0% found this document useful (0 votes)
277 views4 pages

13 PDF

This document discusses reducing second-order partial differential equations to canonical forms through transformations of the independent variables. There are three cases: hyperbolic (real distinct roots), parabolic (repeated real root), and elliptic (complex roots). The transformations determine characteristic curves whose solutions define new variables that put the PDEs into simple forms like uξη=φ, uηη=φ, or uαα+uββ=φ. Examples demonstrate reducing specific PDEs to their canonical forms.

Uploaded by

Amritansh Ranjan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
277 views4 pages

13 PDF

This document discusses reducing second-order partial differential equations to canonical forms through transformations of the independent variables. There are three cases: hyperbolic (real distinct roots), parabolic (repeated real root), and elliptic (complex roots). The transformations determine characteristic curves whose solutions define new variables that put the PDEs into simple forms like uξη=φ, uηη=φ, or uαα+uββ=φ. Examples demonstrate reducing specific PDEs to their canonical forms.

Uploaded by

Amritansh Ranjan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

MODULE 3: SECOND-ORDER PARTIAL DIFFERENTIAL EQUATIONS 7

Lecture 2 Canonical Forms or Normal Forms

By a suitable change of the independent variables we shall show that any equation of the
form
Auxx + Buxy + Cuyy + Dux + Euy + F u + G = 0, (1)

where A, B, C, D, E, F and G are functions of the variables x and y, can be reduced to a


canonical form or normal form. The transformed equation assumes a simple form so that
the subsequent analysis of solving the equation will be become easy.

Consider the transformation of the indpendent variables from (x, y) to (ξ, η) given by

ξ = ξ(x, y), η = η(x, y). (2)

Here, the functions ξ and η are continuously differentiable and the Jacobian

∂(ξ, η) ξx ξy
J= = = (ξx ηy − ξy ηx ) ̸= 0 (3)
∂(x, y) ηx ηy

in the domain where (1) holds.

Using chain rule, we notice that

ux = uξ ξx + uη ηx
uy = uξ ξy + uη ηy
uxx = uξξ ξx2 + 2uξη ξx ηx + uηη ηx2 + uξ ξxx + uη ηxx
uxy = uξξ ξx ξy + uξη (ξx ηy + ξy ηx ) + uηη ηx ηy + uξ ξxy + uη ηxy
uyy = uξξ ξy2 + 2uξη ξy ηy + uηη ηy2 + uξ ξyy + uη ηyy

Substituting these expression into (1), we obtain

Ā(ξx , ξy )uξξ + B̄(ξx , ξy ; ηx , ηy )uξη + C̄(ηx , ηy )uηη = F (ξ, η, u(ξ, η), uξ (ξ, η), uη (ξ, η)), (4)

where

Ā(ξx , ξy ) = Aξx2 + Bξx ξy + Cξy2


B̄(ξx , ξy ; ηx , ηy ) = 2Aξx ηx + B(ξx ηy + ξy ηx ) + 2Cξy ηy
C̄(ηx , ηy ) = Aηx2 + Bηx ηy + Cηy2 .

An easy calculation shows that

B̄ 2 − 4ĀC̄ = (ξx ηy − ξy ηx )2 (B 2 − 4AC). (5)


MODULE 3: SECOND-ORDER PARTIAL DIFFERENTIAL EQUATIONS 8

The equation (5) shows that the transformation of the independent variables does not
modify the type of PDE.

We shall determine ξ and η so that (4) takes the simplest possible form. We now
consider the following cases:
Case I: B 2 − 4AC > 0 (Hyperbolic type)
Case II: B2 − 4AC = 0 (Parabolic type)
Case III: B2 − 4AC < 0 (Elliptic type)
Case I: Note that B 2 − 4AC > 0 implies the equation Aα2 + Bα + C = 0 has two real
and distinct roots, say λ1 and λ2 . Now, choose ξ and η such that
∂ξ ∂ξ ∂η ∂η
= λ1 and = λ2 . (6)
∂x ∂y ∂x ∂y
Then the coefficients of uξξ and uηη will be zero because

Ā = Aξx2 + Bξx ξy + Cξy2 = (Aλ21 + Bλ1 + C)ξy2 = 0,


C̄ = Aηx2 + Bηx ηy + Cηy2 = (Aλ22 + Bλ2 + C)ηy2 = 0.

