Faculty of Engineering

Department of Engineering Mathematics

Method of Characteristics: Worked Examples

———————————————————————————————————————————
Method of Characteristics:
Suppose we are given a first order PDE a(x, y, u)ux + b(x, y, u)yy = c(x, y, u).

I. With x = x(s), y = y(s), u = u(s), the parametric form of the characteristic equations:
dx dy du
= a, = b, =c
ds ds ds

dx a
II. The solutions of = or the family of curves parameterized by s,
dy b

{(x(s), y(s)) : 0 Æ s < Œ}

are called characteristics or characteristics curves of the PDE.

1. Characteristics curves do not intersect and along the characteristics, the PDE becomes an
ordinary differential equation and has solution u = constant.

Worked examples: Solve the following PDEs using the method of characteristics.

1. aux + buy = 0 where a, b are a constants.

dx dy du
= a, = 1, =0
ds ds ds
Integrating first two equations we get x = as + k, y = bs where k is a constant.
Therefore the characteristics of the PDE are straight lines given by bx ≠ ay = k
The solution is u = constant along a characteristic curve but the constant may be different
on different characteristic curves. As k gives us a different characteristic curve, this implies
that u = F (k) where F is an arbitrary function determined by the initial condition. Thus
the solution to the PDE is given by

u(x, y) = F (bx ≠ ay).

Note:
The PDE cux + ut = 0 for ≠Œ < x < Œ is called the one dimensional wave equation or
transport equation. With c is a constant, the solution is u(x, t) = F (x ≠ ct). c is called the
speed of the wave and if c > 0 wave moves to the right and if c < 0 wave moves to the left.
For example if the initial condition is given by u(x, 0) = sin x then F (x) = sin x and the
solution is u = sin(x ≠ ct).

1
2. 4ux ≠ 3uy = 0; u(0, y) = y 3
Proceeding as with previous example with a = 4 and b = ≠3, the solution is u(x, y) =
F (≠3x ≠ 4y). Applying the auxiliary condition(IC or BC),we have u(0, y) = y 3 = F (≠4y).
w
Let w = ≠4y. This gives y = ≠ .
4
w3
) F (w) = ≠ .
64
(≠3x ≠ 4y)3 (3x + 4y)3
Thus the solution u(x, y) = ≠ = .
64 64
3. ux + yuy = 0.
dx dy du
= 1, = y, =0
ds ds ds
Integrating first two equations we get x = s, y = kes where k is a constant.
Therefore the characteristics of the PDE are curves given by y = kex .
The solution is u = constant along a characteristic curve but the constant may be different
on different characteristic curves. As k gives us a different characteristic curve, this implies
that u = F (k) where F is an arbitrary function determined by the initial condition. Thus
the solution to the PDE is given by

u(x, y) = F (ye≠x ).

Note:
The PDE c(x)ux + ut = 0 for ≠Œ < x < Œ is called the non-uniform transport equation
where c depends on the spacial position. For example with c(x) = ≠x, u(x, t) = F (xet ). If
1 1 1
IC is u(x, 0) = then F (x) = and u(x, t) = .
1+x 2 1+x 2 1 + x2 e2t
4. yux ≠ xuy = 0.
dx dy du
= y, = ≠x, =0
ds ds ds
In this case it is better to combine the first two equations and integrate. This gives ydx +
xdy = 0 and thus x2 + y 2 = k.
Therefore the characteristics of the PDE are concentric circles with the centre at the origin.
Integrating the third equation, the solution u(x, y) = constant = F (k) = F (x2 + y 2 ).
5. ≠2ux + 4uy = ≠5u.
dx dy du
= ≠2, = 4, = 5u
ds ds ds
Integrating first two equations we get x = ≠2s, y = 4s + k where k is a constant.
Therefore the characteristics of the PDE are straight lines given by 2x + y = k.
Integrating the third equation, u = F (k)e≠5s where F is an arbitrary function.
Substituting for k = 2x + y gives u(x, y) = F (2x + y)e5x/2 .
6. ut + 2xtux = u with IC u(x, 0) = x.
dt dx du
= 1, = 2xt, =u
ds ds ds 2
Integrating, t = s, x = kes and u = F (k)es ).

2
 2 2
Therefore the characteristics of the PDE are x = ket and u(x, t) = F (xe≠t )et . Applying
2 2
IC gives, F (x) = x and the solution is u(x, t) = xe≠t et = xet≠t .

7. ux + 2xuy = 2xu.
dx dy du
= 1, = 2x, = 2xu
ds ds ds
Integrating the first equation, x = s.
Put x = s in the second and integrate will give, y = s2 + k.
Therefore the characteristics of the PDE are y ≠ x2 = k.
2
Put x = s in the third equation and integrate will give, u = F (k)es .
2
Thus the solution u(x, y) = F (y ≠ x2 )ex .
Suppose we are given u(x, 0) = x2 .
2
Then F (≠x2 )ex = x2 and with ≠x2 = w, F (w) = ≠wew .
2 2
Thus the solution is u(x, y) = (x2 ≠ y)e(y≠x ) ex = (x2 ≠ y)ey .

8. xux + yuy = xe≠u with auxiliary condition, u = 0 on y = x2 .

dx dy du
= x, = y, = xe≠u
ds ds ds
Here it is much easier if we consider the equivalent form of the above equations,
dx du
= x/y, = e≠u
dy dx
Integrating, y = kx and eu = x + F (k) = x + F (y/x).
Therefore the characteristics of the PDE are y = kx.
Applying u = 0 0n y = x2 gives,3 F (x) = 14≠ x =∆ F (y/x) = 1 ≠ y/x.
y
Thus the solution u(x, y) = ln x + 1 ≠ .
x
9. x2 ux + uuy = 1 with u = 0 on x + y = 1.
dx dy du
= x2 , = u, = 1.
ds ds ds
Integrating the first and the third give, ≠1/x = s and u = s + k.
Put u = s + k1 in the second and integrating gives, y = s2 /2 + ks + F (k).
s2 1 1 1 1
) y= + (u ≠ s)s + F (u ≠ s) = 2 + (u + )(≠ ) + F (u + ).
2 2x x x x
Now apply u = 0 on x + y = 1.
1 1 1
1 ≠ x = 2 ≠ 2 + F ( ).
2x x x
1 1 1 1
Thus F ( ) = 2 + 1 ≠ x. Put v = 1/x, then F (v) = v 2 + 1 ≠ .
x x 2 v
u2 x
) y =1+ ≠ .
2 (ux + 1)
In this case we cannot solve this explicitly for u and the solution is called the implicit
solution.
Ô
10. u2 ux + ut = 0; x > 0, t > 0 with u(x, 0) = x and u(0, t) = 0. This is an IBVP.
dx dt du
= u2 , = 1, = 0.
ds ds ds

3
Integrating the second and the third gives, t = s and u = constant.
Thus integrating the first gives x = su2 + k = tu2 + k.
Thus u = ÔF (k) = F (x ≠ tu2Ô).
u(x, 0) = x Ò gives F (x) = x.
) u(x, y) = (x ≠ tu2 ). This is the implicit solution.
Squaring both sides, we can manipulate this solution into an explicit solution.

u2 = x ≠ tu2
x
u2 =
(1 + t)
Ú
x
Thus the solution u = , x > 0, t > 0.
1+t