LHC portal

Why is it taking so darned long to get to the maximum intensity? With 900 bunches as currently, that's only about 25% of the max possible. Put another way, each physics run could return four times more data at max intensity. So what's the hold up?

andrewp wrote:Why is it taking so darned long to get to the maximum intensity? With 900 bunches as currently, that's only about 25% of the max possible. Put another way, each physics run could return four times more data at max intensity. So what's the hold up?

I believe they are slowly building up to full intensity. This is what they usually do after a shutdown. They incrementally increase the number of bunches they are running with, run like that for a while, then they all sit down and check whether they are happy with how its gone and that they're happy to move up to the next level of intensity.

Because of the energies involved in the machine they are (rightly) highly cautious about taking steps up in power unless they are confident everything is OK.

DCWhitworth wrote:I believe they are slowly building up to full intensity. This is what they usually do after a shutdown. They incrementally increase the number of bunches they are running with, run like that for a while, then they all sit down and check whether they are happy with how its gone and that they're happy to move up to the next level of intensity.

The machine operators want ~20 hours of stable beams per step of ~300 bunches. In between, various tests have to be done, so not all time can be spend on stable beams. And then you have weasels, fires, vacuum leaks, some broken sensor at a door, and issues like that.

After 1200 bunches, safety is not the only limit any more: electrons in the beam pipe (accelerated by protons moving by), synchrotron radiation, and a few protons that get lost directly lead to heat in the magnets of the accelerator. Their heat tolerance is limited, which limits the number of bunches. Over time, heating from the electrons (the main component) goes down, which allows to increase the number of bunches slowly. Special "scrubbing" runs with many (~2000) bunches at injection energy (=> heat load not an issue) can help to speed up that process.

Both experiments reached 10/fb today. Compared to the plan they pushed from 2 weeks behind end of may to 3 weeks in front the projection.http://lhc-commissioning.web.cern.ch/lh ... 27June.png
What can we expect for the year and run 2? Optimistic would be something like 40/fb this year. What do you think?

I added the recent progress and extrapolated a bit, including the technical stop and the large beta* days, but not the weasel incident or the POPS PS issue. MD and TS can reduce that a bit, as can a worse machine performance. On the other hand, a fixed SPS beam dump vacuum could increase the luminosity further.

Beware of wild extrapolations!

Wild? Taking the data from one month and assume the same for the rest of the year. I guess it will be not that wrong finally.

sciing wrote:Both experiments reached 10/fb today. Compared to the plan they pushed from 2 weeks behind end of may to 3 weeks in front the projection.http://lhc-commissioning.web.cern.ch/lh ... 27June.png
What can we expect for the year and run 2? Optimistic would be something like 40/fb this year. What do you think?

That's far too high IMO.

Looking at the latest schedule:
https://espace.cern.ch/be-dep/BEDepartm ... e_2016.pdf

Let's say 2fb/week over 7 weeks up to TS2 giving 14 + 11 = 25\fb delivered.
We then have another 5 weeks giving a projected total = 35\fb delievered for 2016

Of course, that's if there are no more mishaps and all the MDs go according to plan. But who would have thought that 2 weeks would go amiss right at the start, together with the SPS dump leak ?

As a guess, I'd say ~30\fb is what's now realistically projected with a risky SPS dump replacement after TS2, but perhaps we'll find out after the next LHCC meeting.

JMc

30/fb is quite pessimistic, isn't? As I said 40/fb is optimistic, so in between 35/fb is realistic;-)

They plan to reduce the emittance a bit, which could increase the luminosity further even without a fix of the SPS issue.

mfb wrote:They plan to reduce the emittance a bit, which could increase the luminosity further even without a fix of the SPS issue.

What are the disadvantages to reducing the emittance?

I remember in 2012 that the beam half life was severely reduced from 20 hours to around 8 hours as the luminosity was increased, by increasing the bunch intensities and reducing the emittance; whereas merely increasing the number of bunches has no effect on beam life time.

JMc

The luminosity lifetime might go down a bit, but for a given time in stable beams the luminosity should always be higher, so no disadvantages (higher pileup in the experiments, but that's what you get if you want more luminosity for a given number of bunch crossings). Just challenging for the preaccelerator chain.

https://lpc.web.cern.ch/lpc-minutes/2016-06-27.htm

Interesting stuff about current and future operations. One thing in particular caught my attention, "Since in the ongoing year there is no possibility to inject trains longer than 144b in the LHC it would be possible to move this AGK towards the end of the orbit, since it is currently being positioned for trains of up to 288 bunches. This possibility is currently being discussed among the experts and the management of the LHC."

This seems like they are not planning on fixing the SPS beam dump system this year.

tomey36 wrote:https://lpc.web.cern.ch/lpc-minutes/2016-06-27.htm

This seems like they are not planning on fixing the SPS beam dump system this year.

There's an interesting paper on BCMS mentioned in the link which is now being used for fills as we speak to increase the luminosity further:
http://emetral.web.cern.ch/emetral/Some ... -02-14.pdf

Cheers,

JMc

What a perfect week so far, 4 fills dumped by OP in a row. Far above 400/pb per day. Already 2.8/fb in this week. So plan is now 24h fills with >500/pb like this week. So a perfect week could now achieve 3-3.5/fb.
Amazing:-)
I am an optimist we will see 20/fb before MD1 (in 17 days)
My old guess of 15/fb recorded (with last year data) before july 15th is already achieved today, 1 week earlier.