The in-progress scientific analyses of this project use data recorded by LHCb experiment [1], one of the four large experiments installed at the LHC (Large Hadron Collider) in international laboratory at CERN, Geneva, Switzerland.

The first stage of the project in 2012 was dedicated to building the research team and to the incipient scientific studies for the implementation of further analyses. The initial team of four researchers mentioned in the project proposal was enlarged with an IT engineer offering technical support (on software deployment and networking), a postdoc with solid and long previous scientific experience in the LHCb collaboration as PhD student followed by a postdoc grant from the Heidelberg University and Max Plack Institute in Germany, a Ms. Sc. student who during the year was admitted as student of the doctoral school at the University of Bucharest and another Ms. Sc. student. At present, the recruiting process is continued even if it proved troublesome and time-consuming both from technical and administrative reasons but also because of the degree and quality of physics education of the interviewed candidates. The PhD. and Ms.Sc. students, preliminary selected based on the submitted applications and recommendations, were required to pass a series of tests designed by members of the group to verify their abilities to understand and use the scientific literature and their programming skills. The postdocs and the candidates for the technical support position were only interviewed. The English language skills were tested for both types of candidates. Following repeated announcements of the available positions, 12 candidates were invited to the final interview/test. Two of the candidates who passed the tests subsequently refused the positions. Other two candidates recalled their applications as they found job offers that they considered more advantageous. The two recruitees received a training program to follow in order to familiarize themselves with the software tools used by the LHCb collaboration. During one of the periodic collaboration meetings, they were accompanied by members of the group to CERN where they enlisted as student members of LHCb, they visited the control room at the experiment site and by attending the meeting's plenary sessions they had the opportunity to learn of the physics topics other members of the collaboration are working on as well as meet some of the group's close collaborators from LHCb international institutes.

The scientific activity focused on strange particle production studies in order to offer a better understanding of the strange hadron production mechanisms in the light of the theoretical models that describe them. The ultimate goal is to contribute to the effort of optimizing the parameters which control the existing models and/or help in designing new models. Thus, some of the time spent for working stages at CERN was dedicated to discussions with the theoreticians involved in creating production/hadronisation models in order to identify the best kinematic observables for such studies. The final conclusions were presented to other LHCb members activating in the same working group at CERN to ensure synergy between the proposed measurements and the evolution of the existing software tools and that sufficient statistics for measurement will be available by including the necessary pre-selection and trigger lines to retain it during raw data processing. Taking into account that the PYTHIA [2] generator, based on the string fragmentation model [2, 3], is currently the most used generator in the field, several meetings were scheduled at CERN with authors of this software package. Apart the already contracted studies - the correlated differential particle production as function of rapidity, y, and transversal momentum, pT, further analyses were proposed and their feasibility was tested in the frame of the LHCb experimental constraints.

One such endeavour is the measurement of the correlation between strange mesons and baryons K and Λ, F(y,pT), produced in the same hadronic jet [3,4]. In this case, y and pT will not be defined in the centre-of-mass system of the colliding pp pair but with respect to the main axis of the jet containing the two particles. This approach comes at a good moment when the jet reconstruction procedures at LHCb reached the level of maturity with enables high precision measurements to be made[5].

As an alternative to defining a reference axis one can use instead of the jet central axis the moving direction of the particle with the largest measured pT in the event. This measurement raises some difficulties for the particular setup of the LHCb experiment as a forward spectrometer because the trajectory with the highest pT detected in the detector acceptance has a smaller probability to the associated to the highest pT particle from the primary pp interaction than in the case of the (barrel-like) detectors covering the whole solid angle ($4 \pi$). Yet, not considering the drop in statistics, the benefits would be important. A very particular and optimal case is the situation when a strange hadron proves to be the reference particle. In this case, the strange hadron may represent the „leading particle” [3,4] in a jet and therefore will contain partons from the original string which are directly produced in the pp interaction before the process of string fragmentation/hadronisation started. These ideas lead to conclusions that emphasize the importance of the already intended studies of strange meson to strange baryon correlations and strange baryon correlations. The analysis of (K⁺ Λ), (K^- anti-Λ) and (Λ anti-Λ) correlated production is an adequate environment for testing correlation models which imply the creation of diquarks and the correlated production of quark-diquark and diquark-diquark, all these partons originating from the same string [3,4,6]. The production of baryons containing more than one strange quark, from here on called multistrange, e.g. Ξ and Ω, correlated to another strange baryon or meson can be also analysed in the context of diquark-quark or diquark--anti-diquark correlations. Another class of models that can be tested is based on a "pop-corn"[2] type de-correlation mechanism. The multistrange baryon study allows testing the production of diquarks containing two strange quarks. It is postulated that the production of such diquarks is very low in the fragmentation process and therefore the study requires special reconstruction techniques to be elaborated in order to highlight this behaviour. These conclusions drawn from the private meetings with CERN theoreticians, show that the LHCb measurements foreseen in the frame of the present project will test the physics processes at diquark scale (femto-meter) directly by studying the strange meson-hadron, meson-baryon and baryon-baryon correlations.

Further discussions with the theoreticians and attending theory seminars inspired the planning of other interesting measurements such as the analysis of the correlated production of strange hadrons with respect to the number of particles produced in the event (the multiplicity of the event) and the angular production correlations of two strange hadrons. We had significant progress in the study of K⁺ K^- correlation. This analysis has the advantage that the strange mesons are produced more abundantly than the strange baryons. The difficult aspect comes from the requirement to select a sample of K^± of high purity which must contain only kaons created promptly in the pp interaction or by decays of very short lived resonances. In order to discriminate K^± from other particle species, the LHCb experiment uses the information provided by two RICH (Ring Imaging Cherenkov) [7] detectors. The default LHCb K^± selection was further tunes to obtain the necessary sample purity. The improvements focused on eliminating the contamination of the sample with other particle species, the fake "ghost" trajectories due to detection errors and the kaons produced in subsequent decays of resonances with medium and high life-time. The studies performed on simulated data show an increase in the purity of the K^± sample from 27% up to 92%-95%. The kaons produced in decays of other mesons and baryons, e.g. ф, represent under 2.5% of the final sample of daughter particles K⁺ + K^-. Taking into account the high purity of the sample this contamination can be neglected in the further study. However, the strong cuts applied allowed only 39% of the initial K^± to be found in the sample after optimization. The measured data used in the preliminary study of the K^± correlated production were registered at a center-of-mass energy of 7 TeV. To ensure that the analysed K^± pair comes from the same pp interaction, we used data samples measured during the first LHC runs (2010) for which the primary interaction rate is low so that most of the events include only particles originating from a single pp interaction.

An important step of the analysis is to correct the observable distributions obtained for the detected particles taking into account the reconstruction efficiency, the acceptance and the difference in particle identification efficiency between real data and simulated data samples. The reconstruction and acceptance efficiency are determined using simulated data in the hypothesis that the interaction between the particles and the detect material si correctly reproduced in these data. The models for generating the primary particles are tuned using measurements from previous experiments, at low interaction energies and at central (pseudo)rapidity. The predictions of these models, extrapolated in energy and pseudorapidity (η) cannot reproduce satisfactorily the η, y and pT distributions obtained from the experimental data. For LHCb the geometric/kinematic acceptance is frequently expressed in terms of η and pT. The main differences between the simulated and experimental data samples are linked to the identification efficiency of the particle species and to the medium/maximum value of the transversal momentum of the K^± mesons. The real data distributions have higher values in pT. To account for these differences various corrections are applied. Two types of analyses of the K⁺ + K^- pairs are of particular interest: the measurement of the production cross-section for a K⁺ + K^- pair and the correlations in the production of kaons with opposite electrical charges (i.e. where the strange quarks may come from a s anti-s pair produced during the fragmentation process). To emphasize this correlation one needs a sample containing uncorrelated K⁺ + K^- pairs. Such a sample can be constructed using K^+/- from different events. Such pairs are given weights so that the kinematic distributions from the primary data samples and these artificially generated uncorrelated data samples match. To compare the results obtained using the two data samples, one has to consider that for events with the same multiplicity, the reconstruction and particle identification efficiency depends on η and pT. Thus, the kaon candidates from the simulated data sample get additional weights in order to obtain similar single particle distributions both in real and simulated data. The ratio between two uni-particle distributions as function of η and pT defines the correlation function (and will be called as such from here on). The distributions are not yet corrected for effects due to the K^+/- tracking and identification efficiencies. Comparing the uncorrected real data and simulated data distributions one would observe that in general the measured distributions are closely followed the simulated ones yet there are areas in the phase-space where the experimental points diverge in modulus with up to 15% from the correlation functions obtained from the simulation. In the transversal momentum difference plots, the simulated distributions are closer to the values obtained from the real data and one may notice the kinematic anticorrelations between kaons from the same event. These samples as well as the calibration samples described below contain a fraction of kaons generated from decays of ф mesons (as mentioned above). These secondary kaons are correlated and this fact is clearly visible in distributions including or excluding them from the analysed statistics. For these data the ф signal is more accentuated due to imposing filtering conditions at the same time for both kaons. This component is trivial to eliminate using constraints on the invariant mass of the K⁺ + K^- pair, yet it is useful for calibration and systematic uncertainties check-ups. The raw distributions using measured data must be corrected applying the reconstruction and identification efficiencies for various particle species for the LHCb detector. One should keep in mind that an important analysis could also focus on flavour/strangeness correlations between kaons produced by the same string (see the Lund model of hadronic fragmentation [2-4]) as well as other kinematic or dynamical correlations e.g., momentum conservation, jets, etc.

In order to obtain the kaon identification efficiency corrections we apply the "Probe and Tag" method which uses the kaons originating from ф mesons decays to calibrate the procedure for evaluating the particle identity. This method, described in detail in [7], must be applied in a particular way in the present case. "K-probe" is the kaon which fulfilled the filtering conditions applied on the reconstruction and identification parameters while the "K-tag" represents an unfiltered kaon which together with the "probe" kaon have a combined invariant mass near the mass of the ф meson. The phase-space for the "tag" kaons is split in (η, pT) cells and the events from the calibration samples containing ф mesons receive a specific weight so that distribution of the number of reconstructed tracks is identical with the ones corresponding to the events from the analysed samples. The method applied to determine the efficiencies is the "sWeights" method [8] which considers the ф meson mass distribution fits as "sPlots" in each cell from the (η, pT) plane. Evaluating the signal to background ratios associated to the mass fits, one obtains for each phase-space cell the kaon identification efficiencies for the RICH detector. At RICH level, in each cell, the filtering for "tag" kaons is identical to the one used for the kaons used in the correlation study. The events containing ф mesons for calibration are filtered using a pre-selection which is applied similarly to real and simulated data.

The huge amount of data produced in pp collisions prohibits the indiscriminate registration of all the data, so trigger lines are employed to detect and veto on the acquisition of events of interest. Even so, allowing each member of an experiment to process all the raw experimental data (at present the LHCb collaboration has 843 members from 60 institutes spread in 16 countries world-wide) would be a waste of time, work force and especially computing power. Therefore, the user-physicists are offered access only to data filtered by the triggering system which were further reconstructed and pre-selected to ensure obtaining enriched samples containing the particles and/or the events to study. Vast re-processing campaigns are foreseen to take place with low frequency (about twice a year) to allow the results of various studies of the detector resulting in new alignments and better detector description to be applied as part of the reconstruction and preselection algorithms. Due to the large quantity of events needed to produce the simulated data samples, this type of data is produced in campaigns well planned in advance. The models used to produce the simulated data need to be implemented in the collaboration simulation software (GAUSS). In conclusion the processing of the experimental data as well as the production of simulated data cannot be done using the computing resources of one single institute, instead computers from researched institutes and universities from various corners of the world and linked in a super-network called GRID and used for these purposes. Access to these global resources that allow parallel processing of data is essential for fulfilling physics measurements.

Such event enriching is not necessary for the study of kaon correlations since these particles are copiously produced in each event. A dedicated trigger line called Minimum Bias controls the registration of events that contain at least one track in the detector acceptance. From this very abundant type of events only a small fraction is measured and stored, and the subsequent preselection reduces the statistics even more proportional to the requirements of the analysis. Such an approach is not possible for the next stages of this project. Therefore we provided the LHCb collaboration with a set of stripping lines which ensure the production of strange baryon enriched event samples. For the Λ baryons two methods are available: (i) using the selected samples to calibrate the proton identification in the RICH detector or (ii) re-enabling the preselection line previously used by the collaboration for the single particle analyses. For the multistrange baryons Ξ and Ω members of our group joined the effort of testing the stripping lines which were implemented before the last reprocessing campaign, also aiming at enhancing their efficiency - one of the necessary conditions in order to perform correlation studies. At present, a study is being performed with the goal to establish (using simulated data) the number of events needed to obtain a given precision of the measurements performed in the correlation studies involving strange baryons. This kind of planning is also very important for the analyses on hadrons containing the "beauty" quark since these are produced with much lower probability. As the LHCb experiment was designed in principal for the study of such particles, we envisage to use the already existing preselection lines, the challenge being to choose the lines corresponding to as loose as possible constraints on the phase-space (given the complexity of these selections which is far superior to the ones applied to filter simple strange particles, this intermediary goal is essential to be obtained). Such lines would allow us to study correlations between opposite charge B mesons. Therefore it is imperious to find selections which use similar methods so that the baryon/meson samples retain this similitude with respect to the kinematic observables and allow closely related correction algorithms to be applied, thus facilitating the evaluation of the systematic uncertainties. A study to identify such preselection lines was also initiated.

One of the purposes of these measurements is to provide information and data in order to optimize the existing hadronisation models and implicitly the generators implementing them but also to create new models and/or new generators. The first step on this road consists in comparing the existing models and tunes with single particle production measurements performed by the experiments at LHC. We tried to identify a reference tune, defined as being the specific tune that best describes not only the observables for the strange hadrons measured by LHCb but at the same time the experimental data from the other compatible experiments at LHC. Therefore, we tested various PYTHIA tunes after consulting the scientific literature on this subject and applying the conclusions of discussions with the authors of this generator. Two different versions of PYTHIA were used, the old generation PYTHIA 6 coded in FORTRAN and the new evolving version PYTHIA 8 coded in C++. A series of PYTHIA 6 tunes pointed as adequate by the generator authors of members of other LHC collaborations were considered: Perugia0 - the reference tune for pre-LHC experiments, AMBT1 - a first tune proposed by ATLAS in 2010 and obtained on data collected at LHC during runs at 7 TeV centre-of-mass energy and using Leading Order Parton Distribution Functions (LO PDF), CMS Z1 - a further optimization of AMBT1 proposed by members of the CMS experiment using the CTEQ5L set of PDFs, Perugia 2011 - the last PYTHIA 6 proposed in 2011 as an improvement of Perugia 2010 using CTEQ5L[9] PDF set. Additionally the default PYTHIA 8 tune at the LHC energies, called "4C", was tested. The comparison between LHC data and these tunes was subsequently extended to include the specific LHCb MC tune[10]. The experimental data measured by LHCb[11], ALICE[12] and CMS[13] at different collision energies were confronted to the distributions obtained using the PYTHIA tunes above as input for the corresponding RIVET[17] plugins (see Annex 4 for a few such comparisons generated using the stand-alone versions of the PYTHIA generators and the RIVET miniframework). Unfortunately, it was found that no one of the tested tunes matches satisfactorily the specific observables for the considered measurements. As a continuation we'll analyse the distributions associated to the various tunes taking into account the main differences between the values of their control parameters in order to interpret the results and the discrepancies from the experiment. This strangeness production based study will contribute to the interpretation of the foreseen correlated production measurements, once they will be finalised. The research of the scientific literature is under way to identify the models of interest for the correlated production of hadrons containing a b quark. We are determining the feasibility of implementing these models in the LHCb simulation software packages and create new RIVET plugins in order to benefit from the existing methods of comparing independently MC tunes and experimental data.

