0000000000061682
AUTHOR
Wolfhard Janke
Scaling Behavior of the 2D XY Model Revisited
Using two sets of high-precision Monte Carlo data for the two-dimensional XY model in the Villain formulation on square L × L lattices, the scaling behavior of the susceptibility χ and correlation length ξ in the vicinity of the Kosterlitz-Thouless phase transition is analyzed with emphasis on multiplicative logarithmic corrections (ln ξ)-2r in the high-temperature phase and (ln L)-2r in the finite-size scaling region, respectively.
Monte Carlo Study of Critical Point Shifts in Thin Films
We report preliminary results of Monte Carlo simulations of critical point shifts in thin slit-like capillaries. By making use of the isomorphism with an Ising model subject to bulk and surface fields and employing a multi-cluster update algorithm with ghost-spin term we obtain the coexistence curve and the behavior at the critical point for various film thicknesses D.
Multicanonical Monte Carlo study and analysis of tails for the order-parameter distribution of the two-dimensional Ising model.
The tails of the critical order-parameter distribution of the two-dimensional Ising model are investigated through extensive multicanonical Monte Carlo simulations. Results for fixed boundary conditions are reported here, and compared with known results for periodic boundary conditions. Clear numerical evidence for ‘‘fat’’ stretched exponential tails exists below the critical temperature, indicating the possible presence of fat tails at the critical temperature. Our work suggests that the true order-parameter distribution at the critical temperature must be considered to be unknown at present.
Cross Correlations in Scaling Analyses of Phase Transitions
Thermal or finite-size scaling analyses of importance sampling Monte Carlo time series in the vicinity of phase transition points often combine different estimates for the same quantity, such as a critical exponent, with the intent to reduce statistical fluctuations. We point out that the origin of such estimates in the same time series results in often pronounced cross-correlations which are usually ignored even in high-precision studies, generically leading to significant underestimation of statistical fluctuations. We suggest to use a simple extension of the conventional analysis taking correlation effects into account, which leads to improved estimators with often substantially reduced …
HIGH-PRECISION MONTE CARLO DETERMINATION OF α/ν IN THE 3D CLASSICAL HEISENBERG MODEL
To study the role of topological defects in the three-dimensional classical Heisenberg model we have simulated this model on simple cubic lattices of size up to 803, using the single-cluster Monte Carlo update. Analysing the specific-heat data of these simulations, we obtain a very accurate estimate for the ratio of the specific-heat exponent with the correlation-length exponent, α/ν, from a usual finite-size scaling analysis at the critical coupling Kc. Moreover, by fitting the energy at Kc, we reduce the error estimates by another factor of two, and get a value of α/ν, which is comparable in accuracy to best field theoretic estimates.
Simplicial Quantum Gravity on a Randomly Triangulated Sphere
We study 2D quantum gravity on spherical topologies employing the Regge calculus approach with the dl/l measure. Instead of the normally used fixed non-regular triangulation we study random triangulations which are generated by the standard Voronoi-Delaunay procedure. For each system size we average the results over four different realizations of the random lattices. We compare both types of triangulations quantitatively and investigate how the difference in the expectation value of the squared curvature, $R^2$, for fixed and random triangulations depends on the lattice size and the surface area A. We try to measure the string susceptibility exponents through finite-size scaling analyses of…
Resummation of anisotropic quartic oscillator. Crossover from anisotropic to isotropic large-order behavior
We present an approximative calculation of the ground-state energy for the anisotropic anharmonic oscillator Using an instanton solution of the isotropic action $\delta = 0$, we obtain the imaginary part of the ground-state energy for small negative $g$ as a series expansion in the anisotropy parameter $\delta$. From this, the large-order behavior of the $g$-expansions accompanying each power of $\delta$ are obtained by means of a dispersion relation in $g$. These $g$-expansions are summed by a Borel transformation, yielding an approximation to the ground-state energy for the region near the isotropic limit. This approximation is found to be excellent in a rather wide region of $\delta$ aro…
INTERFACE TENSION AND CORRELATION LENGTH OF 2D POTTS MODELS: NUMERICAL VERSUS EXACT RESULTS
I briefly review new analytical formulas for the correlation length and interface tension of two-dimensional q-state Potts models and compare them with numerical results from recent Monte Carlo simulation studies.
