Chapter 4
Second-order differential
equations
Second-order linear differential equations play a crucial role in the
study of differential equations for two primary reasons. First, they
possess a well-developed theoretical framework that supports
various systematic solution methods. Moreover, much of this
theory and these methods can be understood at a relatively
elementary mathematical level.
The second reason for their importance is that these equations are
indispensable in numerous applications, spanning a wide array of
disciplines. For instance, in mechanical engineering, they are used
to analyze the vibrations of structures and mechanical systems. In
electrical engineering, they govern the behavior of circuits with
inductance, capacitance, and resistance. In fluid dynamics, they
describe the flow of fluids under various conditions, while in
thermal analysis, they model heat distribution and transfer within
materials. Moreover, second-order differential equations are
pivotal in wave mechanics, where they describe the propagation of
sound, light, and water waves. Their versatility and applicability
make them a cornerstone in the toolkit of scientists and engineers
alike.
A linear second-order differential equation is written in the form
P(t)y + Q(t)y′ + R(t)y = G(t) (4.1)
With a non-zero the right-hand side function G(t), (4.1) is
nonhomogeneous. If the right-hand side is zero, (4.1) is a
homogeneous equation.
Finding solution of (4.1) would conceptually involve two
integrations that will introduce two constants of integrations. To
obtain these two constants, we may assign two conditions at an
initial time t0, such as
y(t0) = y0, and y′(t0) = y1
This will make the problem (4.1) an initial-value problem. This is
common when the independent variable is time. If the independent
variable is a space variable x, we may also assign two conditions at
two boundaries x1 and x2, such as
y(x1) = y1, and y(x2) = y2
This will make the problem (4.1) a boundary-value problem.
To attempt to find a procedure to solve (4.1), we start with a
simpler version of (4.1), then build on that. We consider that the
right-hand side of (4.1) is zero, and the coefficients P(t), Q(t) and
R(t) are constants. In other words, we will begin with an equation
like
ay + by′ + cy = 0 (4.2)
Homogeneous second-order equation with constant
coefficients
If we ponder on equation (4.2), since a, b and c are constants,
equation (4.2) demands that if y = f(t) is to be a solution of (4.2) its
first derivative f′(t) and second derivative f(t) algebraically must
also be of same form f(t) such that after multiplying with constants
a, b and c their sum could cancel out to give us zero. There is one
function that demonstrates this property, and that is the exponential
function ert. Therefore, we assume that our solution will be like ert
where r is yet an undetermined constant. To formalize, let
y = ert (4.3)
y′ = rert (4.4)
y = r2ert (4.5)
Substituting (4.3)-(4.5) in (4.2) we obtain
(ar2 + br + c)ert = 0 (4.6)
To satisfy (4.6) we either choose ert = 0 or ar2 + br + c = 0.
Considering ert = 0 would make our solution (4.3) zero. This
means that this is a trivial solution. Therefore, we disregard this,
and we consider the other possibility.
ar2 + br + c = 0 (4.7)
This is referred to as the characteristics equation for (4.2). This is
a quadratic equation whose roots would be r1 and r2. Therefore, our
solutions would be and . The roots r1 and r2 are referred to
as eigenvalues and the functions and are referred to as
eigenfunctions. These two eigenvalues, being roots of a quadratic
equation, can be of three types: real and distinct, complex
conjugates, and equal. The eigenfunctions behave differently in
each of these three categories. We will investigate this with some
examples.
Real distinct roots
Considering the two distinct real roots being r1 and r2, the general
solution is
We directly embark upon some examples.
Example 4.1 Real roots.
