Soft interactions in Herwig Stefan Gieseke Institut f ur - - PowerPoint PPT Presentation

Soft interactions in Herwig

Stefan Gieseke

Institut f¨ ur Theoretische Physik KIT

work with Patrick Kirchgaeßer and Frash¨ er Loshaj MPI@LHC16 – San Crist´

• bal de las Casas

28 Nov–2 Dec 2016

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 1/21

pp Event Generator

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 2/21

pp Event Generator

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 2/21

pp Event Generator

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 2/21

pp Event Generator

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 2/21

pp Event Generator

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 2/21

pp Event Generator

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 2/21

Underlying event in Herwig++

Semihard UE

• Default from Herwig++ 2.1.

[Herwig++, 0711.3137]

• Multiple hard interactions, pt ≥ pmin

t

.

[B¨ ahr, SG, Seymour, JHEP 0807:076]

• pQCD 2 → 2.
• Similar to JIMMY.
• Good description of harder UE data (“plateau”).

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 3/21

Underlying event in Herwig++

Soft UE

• Default from Herwig++ 2.3.

[Herwig++, 0812.0529]

• Extension to soft interactions pt < pmin

t

.

• Theoretical work with simplest possible extension.

[B¨ ahr, Butterworth, Seymour, JHEP 0901:065]

• “Hot Spot” model.

[B¨ ahr, Butterworth, SG, Seymour, 0905.4671] Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 3/21

Hot Spot model

Fix the two parameters µsoft and σinc

soft in

χtot(

• b,s) = 1

2

• A(
• b;µ)σinc

hard(s;pmin t

)+A(

• b;µsoft)σinc

soft

• from two constraints. Require simultaneous description of σtot

and bel (measured/well predicted), σtot(s) ! = 2

• d2

b

• 1−e−χtot(
• b,s)

, bel(s) ! =

• d2

b b2 σtot

• 1−e−χtot(
• b,s)

.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 4/21

Extending into the soft region

Continuation of the differential cross section into the soft region pt < pmin

t

(here: pt integral kept fixed)

2 4 6 8 10

pt(GeV)

1 2 3 4 5

1/

• (5 GeV) d
• /dpt(1/GeV)

d

soft

dpt

pt e

(p 2

t

p min,2

t

) p min

t

=3 GeV,

=

0.5 GeV

2

p min

t

=5 GeV,

=0.06 GeV

2

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 5/21

Implementation of soft scattering

Soft gluon production with soft pt < pmin

t

spectrum. Colour structure important. Two extreme cases possible. Sensitivity to parameter colourDisrupt = P(disrupt colour lines) Long colour lines appear when swapping outgoing gluons.

R1 R2 g3 g4 g1 g2

g g g g R R

Colour reconnections applied!

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 6/21

So far at the LHC

Soft model is extension of MPI model for Underlying Event and harder aspects of Min Bias events. Herwig 7.0 at 900 GeV and 7 TeV:

[ATLAS, Eur.Phys.J. C71 (2011) 1636]

b b b b b b

Data

b

Hw 7.0 LO ⊕ PS 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Transverse ∑ p⊥ density vs. pclus1

⊥

, √s = 900 GeV d2 ∑ p⊥/dηdφ 1 2 3 4 5 6 7 8 0.6 0.8 1 1.2 1.4 p⊥ (leading particle) [GeV] MC/Data

b b b b b b b b

Data

b

Hw 7.0 LO ⊕ PS 0.5 1 1.5 2 Transverse ∑ p⊥ density vs. pclus1

⊥

, √s = 7 TeV d2 ∑ p⊥/dηdφ 2 4 6 8 10 12 14 0.6 0.8 1 1.2 1.4 p⊥ (leading particle) [GeV] MC/Data

Still reasonably well for moderately soft particles.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 7/21

So far at the LHC

Soft model is extension of MPI model for Underlying Event and harder aspects of Min Bias events. Herwig 7.0 at 900 GeV and 7 TeV:

[ATLAS, Eur.Phys.J. C71 (2011) 1636]

b b b b b b b b b b b b b b b b b b b b b

Data

b

Hw 7.0 LO ⊕ PS 0.2 0.4 0.6 0.8 1 1.2 N density vs. ∆φ, pclus1

⊥

> 3.0 GeV, √s = 900 GeV d2N/dηdφ 0.5 1 1.5 2 2.5 3 0.6 0.8 1 1.2 1.4 |φ| (w.r.t. leading particle) [rad] MC/Data

b b b b b b b b b b b b b b b b b b b b b

Data

b

Hw 7.0 LO ⊕ PS 0.5 1 1.5 2 N density vs. ∆φ, pclus1

⊥

> 3.0 GeV, √s = 7 TeV d2N/dηdφ 0.5 1 1.5 2 2.5 3 0.6 0.8 1 1.2 1.4 |φ| (w.r.t. leading particle) [rad] MC/Data

Still reasonably well for moderately soft particles.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 7/21

The bump

A clear case of abusing a model for the hard UE in forward/diffractive final states. . .

