New Particle Searches at the LHC
• SUSY- Inclusive search- Backgrounds - SUSY parameters
• Resonances in the Drell-Yan mass distribution • Heavy long-lived particles• Universal Extra Dimensions• Black holes
R. Ströhmer RAL-PPD Seminar 9.5.07 2
First challenge: get the LHC operational
Still on course for engineering run fall 2007: system commissioning single beam operations at 450 GeV collisions at 450 x 450 GeV, no ramp, no squeeze low luminosity: ATLAS/CMS commissioning
First collisions at 14 TeV: June 2008 ? after system and beam commissioning 26 weeks of proton-proton physics run in 2008 phase 1: 43 bunches, L ~ 5 x 1030 phase 2: 75 ns, L ~2.5 x 1031 1 x 1032 phase 3: 25 ns, L ~4 x 1032 1 x 1033 cm-2s-1
Integrated luminosity end of 2008: 0.5 - 1 fb-1 ? (e.g.: 1 fb-1 = 120 effective days @ 1032 cm-2s-1)
R. Ströhmer RAL-PPD Seminar 9.5.07 3
And the experiments too: huge challenge
Getting the subdetectors built, tested and installed.Power and signal cables, detector control and monitoringCooling pipes, cryogenic installations, magnets…
CMS: lowered central part (YB0) February 28th , rest soon will run in 2007 without ECAL endcap and pixels rest going well
ATLAS: on a tight schedule to run almost complete in 2007 No TRT at high ||, some muon chambers missing
Both will have reduced trigger/DAQ capabilities initially
R. Ströhmer RAL-PPD Seminar 9.5.07 4
Getting the data flowing…
First individual detectors, then combinedCommissioning the DAQ system with cosmicsSingle beam in LHC: beam halo
Use: debug cabling errors initial alignment first intercalibration: uniformity to few %
Data processing: Grid, Tier-1, Tier-2 etc
Challenge: get processing of HUGE quantities of data going Data Challenges, Calibration Challenges, Computing System Commissioning (ATLAS 2007)
ATLAS: CSC exercise should lead to notes CMS: published physics TDR in summer 06
5
R. Ströhmer RAL-PPD Seminar 9.5.07 5
Particle Searches in a Nutshell
Look for excess of events over Standard Model expectation
- Multiple or high pt leptons, jets and missing transverse energy
If excess is observed:Is the instrumental background understood?
Is the expected SM background understood?
Which models predict the excess?- How can we distinguish between models and
measure model parameters?
eeZ
If no excess is observed:What is the signal efficiency: reconstruction and trigger efficiencies
exclude models or parameter regions for models
R. Ströhmer RAL-PPD Seminar 9.5.07 6
Calorimeter noise has to be understood
Run IIV. Shary CALOR04
ETmiss spectrum contaminated by cosmics,
beam-halo, machine/detector problems, etc.
R. Ströhmer RAL-PPD Seminar 9.5.07 7
Material in front of calorimeter
Affects electrons and photons: energy loss, conversions
R. Ströhmer RAL-PPD Seminar 9.5.07 8
Getting efficiency from data
Use Zµµ to study muons
select one muon with tight requirements
both muons from Z
don’t require quantity you want to test from second muon- tracking efficiency
- muon chamber efficiency
- trigger efficiency
R. Ströhmer RAL-PPD Seminar 9.5.07 9
Top events as test sample
If one sees the expected distribution and rate on can have confidence in the object reconstructions
Isolated lepton pT> 20 GeV
ETmiss > 20 GeV
4 jets pT> 40 GeV
NO b-tag !!
2 jets M(jj) ~ M(W)
3 jets with largest ∑ pT
100 pb-1
Bg: W+jets
R. Ströhmer RAL-PPD Seminar 9.5.07 10
SUSY SearchesMany SUSY searchesare performed in the framework of mSUGRA
Aim of SUSY searches• Find SUSY (or something new)• Measure quantities (e.g. mass differences)• Prove that it is SUSY• Determine model parameters
WMAP: 0.094<Ωχh2<0.129
Excluded by b- > s(CLEO,BELLE)
0,0,10tan 0 A
Favored by gμ−2 at the 2σ levelMuon g−2 coll.
