Final Exam

8.6.2020

1. Use the method of Frobenius to …nd a series solution for the di¤erential equation

x(2 x)y 00 xy 0 + 9y = 0

around the point x0 = 0:

Remark We have to follow the steps suggested by the method of Frobenius, this isnot just a series
method it is a method which guarentees to …nd at least one solution to the equation and can be applied
if the point x0 is a regular singular point of the di¤erential equation. First we have to …nd the indicial
equation and use it to determine
P1 the index s which we use to determine the form of the solution
y(x) and we put y(x) = xs n=0 an xn : Putting this in the given equation we determine the recursive
relations among the coe¢ cients from which we can either determine the function explicitly or at least
su¢ ciently many nonzero terms of the series expansion.
Answer
x0 = 0 is a regular singular point of this equation , because all the coe¢ cients are analytic functions,
the leading coe¢ ent vanishes at x0 and both limits
x 9
lim (x + x0 ) = 0 and lim (x + x0 )2 =0
x!x0 x(2 x) x!x0 x(2 x)

exist. The indicial equation is then

r(r 1) = 0
and it hasPtwo roots r1P= 0 and r2 = 1: So we choose the bigger root and put s = r2 and
1 1
y(x) = xs n=0 an xn = n=0 an xn+1 with a0 =6 0:
We get
P1 P1 P1
(2x x2 ) n=0 an (n + s)(n + s 1)xn+s 2
x n=0 an (n + s)xn+s 1
+9 n=0 an xn+s = 0
P1 P1
n=0 2an (n + 1)nxn n=0 [an (n + 1)n + an (n + 1) 9an ]xn+1 = 0
P1 P1
n=0 2an+1 (n + 2)(n + 1)xn+1 n=0 an [(n + 1)
2
9]xn+1 = 0:

Thus we obtain
2an+1 (n + 2)(n + 1) an [(n + 1)2 9] = 0
and hence
(n + 4)(n 2)
an+1 = an ( )
2(n + 2)(n + 1)
for all n = 0: In particular for n = 2 we get a3 = 0: And then ( ) shows that ak = 0 for all k = 3
5 5
and a1 = 2; a2 = a1 = a0 : Therefore we can …nd the solution function explicitly as a scalar
12 6
multiple of
5
y(x) = x 2x2 + x3 :
6
2. Solve the initial value problem

1 if 0 t < 1
y 00 + 4y 0 = g(t); y(0) = 0; y 0 (0) = 1 where g(t) =
0 if ...... 1 < t

Remark Observe that we have a linear di¤erential equation with constant coe¢ cients and incprinciple
we can apply the general method of getting …rst the general solution of this equation as the sum of the
general solution yh of the corresponding homogeneous di¤erential equation and one particular solution
yp of the given equation. But since the right hand side of the equation is a discontinuous function it
is not a linear combination of functions of the form tm eat cos bt or tm eat sin bt with positive integer m
and real numbers a and b it is not possible to apply UC-method and the other method we could apply
namely the method of variation of parameters to the equation in standard form

y 00 + p(t)y 0 + q(t)y = g(t)

needs the functions p; q and g to be continuous functions (Compare Theorem 7 on page 197 of the
text book). The method we have learned to deal with this kind of problem is the method of Laplace
transform.
Answer
First observe that
g(t) = 1 u(t 1)
where u(t) is the unit step function.
Taking the Laplace transform of both sides of the given equation and putting Y (s) = Lfy(t)](s) we get
s
1 e
s2 Y (s) 1 + 4sY (s) =
s
which shows that
s+1 e s
Y (s) = ;
s(s2 + 4s)

