The Hitchhikers’ Guide to Beam-Beam Based Optimization of ...

Beam-Beam Based Optimization

of the LHC Performance:

A view towards Run-III

S. Fartoukh, N. Karastathis, Y. Papaphilippou

with many thanks to

LCR3 Working Group

Beam-Beam & Luminosity Studies WG

07.05.2019

The Hitchhikers’ Guide to

Large Hadron Collider

• LHC: p-p, I-I collider

07/06/2019 - BBLS N. Karastathis 2

• 2 rings (... to revolve them all)

• 8 arcs (... to bind them)

• The beams share the same region at

the 4 experimental Insertion regions

(IR) and cross at a single point in

space namely the Interaction Point (IP)

(... to smash them all)

• 2 High-Luminosity

(...and the dark matter find, man)

• 2 Low-Luminosity

Our clients’ needs

• The detectors at the experimental insertions are looking for the particle debris generated by the p-p collisions.

07/06/2019 - BBLS N. Karastathis 3

An ATLAS event display from a H WW* eνμν event [1]

Luminosity

07/06/2019 - BBLS N. Karastathis 4

• The instantaneous rate of event production at an IP is linearly correlated to the

delivered luminosity as 𝒅𝑵

𝒅𝒕= 𝝈𝒑𝒑 ⋅ 𝑳

where 𝜎𝑝𝑝 is the cross-section of proton-proton interactions is a characteristic of each

process and depends on the center-of-mass energy 𝑠.

• The luminosity, is defined as the product

of the incident particle (A) flux and the

target (B) area density:

𝐿 = Φ𝐴 ⋅ 𝜌𝐵

• In our case is overlap in space and time of

the colliding bunches [2] normalized by the

the beam parameters

𝐿 = 2𝑐𝑁𝑏1𝑁𝑏2𝑛𝑏𝑓𝑟𝑒𝑣 cos2𝜙

2න

−∞

+∞

න

−∞

+∞

න

−∞

+∞

න

−∞

+∞

𝜌 1 𝑥 𝜌 1 𝑦 𝜌 1 𝑧 + 𝑐𝑡 𝜌 2 𝑥 𝜌 2 𝑦 𝜌 2 𝑧 − 𝑐𝑡 𝑑𝑥𝑑𝑦𝑑𝑧𝑑𝑡

Beam 1Beam 2

𝜎𝑥(1) 𝜎𝑥

(2)𝜎𝑧(2)

IP

φ

Particle Physics Break

Luminosity (II)

07/06/2019 - BBLS N. Karastathis 5

𝑳 =𝑵𝒃𝟏𝑵𝒃𝟐𝒏𝒃𝒇𝒓𝒆𝒗

𝟒𝝅𝝈𝒙𝝈𝒚⋅ 𝑾 ⋅ 𝒆

𝑩𝟐

𝑨 ⋅ 𝑺

• So I could increase my delivered luminosity by simply increasing the injected

bunch intensity, but what can I do if the injection parameters are fixed?

𝑊 = 𝑒1

4𝜎𝑥2⋅𝑑

2

, 𝐴 =sin2

𝜙

2

𝜎𝑥2 ⋅

cos2𝜙

2

𝜎𝑧2 ,

𝐵 =𝑑⋅𝑠𝑖𝑛

𝜙

2

2𝜎𝑥2 , 𝑆 =

1

1+𝜎𝑧𝜎𝑥⋅𝜙

2

2

Beam 1 Beam 2IP

φ’<φ

φ

Beam 1 Beam 2IP

IPBeam 1

Beam 2

For Gaussian profiles

Some tools:

y2 [4] | C [5] | available too

Simulation of an LHC Fill

07/06/2019 - BBLS N. Karastathis 6

μ =𝐿𝜎𝑖𝑛𝑒𝑙𝑛𝑏𝑓𝑟𝑒𝑣

Beam Parameters

We could also work in reverse:

Modify the machine parameters to keep

the luminosity constant “Luminosity Leveling” [3]

Leveling by offsetting the beams is used continuously in the LHCb/ALICE

experiments and in the ATLAS/CMS when needed.

It is an easy and efficient method of keeping the luminosity under control.

Turnaround time [h]

Beam-Beam simulations

07/06/2019 - BBLS N. Karastathis 7

[6]

• A close collaboration of ABP - OP group resulted in a continuous optimization of

the LHC, based on beam-beam DA simulations.

• For example, we saw that improving the working point, we improve DA thus

lifetime!

• Without the WP optimization we would have to increase the crossing-angle to reduced the

effect of the beam-beam (𝜉𝑏𝑏 ∝ 1/𝑑𝑏𝑏2 ) Luminosity

• Even if the DA is ok-ish, optimizing it we could spend additional margin on reducing the

crossing angle event further, at the same initial bunch intensity Luminosity

Beam-Beam simulations (II)

07/06/2019 - BBLS N. Karastathis 8

[6]

• Why optimize only the initial crossing

angle?

• At the regime of the optimized WP the

correlation of the crossing angle and the

bunch intensity in terms of DA is almost

linear for the targeted amplitudes (5,6σ)

• Αs the intensity decays, the margin of DA

increases in the case the half-crossing

angle is fixed.

• A further reduction of the crossing angle

within the fill could slightly buy us some

additional luminosity

• Limiting ourselves in maintaining throughout the fill the initial iso-DA line

would keep the lifetime always above a certain threshold.

• Studying the evolution of beam parameters and correlate with the simulated DA

one could provide some time instances along an operational fill that the crossing

angle would be safe to be reduced by a certain amount. Anti-leveling

LHC 2017

07/06/2019 - BBLS N. Karastathis 9

+100pb/day

Looking a bit further: HL-LHC

07/06/2019 - BBLS N. Karastathis 10

[7]

• The HL-LHC is

planned to be

operated in

leveling mode for

most of the fill

length.

• The leveling

method considered

is adapting the β*

at the ATLAS/CMS

experiments

(ALICE/LHCb

remain in offset

leveling).

LHC 2018 – One step further

07/06/2019 - BBLS N. Karastathis 11

• So in 2018 several development of controls and improving our understanding of

the machine based on our simulations

• The evolution of the

crossing angle as a

function of the bunch

intensity was

automatized.

• Steps in further

squeezing the β* have

been implemented at the

end of each fill.

Run-III – LIU Commissioning

07/06/2019 - BBLS N. Karastathis 12

[8]

• LIU will slowly ramp up the injected bunch intensity throughout the Run-III.

• Certain limitations restrict the

injected bunch intensity into the

LHC and the maximum luminosity

accepted by the two high

luminosity experiments.

So what is Run-III?

07/06/2019 - BBLS N. Karastathis 13

Fact #1: Run-III is a preparatory HL-LHC run period.

Fact #2: LIU is completing its commissioning, will continuously

push higher intensity into the LHC.

Fact #3: The experiments are still there!

• Back to DA simulations:

After optimizing the working point define the function of 𝝈𝒃𝒃𝒏𝒐𝒓𝒎 𝑵𝒃

Force at least 9.43σ 1.17e11 ppb, as in the present LHC

We need to prepare the

mechanics towards HL-

LHC, but not to

compromise the

integrated performance

• Since HL-LHC is based around β* leveling expand 2018 to a full leveling.

• We already have the evolution of the crossing angle as a function of the bunch intensity

implemented. Could we combine the two to gain an additional knob?

Run-III – Adapting β* & X

07/06/2019 - BBLS N. Karastathis 14

These intensities

are not reachable

• Based on the DA simulations and the scaling law acquired for the normalized crossing angle define scaling

law of 𝐍 𝜷∗ under the constraint of 𝐋 = 𝟐 × 𝟏𝟎𝟑𝟖 𝑯𝒛/𝒎𝟐.

• Keep in mind that there are some additional unknowns:

• How well do we preserve emittance from injection to collisions? Take the extremes 1.8μm & 2.5μm

• What is the impact of non-simulated diffusive mechanisms (e-cloud)? Leave additional DA margin & e-

cloud can be mitigated by smaller number of bunches Study for 2736b & 2484b

• The commissioned optics should

be such that they accommodate

the large dynamic range of

initial β*

• Leave additional margin in the start of the leveling when the bunch intensity is high.

