|
|
John
Negele, Christoph Best, Patrick Dreher, Andrew Pochinsky, and Uwe-Jens
Wiese, Massachusetts Institute of Technology
Robert Edwards, Thomas Jefferson National Accelerator Facility
Thomas Lippert, Hartmut Neff, and Klaus Schilling, University of Wuppertal
Research
Objectives
The major focus this work is on understanding the role of instantons and
their associated quark zero modes in nucleon structure, and using the
quark zero modes to calculate the sea quark content of the nucleon.
Computational
Approach
We calculate the low eigenmodes of the Dirac operator using the Arnoldi
method, which has compelling advantages for our work. One advantage is
that since it works in a fixed dimension space, there is no degradation
of orthogonality and corresponding loss or duplication of modes. A second
advantage, when applied to the non-hermitian Dirac operator, is its insensitivity
to the quark mass, which makes it extremely useful near the chiral limit
of low pion mass. We have two complementary implementations. One is an
exploratory code in which we can control the region of eigenvalues at
will. The other uses the robust and well optimized PARPACK package from
ORNL combined with Chebyshev acceleration.
Accomplishments
NERSC resources enabled us to calculate on 400 configurations the low
eigenmodes of the hermitian Dirac operator, which is G5
times the standard Dirac operator. This showed that, contrary to our original
expectation, at the large quark masses relevant to current unquenched
calculation, the eigenmode expansion of the hermitian Dirac operator has
superior convergence properties to that of the standard operator. An unfortunate
property of the hermitian Dirac operator, however, is the need to calculate
new eigenmodes at each quark mass, in contrast to the standard operator,
for which a single set of eigenmodes applies for all masses. Hence, the
computational needs of the project have increased relative to our original
expectation in order to calculate eigenmodes at several masses.
Significance
Ever since the discovery of quarks in the nucleon, tremendous experimental
effort and resources have been devoted to the measurement of the detailed
quark and gluon structure of the nucleon, and theorists have sought to
understand this structure from first principles. In a clever and difficult
series of experiments at Bates and Jefferson Lab which are now finally
coming to fruition, experimentalists have used the interference between
parity-conserving and parity-violating electron scattering amplitudes
to measure the contributions of strange quarks to electric and magnetic
form factors. Given the investment of effort and resources in these fundamental
experiments, it is extremely important to develop the means to calculate
the strange quark content of the nucleon reliably using lattice QCD. We
are developing a new method which can attain a higher level of statistical
accuracy than existing methods, and will provide the essential quark zero
modes necessary for these calculations. In addition to elucidating the
physics for timely parity-violating electron scattering experiments, this
new method should also enable the evaluation of the disconnected diagrams
encountered in deep inelastic electron scattering.
Publications
W. Detmold, W. Melnitchouk, J. W. Negele, D. B. Renner, and A. W. Thomas,
"Chiral extrapolation of lattice moments of proton quark distributions,"
Phys. Rev. Lett. 87, 172001 (2001); hep-lat/0103006.
D. Dolgov, R. Brower, S. Capitani, J. W. Negele, A. Pochinsky, D. Renner,
N. Eicker, T. Lippert, K. Schilling, R. G. Edwards, and U. M. Heller,
"Moments of structure functions in full QCD," Nucl. Phys. Proc.
Suppl. 94, 303 (2001); hep-lat/0011010.
H. Neff, N. Eicker, T. Lippert, J. W. Negele, and K. Schilling, "On
the low fermionic eigenmode dominance in QCD on the lattice," Phys.
Rev. D (in press): hep-lat/0106016.
http://www-ctp.mit.edu/~negele/
|