[ATLAS, Eur.Phys.J. C72 (2012) 1926]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

Hw 7.0 LO ⊕ PS 10−1 1 10 1 10 2 Rapidity gap size in η starting from η = ±4.9, pT > 200 MeV dσ/d∆ηF [mb] 1 2 3 4 5 6 7 8 0.6 0.8 1 1.2 1.4 ∆ηF MC/Data

Bump is artefact. No Diffraction. Poor modeling of soft

• interactions. Colour assignment ad hoc.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 8/21

Outline

Challenge accepted.

• Model for diffractive final states.
• Model for soft particle production.
• Results.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 9/21

Diffraction as part of minimum bias simulation

Diffractive final states directly modeled. Not embedded in MPI approach via cuts through triple pomeron vertices. Therefore change in constraint xσtot(s) ! = 2

• d2

b

• 1−e−χtot(
• b,s)

, where x ≈ 1− σdiff σtot ∼ 20−25% . In min-bias simulation: every event is either

• diffractive, directly modeled from pp initial state.
• non-diffractive, modeled in the MPI picture, parton level.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 10/21

Diffractive final states

Strictly low mass diffraction only. Allow M2 large nonetheless. M2 power-like, t exponential (Regge). pp → (baryonic cluster)+p . Hadronic content from cluster fission/decay C → hh... Cluster may be quite light. If very light, use directly pp → ∆+p . Also double diffraction implemented. pp → (cluster)+(cluster) pp → ∆+∆ . Technically: new MEs for diffractive processes set up.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 11/21

Model for soft particle production in Herwig

Reproduce core properties of soft particle production. “flat in rapidity”, “narrow in pt”. Main idea: “soft interaction = cut pomeron = particle ladder”. Nsoft from MPI model = #ladders. Clusters produced via colour connected quarks and gluons. Adopt to soft interactions in Herwig via remnant decays.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 12/21

Multiperipheral kinematics

[Baker, Ter-Martirosyan 1976]

Average relative momentum fraction x. Leads to flat rapidity distribution of emissions in a single ladder. ∆y ∼ ln 1 x . Choose some constant C, then x ∼ 1/C . N average number of emitted particles. N = 1 lnC ln s m2 p⊥ or m⊥ moderate, unordered.

qi, z pi+1, z = (1 − xi+1)qi ,z qi+1, z = xi+1qi, z p1 p2 p3 p4 p5 pN q1 q2 q3 pA pB q4

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 13/21

Soft particle production model in Herwig

• #ladders = Nsoft.
• N particles from Poissonian, width N.

Model parameter 1/lnC ≡ nladder → tuned.

• xi smeared around x (calculated).
• p⊥ from Gaussian acc to soft MPI model.
• particles are q,g, see figure.

Symmetrically produced from both remnants.

• Colour connections between neighboured particles.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 14/21

Soft particle production model in Herwig

Single soft ladder with MinBias initiating process.

P q ¯ q g g g g Remnant1 Remnant2 PBeam, 1 PBeam, 2 Cluster

Further hard/soft MPI scatters possible.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 14/21

Parameters and tuning

Diffraction plus MPI incl new soft model. Diffractive cross sections adjusted to data. Tuning to Min Bias data: η,p⊥ for various Nch, p⊥(Nch). Usual MPI parameters (pmin

⊥,0 ,b) → pmin ⊥ (√s),

µ2, preco . One additional parameter nladder .

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 15/21

Tuned results

ATLAS Min Bias 7 TeV.

[ATLAS, New.J.Phys. 13 (2011) 053033]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newSoftMPI, χ2/n = 0.28 LHC-MB, χ2/n = 8.15 0.5 1 1.5 2 2.5 3 3.5 4 Charged particle η at 7 TeV, track p⊥ > 500 MeV, for Nch ≥ 6 1/Nev dNch/dη

• 2
• 1

1 2 0.6 0.8 1 1.2 1.4 η MC/Data

Similar to previous results, “harder part of Min Bias”.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 16/21

Tuned results

ATLAS Min Bias 7 TeV.

