Debian Codegraph @

demo/documented/cahn-hilliard/common.txt

raw
1 This example demonstrates the solution of a particular nonlinear
2 time-dependent fourth-order equation, known as the Cahn-Hilliard
3 equation. In particular it demonstrates the use of
4
5 * The built-in Newton solver
6 * Advanced use of the base class ``NonlinearProblem``
7 * Automatic linearisation
8 * A mixed finite element method
9 * The :math:`\theta`-method for time-dependent equations
10 * User-defined Expressions as Python classes
11 * Form compiler options
12 * Interpolation of functions
13
14
15 Equation and problem definition
16 -------------------------------
17
18 The Cahn-Hilliard equation is a parabolic equation and is typically
19 used to model phase separation in binary mixtures. It involves
20 first-order time derivatives, and second- and fourth-order spatial
21 derivatives. The equation reads:
22
23 .. math::
24 \frac{\partial c}{\partial t} - \nabla \cdot M \left(\nabla\left(\frac{d f}{d c}
25 - \lambda \nabla^{2}c\right)\right) &= 0 \quad {\rm in} \ \Omega, \\
26 M\left(\nabla\left(\frac{d f}{d c} - \lambda \nabla^{2}c\right)\right) \cdot n &= 0 \quad {\rm on} \ \partial\Omega, \\
27 M \lambda \nabla c \cdot n &= 0 \quad {\rm on} \ \partial\Omega.
28
29 where :math:`c` is the unknown field, the function :math:`f` is
30 usually non-convex in :math:`c` (a fourth-order polynomial is commonly
31 used), :math:`n` is the outward directed boundary normal, and
32 :math:`M` is a scalar parameter.
33
34 Mixed form
35 ^^^^^^^^^^
36
37 The Cahn-Hilliard equation is a fourth-order equation, so casting it
38 in a weak form would result in the presence of second-order spatial
39 derivatives, and the problem could not be solved using a standard
40 Lagrange finite element basis. A solution is to rephrase the problem
41 as two coupled second-order equations:
42
43 .. math::
44 \frac{\partial c}{\partial t} - \nabla \cdot M \nabla\mu &= 0 \quad {\rm in} \ \Omega, \\
45 \mu - \frac{d f}{d c} + \lambda \nabla^{2}c &= 0 \quad {\rm in} \ \Omega.
46
47 The unknown fields are now :math:`c` and :math:`\mu`. The weak
48 (variational) form of the problem reads: find :math:`(c, \mu) \in V
49 \times V` such that
50
51 .. math::
52 \int_{\Omega} \frac{\partial c}{\partial t} q \, {\rm d} x + \int_{\Omega} M \nabla\mu \cdot \nabla q \, {\rm d} x
53 &= 0 \quad \forall \ q \in V, \\
54 \int_{\Omega} \mu v \, {\rm d} x - \int_{\Omega} \frac{d f}{d c} v \, {\rm d} x - \int_{\Omega} \lambda \nabla c \cdot \nabla v \, {\rm d} x
55 &= 0 \quad \forall \ v \in V.
56
57 Time discretisation
58 ^^^^^^^^^^^^^^^^^^^
59
60 Before being able to solve this problem, the time derivative must be
61 dealt with. Apply the :math:`\theta`-method to the mixed weak form of
62 the equation:
63
64 .. math::
65
66 \int_{\Omega} \frac{c_{n+1} - c_{n}}{dt} q \, {\rm d} x + \int_{\Omega} M \nabla \mu_{n+\theta} \cdot \nabla q \, {\rm d} x
67 &= 0 \quad \forall \ q \in V \\
68 \int_{\Omega} \mu_{n+1} v \, {\rm d} x - \int_{\Omega} \frac{d f_{n+1}}{d c} v \, {\rm d} x - \int_{\Omega} \lambda \nabla c_{n+1} \cdot \nabla v \, {\rm d} x
69 &= 0 \quad \forall \ v \in V
70
71 where :math:`dt = t_{n+1} - t_{n}` and :math:`\mu_{n+\theta} =
72 (1-\theta) \mu_{n} + \theta \mu_{n+1}`. The task is: given
73 :math:`c_{n}` and :math:`\mu_{n}`, solve the above equation to find
74 :math:`c_{n+1}` and :math:`\mu_{n+1}`.
75
76
77 Demo parameters
78 ^^^^^^^^^^^^^^^
79
80 The following domains, functions and time stepping parameters are used
81 in this demo:
82
83 * :math:`\Omega = (0, 1) \times (0, 1)` (unit square)
84 * :math:`f = 100 c^{2} (1-c)^{2}`
85 * :math:`\lambda = 1 \times 10^{-2}`
86 * :math:`M = 1`
87 * :math:`dt = 5 \times 10^{-6}`
88 * :math:`\theta = 0.5`
89
90 With the above input the solution for :math:`c` will look as follows:
91
92 .. image:: ../cahn-hilliard_c.png
93 :scale: 75
94 :align: center