Find solution of the initial-value problem
y + 5y′ + 6y = 0 y(0) = 2 y′(0) = 3
We assume the solution to be of type y = ert, where r are roots of
the characteristics equation
r2 + 5r + 6 = 0
The roots of this equation are r1 = −2 and r2 = −3. Therefore, the
general solution is
y = c1e−2t + c2e−3t (4.8)
To substitute the initial conditions, we will need to differentiate
this
y′ = −2c1e−2t − 3c2e−3t (4.9)
Substituting the condition y(0) = 2 in (4.8) we get
c1 + c2 = 2 (4.10)
Substituting the condition y′(0) = 3 in (4.9) we get
−2c1 − 3c2 = 3 (4.11)
Solving (4.10) and (4.11) we get c1 = 9 and c2 = −7. Substituting
these in (4.8) we get the solution
y = 9e–2t − 7e–3t
The graph of the solution is shown in Figure 4.1.
We notice in Figure 4.1 that the solution has a maximum. In
engineering, the information regarding the maximum is important
for many reasons. This information is often of interest as it is
related to optimization. If the differential equation is related to
flow of some quantity, the maximum also refers to a location
Figure 4.1
where the potential is zero; hence no quantity will flow across this
location. At the point of maximum, the slope of the curve is zero.
Therefore, after substituting the values of c1 and c2 in (4.9), we set
this equal to zero. Therefore
−18e−2t + 21e−3t = 0 (4.12)
Solving this
This gives et = 1.166667. Taking natural logarithm on both sides,
we get t ≈ 0.1541.
Example 4.2 Real roots
Find the solution to the initial-value problem
y + y′ − 2y = 0 y(0) = 4 y′(0) = −5
Let the solution be y = ert, where r are roots of
r2 + r − 2 = 0
Figure 4.2
The roots of this equation are r1 = 1 and r2 = −2. Therefore, the
general solution is
y = c1et + c2e–2t (4.13)
To substitute the initial condition, we differentiate this
y′ = c1et − 2c2e–2t (4.14)
Therefore
y(0) = c1 + c2 = 4
y′(0) = c1 − 2c2 = −5
Solving these two, we obtain c1 = 1 and c2 = 3. Therefore, the
particular solution
y = et + 3e−2t
The solution is shown in Figure 4.2. Once again, we notice there is
a point of minimum. Solving for that, we get t ≈ 0.5972.
Example 4.3 Real roots
Find the solution of the initial-value problem
4y − 8y′ + 3y = 0 y(0) = 2 y′(0) =
Let the solution be y = ert, where r are roots of the characteristics
equation
4r2 − 8r + 3 = 0
The roots of this equation are r1 = and r2 = . The general
solution is
y = c1e3t/2 + c2et/2 (4.15)
Differentiating this
y′ = c1e3t/2 + c2et/2 (4.16)
Applying the initial conditions, we get
c1 + c2 = 2, c1 + c2 =
Solving these, we get c1 = , c2 = . Therefore, the particular
solution is
y = – e3t/2 + et/2
The solution is shown in Figure 4.3.
Complex conjugate roots
Figure 4.3
Next, we will embark upon discussing solutions when the roots of
the characteristics equation are complex conjugates. We foresee
that here we will encounter complex exponentials like e+i, where
i= .
Euler formula
Recalling the Taylor series for ex
(4.17)
We also recall the Taylor series for sine and cosine functions.
(4.18)
(4.19)
To find an expression for the complex exponential, we replace x
with ix in (4.17).
(4.20)
We write (ix)n as inxn. We recognize the fact that i2 = −1, i3 = −i,
i4 = 1, i5 = i, etc. Therefore when n is even there is an integer k
with n = 2k; in this case in = i2k = (−1)k. When n is odd, n = 2k + 1,
so in = i2k+1 = i(−1)k. This suggests separating the terms on the
right-hand side of (4.20) into its real and imaginary parts. The
result is
(4.21)
The first series in (4.21) is the series for cos x and the second series
is the series for sin x. Thus, we have
eix = cos x + i sin x (4.22)
Equation (4.22) is known as the Euler’s formula and is an
indispensable tool in physics and engineering.
For a second-order differential equation, when the eigenvalues of
the characteristics equation are complex conjugates, the solution
would still be of the form (4.3), except that the roots would be
complex. Let us assume that the roots are r1 = +i and r2 = −i.
