Scaling and nonscaling finite-size effects in the Gaussian and the mean spherical model with free boundary conditions
X.S. Chen, V. Dohm

TL;DR
This paper investigates finite-size effects and deviations from finite-size scaling in the Gaussian and mean spherical models with free boundary conditions across dimensions 2<d<4, revealing logarithmic and power-law violations at critical dimensions.
Contribution
It provides a detailed analysis of finite-size effects, including logarithmic and power-law deviations from scaling, in Gaussian and mean spherical models with free boundaries across various dimensions.
Findings
Finite-size scaling valid for d<3 and d>3, with logarithmic deviations at d=3.
Non-logarithmic violation of scaling in susceptibility at d=3 in film geometry.
Power-law violation of scaling for d>3 and universal finite-size scaling for 2<d<3.
Abstract
We calculate finite-size effects of the Gaussian model in a L\times \tilde L^{d-1} box geometry with free boundary conditions in one direction and periodic boundary conditions in d-1 directions for 2<d<4. We also consider film geometry (\tilde L \to \infty). Finite-size scaling is found to be valid for d<3 and d>3 but logarithmic deviations from finite-size scaling are found for the free energy and energy density at the Gaussian upper borderline dimension d* =3. The logarithms are related to the vanishing critical exponent 1-\alpha-\nu=(d-3)/2 of the Gaussian surface energy density. The latter has a cusp-like singularity in d>3 dimensions. We show that these properties are the origin of nonscaling finite-size effects in the mean spherical model with free boundary conditions in d>=3 dimensions. At bulk T_c in d=3 dimensions we find an unexpected non-logarithmic violation of finite-size…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