On the same topic, since we consider advantageous that the results of our measurements can be interpreted and easily used by the theory community and in order to keep up to date with their research as well as to establish new contacts, we organized a workshop in Bucharest at IFIN-HH on 22nd and 23rd of November 2012. We invited both theoreticians and collaborators from the LHCb experiment (or other experiments at the LHC and not only) who are interested in implementing and tuning Monte Carlo generators. The scientific programme (available here) was structured in sessions containing key talks dedicated to different models/generators and the software used to tune and validate them, followed by specific talks about the way these generators are implemented or integrated into the software framework of the experiment and/or used by LHCb collaborators. Generous time was allocated to discussions. These discussions dealt with the way simulation is used by analysts emphasizing their request for ease of access to the software in order to obtain the simulated data, the possibilities one sees in order to optimize the generators making use of the measurements done by the experiments. A clear gain was the opportunity both for the members of the project team and the other experimentalist attendees to be informed of the latest improvements and the plans of development for the main generators used at present in particle physics by the experiments installed at large particle accelerators, namely PYTHIA[2], SHERPA[14], HERWIG[15], EVTGEN[16]. Issues related to generator validation and tuning were also considered and also the current differences between model predictions and experimental points were discussed. The two workshop days were very intense and useful for all the participants. A fruitful exchange of informations took place between theoreticians and experimentalists which we hope will crystalize in future synergy of these topics.

Progress was made in implementing and executing the selection algorithm of baryon particle pairs (Lambda, anti-Lambda) and in charged Kaon pair selection algorithm. These selections were applied on real 2010 data at a total center-of-mass collision energy of 7 TeV, and on the corresponding Monte Carlo samples, samples which are compatible with the 2010 real data. The meson+baryon correlation studies, i.e. charged Kaon plus Lambda hyperon, remain to be implemented in a similar fashion.

In the next two months, we plan to extend the analysis on the 2011 and 2012 data at various LHC proton-proton collision energies: 8 TeV and 2.76 TeV, respectively. In parallel with the previous 2011-2012 data analyses, we will continue the effort for background reduction and reconstruction efficiency estimation. The latter can be obtained directly from the data, by using the calibration decay channels of various particles. With tracking efficiencies close to 100 %, this is especially useful in the case of RICH ("Ring-imaging Cherenkov detectors") detectors particle identification efficiencies, which give the dominant effect on particle reconstruction and identification loss.

The LHCb collaboration gives access to its members at various services, chiefly among which is, besides data access, the Monte Carlo data mass production facility on the GRID. These MC data are generated, digitized and reconstructed by LHCb collaboration experts. Our group has forwarded several requests and is waiting for the finalizing of the corresponding MC production. These requests are mostly for "Minimum Bias" data samples, data which are ideally suited to conduct efficiency and generation studies in case of QCD processes that take place for small parton-parton collision energies when compared to LHC collision energy. These data are also useful in describing the fragmentation/hadronization processes, which are exactly the processes that are directly probed during our correlation studies. Up to now, the MC data used were Monte Carlo data produced for 2010 real data, and small private samples generated and reconstructed by our group using dedicated reconstruction and simulation software packages of the LHCb collaboration. We foresee that we shall have available the new MC samples within the next months, to use in reconstruction efficiency estimation and LHCb detector acceptance unfolding for the 2011-2013 real data.

Given the LHCb policy of data access, which place some restrictions on the access to LHCb reconstructed data, it became imperious to implement and commit a preselection algorithm as part of the official LHCb software. The algorithm will be used on 2011 and 2012 data. This will enable the increase by at least an order of magnitude for the number of correlation candidates (K^-,anti-Λ), (Λ,anti-Λ), (K⁺,K^-) which have the total strangeness number equal to zero. Here the work on the preselection algorithms will go forward till the end of July. The work opened a potential collaboration with the members of another LHCb group, which are interested in the Bose-Einstein and Fermi-Dirac types of correlations between particles found in the same data samples.