Therefore, the solution is
y = b1e(+it + b2e(−it (4.23)
We write (4.23) as
y = et(b1eit + b2e−it) (4.24)
Using Euler’s formula (4.22) and collecting terms
y = et((b1 + b2) cos t + i(b1 − b2) sin t) (4.25)
We define new constant c1 = b1 + b2 and c2 = i(b1 − b2). Therefore,
the final form of solution for complex conjugate roots is
y = et(c1 cos t + c2 sin t) (4.26)
Example 4.4 Complex conjugate roots
Find the solution of the initial-value problem
y + y′ + 9.25y = 0 y(0) = 2 y′(0) = 8
Figure 4.4
Let the solution be of type y = ert, where r are the roots of the
equation
r2 + r + 9.25 = 0 (4.27)
The roots of 4.27 are and . Therefore,
the general solution is
/ (4.28)
Differentiating and substituting the initial conditions, we obtain
c1 = 2 and c2 = 3. Therefore, the particular solution is
/ (4.28)
The solution is shown in Figure 4.4.
Example 4.5 Complex roots
Find the solution to the initial-value problem
16y − 8y′ + 145y = 0 y(0) = −2 y′(0) = 1
Let the solution be of the type y = ert where r are roots of the
characteristics equation
16r2 – 8r + 145 = 0 (4.29)
Figure 4.5
The roots of equation (4.29) are . Therefore, the general
solution is
/ (4.30)
Differentiating and substituting the initial-conditions, we obtain the
particular solution
/ (4.31)
The solution is shown in Figure 4.5.
Example 4.6 Complex roots
Find the solution to the initial-value problem
y + 0.4y′ + 9.04y = 0 y(0) = 0 y′(0) = 3
Let the solution be of type y = ert where r are roots of the
characteristics equation
r2 + 0.4r + 9.04 = 0 (4.32)
Figure 4.6
The roots of equation (4.32) are . Therefore, the
general solution is
. (4.33)
Differentiating and substituting the initial-conditions, we obtain
c1 = 0 and c2 = 1. Therefore, the particular solution
. (4.34)
The solution is shown in Figure 4.6.
Repeated roots
If the characteristics equation has repeated roots, or in other words
equal roots, we have a fundamental problem. We have a second-
order differential equation, but we have one solution. A second-
order differential equation is supposed to have a second solution.
We are just stating it here without demonstrating it. If the
characteristics equation has just a single root r which will lead to a
solution ert, the second solution would be tert, so that the general
solution would be
(4.35)
Figure 4.7
Example 4.7 Repeated roots
Find solution of the initial-value problem
y + y′ + 0.25y = 0 y(0) = 3 y′(0) = −3.5
Let the solution be of type y = ert where r are roots of the
characteristics equation
r2 + r + 0.25 = 0 (4.36)
Equation (4.36) has one root, r = −0.5. Therefore, following (4.35),
the general solution is
. (4.37)
Using the initial conditions, solving for the constants, we obtain
c1 = 3 and c2 = −2. Therefore, the particular solution is
. (4.38)
The graph of the solution is shown in Figure 4.7.
Example 4.8 Repeated roots
Find solution of the initial-value problem
y − y′ + 0.25y = 0 y(0) = 2 y′(0) =
Let the solution be of type y = ert where r are roots of the
characteristics equation
r2 − r + 0.25 = 0 (4.39)
Equation (4.36) has one root, r = 0.5. Therefore, following (4.35),
the general solution is
. (4.40)
Using the initial conditions, solving for the constants, we obtain
c1 = 2 and c2 = . Therefore, the particular solution is
. (4.41)
The solution is shown in Figure 4.8.
We have investigated all possible combinations of solutions of
linear second-order differential equation with constant coefficients.
Figure 4.8
We want to formalize the solution procedure and investigate the
theoretical bases of these solutions.