As
1 A B C
= + 2+
s2 (s + 4) s s s+4
1 1 1 1 1 1 1
implies that B = ;C = and = A + + and hence A = :
4 16 5 4 16 5 16
Furthermore
s+1 1 A B C
= + + 2+
s2 (s + 4) s(s + 4) s s s+4
1 1 1 1 1 1
= ( ) + 2+
4 s s+4 16s 4s 16(s + 4)
3 1 1 1 3 1
= + ;
16 s 4 s2 16 s + 4
So we have
3 1 1 1 3 1 s 1 1 1 1 1 1
Y (s) = + e ( + + )
16 s 4 s2 16 s + 4 16 s 4 s2 16 s + 4
and
3 1 3 4t 1 1 1 4(t 1)
y(t) = + t e ( + (t 1) + e )u(t 1)
16 4 16 16 4 16
3
16 + 14 t 16
3
e 4t if 0 t < 1
y(t) = 1 e 4
+3 :
2 e 4t ( 16 ) if ...... 1 < t
3. Find the general solution of the di¤erential equation

y 00 6y 0 + 9y = e3x + 50 sin x:

Remark A large number of students have found a particular solution to this equation by the so-called
operator method. which I did not mention in my lectures. So actually they should explain what that
method is and why it is possible to apply that method to this concrete problem.
Answer This is a nonhomogeneous linear di¤erential equation with constant coe¢ cients of order
two.The characteristic equation of the corresponding homogeneous equation is

r2 6r + 9 = 0

and it has only one real root with multiplicity 2, namely r = 3; So the general solution of the corre-
sponding homogeneous equation is

yh (t) = C1 e3x + C2 xe3x :

A particular solution yp (t) for the nonhomogeneous equation can be obtained by the UC-method.
According to the rule of the UC-method we have to put

yp (t) = Ax2 e3x + B cos x + C sin x

and get

e3x + 50 sin x = 9(Ax2 e3x + B cos x + C sin x) +

6(3Ax2 e3x + 2Axe3x B sin x + C cos x) +
9Ax2 e3x + 12Axe3x + 2Ae3x B cos x C sin x

e3x + 50 sin x = 2Ae3x + (8B 6C) cos x + (8C + 6B) sin x:

1
So we obtain A = ; 8B 6C = 0; 8C + 6B = 50 and hence B = 3; C = 4: Therefore the general
2
solution of the given equation is
1
y(t) = yh (t) + yp (t) = C1 e3x + C2 xe3x + x2 e3x + 3 cos x + 4 sin x:
2
4. Find the general solution of the di¤erential equation
1 2
(x y + 2y + 3x)dx + (x + 2y)dy = 0

1 2 @M 1 @N
Let M (x; y) = x y + 2y + 3x and N (x; y) = x + 2y: Then we get that = 2x y + 2 and =1
@y @x
and hence the given equation is not exact but as
@M @N
@y @x 2x 1 y + 1 1
= =
N x + 2y x
R
1
depend only on x we see that it has an integrating factor depending only on x, namely =e x dx =
eln x = x: The equation we obtain by multiplying the given equation by = x namely

(y 2 + 2xy + 3x2 )dx + (x2 + 2xy)dy = 0;

is exact. So there exists a function F (x; y) such that

dF (x; y) = Kdx + Ldy

@F @F
and this means that = K and = L where K(x; y) = y 2 + 2xy + 3x2 and L(x; y) = x2 + 2xy:
@x @y
So we get R
F (x; y) = K(x; y)dx = xy 2 + x2 y + x3 + '(y)
and hence
@F
2xy + x2 + '0 (y) = (x; y) = x2 + 2xy
@y
Thus we can take '(y) = 0 and F (x; y) =; xy 2 + x2 y + x3 : Therefore the general solution of the given
equation is
xy 2 + x2 y + x3 = C:

Remark One can see tha the equation is a homogeneous equations because both M / and N are
homogeneous of the same degree 2 / Therefore one can solve this di¤erential equation by applying …rst
the substitution y = vx which transforms the equation to

3x(v 2 + v + 1)dx + (1 + 2v)x2 dv = 0

which is separable. We get

1 + 2v 3
2
dv = dx
1+v+v x
implying that ln[(1 + v + v 2 )x3 ] = K and hence

xy 2 + x2 y + x3 = C: