forked from johannesgerer/jburkardt-f
-
Notifications
You must be signed in to change notification settings - Fork 1
/
multigrid_poisson_1d.html
216 lines (183 loc) · 5.87 KB
/
multigrid_poisson_1d.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
<html>
<head>
<title>
MULTIGRID_POISSON_1D - Multigrid Solver for 1D Poisson Problem
</title>
</head>
<body bgcolor="#EEEEEE" link="#CC0000" alink="#FF3300" vlink="#000055">
<h1 align = "center">
MULTIGRID_POISSON_1D <br> Multigrid Solver for 1D Poisson Problem
</h1>
<hr>
<p>
<b>MULTIGRID_POISSON_1D</b>
is a FORTRAN90 library which
applies a multigrid method to solve the linear system associated
with a discretized version of the 1D Poisson equation.
</p>
<p>
The 1D Poisson equation is assumed to have the form
<pre>
-u''(x) = f(x), for a < x < b
u(a) = ua, u(b) = ub
</pre>
</p>
<p>
Let K be a small positive integer called the mesh index, and let
N = 2^K be the corresponding number of uniform subintervals into which
[A,B] is divided. Assigning a value U(I) to each of the N+1 equally
spaced nodes with coordinate X(I), we approximate the equation by
<pre>
-U(i-1) + 2 U(i) - U(i+1)
------------------------- = f( X(i) ), 1 < I < N+1
h^2
U(1) = ua, U(N+1) = ub.
</pre>
</p>
<p>
It remains to solve the linear system for the desired values of U.
This could be done directly, or iteratively. An iterative method
such as Jacobi, Gauss-Seidel or SOR might be suitable, but experience
shows that the convergence rate of these iterative methods decreases
drastically as the value of K is increased - that is, as a more
refined and accurate answer is sought.
</p>
<p>
The multigrid method defines a nested set of grids, and corresponding
solutions, to the problem, and applies an iterative linear solver.
By transfering information from one grid to a finer or coarser one,
a more rapid convergence behavior can be encouraged.
</p>
<h3 align = "center">
Licensing:
</h3>
<p>
The computer code and data files described and made available on this
web page are distributed under
<a href = "../../txt/gnu_lgpl.txt">the GNU LGPL license.</a>
</p>
<h3 align = "center">
Languages:
</h3>
<p>
<b>MULTIGRID_POISSON_1D</b> is available in
<a href = "../../c_src/multigrid_poisson_1d/multigrid_poisson_1d.html">a C version</a> and
<a href = "../../cpp_src/multigrid_poisson_1d/multigrid_poisson_1d.html">a C++ version</a> and
<a href = "../../f77_src/multigrid_poisson_1d/multigrid_poisson_1d.html">a FORTRAN77 version</a> and
<a href = "../../f_src/multigrid_poisson_1d/multigrid_poisson_1d.html">a FORTRAN90 version</a> and
<a href = "../../m_src/multigrid_poisson_1d/multigrid_poisson_1d.html">a MATLAB version</a>.
</p>
<h3 align = "center">
Related Data and Programs:
</h3>
<p>
<a href = "../../f_src/cyclic_reduction/cyclic_reduction.html">
CYCLIC_REDUCTION</a>,
a FORTRAN90 library which
solves a tridiagonal linear system using cyclic reduction.
</p>
<p>
<a href = "../../f_src/fd1d_bvp/fd1d_bvp.html">
FD1D_BVP</a>,
a FORTRAN90 program which
applies the finite difference method
to a two point boundary value problem in one spatial dimension.
</p>
<p>
<a href = "../../f_src/mgmres/mgmres.html">
MGMRES</a>,
a FORTRAN90 library which
applies the restarted GMRES algorithm to solve a sparse linear system,
by Lili Ju.
</p>
<h3 align = "center">
Reference:
</h3>
<p>
<ol>
<li>
William Briggs, Van Emden Henson, Steve McCormick,<br>
A Multigrid Tutorial,<br>
SIAM, 2000,<br>
ISBN13: 978-0-898714-62-3,<br>
LC: QA377.B75.
</li>
<li>
William Hager,<br>
Applied Numerical Linear Algebra,<br>
Prentice-Hall, 1988,<br>
ISBN13: 978-0130412942,<br>
LC: QA184.H33.
</li>
</ol>
</p>
<h3 align = "center">
Source Code:
</h3>
<p>
<ul>
<li>
<a href = "multigrid_poisson_1d.f">multigrid_poisson_1d.f</a>, the source code.
</li>
<li>
<a href = "multigrid_poisson_1d.sh">multigrid_poisson_1d.sh</a>,
BASH commands to compile the source code.
</li>
</ul>
</p>
<h3 align = "center">
Examples and Tests:
</h3>
<p>
<ul>
<li>
<a href = "multigrid_poisson_1d_prb.f">multigrid_poisson_1d_prb.f</a>,
a sample calling program.
</li>
<li>
<a href = "multigrid_poisson_1d_prb.sh">multigrid_poisson_1d_prb.sh</a>,
BASH commands to compile and run the sample program.
</li>
<li>
<a href = "multigrid_poisson_1d_prb_output.txt">multigrid_poisson_1d_prb_output.txt</a>,
the output file.
</li>
</ul>
</p>
<h3 align = "center">
List of Routines:
</h3>
<p>
<ul>
<li>
<b>MONOGRID_POISSON_1D</b> solves a 1D PDE, using the Gauss-Seidel method.
</li>
<li>
<b>MULTIGRID_POISSON_1D</b> solves a 1D PDE using the multigrid method.
</li>
<li>
<b>CTOF</b> transfers data from a coarse to a finer grid.
</li>
<li>
<b>FTOC</b> transfers data from a fine grid to a coarser grid.
</li>
<li>
<b>GAUSS_SEIDEL</b> carries out one step of a Gauss-Seidel iteration.
</li>
<li>
<b>TIMESTAMP</b> prints out the current YMDHMS date as a timestamp.
</li>
</ul>
</p>
<p>
You can go up one level to <a href = "../f_src.html">
the FORTRAN90 source codes</a>.
</p>
<hr>
<i>
Last revised on 25 November 2011.
</i>
<!-- John Burkardt -->
</body>
<!-- Initial HTML skeleton created by HTMLINDEX. -->
</html>