Herwig++ for BSM Alix Wilcock, IPPP Durham [on behalf of the - - PowerPoint PPT Presentation

▶
herwig for bsm
SMART_READER_LITE
LIVE PREVIEW

Herwig++ for BSM Alix Wilcock, IPPP Durham [on behalf of the - - PowerPoint PPT Presentation

Dec 04, 2023 412 likes •611 views

Herwig++ for BSM Alix Wilcock, IPPP Durham [on behalf of the Herwig++ team] MC4BSM, Fermilab 2015 18/05/2015 Herwig++ details General purpose MC event generator Uses ThePEG (Toolkit for High Energy Physics Event Generation) Native matrix

slide-1
SLIDE 1

Herwig++ for BSM

Alix Wilcock, IPPP Durham

[on behalf of the Herwig++ team]

MC4BSM, Fermilab 2015 18/05/2015

slide-2
SLIDE 2

Herwig++ details

General purpose MC event generator

Uses ThePEG (Toolkit for High Energy Physics Event Generation)

Native matrix elements and interfaces

Powheg and MC@NLO matching

Angular ordered and dipole showers, cluster hadronization, underlying event... Major new release coming soon (more on that later) Currently ∼ 15 collaboration members in CERN, Durham, Karlsruhe, Manchester More info at arXiv:0803.0883 and download at

http://herwig.hepforge.org

slide-3
SLIDE 3

Outline

1 Current BSM simulation 2 SM in Herwig++ 3.0 3 BSM in Herwig++ 3.x (x>0) 4 Summary

slide-4
SLIDE 4

Hard processes in Herwig++

Simulation of BSM hard processes can be done by:

1 Reading in LHE files 2 Using internal helicity amplitudes with:

Hand-coded models Universal FeynRules Output

slide-5
SLIDE 5

Hard processes in Herwig++

Simulation of BSM hard processes can be done by:

1 Reading in LHE files 2 Using internal helicity amplitudes with:

Hand-coded models Universal FeynRules Output

  • 1. Les Houches event files

Generate partonic events in LHA format using an external matrix element generator Input through ThePEG LesHouchesReader class Handles positive and negative fixed weights - NLO compatible

slide-6
SLIDE 6

Hard processes in Herwig++

  • 2. Internal simulation

Automatic determination of MEs for 2 → 2, 1 → 2, 1 → 3 (and some 1 → 4) processes, including spin correlations Based on implementation of the HELAS formalism Interactions are coded as Vertex classes which evaluate e.g ¯ ψcγµ [gLPL + gRPR] ψǫµ Inherit from existing Lorentz structures

slide-7
SLIDE 7

Hard processes in Herwig++

  • 2. Internal simulation

Automatic determination of MEs for 2 → 2, 1 → 2, 1 → 3 (and some 1 → 4) processes, including spin correlations Based on implementation of the HELAS formalism Interactions are coded as Vertex classes which evaluate e.g ¯ ψcγµ [gLPL + gRPR] ψǫµ Inherit from existing Lorentz structures Hand-code models by implementing: Input file specifying particle content Model class that stores/calculates parameters of the model Vertex classes with the new coupling information E.g. MSSM, NMSSM, RPV SUSY, ADD and RS gravitons, UED, leptoquarks, sextet...

slide-8
SLIDE 8

Hard processes in Herwig++

  • 2. Internal simulation

Automatic determination of MEs for 2 → 2, 1 → 2, 1 → 3 (and some 1 → 4) processes, including spin correlations Based on implementation of the HELAS formalism Interactions are coded as Vertex classes which evaluate, e.g ¯ ψcγµ [gLPL + gRPR] ψǫµ Inherit from existing Lorentz structures UFO model converter Automatically convert models in UFO format into a

Herwig++ model

$ufo2herwig /path/directoryName

slide-9
SLIDE 9

Decays of BSM particles

Finite width effects NWA used to separate production and decay Often need to include finite width effects Distribute masses of outgoing particles in hard processes and decays using a weight factor [arXiv:0805.3037] w = 1

π

  • dm2

mΓ(m) (m2−M2)2−m2Γ2(m)

slide-10
SLIDE 10

Decays of BSM particles

Improve simulation of hard radiation using Powheg inspired ME correction [arXiv:1303.4563] Available for 1 → 2 decays involving:

scalar, fermion, vector and tensor∗ particles colour singlets, (anti)fundamental and adjoint reps of SU(3)

ME correction LO 10−7 10−6 10−5 ˜ uL → u ˜ χ0

1

dσ/dpT,2 [fb/GeV] 50 100 150 200 250 300 350 400 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 pT,2 [GeV] Ratio

∗Not for decays involving coloured tensors

slide-11
SLIDE 11

Herwig++ 3.0

Main feature: automated LO and NLO cross sections Also spin correlations in shower, QED radiation. No major changes for BSM

Matchbox framework

Automated NLO calculations Matching to angular ordered/dipole showers via

Powheg and MC@NLO

(Functionality for (N)LO merging) Improved evaluation of shower and scale uncertainties. Easy variation of: Renormalization scale, µR Factorization scale, µF Hard shower scale, µQ (pT veto for shower emissions)

slide-12
SLIDE 12

Herwig++ 3.0

Matrix element merging in Matchbox [J. Bellm, S. Gieseke, S. Pl¨

atzer]

Unitarized approach Smoothly integrated with no extra event files or external codes to run

b b b b b

H++/MatchBox + MadGraph5/OpenLoops (Prel.) µr = µ f = 1/2 − 2 · MW Data

b

LO+PS 2LO 3LO 4LO 1 10 1 10 2 10 3 Inclusive Jet Multiplicity σ(W + ≥ Njet jets) [pb] 1 2 3 4 0.5 1 1.5 2 Njet MC/Data

b b b b b

H++/MatchBox + MadGraph5/OpenLoops (Prel.) µr = µ f = 1/2 − 2 · MW Data

b

2LO+1NLO 3LO+2NLO 4LO+3NLO 1 10 1 10 2 10 3 Inclusive Jet Multiplicity σ(W + ≥ Njet jets) [pb] 1 2 3 4 0.5 1 1.5 2 Njet MC/Data

slide-13
SLIDE 13

Herwig++ 3.0

Main feature: automated LO and NLO cross sections Also spin correlations in shower, QED radiation. No major changes for BSM

Matchbox framework

Automated NLO calculations Matching to angular ordered/dipole showers via

Powheg and MC@NLO

(Functionality for (N)LO merging) Improved evaluation of shower and scale uncertainties. Easy variation of: Renormalization scale, µR Factorization scale, µF Hard shower scale, µQ (pT veto for shower emissions)

slide-14
SLIDE 14

Matchbox Overview.

σNLO =

  • n

dσLO |Mn,0 |Mn,0|2

  • +
  • n
  • dσV
  • |Mn,0, |Mn,1

2Re(Mn,0|Mn,1)

  • +
  • 1

dσA

  • |Mn,0

|Mij

n,0|2

  • +
  • n+1
  • dσPS
  • P(˜

q), D(p⊥) RME(p⊥)

  • − dσA
  • |Mn,0

|Mij

n,0|2

  • +
  • n+1
  • dσR

|Mn+1,0 |Mn+1,0|2

  • − dσPS
  • P(˜

q), D(p⊥) RME(p⊥)

  • Interfaces at amplitude level

– Color bases provided, including interface to ColorFull. [M. Sj¨

  • dahl, S. Pl¨

atzer] – Spinor helicity library and caching facilities. – MadGraph5. [MadGraph & J. Bellm, S. Gieseke, S. Pl¨ atzer, AW] – Some in-house calculations and parts of HJets++. [F. Campanario, T. Figy, S. Pl¨ atzer, M. Sj¨

  • dahl]