Stau=LSP
J. Ellis et al., Phys. B565 (2003) 176.
R. Ströhmer RAL-PPD Seminar 9.5.07 11
Inclusive SUSY Search (Jets + missing Et)
Calculate effective mass from jet pt and missing transverse energy
Background estimates increased by Matrix Element Monte Carlo w.r.t. showering MC prediction
Main backgrounds - Z(νν) + Jets
- W + Jets
- ttbar
Backgrounds have to be estimated or checked with data
R. Ströhmer RAL-PPD Seminar 9.5.07 12
Backgrounds from Data
Replace observed µ by νZ->µµ- clean sample- correct shape- small statistics
W->µν- large statistics- problem: needs clean
W+6jet sample
Measure in Z -> μμ
Use in Z -> νν
R. Ströhmer RAL-PPD Seminar 9.5.07 13
Background Normalization from DataSystematic uncertainties due to: Renormalization scale, factorization scale,
PDF mostly effect normalization and not shape.Same normalization for Z->νν, Z->µµ, W->νµ
Determine normalization from Z->µµ and apply to Z->νν, W->νµ
Test with “pseudo data” using different MC parameters
Z 230 +/- 15 (pseudo-data) 200 +/- 23 (estimation)
190 +/- 14 (pseudo-data) 185 +/- 21 (estimation)
W l
Effective Mass
Effective Mass
Missing ET
Missing ET
Leading Jet PT
Leading Jet PT
ATLAS
preliminary
ATLAS
preliminary ATLAS
preliminary
ATLAS
preliminary
ATLAS
preliminary
ATLAS
preliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 14
QCD BackgroundSignificant part with real missing Et from b- and c- decays
Estimate effect of mismeasured jet energy with fast Monte Carlo- get transfer function from full detector simulation
- get transfer function from data where missing Et points in jet direction
MET
jets
Select events with: EtMiss > 100 GeV, dPhi(EtMiss, jet) < 0.1
Et(estimated)/Et(measured)
ATLASpreliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 15
Inclusive SUSY Search (Jets+1lepton+missing Et)
Strong reduction of background due to lepton requirement
Main background is ttbar - important contribution from blνblν with one missing lepton for MT>100
GeV
R. Ströhmer RAL-PPD Seminar 9.5.07 16
Background from Data
Find second quantity not correlated to missing Et General ideas:
SUSY signal plus bg
A
B C
D
Missing ET
other variable
bg
bg bg
Bg in D = A x C/B
normalize to data
For ttbar-> bqqblν the top mass canbe used as second quantity
Contribution from ttbar->blνblνis under study
R. Ströhmer RAL-PPD Seminar 9.5.07 17
Expected significance
The statistical significances have been studied including background uncertainties with the likelihood ratio method
R. Ströhmer RAL-PPD Seminar 9.5.07 18
Determination of SUSY Parameters
Example:Coannihilation point
Two edges in lepton-pair-mass
Estimate background from eµ events
20.6 fb-1
MC Truth, lRMC Truth, lLMC Reconstructed
ATLAS Preliminary
Full sim.