• Overall, one can define the

𝝈𝒃𝒃 𝑵𝒃 𝜷∗ ȁ𝑳=𝟐×𝟏𝟎𝟑𝟒𝑯𝒛/𝒄𝒎𝟐 and

therefore the evolution of the crossing

angle.

Run-III – Evolution along the fill

07/06/2019 - BBLS N. Karastathis 15

• However, the linear fits are only valid at the optimal WP setting Test the evolution of the WP for the

various commissioned optics under the constraint of the leveled luminosity.

• (.313, .318) looks

optimal in all

configurations.

nb = 2736

nb = 2484• The results suggest a

DA above the 5σ target

for most of the path

• Fixing the WP one can

scan the β* as a

function of the

intensity to find the

leveling path 𝑁𝑏 𝛽 ȁ𝜑(𝛽)for a certain luminosity

target.

Run-III – Fill Profile

07/06/2019 - BBLS N. Karastathis 16

[9, 10]

9.7h @ 𝟏. 𝟑𝟒 ⋅ 𝟏𝟎𝟏𝟏𝐩𝐩𝐛

13.4h

1.2𝟔𝐟𝐛−𝟏

52.6 evt

Run-III – Our Crystal Ball

07/06/2019 - BBLS N. Karastathis 17

(**) Telescopic Optics & CRATS

07/06/2019 - BBLS N. Karastathis 18

• The ATS optics offers the flexibility to reduced the β* while controlling the chromatic aberrations

induced the usual β* squeezing at an IP stops at a certain value (𝛽𝑝𝑟𝑒∗ ) and then the magnets in the

neighboring IRs are used to further reduce the β* at the target IP

• A characteristic of the telescopic squeeze is the tele-index defined as 𝑟 =𝛽𝑝𝑟𝑒∗

𝛽𝑡𝑎𝑟𝑔𝑒𝑡∗

[9,10,11]

(**) = in layman’s terms

• Interestingly enough, by

applying a tele-index, the

variation of the β-beating

wave in the arcs,

modifies the effectiveness

of the octupoles by

enhancing the provided

detuning by ∝1

4(𝑟2 +

1

𝑟2)

550A (with safety

factor of 2)

• This affects the coherent

stability, mainly during the

RAMP, allowing for

additional Landau damping

some telescope is

needed in the RAMP

already successful MDs

during 2018

• In addition the FP experiments prefer the

squeezing of the β* during collisions in

telescopic mode as the transfer matric is

constant.

• The only possible solution of

combining telescope in the RAMP and

collision in telescopic mode, within the

limits of matchability would be to start

colliding from an anti-telescope

i.e. 𝑟 =1

1.8

S. Fartoukh

Beam

-Beam

(**) Run-III – Flat Optics & Triplet Lifetime

07/06/2019 - BBLS N. Karastathis 19

[10, 12]

• So far we were considering colliding in

ATLAS/CMS with 𝛽𝑋𝐼𝑃1 = 𝛽𝑌

𝐼𝑃1 = 𝛽𝑋𝐼𝑃5 =

𝛽𝑌𝐼𝑃5 (round optics)

Slightly better performance (hindered by the leveled

luminosity threshold), but…

• Flat Optics assumes asymmetric β*

aspect ratio 𝑟∗ =𝛽𝑋∗

𝛽ȁȁ∗ ≠ 1, with 𝛽𝑋

∗𝛽ȁȁ∗ =

𝛽𝑟𝑜𝑢𝑛𝑑∗

• It is an alternative to HL-LHC

operation without crab cavities as

increasing the β* aspect ratio rapidly

mitigates the geometric luminosity

loss.

Improvement on the irradiation of the triplet

magnets

Reducing the crossing angle has also a beneficial

effect

the parametric variation of the crossing angle

with the β* keeps the crossing angle as low as

possible for as long as possible!

Some things yet to solve...

07/06/2019 - BBLS N. Karastathis 20

• There is work yet to be done both from our side (optics production and further tests) but also from the side of the experiments.

