Diffractive physics simulation for LHC

1

Diffractive physics simulation for LHC

Marek Taševský (Physics Inst. Prague)In collaboration with Ch.Royon, A.Kupčo, M.Boonekamp

Low-x workshop - Lisbon 29/06 2006

Upgrade of 240 m RP in ATLAS

2

Forward physics in ATLAS

Originally oriented to 1. Luminosity calibration using Roman pots of Totem type

2. Luminosity monitoring using integrated Cerenkov det. LUCID

To access the diffraction physics, the RPs need to be upgradedto detect the diffractive protons and to stay the radiation hardness.

Hard diffraction, Soft diffraction, Double Pomeron Exchange canbe studied using the central detector + RPs.

Participating institutes: Saclay (Ch.Royon, M.Boonekamp, L.Shoeffel, O.Kepka et al.)Prague (A.Kupčo, M.T., V.Juránek, M.Lokajíček)Stony Brook (Michael Rijsenbeek) Cracow (group being formed)

3

Hgap gap

b

b -jet

-jet

h

Double Pomeron Exch. Higgs Production

Exclusive DPE Higgs production pp p H p : 3-10 fbInclusive DPE Higgs production pp p+X+H+Y+p : 50-200 fb

p p

Mh² measured in RP via missing mass as ξ1*ξ2*s bb: Jz=0 suppression of gg->bb bg | WW: bg almost negligible

E.g. V. Khoze et alM. Boonekamp et al.B. Cox et al. …V.Petrov et al.

Advantages of Exclusive:

bb: L1-trigger of “central CMS+220 RP” type extensively studied by CMS/Totemgroup.

WW: Extremely promising for Mh>130 GeV. Relevant triggers already exist.Better Mh resolution for higher Mh.

(Wˉ)

(W+)

4

DPE Higgs event generators

1. DPEMC 2.4 (M.Boonekamp, T.Kucs, Ch.Royon, R.Peschanski)

- Bialas-Landshof model for Pomeron flux within proton - Rap.gap survival probability = 0.03 - Herwig for hadronization - Exclusive+Inclusive processes

2. EDDE 1.2 (V.Petrov, R.Ryutin)

- Regge-eikonal approach to calculate soft proton vertices - Sudakov factor to suppress radiation into rap.gap - Pythia for hadronization

3. ExHuMe 1.3.1 (J.Monk, A.Pilkington)

- Durham model for exclusive diffraction (pert.calc. by KMR) - Improved unintegrated gluon pdfs - Sudakov factor to suppress radiation into rap.gap + rap.gap survival prob.=0.03 - Pythia for hadronization

All three models available in the fast CMS simulation

5

Roman Pot acceptances on CMS side

6

Upgrade of 240 (220) m RP in ATLASMain goal is to extend the forward physics program in ATLAS by veryrich diffraction physics using the existing place for RPs at or close to240 m. Complementary to FP420 program.

o 220 or 240 m? Study the acceptances of RPs using MAD-X (complex program used by beam division)o What type of detector?o What’s the effect of the collimator at Q5 to be put at high lumi?

7

Proton tracking in LHC

MAD-X (beam division)LHC optics v6.5, low β*

Hector (Piotrzkowski, Favereau, Rouby from UCLouvain)LHC optics v6.5, low β*Beam apertures included, kickers switched off Linear approximation for effects of dispersion

FPtrack (Bussey from Uni Glasgow)LHC optics v6.5, low β*Beam apertures and kickers includedExact magnet formulae for effects of dispersion

All MAD-X pictures made by Sasha Kupčo from Prague

8

Elastic events at 240 m RPs

MAD-X HECTOR FPTRACK

Beam 1

9

Diffractive events at 240 m RP

MAD-X HECTOR FPTRACK

Beam 1

10

Diffractive events using MAD-X

220 m 240 m 420 m

• MAD-X tracking• Beam 1• β*=0.55m LHC optics, v6.5 420 has opposite orientation than 220 (240)

11

Diffractive events at 420 m RP

MAD-X FPTRACK

Orientation opposite because of opposite conventions in MAD-X and FPTRACK

Beam 1

12

Beam spots using MAD-X

Numbers agree withthose based on σ(s) = √β*(s)ε

Simulated parameters:

Trans.vtx position σx,y = 16 μm

Beam en. spread σE = 0.77 GeV

Beam divergence σθx,θy = 30

μrad

13

Acceptance for 220 m RPs (beam1)

