[Paper Review] Measuring interface tensions in 4d SU(N) lattice gauge theories

Biagio Lucini, Philippe de Forcrand+1 more|ArXiv.org|Sep 27, 2004

Theoretical and Computational Physics6 references4 citations

TL;DR

This paper introduces a novel algorithm to measure order-order interface tensions in 4d SU(N) lattice gauge theories, leveraging a modified snake algorithm with variance reduction to efficiently compute free energy ratios. The method enables accurate extraction of interface tensions, showing Casimir scaling for SU(4) down to T=1.2Tc and confirming that prior underestimates in SU(3) were due to finite-size effects.

ABSTRACT

We propose a new algorithm to compute the order-order interface tension in SU(N) lattice gauge theories. The algorithm is trivially generalizable to a variety of models, e.g., spin models. In the case N=3, via the perfect wetting hypothesis, we can estimate the order-disorder interface tension. In the case N=4, we study the ratio of dual k-tensions and find that it satisfies Casimir scaling down to T=1.2 T_c.

Motivation & Objective

• To develop a robust, efficient method for measuring interface tensions in 4d SU(N) lattice gauge theories.
• To overcome the severe overlap problem in direct interface tension measurements using a factorized ratio approach.
• To quantify finite-size effects and establish reliable lattice size criteria for accurate tension extraction.
• To test Casimir scaling of dual k-tensions in SU(4) down to T=1.2Tc.
• To re-evaluate the order-disorder interface tension in SU(3) using improved finite-size control.

Proposed method

• The method uses a factorized ratio of partition functions, Z(k)/Z(k-1), to compute the free energy cost per plaquette, avoiding direct computation of Z(A)/Z(0).
• Each ratio Z(k)/Z(k-1) is measured via Monte Carlo averages of Tr(ζ₁Πₖ) in a system with k−1 plaquettes flipped, enabling single-simulation efficiency.
• Finite-size corrections from Gaussian interface fluctuations are removed using a Lüscher-like correction involving the Dedekind η-function.
• The interface is modeled as a periodic, pinned 2D surface of size r×L, with r chosen to minimize finite-size effects.
• Variance reduction techniques are applied to improve statistical precision in the ratio measurements.
• The algorithm is generalizable to spin models and other systems with similar interface structures.

Experimental results

Research questions

• RQ1 Can a new algorithm be developed to measure order-order interface tensions in 4d SU(N) gauge theories with improved efficiency and accuracy?
• RQ2 To what extent do finite-size effects bias previous measurements of interface tensions, particularly in SU(3)?
• RQ3 Does Casimir scaling hold for dual k-tensions in SU(4) down to T=1.2Tc?
• RQ4 Can the perfect wetting hypothesis provide a reliable lower bound for the order-disorder interface tension in SU(3)?
• RQ5 What lattice size is required to achieve reliable results without significant finite-size bias?

Key findings

• For SU(3), the order-disorder interface tension is found to be significantly larger than previous histogram method results, with the discrepancy attributed to finite-size effects in earlier simulations.
• Finite-size effects systematically underestimate the interface tension unless L√σ ≳ 7, a condition confirmed by simulations at L=16, 24, and 32.
• In SU(4), the ratio of dual k-tensions σₖ/σ₁ is consistent with Casimir scaling, σₖ/σ₁ = k(N−k)/(N−1), down to T=1.2Tc.
• The ratio σ₂/σ₁ for SU(4) at T=1.2Tc is measured as 1.33(3), close to the perturbative prediction of 4/3=1.333.
• A broad plateau in the effective tension as a function of interface width r confirms that higher-order corrections are negligible beyond r∼ξ.
• The algorithm successfully mitigates the overlap problem and achieves O(L⁴) efficiency gain over standard snake methods.