Page 369 - Numerical Methods for Chemical Engineering
P. 369

358     7 Probability theory and stochastic simulation



                   a


                      1                                    1
                      1                                    1
                     er                                   er
                      2                                    2
                     site n  2                            site n  2

                     attice                               attice





                             1  1  2  2                           1  1  2  2
                                attice  site n er                    attice  site n er

                   Figure 7.14 Sample states from Monte Carlo simulation of a 2-D Ising lattice at a temperature above
                   the Curie point, with points showing spin-up sites: (a) in the absence of an external field, the lattice
                   has a mixture of spin-up and spin-down sites; (b) with H < 0, lattice sites are mostly spin-up with a
                   few spin-down sites due to thermal fluctuations (N = 50,µ = 1, H = 0, J = 1, K b T = 5.)


                   Field theory and stochastic PDEs
                   Above, we have considered only ordinary SDEs, but many models in statistical physics are
                   of the form of stochastic PDEs. Let us consider a system such as the Ising lattice that at high
                   temperatures is disordered, but in which as the temperature is lowered, some form of order
                   emerges below a critical temperature T c . We define an order parameter ϕ such that ϕ = 0
                   denotes a completely disordered state. We have defined such a spin order parameter for the
                   lattice model (7.244); however, the concept is quite general. For example, when modeling
                   the phase separation of two species A and B, we could take ϕ to be the difference in the
                   local densities of each species, ϕ = ρ A − ρ B .
                     Commonly, near T c , the system experiences significant long-range fluctuations in ϕ (you
                   can observe this with the Ising MC program). Modeling the system on length scales large
                   compared to the lattice size, we define a local order parameter field ϕ(x) that characterizes
                   the local degree of order at x.As ϕ is presumed to be small near T c (the order is just beginning
                   to emerge), we approximate the local free energy density f (x) as a Taylor series in ϕ(x),
                   and write the total free energy Hamiltonian of the system as the Landau phenomological
                   free energy model:
                                '                                          1
                                               1       2        4  1      2
                       H[ϕ(x)] =    −w(x)ϕ(x) + r[ϕ(x)] + u[ϕ(x)] + c|∇ϕ|   dx       (7.245)
                                               2                   2
                   For an Ising lattice,

                           k b              J2 d      k b T     Hµ         2−d
                                        ∗
                                   ∗
                       r =   (T − T )  T =       u =        w =      c = Ja          (7.246)
                           a d               k b     12a d       a d
                   where a is the lattice spacing and d is the dimension of the lattice. If we neglect any
   364   365   366   367   368   369   370   371   372   373   374