• 0.02 steps in ξ• t=0.0 and 0.05 GeV²• 2x2 cm² detector has acceptance of upto ξ ~ 0.16

• Detailed look at low ξ• 0.005 steps in ξ• σx = 96 μm

14

Acceptance for 240 m RPs (beam1)

• 0.02 steps in ξ• t=0.0 and 0.05 GeV²• 2x2 cm² detector has acceptance of upto ξ ~ 0.14

• Detailed look at low ξ• 0.005 steps in ξ• σx = 125 μm

15

Hit maps at 216 and 224 m RPs

- Test of the idea of using a displacement for the L1 trigger to suppress beam halo.- No uniform shift direction between 216 and 224 m

16

First study of pile-up at 240 m RP

Pile-up generated by Pythia msel=2 (diffr.+non-diffr. processes, but DPE missing)

Assume 2x2 cm²active volumeand 20 σ for distance from beam

2.5% of PU protons seen inRP at 240 m -> expect about 1 PU event/BX at highest luminosity

Beam 1

17

Roman Pots

Idea: follow Totem design

Acceptance studies showed the horizontal pots only are needed

Restudy the supports

18

Detectors- Very good space resolution (~ μm) useful to distinguish halo from

signal.- Very good timing resolution (~ O(10) ps) useful to distinguish pile-up vertices from signal ones.- Good readout time (5 ns)

Si strips / Micromegas:1) Si strips: Advantage: fast readout (5 ns) Disadvantage: sensitivity of Si signal to EM noise Timing: provided by Cerenkov counters

2) Micro-Mesh-Gaseous Structure: Advantage: timing resolution of 1 ns, space resolution of 15 μm good behaviour in radiative env. insensitive to EM noise Disadvantage: gas circulation in tunnel (safety problem? the volume will be very small)

19

Summary

1) The project has started this year. All institutes need to submit it by October.

2) MAD-X installed and working in Prague: Acceptance studies ongoing. Fast simulation existing (beam parameters smeared, the

detector ones to follow)

3) Roman Pots will use the Totem design. Need to decide what type

of sensitive detectors to put in. They need to be included in ATLAS L1 trigger.

4) This project is complementary to FP420 and a natural follow-up

of existing luminosity program in ATLAS. The forward physics programmes move ahead on both sides, CMS and ATLAS.

20

BACKUP SLIDES

21

Difference between DPEMC and (EDDE/ExHuMe) is an effect of Sudakov suppression factor growing as the available phase space

forgluon emission increases with increasing mass of the central system

Models predict different physics potentials !

22

Effect of pile-up events What is the number of fake signal events per bunch crossing

(Nfake/BX)

caused by PU events?

Selection criteria for signal events (Higgs in DPE):

[2 protons in RPs, each on opposite side] x [Jet cuts] x [Mass window]

For the moment (till I get the final results), assume we can factorize the

task the above way:

Nfake = NRP * [Jet cuts] * [Mass window]

Estimate of NRP: 1.Rough-but-Fast 2.Precise-but-Slow All RP acceptances are taken as means.

23

Phojet generation of PU events

All processes 118 mb Non-diff.inelastic 68 mb Elastic 34 mb Single Diffr.(1) 5.7 mb Single Diffr.(2) 5.7 mb Double Diffr. 3.9 mb DPE 1.4 mb

Number of pile-up events per bunch crossing (BX) Ξ NPU = Lumi x cross section x bunch time width = LHC bunches/filled

bunches =

1034cm-2s-1x104cm2/m2 x 10-28m2/b x 110mb x 10-3b/mb x 25*10-9sX 3564/2808 ~ 35

5*1033 ~ 17.6 , 2*1033 ~ 7.0, 1*1033 ~ 3.5, 1*1032 ~ 0

24

NRP estimate – precise method

Mix PU events with signal or bg – using FAMOS- Sum RP acceptances over all possible proton pairs in all PU

events in one BX and then look at mean over all signal or bg

events. NPU properly smeared using Poisson dist.

E.g. NRP420 = <Σi

NPU(n) ΣjNPU(n) AL

420(i)xAR420(j)>n=5k signal or bg

events Mean nr.of PU events with 2 p’s seen in opposite 420

RPs

<NPU> NRP420 NRP

220 NRPcomb NRP

totalcombinatorics

3.5 0.003 0.016 0.016 0.034

7.0 0.010 0.045 0.053 0.10

17.6 0.045 0.220 0.280 0.54

25.0 0.080 0.420

0.560 1.03

35.0 0.155 0.807 1.040 2.00