Thus, (5) reduces to


B̄ 2 = (B 2 − AC)(ξx ηy − ξy ηx )2 > 0
as B 2 −4AC > 0. Note that (6) is a first-order linear PDE in ξ and η whose characteristics
curves are satisfy the first-order ODEs
dy
+ λi (x, y) = 0, i = 1, 2. (7)
dx
Let the family of curves determined by the solution of (7) for i = 1 and i = 2 be

f1 (x, y) = c1 and f2 (x, y) = c2 , (8)

respectively. These family of curves are called characteristics curves of PDE (5). With
this choice, divide (4) throughout by B̄ (as B̄ > 0) and use (7)-(8) to obtain
∂2u
= ϕ(ξ, η, u, uξ , uη ), (9)
∂ξ∂η
which is the canonical form of hyperbolic equation.
EXAMPLE 1. Reduce the equation uxx = x2 uyy to its canonical form.

Solution. Comparing with (1) we find that A = 1, B = 0, C = −x2 .

The roots of the equations Aα2 + Bα + C = 0 i.e., α2 + x2 = 0 are given by λi = ±x.


The differential equations for the family of characteristics curves are
dy
± x = 0.
dx
MODULE 3: SECOND-ORDER PARTIAL DIFFERENTIAL EQUATIONS 9

whose solutions are y + 12 x2 = c1 and y − 12 x2 = c2 . Choose

1 1
ξ = y + x2 , η = y − x2 .
2 2
An easy computation shows that

ux = uξ ξx + uη ηx ,
uxx = uξξ ξx2 + 2uξη ξx ηx + uηη ηx2 + uξ ξxx + uη ηxx
= uξξ x2 − 2uξη x2 + uηη x2 + uξ − uη ,
uyy = uξξ ξy2 + 2uξη ξy ηy + uηη ηy2 + uξ ξyy + uη ηyy ,
= uξξ + 2uξη + uηη .

Substituting these expression in the equation uxx = x2 uyy yields

4x2 uξη = (uξ − uη )


1
or 4(ξ − η)uξη = (uξ − uη )
4(ξ − η)
1
or uξη = (uξ − uη )
4(ξ − η)

which is the required canonical form.

CASE II: B 2 − 4AC = 0 =⇒ the equation Aα2 + Bα + C = 0 has two equal roots, say
dy
λ1 = λ2 = λ. Let f1 (x, y) = c1 be the solution of dx + λ(x, y) = 0. Take ξ = f1 (x, y) and
η to be the any function of x and y which is independent of ξ.

As before, Ā(ξx , ξy ) = 0 and hence from equation (5), we obtain B̄ = 0. Note that
C̄(ηx , ηy ) ̸= 0, otherwise η would be a function of ξ. Dividing (4) by C̄, the canonical form
of (2) is
uηη = ϕ(ξ, η, u, uξ , uη ). (10)

which is the canonical form of parabolic equation.


EXAMPLE 2. Reduce the equation uxx + 2uxy + uyy = 0 to canonical form.

Solution. In this case, A = 1, B = 2, C = 1. The equation α2 + 2α + 1 = 0 has equal


dy
roots λ = −1. The solution of dx − 1 = 0 is x − y = c1 Take ξ = x − y. Choose η = x + y.
proceed as in Example 1 to obtain uηη = 0 which is the canonical form of the given PDE.

CASE III: When B 2 − 4AC < 0, the roots of Aα2 + Bα + C = 0 are complex. Following
the procedure as in CASE I, we find that

uξη = ϕ1 (ξ, η, u, uξ , uη ). (11)


MODULE 3: SECOND-ORDER PARTIAL DIFFERENTIAL EQUATIONS 10

The variables ξ, η are infact complex conjugates. To get a real canonical form use the
transformation
1 1
α = (ξ + η), β = (ξ − η)
2 2i
to obtain
1
uξη = (uαα + uββ ), (12)
4
which follows from the following calculation:
1 1
uξ = uα αξ + uβ βξ = uα + uβ
2 2i
1 1
uξη = (uαα αη + uαβ βη ) + (uβα αη + uββ βη )
2 2i
1
= (uαα + uββ ).
4
The desired canonical form is

uαα + uββ = ψ(α, β, u(α, β), uα (α, β), uβ (α, β)). (13)

EXAMPLE 3. Reduce the equation uxx + x2 uyy = 0 to canonical form.

Solution. In this case, A = 1, B = 0, C = x2 . The roots are λ1 = ix, λ2 = −ix.


Take ξ = iy + 12 x2 , η = −iy + 12 x2 . Then α = 21 x2 , β = y. The canonical form is
1
uαα + uββ = − uα .

Practice Problems

1. Reduce the following equations to canonical/normal form:

(A) 2uxx − 4uxy + 2uyy + 3u = 0.


(B) uxx + yuyy = 0.
(C) uxy + ux + uy = 2x.

2. Show that the equation

uxx − 6uxy + 12uyy + 4ux − u = sin(xy)

is of elliptic type and obtain its canonical form.

3. Determine the regions where Tricomi’s equation uxx + xuyy = 0 is of elliptic,


parabolic, and hyperbolic types. Obtain its characteristics and its canonical form in
the hyperbolic region.

You might also like