Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Standard view
Full view
of .

0 of .
Results for:
P. 1
1diferen Method

# 1diferen Method

Ratings: (0)|Views: 132 |Likes:

### Availability:

See more
See less

10/20/2011

pdf

text

original

Chapter 3
Finite Difference Methods
“Those who know others are clever; those who know themselves have discern-ment; thosewhoovercomeothershaveforce; thosewhoovercomethemselvesarestrong; those who know contentment are rich; those who persevere are people of  purpose.”
Paraphrase of Lao Tzu
3.1 Introduction
It is rare for real-life EM problems to fall neatly into a class that can be solved bythe analytical methods presented in the preceding chapter. Classical approaches mayfail if [1]:the PDE is not linear and cannot be linearized without seriously affecting theresultthe solution region is complexthe boundary conditions are of mixed typesthe boundary conditions are time-dependentthe medium is inhomogeneous or anisotropicWhenever a problem with such complexity arises, numerical solutions must be em-ployed. OfthenumericalmethodsavailableforsolvingPDEs, thoseemployingﬁnitedifferences are more easily understood, more frequently used, and more universallyapplicable than any other.Theﬁnitedifferencemethod(FDM)wasﬁrstdevelopedbyA.Thom[2]inthe1920sunder the title “the method of squares” to solve nonlinear hydrodynamic equations.Sincethen,themethodhasfoundapplicationsinsolvingdifferentﬁeldproblems. Theﬁnite difference techniques are based upon approximations which permit replacingdifferential equations by ﬁnite difference equations. These ﬁnite difference approx-imations are algebraic in form; they relate the value of the dependent variable at a

© 2001 by CRC PRESS LLC

point in the solution region to the values at some neighboring points. Thus a ﬁnitedifference solution basically involves three steps:(1) dividing the solution region into a grid of nodes(2) approximating the given differential equation by ﬁnite difference equivalentthat relates the dependent variable at a point in the solution region to its valuesat the neighboring points(3) solving the difference equations subject to the prescribed boundary conditionsand/or initial conditionsThe course of action taken in three steps is dictated by the nature of the problembeing solved, the solution region, and the boundary conditions. The most commonlyused grid patterns for two-dimensional problems are shown inFig. 3.1. A three-dimensional grid pattern will be considered later in the chapter.
Figure 3.1Common grid patterns: (a) rectangular grid, (b) skew grid, (c) triangular grid,(d) circular grid.
3.2 Finite Difference Schemes
Before ﬁnding the ﬁnite difference solutions to speciﬁc PDEs, we will look at howone constructs ﬁnite difference approximations from a given differential equation.This essentially involves estimating derivatives numerically.Given a function
f(x)
shown inFig. 3.2,we can approximate its derivative, slope or the tangent at P by the slope of the arc PB, giving the
forward-difference
formula,

© 2001 by CRC PRESS LLC

Figure 3.2Estimates for the derivative of
f(x)
at
using forward, backward, and centraldifferences.
(x
o
)
f(x
o
+
x)
f(x
o
)x
(3.1)or the slope of the arc AP, yielding the
backward-difference
formula,
(x
o
)
f(x
o
)
f(x
o
x)x
(3.2)or the slope of the arc AB, resulting in the
central-difference
formula,
(x
o
)
f(x
o
+
x)
f(x
o
x)
2
x
(3.3)We can also estimate the second derivative of
f(x)
at P as

(x
o
)
(x
o
+
x/
2
)
(x
o
x/
2
)x
=
1
x
f(x
o
+
x)
f(x
o
)x
f(x
o
)
f(x
o
x)x
or

(x
o
)
f(x
o
+
x)
2
f(x
o
)
+
f(x
o
x)(x)
2
(3.4)Any approximation of a derivative in terms of values at a discrete set of points iscalled
ﬁnite difference
approximation.

© 2001 by CRC PRESS LLC