SuperKEKB/Belle II start up: First collisions

[mfb]

This year we have two big particle physics accelerators running. SuperKEKB is an upgrade of KEKB, and the Belle II experiment at the accelerator is an upgrade of - you guessed it - Belle. It is located at KEK*, close to Tokyo in Japan, and it is starting up for the first time now.

SuperKEKB is a B-factory, an accelerator designed to produce a large number of hadrons with a b-quark. It collides electrons and positrons at a few GeV energy, just right to produce pairs of b-mesons with nothing else in the event. The concept is the same as with KEKB/Belle, but this time the collision rate and the total number of collisions over the lifetime of the experiment will be a factor 50 larger and the experiment can measure them better than before.

A factor 50 in statistics means all the previous analyses can be improved significantly, and many new analyses can be done that were not useful previously. In particular, if the anomalies seen in B -> s µµ are real instead of statistical fluctuations, Belle II will clearly measure them.

First beams were injected in the electron ring two weeks ago, now the beam can circulate there for hours. The positron ring had first injections a few days ago, but the accelerator experts are still working on making the beam circulate for a longer time. You can see the status live online:
Last 24 hours
Last 2 hours

First collisions might happen in April.

* KEK = "Kō Enerugī Kasokuki Kenkyū Kikō", high energy accelerator research lab

Related: LHC starts up in 2018

 

[dukwon]

Some slides about the schedule and luminosity goals:

Due to the climate in Japan, KEK has a summer shutdown (rather than winter), which will begin in the middle of July this year. It's usually earlier.

 

[mfb]

The beam is now routinely circulating in both rings. Currently the beam can be kept for several hours. With the conditions needed for collisions this will drop significantly, and the accelerator will frequently need more electrons and positrons injected into the rings to keep the collision rate high. First collisions could happen quite soon.

This is basically a completely new accelerator and large parts of the detector are new as well. Not everything will work as expected initially, but data-taking should become smoother over time.

 

[mfb]

The last weeks the beam optics have been studied and the accelerator experts are confident they can go to collisions this week or the next week.

There will be a live stream starting Friday noon and running for up to one week. Note: This is Thursday evening in the US and the night to Friday in Europe.
KEK news
Direct link to stream

 

[dukwon]

From that press release:

the world's most-powerful accelerator

Hmmmm, by what metric? Luminosity?

 

[George Jones]

Hmmmm, by what metric? Luminosity?

Here is a naive guess, and since I know very little about this, this guess probably is easily shot down.

[(collision rate) times (energy per collision)] has units of power, so if "most powerful" is taken literally, the accelerator for which this quantity is largest is the most powerful accelerator.

 

[dukwon]

The SuperKEKB max bunch crossing rate is about 6 times higher than the LHC, but the LHC has many more collisions per crossing and much much more energetic collisions.

 

[Vanadium 50]

Hmmmm, by what metric? Luminosity

I suspect that this was originally written in Japanese and translated into English. Someone who knows Japanese would need to read the original.

 

[mfb]

Hmmmm, by what metric? Luminosity?

That headline is a bit strange. It will reach the highest luminosity at some point but it won't get any other obvious record.

[(collision rate) times (energy per collision)] has units of power, so if "most powerful" is taken literally, the accelerator for which this quantity is largest is the most powerful accelerator.

About 3 kW in ATLAS and CMS at the LHC, tens of microwatts for SuperKEKB.

Here is the original Japanese headline: 生放送・第二弾《世界最強加速器SuperKEKB》電子と陽電子の初衝突を見守ろう！

 

[fresh_42]

Google makes it: "World's Strongest Accelerator SuperKEKB"

 

[dukwon]

The "status" reported on the monitoring plots has been "collision study" for about 12 hours now, but the luminosity graph looks empty. Have they actually had any collisions yet?

 

[mfb]

Now it is "Linac study".
No collisions yet.

I don't understand the Japanese chat, my translation tool doesn't understand it either, but it looks like the discussion is all over the place about black holes and whatever.

 

[mfb]

Collisions!
KEK news

Now the target will be to understand the detector better and to increase the collision rate in the next months. Belle (the previous experiment) collected a large dataset already, it will take a while to surpass this.

Event display from the image archive there:

 

[dukwon]

Belle (the previous experiment) collected a large dataset already, it will take a while to surpass this.

Won't take that long. Belle had 0.7 ab⁻¹ at the ϒ(4S). They aim to have at least that by the end of 2019.

 

[mfb]

I consider 1.5 years "a while".

When the LHC accelerated protons to 3.5 TeV in 2010 the first collisions were immediately very interesting as no previous accelerator could reach this energy. This is different here. There are some studies that can be done with low luminosity but high trigger rate (mainly looking for somewhat frequent processes that didn't have high priority triggers in Belle / BaBar - dark photons for example), but most analyses will need a dataset that is (at least) similar to Belle. This won't happen in 2018. In addition, a new machine and detector means more unforeseen issues - in 2018 more of the time will be spent on testing and debugging compared to the LHC.

 

[mfb]

The accelerator is alternating between scrubbing (reducing the residual gas in the beam pipe), machine studies to improve the performance and luminosity runs (collisions). The highest luminosity I saw was 4*10³²/(cm²*s), about 2% of what KEKB achieved, and about 0.04% of the goal. As comparison: The LHC needed 4 months to reach 0.04% of its design luminosity in 2010.
With this low collision rate the focus is on improving the machine performance and understanding the detector better to prepare everything for later when the rate is higher.

 

[mfb]

They now routinely reach a bit more than 10³³/(cm²s), 5% of what KEKB achieved. The improvement came from slightly higher beam currents and better focusing of the beam. Both will be improved more to increase the luminosity further.
I didn't see a number for the integrated luminosity but it should be somewhere in the tens of 1/pb or 0.0x/fb. Not enough to improve previous measurements, but enough for a lot of calibration work. The relative orientation of all the detector components has to be determined as accurately as possible, the efficiency and resolution of everything has to be measured, ... all these things have simulations to estimate them but the simulations are never perfect.

 

[mfb]

Forgot to update this thread.

Phase 2 ended in July. The peak luminosity achieved was 5*10³³/(cm²s), about 1/4 of the KEKB record. Lower than the original target of 10³⁴/(cm²s) due to various challenges that appeared with the new machine. SuperKEKB uses a so-called "nano-beam scheme": The beam is collimated to an extremely narrow line. At the interaction point the beams will be just 50 nm in the vertical direction and about 10 μm in the horizontal direction. As comparison: the bunches are a few millimeters long. Slide 10 in this presentation has an illustration. This hasn't been achieved yet - I didn't find values but as far as I understand the vertical size was somewhere at 1 μm already, smaller than any beam before.

Currently the inner detector gets installed, in early 2019 phase 3 will begin with the full detector and increasing luminosity.