High-Temperature Series Analysis of the Free Energy and Susceptibility of the 2D Random-Bond Ising Model
We derive high-temperature series expansions for the free energy and susceptibility of the two-dimensional random-bond Ising model with a symmetric bimodal distribution of two positive coupling strengths J_1 and J_2 and study the influence of the quenched, random bond-disorder on the critical behavior of the model. By analysing the series expansions over a wide range of coupling ratios J_2/J_1, covering the crossover from weak to strong disorder, we obtain for the susceptibility with two different methods compelling evidence for a singularity of the form $\chi \sim t^{-7/4} |\ln t|^{7/8}$, as predicted theoretically by Shalaev, Shankar, and Ludwig. For the specific heat our results are less…
Monte Carlo study of asymmetric 2D XY model
Employing the Polyakov-Susskind approximation in a field theoretical treatment, the t-J model for strongly correlated electrons in two dimensions has recently been shown to map effectively onto an asymmetric two-dimensional classical XY model. The critical temperature at which charge-spin separation occurs in the t-J model is determined by the location of the phase transitions of this effective model. Here we report results of Monte Carlo simulations which map out the complete phase diagram in the two-dimensional parameter space and also shed some light on the critical behaviour of the transitions.
Monte-Carlo Study of Pure-Phase Cumulants of 2D q-State Potts Models
We performed Monte Carlo simulations of the two-dimensional q-state Potts model with q=10, 15, and 20 to study the energy and magnetization cumulants in the ordered and disordered phase at the first-order transition point $\beta_t$. By using very large systems of size 300 x 300, 120 x 120, and 80 x 80 for q=10, 15, and 20, respectively, our numerical estimates provide practically (up to unavoidable, but very small statistical errors) exact results which can serve as a useful test of recent resummed large-q expansions for the energy cumulants by Bhattacharya `et al.' [J. Phys. I (France) 7 (1997) 81]. Up to the third order cumulant and down to q=10 we obtain very good agreement, and also the…
Multicanonical multigrid Monte Carlo method.
To further improve the performance of Monte Carlo simulations of first-order phase transitions we propose to combine the multicanonical approach with multigrid techniques. We report tests of this proposition for the d-dimensional ${\mathrm{\ensuremath{\Phi}}}^{4}$ field theory in two different situations. First, we study quantum tunneling for d=1 in the continuum limit, and second, we investigate first-order phase transitions for d=2 in the infinite volume limit. Compared with standard multicanonical simulations we obtain improvement factors of several, and of about one order of magnitude, respectively.
Multi-overlap simulations of free-energy barriers in the 3D Edwards–Anderson Ising spin glass
We report large-scale simulations of the three-dimensional Edwards‐Anderson Ising spin-glass model using the multi-overlap Monte Carlo algorithm. We present our results in the spin-glass phase on free-energy barriers and the non-trivial finite-size scaling behavior of the Parisi order-parameter distribution. © 1999 Elsevier Science B.V. All rights reserved.
Monte Carlo Simulations of Spin Systems
This chapter gives a brief introduction to Monte Carlo simulations of classical O(n) spin systems such as the Ising (n = 1), XY (n = 2), and Heisenberg (n = 3) models. In the first part I discuss some aspects of the use of Monte Carlo algorithms to generate the raw data. Here special emphasis is placed on nonlocal cluster update algorithms which proved to be most efficient for this class of models. The second part is devoted to the data analysis at a continuous phase transition. For the example of the three-dimensional Heisenberg model it is shown how precise estimates of the transition temperature and the critical exponents can be extracted from the raw data. I conclude with a brief overvi…
Ising Spins on 3D Random Lattices
We perform single-cluster Monte Carlo simulations of the Ising model on three-dimensional Poissonian random lattices of Voronoi/Delaunay type with up to 128 000 sites. For each lattice size quenched averages are computed over 96 realizations. From a finite-size scaling analysis we obtain strong evidence that the critical exponents coincide with those on regular cubic lattices.
Single-cluster Monte Carlo study of the Ising model on two-dimensional random lattices.
We use the single-cluster Monte Carlo update algorithm to simulate the Ising model on two-dimensional Poissonian random lattices with up to 80,000 sites which are linked together according to the Voronoi/Delaunay prescription. In one set of simulations we use reweighting techniques and finite-size scaling analysis to investigate the critical properties of the model in the very vicinity of the phase transition. In the other set of simulations we study the approach to criticality in the disordered phase, making use of improved estimators for measurements. From both sets of simulations we obtain clear evidence that the critical exponents agree with the exactly known exponents for regular latti…
Spin Glasses on Thin Graphs
In a recent paper we found strong evidence from simulations that the Isingantiferromagnet on ``thin'' random graphs - Feynman diagrams - displayed amean-field spin glass transition. The intrinsic interest of considering such random graphs is that they give mean field results without long range interactions or the drawbacks, arising from boundary problems, of the Bethe lattice. In this paper we reprise the saddle point calculations for the Ising and Potts ferromagnet, antiferromagnet and spin glass on Feynman diagrams. We use standard results from bifurcation theory that enable us to treat an arbitrary number of replicas and any quenched bond distribution. We note the agreement between the f…
High-temperature series analysis of the p-state Potts glass model on d-dimensional hypercubic lattices
We analyze recently extended high-temperature series expansions for the “Edwards-Anderson” spin-glass susceptibility of the p-state Potts glass model on d-dimensional hypercubic lattices for the case of a symmetric bimodal distribution of ferro- and antiferromagnetic nearest-neighbor couplings \(\). In these star-graph expansions up to order 22 in the inverse temperature \(\), the number of Potts states p and the dimension d are kept as free parameters which can take any value. By applying several series analysis techniques to the new series expansions, this enabled us to determine the critical coupling Kc and the critical exponent \(\) of the spin-glass susceptibility in a large region of …
Numerical test of finite-size scaling predictions for the droplet condensation-evaporation transition
We numerically study the finite-size droplet condensation-evaporation transition in two dimensions. We consider and compare two orthogonal approaches, namely at fixed temperature and at fixed density, making use of parallel multicanonical simulations. The equivalence between Ising model and lattice gas allows us to compare to analytical predictions. We recover the known background density (at fixed temperature) and transition temperature (at fixed density) in the thermodynamic limit and compare our finite-size deviations to the predicted leading-order finite-size corrections.
The Ising transition in 2D simplicial quantum gravity - can Regge calculus be right?
We report a high statistics simulation of Ising spins coupled to 2D quantum gravity in the Regge calculus approach using triangulated tori with up to $512^2$ vertices. For the constant area ensemble and the $dl/l$ functional measure we definitively can exclude the critical exponents of the Ising phase transition as predicted for dynamically triangulated surfaces. We rather find clear evidence that the critical exponents agree with the Onsager values for static regular lattices, independent of the coupling strength of an $R^2$ interaction term. For exploratory simulations using the lattice version of the Misner measure the situation is less clear.
Convergent Strong-Coupling Expansions from Divergent Weak-Coupling Perturbation Theory
Divergent weak-coupling perturbation expansions for physical quantities can be converted into sequences of uniformly and exponentially fast converging approximations. This is possible with the help of an additional variational parameter to be optimized order by order. The uniformity of the convergence for any coupling strength allows us to take all expressions directly to the strong-coupling limit, yielding a simple calculation scheme for the coefficients of convergent strong-coupling expansions. As an example, we determine these coefficients for the ground state energy of the anharmonic oscillator up to 22nd order with a precision of about 20 digits.
Multibondic cluster algorithm for Monte Carlo simulations of first-order phase transitions.
Inspired by the multicanonical approach to simulations of first-order phase transitions we propose for $q$-state Potts models a combination of cluster updates with reweighting of the bond configurations in the Fortuin-Kastelein-Swendsen-Wang representation of this model. Numerical tests for the two-dimensional models with $q=7, 10$ and $20$ show that the autocorrelation times of this algorithm grow with the system size $V$ as $\tau \propto V^\alpha$, where the exponent takes the optimal random walk value of $\alpha \approx 1$.
Measure dependence of 2D simplicial quantum gravity
We study pure 2D Euclidean quantum gravity with $R^2$ interaction on spherical lattices, employing Regge's formulation. We attempt to measure the string susceptibility exponent $\gamma_{\rm str}$ by using a finite-size scaling Ansatz in the expectation value of $R^2$. To check on effects of the path integral measure we investigate two scale invariant measures, the "computer" measure $dl/l$ and the Misner measure $dl/\sqrt A$.
Multioverlap Simulations of the 3D Edwards-Anderson Ising Spin Glass
We introduce a novel method for numerical spin glass investigations: Simulations of two replica at fixed temperature, weighted such that a broad distribution of the Parisi overlap parameter $q$ is achieved. Canonical expectation values for the entire $q$-range (multi-overlap) follow by re-weighting. We demonstrate the feasibility of the approach by studying the $3d$ Edwards-Anderson Ising ($J_{ik}=\pm 1$) spin glass in the broken phase ($\beta=1$). For the first time it becomes possible to obtain reliable results about spin glass tunneling barriers. In addition, as do some earlier numerical studies, our results support that Parisi mean field theory is valid down to $3d$.
Monte Carlo simulation of dimensional crossover in the XY model.
We report Monte Carlo simulations of Villain's periodic Gaussian XY model on ${\mathit{L}}^{2}$\ifmmode\times\else\texttimes\fi{}N lattices of film geometry (L\ensuremath{\gg}N) with up to N=16 layers, employing the single-cluster update algorithm combined with improved estimators for measurements. The boundary conditions are periodic within each layer and free at the bottom and top layer. Based on data for the specific heat, the spin-spin correlation function, and the susceptibility in the high-temperature phase we study the crossover from three- to two-dimensional behavior as criticality is approached. For the transition temperatures, determined from Kosterlitz-Thouless fits to the correl…
Fixed versus random triangulations in 2D Regge calculus
Abstract We study 2D quantum gravity on spherical topologies using the Regge calculus approach with the dl l measure. Instead of a fixed non-regular triangulation which has been used before, we study for each system size four different random triangulations, which are obtained according to the standard Voronoi-Delaunay procedure. We compare both approaches quantitatively and show that the difference in the expectation value of R2 between the fixed and the random triangulation depends on the lattice size and the surface area A. We also try again to measure the string susceptibility exponents through a finite-size scaling Ansatz in the expectation value of an added R2 interaction term in an a…
Z2-Regge versus standard Regge calculus in two dimensions
We consider two versions of quantum Regge calculus: the standard Regge calculus where the quadratic link lengths of the simplicial manifold vary continuously and the ${Z}_{2}$ Regge model where they are restricted to two possible values. The goal is to determine whether the computationally more easily accessible ${Z}_{2}$ model still retains the universal characteristics of standard Regge theory in two dimensions. In order to compare observables such as the average curvature or Liouville field susceptibility, we use in both models the same functional integration measure, which is chosen to render the ${Z}_{2}$ Regge model particularly simple. Expectation values are computed numerically and …
Ising model universality for two-dimensional lattices
We use the single-cluster Monte Carlo update algorithm to simulate the Ising model on two-dimensional Poissonian random lattices of Delaunay type with up to 80\,000 sites. By applying reweighting techniques and finite-size scaling analyses to time-series data near criticality, we obtain unambiguous support that the critical exponents for the random lattice agree with the exactly known exponents for regular lattices, i.e., that (lattice) universality holds for the two-dimensional Ising model.
Recent Developments in Monte-Carlo Simulations of First-Order Phase Transitions
In the past few years considerable progress has been made in Monte Carlo simulations of first-order phase transitions and in the analysis of the resulting finite-size data. In this paper special emphasis will be placed on multicanonical simulations using multigrid update techniques, on numerical estimates of interface tensions, and on accurate methods for determining the transition point and latent heat.
Multicanonical Simulations of the Tails of the Order-Parameter Distribution of the Two-Dimensional Ising Model
We report multicanonical Monte Carlo simulations of the tails of the order-parameter distribution of the two-dimensional Ising model for fixed boundary conditions. Clear numerical evidence for "fat" stretched exponential tails is found below the critical temperature, indicating the possible presence of fat tails at the critical temperature.
Canonical versus microcanonical analysis of first-order phase transitions
Abstract I discuss the relation between canonical and microcanonical analyses of first-order phase transitions. In particular it is shown that the microcanonical Maxwell construction is equivalent to the equal-peak-height criterion often employed in canonical simulations. As a consequence the microcanonical finite-size estimators for the transition point, latent heat and interface tension are identical to standard estimators in the canonical ensemble. Special emphasis is placed on various ways for estimating interface tensions. The theoretical considerations are illustrated with numerical data for the two-dimensional 10-state Potts model.
Monte Carlo study of cluster-diameter distribution: An observable to estimate correlation lengths
We report numerical simulations of two-dimensional $q$-state Potts models with emphasis on a new quantity for the computation of spatial correlation lengths. This quantity is the cluster-diameter distribution function $G_{diam}(x)$, which measures the distribution of the diameter of stochastically defined cluster. Theoretically it is predicted to fall off exponentially for large diameter $x$, $G_{diam} \propto \exp(-x/\xi)$, where $\xi$ is the correlation length as usually defined through the large-distance behavior of two-point correlation functions. The results of our extensive Monte Carlo study in the disordered phase of the models with $q=10$, 15, and $20$ on large square lattices of si…
Three-Dimensional 3-State Potts Model Revisited With New Techniques
We report a fairly detailed finite-size scaling analysis of the first-order phase transition in the three-dimensional 3-state Potts model on cubic lattices with emphasis on recently introduced quantities whose infinite-volume extrapolations are governed `only' by exponentially small terms. In these quantities no asymptotic power series in the inverse volume are involved which complicate the finite-size scaling behaviour of standard observables related to the specific-heat maxima or Binder-parameter minima. Introduced initially for strong first-order phase transitions in q-state Potts models with ``large enough'' q, the new techniques prove to be surprisingly accurate for a q value as small …
Standard and Z2-Regge theory in two dimensions
Abstract We qualitatively compare two versions of quantum Regge calculus by means of Monte Carlo simulations. In Standard Regge Calculus the quadratic link lengths of the triangulation vary continuously, whereas in the Z2-Regge Model they are restricted to two possible values. The goal is to determine whether the computationally more easily accessible Z2 model retains the characteristics of standard Regge theory.
Softening Transitions with Quenched 2D Gravity
We perform extensive Monte Carlo simulations of the 10-state Potts model on quenched two-dimensional $\Phi^3$ gravity graphs to study the effect of quenched connectivity disorder on the phase transition, which is strongly first order on regular lattices. The numerical data provides strong evidence that, due to the quenched randomness, the discontinuous first-order phase transition of the pure model is softened to a continuous transition.
Error estimation and reduction with cross correlations
Besides the well-known effect of autocorrelations in time series of Monte Carlo simulation data resulting from the underlying Markov process, using the same data pool for computing various estimates entails additional cross correlations. This effect, if not properly taken into account, leads to systematically wrong error estimates for combined quantities. Using a straightforward recipe of data analysis employing the jackknife or similar resampling techniques, such problems can be avoided. In addition, a covariance analysis allows for the formulation of optimal estimators with often significantly reduced variance as compared to more conventional averages.
Multicanonical Monte Carlo simulations
Canonical Monte Carlo simulations of disordered systems like spin glasses and systems undergoing first-order phase transitions are severely hampered by rare event states which lead to exponentially diverging autocorrelation times with increasing system size and hence to exponentially large statistical errors. One possibility to overcome this problem is the multicanonical reweighting method. Using standard local update algorithms it could be demonstrated that the dependence of autocorrelation times on the system size V is well described by a less divergent power law, τ∝Vα, with 1<α<3, depending on the system. After a brief review of the basic ideas, combinations of multicanonical reweighting…
MC Study of the p-state Mean-Field Potts Glass
The p-state mean-field Potts glass with ±J-couplings is studied by Monte Carlo (MC) simulations, both for p = 3 and p = 6 states. At the exactly known glass transition temperature Tc, the moments q( k ) of the spin glass order parameter satisfy for p = 3 a simple scaling behavior, q( k ) \({q^{\left( k \right)}}\alpha {N^{ - k/3}}{\tilde f_k}\left\{ {{N^{1/3}}\left( {1 - T/{T_c}} \right)} \right\},k = 1,2,3,...\). The specific-heat maxima exhibit a similar behavior, c V max α const — N -l/3, while the approach of the maxima positions T max to T c as N → ∞ is non-monotonic. For p = 6 the results are compatible with the expected result of a quite peculiar first-order phase transition. The spe…
Numerical tests of conjectures of conformal field theory for three-dimensional systems
The concept of conformal field theory provides a general classification of statistical systems on two-dimensional geometries at the point of a continuous phase transition. Considering the finite-size scaling of certain special observables, one thus obtains not only the critical exponents but even the corresponding amplitudes of the divergences analytically. A first numerical analysis brought up the question whether analogous results can be obtained for those systems on three-dimensional manifolds. Using Monte Carlo simulations based on the Wolff single-cluster update algorithm we investigate the scaling properties of O(n) symmetric classical spin models on a three-dimensional, hyper-cylindr…
Scaling property of variational perturbation expansion for a general anharmonic oscillator with xp-potential
We prove a powerful scaling property for the extremality condition in the recently developed variational perturbation theory which converts divergent perturbation expansions into exponentially fast convergent ones. The proof is given for the energy eigenvalues of an anharmonic oscillator with an arbitrary $x^p$-potential. The scaling property greatly increases the accuracy of the results.