Page 14 of 35

Re: Current Events Discussion

Posted: Fri May 14, 2010 10:42 pm
by jmayes
So on the last record breaking run there was a notch in the 12hr chart, anyone know what happened here?

Re: Current Events Discussion

Posted: Sat May 15, 2010 12:49 am
by LarryS
According to the LHC OP Log the "FBCT got stuck"; that is, the Fast Beam Current Monitor stopped sending data and needed to be rebooted.

Re: Current Events Discussion

Posted: Sat May 15, 2010 8:54 am
by PhilG
They have just succeeded in ramping "nominal bunches" to 3.5 TeV. This means bunches with 100 billion protons which is 5 times the number used per bunch in the latest physics runs.

This is exciting because luminosity is increased by the square of this number. In fact if they used nominal bunches on a physics run right now they could have a new luminosity record even without the squeezing and multiple bunches.

On this occassion they lost a large part of the beams during ramping but it was still a good step. It seems that using such high intensities can lead to problems with instabilities and beam spreading, but they nearly have it cracked.

Looking forward to seeing physics with squeezed nominal bunches, hope that is not too far off now.

Re: Current Events Discussion

Posted: Sat May 15, 2010 12:12 pm
by Bornerdogge
6 bunches/beam and ramping! According to what PhilG said, we should see a nice luminosity increase in this fill!

Re: Current Events Discussion

Posted: Sat May 15, 2010 1:03 pm
by PhilG
Six bunches (plus pilot), should give a 50% luminosity increase on previous physics run.

If they could do this with nominal bunch intensity it would be about another 25 times as much, but that's for later.

Re: Current Events Discussion

Posted: Sat May 15, 2010 7:09 pm
by Chuli
As i'm seeing on the LHC1 page, they made a new luminosity record :)

Re: Current Events Discussion

Posted: Sat May 15, 2010 7:21 pm
by PhilG
Luminosity now is about twice previous record

Re: Current Events Discussion

Posted: Sat May 15, 2010 7:36 pm
by godzila
Just wondering - how far are we to get "usable" luminosity (ie. a meaningful amount of collisions) ?

Re: Current Events Discussion

Posted: Sat May 15, 2010 7:45 pm
by PhilG
It's always possible that something new could be discovered at low luminosity because the energy is higher than previous accelerators.

More likely they need to get the luminosity above the levels at the Tevatron which is currently running with 6000 times more luminosity than the latest runs at the LHC. With good luck it may not take long for the LHC to surpasses that.

That's assuming you mean "usable" in the sense that new physics can be discovered. Current luminosity is usable for calibration and rediscovering old physics.

Re: Current Events Discussion

Posted: Sat May 15, 2010 7:46 pm
by Tau
And still, they could have more hits per rotation per experiment instead of three, so I an sure there must be more to these bunch numbers, otherwise I don't get it. Probably it has to do with bunch scheduling in SPS. This could also explain why they did four bunches recently.
As you can see in http://blog.vixra.org/2010/04/24/lhc-ac ... zed-beams/ it is possible to get two hits per rotation per experiment, so it should be possible to get four with six bunches, by combining this pattern twice (with say 2000 spacing for the insertion kicker):
b1: 1, 2001, 8941, 10941, 17851, 19851
b2: 1, 2001, 8911, 10911, 17851, 19851
(This pattern also has every bunch involved in two collisions per round: this is not the case for the actual pattern used, where some bunches have one and others have two.)
We'll see how things develop in the future.

Later addition: My dream pattern with six bunches would be:
b1: 1,61,8911,8941,17851,26791
b2: 1,61,8911,8941,17851,26761
This horribly complicated pattern has two practical problems because of the small distance between some bunches:
- crossing angle needed to prevent "parasitic" collisions
- SPS needs to produce pairs of bunches with both 30*2.5 and 60*2.5 ns spacing
But, it would give 5 collisions per rotation for ATLAS and CMS, and 4 for LHCb and ALICE (and the number of collsions varies from 1 to 3 over the bunches).

Re: Current Events Discussion

Posted: Sat May 15, 2010 7:57 pm
by PhilG
I think one advantage of the current filling pattern is that it does not produce any displaced collisions where the bunches would meet near an experiment and not at exactly the right collision point. The filling patterns that give more collisions tend to have displaced collisions too and maybe those are a nuisance at this stage.

Re: Current Events Discussion

Posted: Sat May 15, 2010 10:00 pm
by Bornerdogge
Interesting, although they run with doubled event rate, CMS stores 5 times less events than before (10% now, 50% last weekend)... "Better" events I guess?

Parasitic collisions

Posted: Mon May 17, 2010 6:28 am
by Tau
PhilG: I now wrote a program to check the bunch patterns, and I see that parasitic collisions indeed are a problem. Here are the results on the patterns we saw in the LHC until now:
- Two bunch pattern: 1 collision, no parasitics
- Three bunch pattern: 2 collisions, with parasitic
- Four bunch pattern: 2 collisions, with parasitic
- Six bunch pattern: 2 collisions, no parasitics
- My six bunch: 3 collisions, with parasitic.
("Collision" means per rotation per experiment; "with parasitic" means there are collisions 30 buckets or 11 m from the experiment)
This would conclude that even a parasitic collision 11 m from the experiment is significant...
(edited because I said 22 m first, ignoring the fact that both beams move together).

Re: Current Events Discussion

Posted: Mon May 17, 2010 8:11 am
by Kasuha
Today's status report, the luminosity graph and the "can we extrapolate?" remark made me laugh... :lol: Too bad we cannot really extrapolate it the "obvious" way, there's only so much that can be added to luminosity. But the progress was great anyway, good job!

Re: Parasitic collisions

Posted: Mon May 17, 2010 1:46 pm
by jmayes
Tau wrote:PhilG: This would conclude that even a parasitic collision 11 m from the experiment is significant...
(edited because I said 22 m first, ignoring the fact that both beams move together).
It was my understanding that while the beams become very close (almost parallel to each other) they don't actually cross until they reach nearly the very center of the experiment. I also understand there are steering magnets at both sides of most of the experiments that can move the actual crossing point. If this is the case then the parasitic collisions will not take place.