Why are the Tevatron fill times so long?

[jason12345]

The Tevatron is being run with fill times of between 15-20 hours, with a beam half life of 6 hours and machine turnaround times of around 5 hours. Theoretically, the optimum fill time should be around 8 hours, so why are they using fill times so much longer?

 

[Vanadium 50]

Because it's not a very steep optimum, and sometimes things break when you try and shoot again, so there's always a risk it will take longer. If you're running, it's a good idea to keep running as long as you can.

 

[jason12345]

Because it's not a very steep optimum, and sometimes things break when you try and shoot again, so there's always a risk it will take longer. If you're running, it's a good idea to keep running as long as you can.

Have you worked on an accelerator where this has been a reason? If not, I'll send an email asking about this.

The Tevatron is working so well with these consistent long fills that it'd be interesting to find out just what the difference in integrated luminosity is between 10 and 20 hour fills.

 

[Vanadium 50]

Take a look at http://www.c-ad.bnl.gov/pac2011/proceedings/talks/moocn2_talk.pdf" , slide 18, "Optimizing the Store Duration". The red line is relevant. It peaks at 23 hours, but a 6 hour shift makes only a 10% effect.

Also look at slide 15, where it points out that almost half of the magnet quenches (57 vs. 68) occur when putting a store in. That's a good reason to avoid doing this more often than necessary.

 

[jason12345]

Take a look at http://www.c-ad.bnl.gov/pac2011/proceedings/talks/moocn2_talk.pdf" , slide 18, "Optimizing the Store Duration". The red line is relevant. It peaks at 23 hours, but a 6 hour shift makes only a 10% effect.

Also look at slide 15, where it points out that almost half of the magnet quenches (57 vs. 68) occur when putting a store in. That's a good reason to avoid doing this more often than necessary.

Thanks for the link to the slides, looks interesting. Slide 18 doesn't show the the integral luminosity for 0 - 15 hours where you can say if a 6 or a 15 hour fill will make much of a difference, which is what I'm after.

Slide 15 is for 1200 stores, 154 quenches, which means close to 1 in 20 quenches occur when putting a store in. To me, that shouldn't have much of an effect on the fill time chosen.

Alexander Valishev kindly emailed me:

Luminosity life time is not constant but increases as the beams
decay and disperse. See e.g.
http://accelconf.web.cern.ch/AccelConf/p05/PAPERS/TPAP038.PDF

For a description of the store length optimization refer to
http://accelconf.web.cern.ch/AccelConf/PAC2009/papers/mo4rac03.pdf

The antiproton production rate is a factor in deciding the fill time.