[ATLAS, New.J.Phys. 13 (2011) 053033]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newSoftMPI, χ2/n = 0.18 LHC-MB, χ2/n = 88.22 0.5 1 1.5 2 2.5 Charged particle η at 7 TeV, track p⊥ > 500 MeV, for Nch ≥ 1 1/Nev dNch/dη

• 2
• 1

1 2 0.6 0.8 1 1.2 1.4 η MC/Data

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newSoftMPI, χ2/n = 0.15 LHC-MB, χ2/n = 12.67 1 2 3 4 5 6 7 Charged particle η at 7 TeV, track p⊥ > 100 MeV, for Nch ≥ 2 1/Nev dNch/dη

• 2
• 1

1 2 0.6 0.8 1 1.2 1.4 η MC/Data

Also soft rates well described.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 16/21

Tuned results

ATLAS Min Bias 7 TeV.

[ATLAS, New.J.Phys. 13 (2011) 053033]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newSoftMPI, χ2/n = 2.94 LHC-MB, χ2/n = 84.74 10−6 10−5 10−4 10−3 10−2 10−1 1 Charged particle p⊥ at 7 TeV, track p⊥ > 500 MeV, for Nch ≥ 1 1/Nev 1/2πp⊥ dσ/dηdp⊥ 1 10 1 0.6 0.8 1 1.2 1.4 p⊥ [GeV] MC/Data

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newSoftMPI, χ2/n = 3.04 LHC-MB, χ2/n = 75.22 10−5 10−4 10−3 10−2 10−1 1 10 1 Charged particle p⊥ at 7 TeV, track p⊥ > 100 MeV, for Nch ≥ 2 1/Nev 1/2πp⊥ dσ/dηdp⊥ 10−1 1 10 1 0.6 0.8 1 1.2 1.4 p⊥ [GeV] MC/Data

Tails? Still within 1σ.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 16/21

More results

CMS, NSD analysis 7 TeV

[CMS, PRL 105 (2010) 022002]

b b b b b b b b b b b b b b

Data

b

MB-7, χ2/n = 1.32 H7.0, χ2/n = 23.85 10−1 1 10 1 Charged hadron p⊥ for |η| = 1.9 at √s = 7 TeV d2Nch /dη dp⊥ [(GeV/c)−1] 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0.6 0.8 1 1.2 1.4 p⊥ [GeV/c] MC/Data

Lowest bin → potential to be tunable.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 17/21

The bump plot, ∆ηF

[CMS, PRD 92 (2015) 012003]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newSoftMPI, χ2/n = 0.35 LHC-MB, χ2/n = 796.13 10−1 1 10 1 10 2 Rapidity gap size in η starting from η = ±4.7, pT > 200 MeV dσ/d∆ηF [mb] 1 2 3 4 5 6 7 8 0.6 0.8 1 1.2 1.4 ∆ηF MC/Data Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 18/21

Individual contributions to ∆ηF

[ATLAS, Eur.Phys.J. C72 (2012) 1926]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

LHC-MinBias LHC-Diffraction 10−1 1 10 1 10 2 Rapidity gap size in η starting from η = ±4.9, pT > 200 MeV dσ/d∆ηF [mb] 1 2 3 4 5 6 7 8 0.6 0.8 1 1.2 1.4 ∆ηF MC/Data Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 19/21

Charged particle multiplicity

CMS, NSD analysis 7 TeV

[CMS, PRL 105 (2010) 022002]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

MB-7, χ2/n = 12.50 H7.0, χ2/n = 23.31 10−5 10−4 10−3 10−2 10−1 Charged hadron multiplicity, |η| < 2.4, √s = 7 TeV Pn 20 40 60 80 100 120 140 160 180 0.6 0.8 1 1.2 1.4 n MC/Data

Large discrepancies, tail in particular. Low n → “NSD”?

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 20/21

Charged particle multiplicity

CMS, NSD analysis 7 TeV

[CMS, PRL 105 (2010) 022002]

b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b

Data

b

newCR, χ2/n = 0.84 softMPI, χ2/n = 12.50 H7.0, χ2/n = 23.31 10−5 10−4 10−3 10−2 10−1 Charged hadron multiplicity, |η| < 2.4, √s = 7 TeV Pn 20 40 60 80 100 120 140 160 180 0.6 0.8 1 1.2 1.4 n MC/Data

CR model with Baryons (preliminary teaser).

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 20/21

Conclusions

• GIGO . . .
• Completely new Min Bias scattering with diffraction and

soft particle production.

• Data well described.
• Nch distributions tricky → outlook colour reconnection.

Stefan Gieseke · MPI@LHC 2016 · 28 Nov – 2 Dec 2016 21/21