arXiv subjects
Publications and source records attributed to M. Pepe.
We calculate up to four loops the free energy of the two-dimensional (2D) O(n) nonlinear sigma-model regularized on the lattice with the 0-loop and 1-loop Symanzik improved actions. An effective coupling constant based on this calculation is defined.
We analyze the percolation properties of certain clusters defined on configurations of the 2--dimensional Heisenberg model. We find that, given any direction \vec{n} in O(3) space, the spins almost perpendicular to \vec{n} form a percolating cluster. This result gives indications of how the model can avoid a previously conjectured Kosterlitz-Thouless phase transition at finite temperature T.
Several Wilson loops on several lattice sizes are computed in Perturbation Theory via a stochastic method. Applications include: Renormalons, the Mass Term in Heavy Quark Effective Theory and (possibly) the beta-function.
We present complete three loop results and preliminary four loop results for the 2D O(n) nonlinear sigma model with 0-loop and 1-loop Symanzik improved actions. This calculation aims to test the improvement in the numerical precision that the combination of Symanzik actions and effective couplings can give in Monte Carlo simulations.
We give the correct analytic expression of a finite integral appearing in the four-loop computation of the renormalization-group functions for the two-dimensional nonlinear sigma-model on the square lattice with standard action, explaining the origin of a numerical discrepancy. We revise the numerical expressions of Caracciolo and Pelissetto for the perturbative corrections of the susceptibility and of the correlation length. For the values used in Monte Carlo simulations, N=3, 4, 8, the second perturbative correction coefficient of the correlation length varies by 3%, 4%, 3% respectively. Other quantities vary similarly.
A new method to compute observables at many values of the parameters λfor a model with lattice action {\cal{S}}(ϕ, λ) is described. After fixing a reference set λ^r of parameters, a single simulation is carried out by using a ``reference action'' {\cal{S}} (ϕ^r, λ^r) to generate configurations of the field ϕ^r. Then a suitable analytic transformation is performed from the configurations of ϕ^r to the ones corresponding to the action {\cal{S}} (ϕ, λ). Such a transformation allows to obtain the observables for values of the parameters λclose to λ^r. I present studies on the reliability of the algorithm in the case of the ϕ^4 model in 2 dimensions.
First attempts in the application of Numerical Stochastic Perturbation Theory (NSPT) to the problem of pushing one loop further the computation of SU(3) (SU(2)) pertubative beta function (in different schemes) are reviewed and the relevance of such a computation is discussed. Other issues include the proposal of a different strategy for gauge-fixed NSPT computations in lattice QCD.
A review of new developments in numerical stochastic perturbation theory (NSPT) is presented. In particular, the status of the extension of the method to gauge fixed lattice QCD is reviewed and a first application to compact (scalar) QED is presented. Lacking still a general proof of the convergence of the underlying stochastic processes, a self-consistent method for testing the results is discussed.