Monte Carlo study of the random-field Ising model

Physical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics

Volumn 53, Issue 1, 1996, Pages 393-404

  Newman, M E J ^a   Barkema, G T ^b  

^a Department of Obstetrics Gynecology   (United States)

^b INSTITUTE FOR ADVANCED STUDY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0000021025     PISSN: 1063651X     EISSN: None     Source Type: Journal    
DOI: 10.1103/PhysRevE.53.393     Document Type: Article

References (41)

1. frenchspacing T. Nattermann and J. Villain, Phase Transitions 11, 5 (1988).
2. frenchspacing G. Parisi and N. Sourlas, Phys. Rev. Lett. 43, 744 (1978).
3. frenchspacing G. Grinstein and S. K. Ma, Phys. Rev. Lett. 49, 685 (1982);
4. G. Grinstein and S. K. Ma, Phys. Rev. B 28, 2588 (1983).
5. frenchspacing J. Z. Imbrie, Phys. Rev. Lett. 53, 1747 (1984);
6. J. Z. Imbrie, Commun. Math. Phys. 98, 145 (1985).
7. frenchspacing J. Bricmont and A. Kupiainen, Phys. Rev. Lett. 59, 1829 (1987).
8. frenchspacing K. Dahmen and J. P. Sethna (private communication).
9. frenchspacing S. Fishman and A. Aharony, J. Phys. C 12, L729 (1979).
10. frenchspacing J. Cardy, Phys. Rev. B 29, 505 (1985).
11. frenchspacing D. P. Belanger, A. R. King, V. Jaccarino and J. L. Cardy, Phys. Rev. B 28, 2522 (1983).
12. frenchspacing M. Hagen, R. A. Cowley, S. K. Satija, H. Yoshizawa, G. Shirane, R. J. Birgeneau and H. J. Guggenheim, Phys. Rev. B 28, 2602 (1983).
13. frenchspacing D. S. Fisher, Phys. Rev. Lett. 56, 416 (1986).
14. frenchspacing A. J. Bray and M. A. Moore, J. Phys. C 18, L927 (1985).
15. frenchspacing M. E. J. Newman, B. W. Roberts, G. T. Barkema and J. P. Sethna, Phys. Rev. B 48, 16533 (1993).
16. frenchspacing I. Dayan, M. Schwartz and A. P. Young, J. Phys. A 26, 3093 (1993).
17. frenchspacing A. P. Young and M. Nauenburg, Phys. Rev. Lett. 54, 2429 (1986).
18. frenchspacing A. T. Ogielski and D. A. Huse, Phys. Rev. Lett. 56, 1298 (1986).
19. frenchspacing H. Rieger and A. P. Young, J. Phys. A 26, 5279 (1993).
20. frenchspacing H. Rieger (unpublished).
21. frenchspacing A. M. Ferrenberg and D. P. Landau, Phys. Rev. B 44, 5081 (1991).
22. frenchspacing N. Metropolis, A. Rosenbluth, M. Rosenbluth, A. Teller and E. Teller, J. Chem. Phys. 21, 1087 (1953).
23. frenchspacing U. Wolff, Phys. Rev. Lett. 62, 361 (1989).
24. frenchspacing R. H. Swendsen and J. S. Wang, Phys. Rev. Lett. 58, 86 (1987).
25. frenchspacing Vl. S. Dotsenko, W. Selke and A. L. Talapov, Physica A 170, 278 (1991).
26. frenchspacing J. J. Binney, N. J. Dowrick, A. J. Fisher, and M. E. J. Newman, The Theory of Critical Phenomena (Oxford University Press, Oxford, 1992).
27. It will be noticed that this acceptance ratio can become larger than one if enough of the spins are aligned opposite their local fields. This problem can trivially be overcome by the standard procedure of dividing the acceptance ratios for a move to proceed in either direction by the larger of the two ratios.
28. frenchspacing G. T. Barkema and J. F. Marko, Phys. Rev. Lett. 71, 2070 (1993).
29. The distance can be defined in various different ways. In their simulations, Barkema and Marko citebarkema defined it using a ``Manhattan'' formula in which the distance between two points (x1,y1) and (x2,y2) on a two dimensional square lattice is x1-x2+y1-y2 (with the obvious generalization in higher dimensions). However, the exact definition of the distance is not important for the working of the algorithm.
30. In their simulations with the LCF, Barkema and Marko used a length scale that was tuned to the correlation length of the equilibrated system at the temperature they were studying. However, this is not a shrewd idea in the present case, since the rapid decrease of the correlation length below Tc means that the algorithm would become equivalent to the Metropolis algorithm again at temperatures only a little below the phase transition.
31. frenchspacing A. M. Ferrenberg and R. H. Swendsen, Phys. Rev. Lett. 61, 2635 (1988);
32. A. M. Ferrenberg and R. H. Swendsen, Phys. Rev. Lett. 63, 1195 (1989).
33. In fact, the speed with which the iteration converges seems to depend on how well the samples that we have taken cover the important ranges of energy. If we try to get the method to converge for values of J(n) for which the energies of the most likely states fall far away from any of the energies sampled during our Monte Carlo runs, convergence will be slow, and even after convergence the method will not give a very good estimate of the partition function. For this reason, the method is not suitable for extrapolating far beyond the regime in which the simulations were performed or for interpolating between simulations with very different values of the parameters J(n).
34. To be strictly equivalent, we would have to perform our collapse in a range centered around the maximum of the scaling function for the specific heat. But since the difference between the position of this maximum and the positions xc that we have used in our collapses are much smaller than the smallest values of Δ x, we can ignore this point.
35. frenchspacing S. Wiseman and E. Domany (unpublished).
36. frenchspacing B. Efron, SIAM Rev. 21, 460 (1979).
37. frenchspacing J. P. Sethna, K. Dahmen, S. Kartha, J. D. Shore and J. A. Krumhansl, Phys. Rev. Lett. 70, 3347 (1993).
38. frenchspacing A. Maritan, M. Cieplak, M. R. Swift and J. R. Banavar, Phys. Rev. Lett. 72, 946 (1994).
39. frenchspacing O. Perković (private communication).
40. frenchspacing A. T. Ogielski, Phys. Rev. Lett. 57, 1251 (1986).
41. frenchspacing M. Gofman, J. Adler, A. Aharony, A. B. Harris and M. Schwartz, Phys. Rev. Lett. 71, 1569 (1993).