Superposition, Linear combination, independence,
Wronskian
We have been discussing the solution of the differential equation
(4.2)
ay + by′ + cy = 0 (4.2)
Let us define the equation (4.2) as f(y) and write (4.2) as
f(y) = ay + by′ + cy = 0 (4.42)
If a function y1 is a solution of (4.42), we will write
f(y1) = ay1 + by1′ + cy1 = 0 (4.43)
Now, if y1 and y2 are both solutions to (4.42), we are suggesting
that a linear superposition of these two solutions, c1y1 + c2y2,
where c1 and c2 are non-zero constants, would also be a solution to
(4.42). To prove this, we write
f(c1y1 + c2y2) = a(c1y1 + c2y2)
+ b(c1y1 + c2y2)′ + c(c1y1 + c2y2)
or f(c1y1 + c2y2) = a(c1y1 + c2y2)
+ b(c1y1′ + c2y2′) + c(c1y1 + c2y2)
or f(c1y1 + c2y2) = (ac1y1 + ac2y2)
+ (bc1y1′ + bc2y2′) + (cc1y1 + cc2y2)
Now collecting terms for c1 and c2
f(c1y1 + c2y2) = c1(a y1 + by1′ + cy1)
+ c2(a y2 + by2′ + cy2)
f(c1y1 + c2y2) = c1 f(y1) + c2 f(y2)
Since y1 and y2 are both solutions of (4.42), both f(y1) and f(y2) are
zero. Therefore
f(c1y1 + c2y2) = 0
We have been using this property in finding all our ‘general
solutions’. We found two separate solutions y1 and y2, and
proposed – without proof – that c1y1 + c2y2 is the general solution.
Now we have proven the proposition.
To move further, let us revisit Example 1.1 in chapter 01.
Example 1.1
For the differential equation y″ + y = 0, y1 = c1sin x, y2 = c2cos x,
y3 = c3sin x + c4cos x are all solutions of the differential equation.
We ask an important question. Are these three solutions
independent, or are these combinations of each other? We want to
generalize the question and ask if there is a set of functions y1, y2, .
. . yn those are solutions of a differential equation, are they
independent?
Let us assume that y0, y1, y2, . . ., yn are solutions of a linear
differential equation. Then any linear combination of these
functions
a0y0 + a1y1 + a2y2 + a3y3 + . . . + anyn
where a0, a1, a2, . . ., an, are constants, is also a solution of the same
differential equation.
If a linear combination
a0y0 + a1y1 + a2y2 + a3y3 + . . . + anyn = 0 (4.44)
where not all constants are zero, then the functions y0, y1, y2, . . ., yn
are dependent. If the only way to satisfy equation (4.44) is a0 = a1
= a2 = . . . = an = 0, then the functions y0, y1, y2, . . ., yn are
independent.
To ascertain if equation (4.44) has a non-zero solution, we need a
system of linear equations. To that end, we differentiate
equation (4.44) (n – 1) times to obtain the other (n – 1) equations.
(4.44.1)
(4.44.2)
( ) ( ) ( ) ( ) (4.44.3)
( ) ( ) ( ) (4.44.n−1)
( )
Equation (4.44)–(4.44.n−1) forms a homogeneous system of linear
equation of order n. This system will have a non-trivial solution for
a0, a1, a2, . . ., an only if the system is forced to become dependent.
This means that we must ensure that the rank of the system is not
n. Therefore,
(4.45)
( ) ( ) ( ) ( )
The determinant in equation (4.45) is called the Wronskian,
W(y0, y1, y2, . . ., yn).
To determine if a given set of functions y0, y1, y2, . . ., yn is
dependent or independent, we differentiate the functions (n – 1)
times. The Wronskian determinant is evaluated. If the Wronskian is
zero, this means that the system of linear equation (4.44)–
(4.44.n−1) will have a non-trivial solution. This means that the set
of functions y0, y1, y2, . . ., yn is dependent. On the other hand, if the
Wronskian is non-zero, this means that the only solution the
system of linear equation (4.44)–(4.44.n−1) will have is a trivial
solution. This means that the set of functions y0, y1, y2, . . ., yn is
independent.
Example 4.9
Find the dependency of the functions y1 = e2t and y2 = e−3t/2.
We evaluate the Wronskian
Since the Wronskian is not equal to zero, the functions are not
dependent.