01,
02
~ llll LR 264 154 , 255 137
~ ~
GeV
R. Ströhmer RAL-PPD Seminar 9.5.07 19
Spin Measurement
01
02
~ ~ ~ ~ llqllqqq nearRL
First emitted lepton (“near”)0 1/2 0
More quarks thanantiquarks (pp collisions)Remaining asymmetry:
LHCC5:LHCC5:mm00 =100 GeV =100 GeV
mm1/21/2 =300 GeV =300 GeV
AA00 =-300 GeV =-300 GeV
tan(tan(ββ) =2.1 ) =2.1 sign(sign(μμ)=+)=+
M(qlfar)quark
antiquark
ql-
ql+ql-
ql+
M(qlnear)
antiquark
quark
No spin correlations,no asymmetry
After selection
Parton level x 0.6
L=500 fb-1
ATLAS Fast Simulation
ql+
L=500 fb-1
SPS1a Non-zero M(ql) asymmetry may be observed with 30fb-1
ql-
A. J. Barr Phys.Lett.B596: 205-212,2004
Spin:
R. Ströhmer RAL-PPD Seminar 9.5.07 20
RS Gravitons & Heavy Bosons
pp
θ
gravitone
e
CharacterisationMeasure spin
- G* Spin 2- Z’ Spin 1
5 sigma discovery
eeZATLASpreliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 21
Effect of initial alignment
generator initialalignment
CMS
1 TeV Z’
Track-based alignment using minimum bias, Zee,
initial alignment
alignment after few fb-1
R. Ströhmer RAL-PPD Seminar 9.5.07 22
Search for Heavy Stable Particles
Predicted in various modelsLong lived stau as next to lightest particle in GMSB
R-hadrons in Split-SUSY- Colored SUSY particle
hadronizes (e.g. gluino)
Determination of mass from momentum and velocity (β)
β can be determined in the range from 0.6 – 0.8 by:
- Energy loss in the tracker- Time of flight in the muon
system
CMS preliminary
CMS preliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 23
Search for Heavy Stable Particles
R–hadrons have hadronic interactionEnergy/momentum mostly carried by SUSY particle
Hadronic interactions will change the charge of the R-hadron
ATLASpreliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 24
Search for Heavy Stable Particles
Event selection β(dE/dx) < 0.85
- to exclude MIPs
0.6 < β(dE/dx) < 0.80.6 < β(TOF) < 0.8 m(dE/dx) > 30 GeV
- to reject slow standard model particles
number of tracker hits > 10- to eliminate fake tracks and
optimize the quality of dE/dx
pt cut at:- 150 GeV (300 GeV gluino)- 200 GeV (600 GeV gluino)- 80 GeV (152.3 GeV stau)
expect< 25 BG events at L= 500 pb-1
- (zero unweighted MC events)
Combined β resolution for stau
CMS preliminary
gluino 300 GeV
gluino 600 GeV stau 152.3 GeV
500 pb-1
500 pb-1
30 pb-1
R. Ströhmer RAL-PPD Seminar 9.5.07 25
Search for Long-Lived Neutralinos
Event selectionIsolated photon pt>80 GeV
4 Jets pt> 50 GeV
missing energy > 160 GeV- not in Jet direction (Δφ>20o)
• Lifetime reconstructionphotons from neutralino with finite lifetime are not pointing to primary vertex.
shape of energy deposition in calorimeter depends on photon direction.
sensitivity to log(c)
CMSpreliminary
CMSpreliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 26
Universal Extra Dimensions (UED)
mass- degenerate spectrum
Conservation of KK parity (-1)n
- n=1 similar to SUSY but Spin(KK) = Spin(SM)
2nd excitation can be singly produced
Pair production of g1 g1 ,q1 g1 and q1 q1
Signal: 4 leptons (2 pairs OSSF), jets, andmissing energy
Long decay chains
R. Ströhmer RAL-PPD Seminar 9.5.07 27
Z veto: one OSSFwith M<5 GeV or M>80 GeV
Universal Extra Dimensions (UED)
µ: pt> 5 GeV, |η|<2.4e: pt> 7 GeV, |η|<2.5
CMSpreliminary
CMSpreliminary
R. Ströhmer RAL-PPD Seminar 9.5.07 28
Spectacular States : Micro Black Holes
Large EDsMicro black hole decaying via Hawking radiation
- Photons + Jets + …
We will certainly know something funny is happening
- Large multiplicities- Large ET- Large missing ET- Highly spherical
compared to BGs
Theory uncertainty limits interpretation
- Geometrical information difficult to disentangle
CMSpreliminary
Invariant mass [GeV/c2]
sphericity
R. Ströhmer RAL-PPD Seminar 9.5.07 29
Conclusions
ATLAS & CMS have significant discovery potential for physics beyond the standard model
New physics could already show up in early data
In order to claim a discovery on needs to understand the background - detector performance- standard model processes
discovery of “something” is only first step, the second is to distinguish between models and determine parameters