Debian Codegraph @

demo/documented/navier-stokes/common.txt

raw
1 This demo solves the incompressible Navier-Stokes equations. It
2 illustrates how to:
3
4 * Implement a splitting method where different fields are coupled
5 via a set of partial differential equations
6 * Use different iterative solvers and different preconditioners for
7 different steps of the solution process
8 * Use time-dependent Expressions
9 * Set boundary conditions based on geometric constraints
10 * Perform time-stepping
11
12 Equation and solution method
13 ----------------------------
14
15 We consider the incompressible Navier-Stokes equations on a domain
16 :math:`\Omega \subset \mathbb{R}^2`, consisting of a pair of momentum
17 and continuity equations:
18
19 .. math::
20
21 \dot{u} + \nabla u \cdot u - \nu \Delta u + \nabla p &= f, \\
22 \nabla \cdot u = 0.
23
24 Here, :math:`u` is the unknown velocity, :math:`p` is the unknown
25 pressure, :math:`\nu` is the kinematic viscosity, and :math:`f` is a
26 given source. There exist many solution strategies for the
27 incompressible Navier-Stokes equations. One of the oldest is the
28 splitting method proposed by Chorin (1968) and Temam (1969), often
29 refered to as Chorin's method.
30
31 In Chorin's method, one first ignores the pressure in the momentum
32 equation and computes the *tentative velocity* :math:`u_h^{\star}`
33 according to:
34
35 .. math::
36
37 \langle (u_h^{\star} - u_h^{n-1}) / \Delta t_n, v \rangle +
38 \langle \nabla u_h^{n-1} \cdot u_h^{n-1}, v \rangle +
39 \langle \nu \nabla u_h^{\star}, \nabla v \rangle =
40 \langle f, v \rangle .
41
42 Here, :math:`\langle \cdot, \cdot \rangle` denotes the
43 :math:`L^2(\Omega)` inner product. Subsequently, the velocity is
44 projected to the space of divergence free fields. This step may be
45 obtained by subtracting the variational problem for the tentative
46 velocity :math:`u_h^{\star}` from the incompressible Navier-Stokes
47 equations and using the continuity equation. One obtains the following
48 pair of equations for computing the velocity :math:`u_h^n` and
49 pressure :math:`p_h^n` at time :math:`t = t_n` based on the tentative
50 velocity :math:`u_h^{\star}`:
51
52 .. math::
53
54 \langle \nabla p^n, \nabla q \rangle &=
55 - \langle \nabla \cdot u_h^{\star}, q \rangle / \Delta t_n, \\
56 \langle u_h^n, v \rangle &=
57 \langle u_h^{\star}, v \rangle - \Delta t_n \langle \nabla p^n, v \rangle.
58
59 Problem definition
60 ------------------
61
62 In this demo, we solve the incompressible Navier-Stokes equations on
63 an L-shaped domain. The L-shape is the subset of the unit square
64 obtained by removing the upper right quadrant.
65
66 The flow is driven by an oscillating pressure
67 :math:`p_{\mathrm{in}}(t) = \sin 3t` at the inflow :math:`y = 1` while
68 the pressure is kept constant :math:`p_{\mathrm{out}} = 0` at the
69 outflow :math:`x = 1`.
70
71 At the inflow and outflow, we impose free flow ("do nothing") boundary
72 conditions for the velocity and no-slip boundary conditions on the
73 remaining boundary.
74
75 The (kinematic) viscosity is set to :math:`\nu = 0.01`, the time step
76 is :math:`\Delta t = 0.01` and the length of the time interval is
77 :math:`T = 3`. The solution is computed using continuous vector-valued
78 piecewise quadratics for the velocity, and continuous scalar piecewise
79 linears for the pressure (the Taylor-Hood elements).
80
81 With the above input the solution for the velocity :math:`u` and pressure
82 :math:`p` will look as follows:
83
84 .. image:: ../navier-stokes_u.png
85 :scale: 75
86 :align: center
87
88 .. image:: ../navier-stokes_p.png
89 :scale: 75
90 :align: center
91