Nonhomogeneous second-order equation, Method of
undetermined coefficients
A second order nonhomogeneous equation structurally looks like
ay + by′ + cy = f(t) (4.46)
It is understood that the function f(t) can have innumerable
different shapes. Obviously, it is wrong to think that a solution can
be obtained for all functions f(t). We will attempt to obtain solution
for equation (4.46) only when the function f(t) is (i) an exponential,
(ii) combinations of sines and cosines, and (iii) a polynomial.
Towards the end of this course, when we look into Laplace
Transform, we will come back to handle more complicated types
of f(t).
Fo the time being we see the solution procedure through examples.
Example 4.10
Find the general solution of the initial-value problem
y″ + 5y′ + 6y = e2t y(0) = 1 y′(0) = 0 (4.47)
To find a solution to equation (4.47), we first find the solution of
the corresponding homogeneous equation
y″ + 5y′ + 6y = 0 (4.48)
The argument for this is as follows. Suppose by solving
equation (4.47), we obtain the solution to be h(t), and by solving
equation (4.48) we obtain the solution to be g(t). Now if we
substitute h(t) on left-hand side of equation (4.47), we will obtain
e2t on the right-hand side. On the other hand, if we substitute g(t)
on the left-hand side, we will obtain 0. This means g(t) contributes
only 0 to the right-hand side, and a combination of h(t) + g(t) will
also result in e2t. Therefore, g(t) could also be part of the solution.
The solution to equation (4.48) is
yh = c1e−2t + c2e−3t (4.49)
This solution is referred to as the homogeneous solution. To
obtain the solution of the nonhomogeneous equation (with the
right-hand side), there are two methods that are used widely for
this purpose. The first one is the Method of Undetermined
Coefficients, and the second is the Method of Variation of
Parameters. We will concentrate on the method of undetermined
coefficients.
In a nutshell, the method involves assuming a solution that will
correspond to the particular type of f(t) in equation (4.47). The
assumed solution will contain certain undetermined coefficients.
The assumed solution is substituted into equation (4.47). This will
hopefully determine the undetermined coefficients.
For f(t) = e2t in equation (4.47), we assume the solution to be
yp = Ae2t. This solution is referred to as the particular solution.
We substitute yp into equation (4.47). For this purpose, we
obtain , and . Substituting these into
equation (4.47), we obtain
4Ae2t + 10Ae2t + 6Ae2t = e2t
or 20Ae2t = e2t
or A =
Therefore, the particular solution is .
Finally, the general solution is
(4.50)
To find the constants of integration, we differentiate (4.50).
(4.51)
Figure 4.9 Long-term behavior on the left, and short-term
behavior on the right.
Substituting y(0) = 1 in (4.50), we obtain
Substituting y′(0) = 0 in (4.51), we obtain
Solving these two, we obtain and . Therefore, the
specific solution is
The solution is shown in Figure 4.9. The solution has a noteworthy
behavior. Because of the e2t term in the solution, the curve rise
dramatically. The behavior is shown in Figure 4.9 on the left. In
this dramatic rise, we miss the behavior of the solution at smaller
values of t. This behavior is shown in Figure 4.9 on the right. We
notice that the solution reduces initially, reaches a low and then
rises. As mentioned earlier, this kind of behavior is extremely
useful in engineering analysis of systems.
Example 4.11
Find the general solution of the initial-value problem
y″ + 5y′ + 6y = 3 sin 2t y(0) = 1 y′(0) = 0 (4.51)
We have already found the solution to the homogeneous equation
of (4.51).