• Some on-going discussions include

[10, 13, 14]

ATLAS, V CMS, HThe variation of the crossing angle and β*

significantly shrinks the longitudinal

luminous region, thus increasing the pileup

density.

2018 BCMS 2017 8b4e

The bunch-by-bunch luminosity variation

along the train might impact the leveling

(on luminosity? on pileup? Maximum or

average?)

5x48 bpi, 2736 collisions in ATLAS/CMS, 2250/2376 in Alice/LHCb

Baseline: Pure BCMS

Backup: Mixed BCMS with 8b+4e (56b) inserts

What is the impact on using mixed filling

schemes?

G. Iadarola

I. Efthymiopoulos

Summary

07/06/2019 - BBLS N. Karastathis 21

• The estimate of the performance of a collider is given by its luminosity, which depends both on the

beam and machine parameters.

• For fixed injected beam parameters, modifying the crossing-angle, the β* and the parallel

separation changes the delivered luminosity at the IPs luminosity constant (leveling)

• Beam-beam DA simulations were used extensively to identify the optimal path to modifying the

machine parameters along the fill, and after it was proved safe, the technique was pushed into

nominal operation during Run-II.

• In the future, during Run-III the LIU will commission the upgrades and will gradually increase the

delivered bunch intensity LHC must adapt, and prepare for the HL-LHC era without hindering

the its integrated performance.

• Building upon the knowledge and tools developed already, Run-III operation could be:

• Injection : the same

• Combined Ramp and Anti-Telescopic Squeeze

• Collide & Squeeze : in telescopic mode β* && xing in IP1/5, offset in IP2/8

• A plan for the full cycle of the Run-III is in place and the optics are being finalized. Of course the

final decision on the operational settings and LHC deliverables falls on the management.

Some references

07/06/2019 - BBLS N. Karastathis 22

1. ATLAS Collaboration, “Observation of a new particle in the search for the Standard Model Higgs boson with the ATLAS detector at the LHC”,

Phys. Lett. B 716 (2012) 1-29

2. B. Muratori, “Luminosity and luminous region calculations for the LHC”, LHC-PROJECT- NOTE-301

3. B. Muratori, T. Pieloni, “Luminosity leveling techniques for the LHC”, Proceedings of the ICFA Mini-Workshop on Beam–Beam Effects in

Hadron Colliders, CERN, Geneva, Switzerland, 18–22 March 2013 CERN–2014–004 (CERN, Geneva, 2014)

4. “beamCal” : https://github.com/nkarast/beamCal

5. ”BeamParameterEvo”: https://github.com/nkarast/BeamParameterEvo

6. N. Karastathis, K. Fuchsberger, M. Hostettler, Y. Papaphilippou, and D. Pellegrini, “Crossing angle anti-levelling at the LHC in 2017”,

Journal of Physics: Conference Series 1067 022004 (2018)

7. E. Metral et al, “Update of the HL-LHC operational scenarios for proton operation”, CERN-ACC-NOTE-2018-0002

8. G. Rumolo et al, “What to expect from the Injectors during Run-III”, Proceedings of 9th Evian LHC Operations Workshop, 30 Jan– 1 Feb.

2019

9. N. Karastathis, S. Fartoukh et al., “Report from the LHC Run-III Configuration Working Group”, Proceedings of 9th Evian LHC Operations

Workshop, 30 Jan– 1 Feb. 2019

10. S. Fartoukh, N. Karastathis et al., LHC Machine Committee, 6 Mar 2019, 803742 (GPN ONLY)

11. S. Fartoukh, “Achromatic telescopic squeezing scheme and applications to the LHC and its luminosity upgrade”, Phys. Rev. ST. Accel.

Beams 16, 111002, 19 Nov 2013

12. S. Fartoukh, N. Karastathis, M. Solfaroli Camillocci, R. Tomas Garcia, “About flat telescopic optics for the future operation of the LHC”,

CERN ATS Reports, CERN-ACC-2018-0018

13. N. Karastathis, S. Fartoukh, et al, LHC Program Coordination Meeting, 25 Mar 2019, 806576

14. LHC Program Coordination Meeting, 20 May 2019, 820221

Tune along the fill (additional)

07/06/2019 Document reference 24