Dr Michael Teper

Abstract:

We carry out a preliminary calculation of the glueball mass spectrum and the effective string tension in SU(2) lattice gauge theory with 't Hooft's twisted boundary conditions. We make semi-quantitative comparisons with recent analytic results in small volumes and with previous Monte Carlo calculations for periodic boundary conditions. We find a reasonable measure of agreement between the form of our results and those of the analytic calculations. © 1989.

Abstract:

We show that the SU(2) lattice topological susceptibility, as calculated using the Phillips and Stone algorithm, is divergent in the continuum limit, when one uses the standard plaquette action. We discuss remedies and implications. © 1989.

Abstract:

We study the glueball spectrum for pure gauge SU(3) using lattice-gauge-theory simulation. We investigate a range of β-values from 5.9 to 6.2, a range of sizes of spatial lattice from 103 to 203, all JPC possibilities, states of different momentum, etc. We are able to conclude that our results are relevant to the continuum limit of pure gauge QCD. We obtain glueball masses (in units of Keff which is about 0.44 GeV.): 3.5(2) and 6.3(4) for 0++, 5.3(2) for 2++, 5.3(2) for 0-+, 6.9(5) for 1+- and 6.6(10) for 2+-. © 1989.

Abstract:

We implement the algorithm of Phillips and Stone on a hypercubic, periodic lattice and show that at currently accessible couplings the SU(2) topological charge so calculated is dominated by short-distance fluctuations. We propose and test an improvement to rid the measure of such lattice artifacts. We find that the improved algorithm produces a topological susceptibility that is consistent with that obtained by the alternative cooling method, thus resolving the controversial discrepancy between geometric and cooling methods. We briefly discuss the reasons for this and point out that our improvement is likely to be particularly effective when applied to the case of SU(3). © 1989.

Abstract:

We calculate the lowest glueball masses and the string tension for both Manton's action and for Symanzik's tree-level improved action. We do so on large lattices and for small lattice spacings using techniques recently employed in an extensive investigation of the Wilson plaquette action. Comparing all these results we find that the ratios of the lightest masses are universal to a high degree of accuracy. In particular, we confirm that on large volumes the tensor glueball is heavier than the scalar glueball: m[2+]{reversed tilde equals}1.5 m[0+]. We repeat these calculations for larger lattice spacings and find that the string tension follows 2-loop perturbation theory more closely in the case of these alternative actions than in the case of the standard plaquette action. Our attempt to repeat the analysis with Wilson's block-spin improved action foundered on the strong breakdown of positivity apparent in the calculated correlation functions. In all cases which we were able to study the observed violations of scaling are in the same direction. This suggests that the causes of the scaling violations observed with Wilson's plaquette action are "semi-universal". It also weakens the implication of the observed universality for the question of how close we are to the continuum limit. © 1988.