yh = c1e−2t + c2e−3t (4.52)
For the particular solution, for the right-hand side given in (4.51),
we assume
yp = A sin 2t + B cos 2t (4.53)
To substitute into equation (4.51), we differentiate equation (4.53)
(4.54)
(4.55)
Substituting , and into equation (4.51), we obtain
(−4A sin 2t − 4B cos2t) + 5(2A cos 2t − 2B sin2t)
+ 6(A sin 2t + B cos 2t) = 3 sin 2t
Collecting coefficients of sine and cosine terms separately, we
obtain
(2A − 10B) sin2t + (10A + 2B) cos 2t = 3 sin 2t
Equating coefficients of sine and cosine terms on both sides,
2A − 10B = 3 and 10A + 2B = 0
Solving these, we obtain and . Therefore, the
particular solution is
The general is
(4.56)
To apply the initial conditions, we will need to differentiate (4.56)
(4.57)
Applying the condition from (4.51) to (4.56) and (4.57) we get
and
Solving these two, we obtain and . Therefore,
the final solution is
(4.58)
The solution is shown in Figure 4.10.
Figure 4.10
Example 4.12
Find the general solution of the initial-value problem
y″ + 5y′ + 6y = 3t3 + 4t2 – 2t + 4 y(0) = 1 y′(0) = 0 (4.59)
We have already found the solution to the homogeneous equation
of (4.51).
yh = c1e−2t + c2e−3t (4.52)
For a polynomial of form f(t), we will start the particular solution
with a polynomial of same order as f(t). Therefore, we will assume
yp = At3 + Bt2 + Ct + D (4.60)
Differentiating this, we obtain
Substituting , and into equation (4.59), we obtain
(6At + 2B) + 5(3At2 + 2Bt + C)
+ 6(At3 + Bt2 + Ct + D) = 3t3 + 4t2 – 2t + 4
Collecting terms of powers of t, we obtain,
6At3 + (15A + 6B)t2 + (6A + 10B + 6C)t
+ (2B + 5C + 6D) = 3t3 + 4t2 – 2t + 4
Equating coefficients of t on both sides
For t3; 6A = 3. Therefore,
For t2, 15A + 6B = 4. Therefore
For t, 6A + 10B + 6C = −2. Therefore
Finally for constant, 2B + 5C + 6D = 4. Therefore
Therefore, the general solution is
Figure 4.11 Long-term behavior on the left, and short-term
behavior on the right.
The find the constants, we differentiate this
Substituting the initial conditions, we obtain
and
Solving these two, we obtain and . Therefore, the
final solution is
The solution is shown in Figure 4.11. Once again, the behavior of
the solution for lower values of t has been shown separately on the
right.
Example 4.13
Find the general solution of the initial-value problem
y″ + 5y′ + 6y = 4e−t cos 2t y(0) = 1 y′(0) = 0 (4.61)
We have already found the solution to the homogeneous equation
of (4.51).
yh = c1e−2t + c2e−3t (4.52)
For this particular f(t), we will assume the particular solution as
(4.62)
We differentiate (4.62) and collect in terms of e−t sin 2t and
e−t cos 2t
Substituting into the differential equation, and collecting terms of
e−t sin 2t and e−t cos 2t
Equating coefficients from both sides, we get
−2A − 6B = 0 and 6A – 2B = 4
Solving these two, we obtain and . Therefore, the
general solution is
Figure 4.12 General behavior of solution, and long-term
behavior on the right.
Differentiating this, and after substituting the initial-condition, we
obtain the final solution
The solution is shown in Figure 4.12.
Example 4.14
Find the general solution of the initial-value problem
y″ + 5y′ + 6y = 4e−2t y(0) = 1 y′(0) = 0 (4.63)
We have already found the solution to the homogeneous equation
of (4.51).
yh = c1e−2t + c2e−3t (4.52)
For this particular f(t), we will assume the particular solution as
Let us proceed the way we did in example 4.10, and consider the
particular solution to be yp = Ae–2t. We obtain the first derivative,
, and the second derivative .
Substituting into equation (14), we obtain
4Ae–2t + 5(–2Ae–2t) + 6Ae–2t = 4e–2t
or 0 = 4e–2t
Strictly speaking, we should not be surprised that the left-hand side
has turned to be 0. This is because, we have taken the particular
solution to be Ae–2t. But e–2t is already one of the eigenfunctions of
the homogeneous part. Therefore, we have a case of repeated roots
here. So, instead of assuming e–2t, we must assume te–2t as the
particular solution.
