Search for Intermittency in Central Collisions of OXYGEN-16 + Copper at 14.6 A-Gev

Central collisions between 14.6 A -Gev/c oxygen projectiles and copper nuclei are analyzed for evidence of intermittent fluctuations. Since the principal detector system used in this analysis (the E802 Target Multiplicity Array at BNL) measures only the laboratory angles theta and phi of charged tracks, simulations are performed (using the alpha model of intermittency) to confirm that intermittent fluctuations in rapidity can be detected by measuring particle densities in pseudorapidity. It is shown, however, that if fluctuations in rapidity are intermittent the strength of these fluctuations will likely be underestimated when pseudorapidities rather than true rapidities are measured, especially when non-intermittent sources of target spectators are present. Comparisons are also made between central ^{16 }O + Cu data and detailed simulations based on the LUND event generator FRITIOF. To achieve a more accurate imitation of actual data than can be obtained with FRITIOF alone, the kinematics of target spectator nucleons are added to the FRITIOF events. Full events (including spectators) are then passed through the GEANT package for simulation of secondary physical effects and detector response. It is found that this LUND-based simulation is able to predict the measured inclusive spectra quite well. Central ^{16}O + Cu data is then analyzed for intermittency using techniques developed by Bialas and Peschanski. It is found that the scaled factorial moments for 2 <=q i <=q 6 exhibit a power-law dependence on the resolution. On log-log plots, the moments are observed to have good linearity with no apparent tendency to saturate or level off down to deltaeta = 0.1. It is also found that the magnitude of the slope parameters varphi_{i} increase with the moment order i. All of these observations are consistent with intermittency, and it is shown that these effects cannot be attributed to the inclusive shape of events. Correlations between different bins of eta are also measured; though some correlations are found they are so weak that the results cannot be considered conclusive. Finally, a large sample of events generated with the LUND-based simulation are also analyzed using the same techniques, and it is found that in this case the simulation fails to reproduce the effects observed in the data.