Interfaces at squared amplitude level

– Dedicated interfaces. [HEJ & S. Pl¨ atzer] [nlojet++ & J. Kotanski, J. Katzy, S. Pl¨ atzer] – BLHA2. [GoSam & J. Bellm, S. Gieseke, S. Pl¨ atzer, C. Reuschle] [NJet & S. Pl¨ atzer] [OpenLoops & J. Bellm, S. Gieseke] [VBFNLO & K. Arnold, S. Gieseke, S. Pl¨ atzer]

Matchbox infrastructure

based on [S. Pl¨ atzer & S. Gieseke – Eur.Phys.J. C72 (2012) 2187] – Process generation and bookkeeping, integration. – Automated Catani-Seymour dipole subtraction. – Diagram-based mutli-channel phase space.

Shower plugins

matching details & uncertainties [in preparation] – Dipole shower D(p⊥). – Angular ordered shower P(˜ q). – ME correction RME(p⊥), including adaptive sampling.

slide-15
SLIDE 15

Matrix-element corrections for BSM processes

Idea: use Matchbox framework and interfaces to add higher order corrections for BSM production processes Generally limited by absence of virtual matrix elements → Powheg style matrix-element correction Correct hardest parton shower emission using NLO real-emission contribution But total cross section and inclusive observables still only at LO (not NLO) Test case: pp → ˜ q˜ q∗ in MSSM using Matchbox and MadGraph 5 amplitudes

slide-16
SLIDE 16

Top squark pair production: before

ATLAS search for direct production of the top squark in events with missing ET and two b-jets [arXiv:1308.2631] ˜ t1 → b˜ χ+

1 → bf ¯

f ′ ˜ χ0

1 with m˜ χ+

1 − m˜

χ0

1 = 5 GeV

Original signal simulated with MadGraph + PYTHIA 6 (with MLM merging)

200 300 400 500 600 700

m˜

t1 [GeV]

100 200 300 400 500 600

m˜

χ0

1 [GeV] ATLAS-SUSY-2013-05 ∆m ˜

χ+

1 − ˜

χ0

1 = 5 GeV

˜ t1 → b ˜ χ+ 1 forbidden

  • L = 20.1fb−1

√s = 8TeV ATLAS result Herwig++

slide-17
SLIDE 17

Top squark pair production: after

ATLAS search for direct production of the top squark in events with missing ET and two b-jets [arXiv:1308.2631] ˜ t1 → b˜ χ+

1 → bf ¯

f ′ ˜ χ0

1 with m˜ χ+

1 − m˜

χ0

1 = 5 GeV

Original signal simulated with MadGraph + PYTHIA 6 (with MLM merging)

200 300 400 500 600 700

m˜

t1 [GeV]

100 200 300 400 500 600

m˜

χ0

1 [GeV] ATLAS-SUSY-2013-05 ∆m ˜

χ+

1 − ˜

χ0

1 = 5 GeV

˜ t1 → b ˜ χ+ 1 forbidden

  • L = 20.1fb−1

√s = 8TeV ATLAS result Herwig++

slide-18
SLIDE 18

General squark pair production: pp → ˜ q˜ q∗ for ˜ q = ˜ t

New potentially divergent qg-initiated contributions if m˜

g > m˜ q

q g ˜ q ˜ q∗ q g q ˜ q∗ q ˜ q q g q ˜ q∗ ˜ q

Subtract resonant contribution from real-emission correction

(treat instead as qg → ˜ q˜ g with ˜ g → q˜ q∗)

ME correction with DS no ME correction 10−9 10−8 10−7 10−6 10−5 10−4 pp → ˜ uL ˜ u∗

L

dσ/dpT, ˜

uL ˜ u∗

L [fb/GeV]

500 1000 1500 2000 2500 3000 3500 0.6 0.8 1 1.2 1.4 pT, ˜

uL ˜ u∗

L [GeV]

Ratio

slide-19
SLIDE 19

Summary

Herwig++ provides a flexible tool for BSM simulation

Easy to add new models using UFO converter Automated NLO calculations for SM processes coming in

Herwig++ 3.0

Improvement to simulation of hard radiation in BSM processes coming soon Thanks for your attention