English

Self-similar but not conformally invariant traces obtained by modified Loewner forces

Statistical Mechanics 2022-02-16 v1

Abstract

The two-dimensional Loewner exploration process is generalized to the case where the random force is self-similar with positively correlated increments. We model this random force by a fractional Brownian motion with Hurst exponent H12HBMH\geq \frac{1}{2}\equiv H_{\text{BM}}, where HBMH_{\text{BM}} stands for the one-dimensional Brownian motion. By manipulating the deterministic force, we design a scale-invariant equation describing self-similar traces which lack conformal invariance. The model is investigated in terms of the "input diffusivity parameter" κ\kappa, which coincides with the one of the ordinary Schramm-Loewner evolution (SLE) at H=HBMH=H_{\text{BM}}. In our numerical investigation, we focus on the scaling properties of the traces generated for κ=2,3\kappa=2,3, κ=4\kappa=4 and κ=6,8\kappa=6,8 as the representatives, respectively, of the dilute phase, the transition point and the dense phase of the ordinary SLE. The resulting traces are shown to be scale-invariant. Using two equivalent schemes, we extract the fractal dimension, Df(H)D_f(H), of the traces which decrease monotonically with increasing HH, reaching Df=1D_f=1 at H=1H=1 for all κ\kappa values. The left passage probability (LPP) test demonstrates that, for HH values not far from the uncorrelated case (small ϵHHHBMHBM\epsilon_H\equiv \frac{H-H_{\text{BM}}}{H_{\text{BM}}}) the prediction of the ordinary SLE is applicable with an effective diffusivity parameter κeff\kappa_{\text{eff}}. Not surprisingly, the κeff\kappa_{\text{eff}}'s do not fulfill the prediction of SLE for the relation between Df(H)D_f(H) and the diffusivity parameter.

Keywords

Cite

@article{arxiv.2105.02694,
  title  = {Self-similar but not conformally invariant traces obtained by modified Loewner forces},
  author = {S. Tizdast and Z. Ebadi and J. Cheraghalizadeh and M. N. Najafi and José S. Andrade and Hans J. Herrmann},
  journal= {arXiv preprint arXiv:2105.02694},
  year   = {2022}
}