Latest Results from the COMPASS Experiment

In this paper the latest results from the COMPASS experiment are presented. We show results from longitudinally and transversely polarised targets off which high energy muons are scattered. In addition the future plans of COMPASS as well as results of the beam test runs are also presented.

Test of the Bjorken Sum Rule g V can be obtained from neutron beta decay, g A g V = 1.2694 ± 0.0028   Contribution from 3 processes to the observed asymmetry is assumed: . A LL from Low Q 2 , high-p T Data cont.
COMPASS preliminary A LL s are compared with various ∆g /g hypotheses Disagreement is seen between proton data and theory Possibly more resummations needed in the theory

Charged Hadron Multiplicities
Hadron multiplicities dN h /dN DIS , give access to FF COMPASS extracted preliminary charged hadrons, π and K multiplicities in grid of (x, y , z)

Kaon Multiplicity Sum
Kaon multiplicity sum gives an easy access to S R D K S (z)dz For the iso-scalar target: , where:

Gluon Sivers Measurement
Idea & analysis method similar as in ∆g /g High-pT: three processes are considered LP,QCDC, PGF select high-pT sample in order to increase contribution from PGF. extract simultaneously the azimuthal asymmetries for the 3 processes Data 2003-2004 deuteron target Cuts: Q 2 > 1 GeV/c, p T,1,(2) > 0.7 (0.4) GeV/c, z1,2 >0.1 Sivers effect for gluon as well as other asymmetries are compatible with zero, within experimental errors

Extraction of Transversity h 1
Ingredients: 2h π + π − asymmetries for proton and deuteron data Certain symmetry assumptions about dihadron FF Extracted results from 2h case are compared with fit to the 1h Collins asymmetry

Transverse Momentum Dependent Multiplicities
Both intrinsic k T of quarks in the nucleon as well as p ⊥ of the fragmentation needs to be better understood Hadron multiplicities were extracted in 4D (x, Q 2 , z, p 2 T ) binning Main features: the 2-exp fits give reasonable fits to the data, 2nd exp become dominant event as low as p 2 T ≈ 0.6 GeV 2

1
e, µ: calorimeters and muon filters identification: π, K , p (RICH) p > 2, 9, 18 GeV/c respectively New Results on g p In 2011 COMPASS collected data with 200 GeV /c muon beam The increased beam energy (160 GeV /c → 200 GeV /c ) allows to access lower region of x for Q 2 > 1 (GeV/c) 2 COMPASS extracted data in x, Q 2 grid for both A p 1 and g p 1 Good agreement is seen between 2007 and 2011 data Next step: SIDIS asymmetries for h ± , K ± , π ± from All-p T Method

[→→
the fraction of the processes (R i ) and partonic cross-section asymmetries (a i LL ) are obtained from MC and parametrized by NN Idea: larger pT → larger R PGF → larger sensitivity to ∆g /g used for PLB 718 (2013) 922, now with a different method Presently, the A LO 1 and ∆g /g are extracted simultaneously from the same data set low pT data are needed as they are clean source of LP certain systematic uncertainties are reduced w.r.t.previous method some consistency tests of the model used for estimate of Rs, a i LL s are possible.∆g /g = 0.113 ± 0.038 ± 0.035, µ 2 = 3 (GeV/c) 2 , xg = 0.10 In the measured xg range ∆g /g maybe positive, similar conclusion is reached in recent DSS fits, which include RHIC data.see talk by K. Kurek for more detailsA LL from Low Q 2 , high-p T DataThe same idea as before: larger pT → larger R PGF → larger sensitivity to ∆g /gIn the low Q 2 region there is an additional complication: resolved photon processes NLO calculations in collinear pQCD by M. Stratmann, B. J?er and W. Vogelsang(EPJC 44 (2005)  533) see talk by K. Kurek for more details of g1 and SIDIS π is negative in the whole x region However, SIDIS K data prefer zero or positive value at moderate x values Impact of Kaon data strongly dependent upon the choice of strange FF -D K S LSS group reported that problem disappears if HKNS FF set is used instead of DSS.deuteron data of COMPASS, Q

→
see talk by F. Kunne for more details Longitudinal Structure of the Nucleon be seen Problem: COMPASS and HERMES data have different shape of the M K + +K − distribution as well as results do not agree for x ≈ 0.1 − 0Values of x and Q 2 Low x , high parton densities COMPASS (fixed target) strong correlation between x and Q 2 Q 2 < 1 GeV/c -non perturbative region Some models that allow a smooth extrapolation to the low-Q 2 and high-Q 2 regions (J.Kwiecinski et al., B.I. Ermolaev et al.) can be confronted with data More than 10 fold increase of the precision w.r.t.previous experiments Clear non-zero signal is seen for the proton data even at low x , contrary to the deuteron data, where the result is consistent with zero.related with partons intrinsic k T COMPASS, hep-ex/1408.4405,(subm.PLB) Non zero effect observed for π + and K +

→→
see talk by K. Kurek for more details Two hadron asymmetry gives access to the interference FF They were extracted for h + h − pairs as well as for π + π − , K + K − , π + K − , K − π + Good agreement with the theory predictions see talk by G. Sbrizzai for more details correlation between 2h asymmetry and Collins asymmetry is observed More general cross-section formula was written lp → 2h + X interaction It is interesting to study the asymmetries as a function of ∆Φ (φ + h − φ − h ) It turns out that one can predict 2h asymmetry from the 1h Collins asymmetry.Expected ratio approx 4/π → see talk by F. Bradamante for more details Beyond Sivers & Collins I Six other A LT and A UT asymmetries were measured Different combination of PDF, TMD and FF involved Asymmetries comparable with 0, but A sinφ S UT A sinφ S UT related to Sivers f 1⊥ and transversity h1 related to Boer-Mulders PDF and Cahn effect (i.e.kinematic effect due quark k T )

→→
published p 2 T dependent multiplicities from 2004 data in EPJC 73 (2013) 2531 However, issues in this analysis were detected, which can affect the overall x, y , z normalization of multiplicities up to 40%, but the shape as a function of p 2 T are not significantly affected.Erratum in preparationMulti-Dimensional AnalysesRapid theory advances asks for multidimensional analyses... Example: Sivers effect and the matter of TMDs Q 2 evolution see talk by B. Parsamyan for more details Polarized Drell-Yan Measurement @ COMPASS Polarized DY can give complementary information about TMDs to the SIDIS case It is expected that e.g .Sivers PDF changes sign in DY w.r.t SIDIS case Unpolarised studies will also be performed, Boer-Mulders function will be extracted DY will be measured on different Nuclear targets as well as different (π − , K − ) beams COMPASS will measure DY events for M µ + µ − ∈ (4 − 9) GeV/c 2 Large hadron absorber installed in the spectrometer see talk by B. Parsamyan for more details (polarised) → see talk by W.-C. Chang for more details (unpolarised)Expected Experimental Errors and Theory Expectations COMPASS measurement is in valence region, non zero asymmetries are expected For example: the expected theory predictions (for Sivers) are several σ away from the zero