Therefore, we assume
yp = Ate–2t
Then,
Substituting these expressions into equation (14), we obtain
( ) + 5( )
6( ) 4
The terms involving te–2t cancels out completely. Therefore,
collecting terms of e–2t, we obtain
Ae–2t = 4e–2t
or A = 4
Therefore, the particular solution is yp = 4te–2t, and the general
solution is
y = (c1 + 4t)e–2t + c2e–3t
Solving for the constant of integration, we obtain the final solution
y = (−1 + 4t)e–2t + 2e–3t
The solution is shown in Figure 4.13.
Cauchy–Euler equation
Figure 4.13
Cauchy-Euler equations are a special class of linear second-order
equations with variable coefficients. The general form of the
equation is
at2y + bty′ + cy = 0 (4.53)
To solve equation (4.53), we proceed the same way as we argued a
solution for equation (4.2). We ask what must be the form of y for
solution of (4.53) so that when it is differentiated and multiplied
with t, ty′ will give us the same form as y, and differentiated twice
then multiplied by t2, t2y will give us the same form as y. It may
sound complicated at first, but any function y = tm will fulfill these
requirements. Therefore, we will assume that our solution be of
form y = tm. Then y′ = mtm−1 and y = m(m−1)tm−2. We substitute all
these into (4.53). We obtain
at2m(m−1)tm−2 + btmtm−1 + ctm = 0
Simplifying and collecting terms, we obtain
[am2 + (b − a)m + c]tm = 0
We use the same argument that tm = 0 would give us the trivial
solution. Therefore, we will set
am2 + (b − a)m + c (4.54)
Equation (4.54) is the characteristics equation or the indicial
equation. Once again, this is a quadratic equation, and will have
three kinds of roots.
Example 4.15 Distinct real roots
Find the solution of
t2y − 4ty′ + 6y = 0 y(1) = 0.4 y′(1) = 0
The indicial equation is
m2 + (−4 − 1)m + 6 = 0
Figure 4.14 General behavior of solution, and short-term
behavior on the right.
m2 − 5m + 6 = 0
This leads to two distinct real roots, m1 = 3 and m2 = 2. Therefore,
the general solution is
y = c1t3 + c2t2
To substitute the initial conditions, we need to differentiate this.
Therefore,
y′ = 3c1t2 + 2c2t
Substituting the initial conditions, we obtain
c1 + c2 = 0.4
3c1 + 2c2 = 0
Solving these two, we obtain c1 = −0.8 and c2 = 0.8. Therefore, the
final solution is
y = −0.8 t3 + 0.8 t2
The solution is shown in Figure 4.14.
Example 4.16 Repeated roots
Find the solution of
t2y − 5ty′ + 9y = 0 y(1) = 1 y′(1) = 0
Figure 4.15
The indicial equation is
m2 + (−5 − 1)m + 9 = 0
m2 − 6m + 9 = 0
This leads to a case of repeated roots with, m = 3. Therefore, the
general solution is
y = (c1 + c2 ln t) t3
To substitute the initial conditions, we need to differentiate this.
Therefore,
y′ = 3t2(c1 + c2 ln t) + c2 t2
Substituting the initial conditions, we obtain c1 = 1 and c2 = −3.
Therefore, the final solution is
y = (1 – 3 ln t) t3
The solution is shown in Figure 4.15.
Example 4.17 Complex conjugate roots
Find the solution of
9t2y + 3ty′ + y = 0 y(1) = 1 y′(1) = 0
Figure 4.16
The indicial equation is
m2 + (3 − 9)m + 1 = 0
m2 − 6m + 1 = 0
This leads to a case of complex conjugate roots with,
and . Therefore, the general solution is
The algebra is a bit messy. But after differentiating and substituting
the initial conditions, the final solution is
The solution is shown in Figure 4.16.
Exercise
Boyce, DiPrima, Meade, p 112, problems 1-12, page 122,
problems 1-5, page 128, problems 5-15, p 135, problems 1-11, p
144, problems 1-15, p 223, problems 1-11.