1. HERA-LHC Workshop, DESY, 22th March 2005
Forward studies with Cascade at LHC energies E. Rodrigues, Niels Tuning NIKHEF, Amsterdam HERA-LHC Workshop, DESY
* LHCb experiment, the “must know”
* Why Cascade might be useful for LHCb?
* Comparisons with Pythia
- general event kinematics and variables
- tracks and jets
* Conclusions
NB: Updated plots after
the presentation!
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

2. HERA-LHC Workshop, DESY, 22th March 2005
LHCb experiment
LHCb
Goal: B-physics studies
CP violation
rare B-decays
Acceptance: 1.8 < h < 4.9
Luminosity: 2·1032 cm-2 s-1
Nr of B’s /year: 1012
20 m
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

3. HERA-LHC Workshop, DESY, 22th March 2005
LHCb experiment
Unique forward spectrometer
at high energy!
Excellent tracking
Excellent PID
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

4. HERA-LHC Workshop, DESY, 22th March 2005
How forward is LHCb?
How forward is LHCb?:
B hadrons mainly produced in forward region(s)
both B’s tend to be correlated
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

5. HERA-LHC Workshop, DESY, 22th March 2005
Why cascade?
DGLAP (Pythia) vs CCFM (Cascade):
Differences expected in forward region
Cascade and LHCb:
Estimate uncertainty of PYTHIA predictions:
Track multiplicity in LHCb affects:
Trigger
Tracking
Tagging
Study b-jets:
Jet-charge: b orb tagging
Jet reconstruction: bb invariant mass
Study QCD evolution itself:
Validity of DGLAP vs CCFM
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

6. DGLAP and CCFM evolution
DGLAP vs CCFM:
If ln(1/x) terms are large: DGLAP evolution
expected to fail
 Differences expected in forward region
DGLAP evolution suppressed when small phase space
for Q2 evolution
 select jets with Q2 ~ E2T,jet
CCFM evolution enhanced when large phase space
for x evolution
 select jets with xjet >> xBjorken (xjet=Ejet/Eproton )
DGLAP:
kT ordering:
kTn>>kTn-1>>kT1
BFKL:
x ordering:
xn<<xn-1<<x1
CCFM:
angular ordering:
θn<<θn-1<<θ1
NB: Cascade needs pdf’s: f(x,Q2,kT), ie. unintegrated over kT
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

7. HERA-LHC Workshop, DESY, 22th March 2005
Outline
In the following we will show CASCADE vs PYTHIA comparisons … in minimum bias:
Event kinematics
x1, x2, Q2
Event topology
multiplicity, E, pT, etc. (of tracks)
Jets
Multiplicity, ET, etc
PYTHIA “6.2 series” with LHCb tune (cf. P. Szczypka’s talk)
Cascade out of the box
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

8. HERA-LHC Workshop, DESY, 22th March 2005
Pythia versus Cascade
Only unintegrated gluons available in CASCADE
Compare CASCADE with following subprocesses in PYTHIA:
Inclusive QCD: (MSEL=2)
fg fg 50%
ggff 15%
gggg 25% > named “Pythia gluon” in the following …
90%
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

9. HERA-LHC Workshop, DESY, 22th March 2005
Event kinematics (1)
PYTHIA:
Q2 = PARI(18) = PT2 of hard process ?
x1,x2 = PARI(33), PARI(34) ?
(as used in the pdf’s)
CASCADE:
Q = pT ? [outgoing parton]
x1,x2 = (E+pz)/2Eproton ? [ingoing parton]
NB: x1<x2
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

10. HERA-LHC Workshop, DESY, 22th March 2005
Event kinematics (2)
PYTHIA
CASCADE
Large overlap with HERA kinematic region
 CASCADE: more equal x values
NB: x1<x2
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

11. HERA-LHC Workshop, DESY, 22th March 2005
Event distributions
Event distributions:
Multiplicity in 2<η<5
Acoplanarity of outgoing partons (Df)
<pT> of particles in 2<η<5
Max pT in 2<η<5 Df
CASCADE:
lower multiplicity
PYTHIA:
more back-to-back
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

12. HERA-LHC Workshop, DESY, 22th March 2005
Track distributions
Track distributions: η pT
No dramatic difference in 2<η<5 … still Pythia tracks less forward
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

13. HERA-LHC Workshop, DESY, 22th March 2005
Jets
Jets finding
In the laboratory frame
Using the KTCLUS algorithm in inclusive mode
Jets selection
distributions shown only for jets in the LHCb acceptance
ET,jets > 10 GeV … and also loose sample with ET,jets > 1 GeV
NB: following distributions correspond to ~ 1ms of LHCb running…
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

14. Jet distributions ET,jets > 1 GeV
Loose selection
ET,jets > 1 GeV
CASCADE:
lower jet multiplicity …(?)
-> ET and η dists.
are similar
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

15. Jet distributions ET,jets > 10 GeV
Harder selection
ET,jets > 10 GeV
CASCADE:
many events with no jets
-> ET and η dists.
are similar
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

16. Highest ET jet – select phase space?
Forward jets selection
E2T,jet1  Q2
to suppress DGLAP
xjet1 >> xBjorken
to enhance BFKL
 non-negligible phase space for low x phenomena
(e.g. cuts in HERA forward jet analyses)
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

17. HERA-LHC Workshop, DESY, 22th March 2005
Conclusions
LHCb can be interesting for CASCADE
Investigate DGLAP vs CCFM
CASCADE can be interesting for LHCb
Investigate uncertainties of present Monte Carlos
Interesting first comparison:
PYTHIA and CASCADE not wildly different, despite their different philosophy, though differences are present
More studies to follow in selected regions of phase space
where differences might be enhanced (e.g. multiplicities, …)
Some issues risen:
Are the comparisons fair? E.g. choice of Pythia sub-processes
Are the x, Q2 definitions consistent between MCs?
Caveats:
CASCADE out of the box:
parameters need tuning and investigation …
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

18. Why do we care about multiplicity?
Backup slide
Why do we care about multiplicity?
Trigger: CPU time
Multiplicity affects:
Trigger
Tracking
Tagging
Plots from:
Massi F.-L. Apr. ’03
Jeroen v.T. Thesis
Marco Musy: May ‘03
Tracking: eff and ghost rate
 Relative multiplicity
Track multiplicity
Nlong< εeff = 5.4 ±0.3
Nlong> εeff = 3.9 ±0.3
Tagging: εeff
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005

19. Reminder (slide from March 2004…)
LHCb:
can we use the reconstructed b-jet for inclusive triggering (at high level trigger)?
b-jet (charge) useful for flavour tagging?
what kind of jet algorithms are best to use in (very) forward region?
- is KTCLUS good (enough) ? Alternatives?
=> where can HERA help?
LHCb can simulate and make studies
HERA can also analyse data and make comparisons
HERA measurements: forward jet production, particles/energy flow in forward region …
… ongoing analyses / plans in ZEUS/H1?
-> areas where discussions / feedback / between both groups can be profitable …
Any interests from HERA on addressing these points?
Eduardo & Niels
HERA-LHC Workshop, DESY, 22th March 2005