possible candidate (although it is noted with the use of Tetrodes rather than klystrons things might be expected to be quieter).
Cleaning the abort gap is possible. It is monitored using sync. light by the accelerator. The loss rate from the abort gap is also monitored by CDF. Beam in the abort gap makes the experiments very nervous because they get hit by some of the contents on an abort. Note that it is the small amplitude kick on the rising edge that pushes beam into experiments.
Some debate between machine and CDF about measuring losses from abort gap rather than contents.
Halo
- Each experiment has different problems. CDF don't worry about halo so much any more after addition of extra shielding. D0 still suffer.
- Prompt losses - here I assume protons lost upstream of experiments causing showers
- Use of nearby (tertiary in LHC parlance) collimators is not obvious. Pushing them in produces background. Presumably they bite into the tertiary halo and generate showers. Balance between protection and cleanliness to be anticipated.
Halo character
Not totally clear what is populating halo. Beam is breathing: magnet vibrations, quad movements, insertion quad movements with temperature variations "breathing the beam against the collimators".
Anything slow, such as beam-beam, can be mopped up with collimators.
Spikes look common. Fast diagnostics (60 Hz fast time plots) are crucial.
Orbit
Definite orbit control short comings. Clear that properly optimized orbit (position and angle) though the detector is important. Things like:
- big angle through D0 - eating aperture. Beam not centred in detector. Ideally centre with zero angle.
- D0 (and CDF_ have sunk. Importance of detector-accelerator alignment. Quad alignment
- RHIC had similar problems - check out talk by Angelika Drees at Fermilab
Beam loss
Worry about radiation field in tracking volume, in collision hall - concerns include radiation damage to silicon, and effect of radiation field on electronics. Note that the experiments didn't properly vet for radiation hardness. Typical SEU effects: lock-up necessitating power cycles.
Worries include:
- Losses during the injection cycle
- Losses during squeeze. Because of helix manipulations, beam separation drops to 3 sigma - beam-beam causes beam blow-up, multi-turn effect and beam losses in IR (D0 at the moment). NB: collimators are not in during ramp and squeeze. Aperture limit is separators either side of IP.
- Sudden beam loss incidents over a very short time span - leading to killing silicon. Such events include: pre-fires, messy aborts (NB Tev aborts on quenches not losses)
Vacuum
Defintely an issue. Locally bad vacuum clearly can cause problems. E.g. F sector graphite and halo induced heating.
EXPERIMENTS
Instrumentation
Clear that experiments have done well to
- Protect themselves
- Provide additional diagnostics
- CDF have fast scintillators in front of triplets which give data on the abort gap, the halo and proton losses. Very, very useful. What particles? Where are they coming from?
- Dedicated BLMs feeding to abort channel
- Diamond BCMs using new BLM electronics feeding abort loop (1 legitimate dump since installation).
Other measures include appropriate interlocks (if roman pots not where they should be - switch off silicon)
Feed machine with a lot of data including transverse beam position. RT calulation of shape of luminous region - including hour glass.
Culture
Very important to build up relationship with IPs. Cross-domain problem solving. (Note existence of fixed target culture at Fermilab before the Tevatron.)
Monitoring
CDF pulls in a lot of stuff. BLMs, RF (klystrons, modulators, abort gap, vacuum, separators). Active participation in accelerator monitoring. Try to avoid knock-on in sudden MP events.
Seismometers. Quads - traffic vibrations.
CDF
- Shielding
- muons - hard brem - into EM calorimeter - 30- 80% of triggers. Low angle elastic scattering from gas - bad vacuum - adjacent roman pots acting as collimators. Gap in shielding.
- Scattering from collimators
- Cleaning up vacuum had a huge effect.
- Improved control of accelerator beam parameters (coupling etc.) has had a big effect. (BGs an order of magnitude lower - although not sure why).
- St. Catherine's day massacre: sagging bellows plus bad orbit - showering into ff hadron cal - glowed like the sun - lots of dead electronics.
- A few years into the run a collimator was added before the low betas on the proton (higher intensity) side of CDF. It
was beefed up a year later. This definitely helped in cases of kicker pre-fires or messy aborts, though I think there were still some damaging
incidents after it was put in There is a paper at
http://beamdocs.fnal.gov/AD-public/DocDB/ShowDocument?docid=2564
with more information on this subject.
LINKS
c/o Jim Patrick...
Some talks by Rick Tesarek on losses and monitoring at CDF are:
Latest version: http://ncdf67.fnal.gov/~tesarek/halo/BeamMonitors.pdf
Earlier versions given at various LHC radiation monitoring workshops:
http://lhc-expt-radmon.web.cern.ch/lhc-expt-radmon/meetings/2004-01/presentations_2004_01.htm
http://indico.cern.ch/conferenceDisplay.py?confId=a044378 (near bottom)
http://indico.cern.ch/conferenceDisplay.py?confId=a044378 (towards bottom)
Links from Rick Tesarek
Notes - Mike Lamont December 2007. Thanks to Jerry Annala, Dean Still, Jim Patrick, Rick Tesarek.