News on mean pion multiplicity from NA 61 / SHINE

NA61/SHINE is a large acceptance fixed target experiment at the CERN SPS which studies final hadronic states in interactions between various particles and nuclei [1]. The main topic of this contribution are preliminary results for mean negatively charged pion multiplicities 〈π−〉 from central Ar+Sc and Be+Be collisions. The data were taken recently by the NA61/SHINE collaboration for a wide range of beam momenta. Measured rapidity distributions dn dy were extrapolated to unmeasured regions to obtain total multiplicities 〈π−〉. A new scheme to calculate the mean number of wounded nucleons 〈W〉 utilizing the EPOS MC model is described. Using data from other experiments, a comparison of 〈π〉 〈W〉 for different collisions and beam momenta is discussed. 1 π− rapidity distributions The starting point of the analysis described herein are double differential spectra d 2n dydpT of negatively charged hadrons, where y and pT are rapidity and transverse momentum of partcles, respectively. Centrality was determined by selecting the 5% of collisions with the smallest forward going energy as measured by the Projectile Spectator Detector (PSD) [1]. The spectra were obtained from reconstructed tracks applying a series of quality cuts. In order to correct for trigger and reconstruction inefficiencies, one needs to apply a Monte Carlo correction. To this end, the EPOS MC [2] is used in NA61/SHINE. A large statistics of ion collisions is generated and particles are accumulated in bins n j gen in transverse momentum pT versus rapidity y. The generated data undergo the regular reconstruction procedure. Selecting negatively charged pions results in the distribution n j sel. The correction factor c i, j is then calculated as the ratio of the two Monte-Carlo generated spectra ci, j = n j gen/n i, j sel. The final experimental spectra are obtained as n i, j = n j datac i, j. In order to estimate the mean π− multiplicity in the full acceptance, one needs to extrapolate the experimental data to unmeasured regions. The extrapolation process consisted of two steps extrapolation in transverse momentum pT for each bin of rapidity y and extrapolation of dn/dy in rapidity. For the latter a sum of two Gaussian functions was fitted, g(y) = gT(y) + gP(y), where gT(y) = A0Arel σ0 √ 2π exp − (y − y0) 2σ0  , gP(y) = A0 σ0 √ 2π exp − (y + y0) 2σ0  In order to calculate the mean negatively charged pion multiplicity 〈π−〉, the following formula was utilized: ?e-mail: mnaskret@cern.ch ar X iv :1 71 1. 00 31 8v 1 [ he pex ] 1 N ov 2 01 7 〈π−〉 = ∫ ymin −4 g(y)dy + ymax ∑


π − rapidity distributions
The starting point of the analysis described herein are double differential spectra d 2 n dydp T of negatively charged hadrons, where y and p T are rapidity and transverse momentum of partcles, respectively.To identify negatively charged pions, the "h − " method (see details in Ref. [2]) is used.Centrality was determined by selecting the 5% of collisions with the smallest forward going energy as measured by the Projectile Spectator Detector (PSD) [1].The acceptance is as high as 2 GeV/c in p T and 3.5 in y.
The spectra were obtained from reconstructed tracks applying a series of quality cuts.In order to correct for trigger and reconstruction inefficiencies, one needs to apply a Monte Carlo correction.To this end, the EPOS MC [3] is used in NA61/SHINE.A large statistics of ion collisions is generated and particles are accumulated in bins n i, j gen in transverse momentum p T versus rapidity y.The generated data undergo the regular reconstruction procedure.Selecting negatively charged pions results in the distribution n i, j sel .The correction factor c i, j is then calculated as the ratio of the two Monte-Carlo generated spectra c i, j = n i, j gen /n i, j sel .The final experimental spectra are obtained as n i, j = n i, j data c i, j .In order to estimate the mean π − multiplicity in the full acceptance, one needs to extrapolate the experimental data to unmeasured regions.The extrapolation process consisted of two stepsextrapolation in transverse momentum p T for each bin of rapidity y and extrapolation of dn/dy in rapidity.For the latter a sum of two Gaussian functions was fitted, g(y) = g T (y) + g P (y), where In order to calculate the mean negatively charged pion multiplicity π − , the following formula was utilized: Thus the final result is the sum over measured values of dn/dy in the acceptance region and the integral over the Gaussian fits outside.Statistical uncertainties were calculated and systematic uncertainties were assumed to be 5% based on the previous NA61 analysis of p+p collisions [4].

The mean number of wounded nucleons
The number of wounded nucleons can not be measured experimentally in NA61/SHINE.It has to be calculated using Monte Carlo models.Two models were used to perform calculations -Glissando 2.73 [5] based on the Glauber model and EPOS 1.99 (version CRMC 1.5.3)[3] using a parton ladder model.Glissando provides values that are consistent with previous measurements and applicable to the wounded nucleon model [6].EPOS, on the other hand, allows for more detailed centrality analysis and event selection.It is possible to reproduce Glauber-based W values in EPOS and they are in good agreement with Glissando as shown in Fig. 1, where both -statistical and systematic uncertainties are calculated and combined.Therefore, Glauber-based EPOS values are used in later considerations.

Results
Preliminary π − rapidity spectra for Ar+Sc and Be+Be collisions are presented in Fig. 2 Preliminary results on π − and W were calculated according to procedures described in sections 1 and 2. These are presented in Table 1 for Ar+Sc and Be+Be collisions.
In order to compare results obtained for different systems, an isospin correction should be taken into account.To this end phenomenological formulas are used The correction is only applied to measurements where its effect is the strongest.This assumption is based on the compilation of the world data presented in [7] and the model presented therein.Where needed, the data is corrected for slight differences in beam momentum.Applying this correction one  onset of deconfinement is predicted by the SMES [8] model due to the larger number of effective degrees of freedom in comparison to the hadron resonance gas (HRG) model.As for the NA61 Ar+Sc, Be+Be and p+p data only the value of π − was obtained, the multiplicities of π + and π 0 are approximated by multiplying the previously isospin asymmetry corrected π − multiplicities by a factor 3: π = 3 π − .This approach is motivated by the fact that the NA61/SHINE acceptance is the largest for π − .The energy dependence of the π / W ratio is presented in Fig. 4. The Ar+Sc results follow p+p for low energies whereas for high energies they follow Pb+Pb.No simple systematics are observed at low SPS energies.This might be caused by different physics processes and/or systematic bias in the estimate of W . Full simulation of the fragmentation process and PSD response is needed in order to reduce the latter uncertainty.

Figure 1 .
Figure 1.Average number of wounded nucleons W calculated by Glissando and EPOS "a la Glauber".

Table 1 .
Preliminary π − rapidity spectra for the 5% most central Ar+Sc and Be+Be collisions.Preliminary reslults on π −and W for the 5% most central Ar+Sc and Be+Be collisions.