Concerning the production studies of particles with associated strangeness number higher than unit, i.e. baryons like Xi and Omega, the preselection algorithms mentioned above were implemented and runs were executed already on 2011 and 2012 data. The found candidates - at 7 TeV, 2.76 TeV and 8 TeV proton-proton collision energy - were analysed and the signal was separated from the background. The next phase of this analysis will include a detailed analysis of each event in which Xi and Omega candidates were found and searches will be conduced for strange baryons and mesons: Lambda and charged Kaons, or other Xi/Omega candidates. These studies will help gauge the feasibility of initiating a multi-strangeness correlation study between Omega/Xi particles and other or same strange species of particles.

In addition to the previous program, an alternative program is being implemented. An extra constraint will be imposed in preselection: the events should contained a well defined jet of particles and the strange mesons and baryons have to be part of it. The jet requirement can be replaced alternatively by the requirement of a high transverse momentum particle in conjunction with the strange hadron pair. In this case, the energy cut in the transverse domain will be placed between 3 and 5 GeV. The studies will encompass the kinematic parameters of the strange hadrons with respect to the kinematic parameters of the jet or of the high-pT particle in the event. The jet structure and presence will pinpoint a well defined hadronization event within the LHCb phase space. The observables measured for a jet with two correlated strange particles will provide insights into the fragmentation process which generated them. If the second preselection requiring the presence of a highly energetic particle in the event was already tested, the jet reconstruction with its implementation within the preselection LHCb algorithms together with the strange particle pair constraint is currently being implemented. Given the high complexity in terms of the jet physics and of detector quantities involved, we hope to use the latter preselection algorithm by the end of this year.

At the end of this report on the status of the QCD studies at low energy which generate correlations between strange particles through s-sbar parton correlations, we underline the immediate scope of the analysis being the comparison between the LHCb measurements of correlation observables and the quantities obtained from the PYTHIA Monte Carlo generator. The eventual observed differences have the potential to allow for an optimization/retuning of hadronization processes within the PYTHIA code. In this sense we shall benefit from the the recent addition to the software functionality of interfacing the RIVET package - "Robust Independent Validation of Experiment and Theory" is a toolkit for the validation of Monte Carlo event generators RIVET web link - to the LHCb software simulation code. The latter interface developing and implementing is a main responsibility for one of our team members. Additionally the RIVET package will be used in conjunction with the "PROFESSOR" tuning tool for Monte Carlo event generators PROFESSOR web link to perform various similar tunes of PYTHIA with and without LHCb simulation software. The previous work shall be done in direct collaboration with an external scientific peer, member of the LHCb collaboration. The final outcome of the different tuning scheme will be to reduce the difference between the generator results and real data measurements, with the latter extrapolated back to the generation level form the initial raw detector measurements. This will be done with a proper choice of control parameters for tuning of PYTHIA generator. Since some of the parameters mentioned previously are the ones directly controlling the strange parton correlations and the strange meson versus strange baryon ratios, this will allow to interpret the LHCb correlation results within these theoretical production and hadronization models and parametrizations.

By the end of this year, we expect to have obtained from the 2010, 2011 and 2012 data, the raw measurements of the correlation functions and distributions and extrapolate them to the moment just after the hadronization. In computing the initial distributions of the correlated strange particles, we shall make use of the LHCb reconstruction efficiency and acceptance mapping which are currently being computed for the candidates and the events of interest. E.g., when the magnetic effects on the track curvatures are taken separately, the LHCb geometrical acceptance is typical for a single-arm-spectrometer, with the polar angle limits with respect to beam-axis of about 15 to 300 (250) milliradians - 300 for bending plane limit and 250 for non-bending plane limit, respectively.

Stripping v20r0p2 data stripping lines

The following stripping lines were proposed and implemented by the Electro-Weak(EW) physics group. As members of the EW subgroup working on soft QCD our team corrected, adjusted and released the lines in red frame in the table below (the table was adapted from this restricted access page - a snapshot of the page in .PDF format is available here*).

Soft QCD

Proton Ion

MC Sample Generation Requests