2. Jorgen D’Hondt University of Brussels - DELPHI Collaboration l WW  qqQQ events collected at LEP2 l Colour Reconnection effect at LEP2 l Model dependent measurement(s) l Preliminary conclusions XXXII International Symposium on Multiparticle Dynamics 1/12

3. Known example : hadronic decay of B mesons B  J/  + X B  J/  + X Main interest : Probe the interaction of two fragmenting strings Causes largest systematic error on W-mass G.Gustafson, U.Petterson and P.Zerwas, Phys.Lett. B209 (1988) 90 e-e- e+e+ 2/12 DELPHI detector at LEP2 4 jet topology BR.exp ~1% BR.fullCR ~3-5% BR.noCR ~0.3-0.5%

4. 0.1  W decay vert. ~ 0.1 fm distance 1  hadronisation scale ~ 1 fm  large space-time overlap  suppressed in parton shower :  due to group structure of QCD at least 2 gluons must be emitted  W width restrict the energy range of primary gluons from q 1 Q 4 and q 2 Q 3 ~ (  s 2 /M W )  W /N c 2  CR effect could occur in the confined region ???  s <<1 s1s1 Models to emulate the CR effect T.Sjöstrand and V.A.Khoze, Z.Phys. C62 (1994) 3/12

5.  PYTHIA : string reconfiguration if they overlap or cross in space-time –SK1 p CR = 1-exp(-  ·O) –SK1 (lateral flux tube) : via event string overlap O : p CR = 1-exp(-  ·O) –SK2 (vortex line with core) : reconnection if cores cross –SK2’ : SK2 + only if string length is reduced  ARIADNE : rearrangement of colour dipoles to reduce the string length (mass) –AR2 : only after soft gluon radiation (E g <  W ) –AR3 : allowed everywhere (also in pertubative phase)  HERWIG : rearrangement of colour dipoles changing the size of the clusters CR implemented in existing fragmentation models : For each model one can study the effect on the structure of the WW events with Monte Carlo simulation and compare it with data. SK1 m W shift P CR K ~ 0.66 Latest preliminary predictions for the W-mass 50 PYTHIA (SK1) ~ 50 MeV/c 2 70 ARIADNE (AR2) ~ 70 MeV /c 2 40 HERWIG ~ 40 MeV/c 2 30 Statistical uncertainty (LEP2) ~ 30 MeV /c 2 4/12

6. Particle flow W-mass measurements CR is an important systematic uncertainty on m W measured in WW  qqqq events, but we designed another m W estimator which does not have this feature... Design an observable to measure a possible enhancement of particles in the inter-W regions… New!   design an observable sensitive to for example  (SK1) CRCR The correlation between both observables was found to be negligibly small simply count particles (background sensitive) jet kinematics of the event (background insensitive) Eff. ~ 12%, Pur. ~ 87% Eff. ~ 90%, Pur. ~ 71%...all results are preliminary sensitivity ~ 2.7  (for full SK1) sensitivity ~ 4.3  (for full SK1) 5/12

7. # WW  4q selected (m W observable) ‘00 ‘99 ‘98 ‘97 ‘96 design limit  WW  18 pb indirect Up to 5000 WW  qqQQ events expected by DELPHI background WW event up to 208.8 GeV 6000 W-mass : # ~ 6000 (used data ‘98-’00) 720 Particle flow : # ~ 720 (used data ‘97-’00) 6/12

8. m W is assumed to be unknown but it must be invariant for different estimators m W (std)-m W (cone) from MC as function of  Correlation between W mass estimators ~ 83 %  uncertainty on the difference is small SK1 Influence on W mass estimator (MeV/c 2 ) for different values of   (SK1) MC prediction DELPHI preliminary 7/12

9. Data @189-209 : MeV/c 2 Monte Carlo prediction if there is Colour Reconnection  centr. 1 sigma 2 sigma different behaviour MC prediction SK1 first time this is done with LEP data !! 22 8/12

10. P i Project all charged particles P i in the plane ( jet 1, jet 2) A C D B Rescale the angles to have an equal amount of phase-space between the jets :   J,i   J,i. (1/  JK ) ABCD Inside W Between W’s Inside W and Between W’s Z  qq(  ) WW  qql data at 189 GeV #( A+C ) #( B+D ) () dd d 1 N ev #( A+C ) #( B+D ) SK1 100% no CR SK1 100% 0.20.8 9/12

11. In the systematic error, were considered: l Bose-Einstein effects (2%) l Fragmentation modelling (1%) l Background subtraction/modelling (0.5%) l Generators/Tunings (0.3%) = 0.900  0.031  0.021 Data @183-209 : R Monte Carlo prediction : ( extrapolated to 189 GeV ) 10/12

12. no correlation assumed between the two measurements (good approximation)  68% CL for  [ 0.3, 3.3 ] central value1.3 22 P CR 68% CL for P CR [ 14%, 75% ] central value44% 11/12

13. 12/12 WW  qqQQ l Colour reconnection models are investigated in WW  qqQQ events l Two uncorrelated observables are designed : R z Integrating the particle flow between jets ( R )  m W [std,cone]  Using the kinematics of the jets (  m W [std,cone] ) small amount of Colour Reconnection l Preliminary results prefer a small amount of Colour Reconnection (also the LEP combined measurement, cfr. talk of Nigel Watson) ARIADNE model l The observables are not sensitive to the ARIADNE model !? Preliminary shift on the W-mass 50 PYTHIA (SK1) ~ 50 MeV/c 2 70 ARIADNE (AR2) ~ 70 MeV/c 2 40 HERWIG ~ 40 MeV/c 2 30 Statistical uncertainty (LEP2) ~ 30 MeV/c 2 Therefore one cannot decrease the systematic uncertainty on m W More work needed on ARIADNE !!