Boyce/DiPrima 10th ed, Ch 3.

4:
Repeated Roots; Reduction of Order
Elementary Differential Equations and Boundary Value Problems, 10th edition, by William E. Boyce and Richard C. DiPrima, ©2013 by John Wiley & Sons, Inc.

• Recall our 2nd order linear homogeneous ODE

ay  by  cy  0
• where a, b and c are constants.
• Assuming an exponential soln leads to characteristic equation:
y(t )  e rt  ar 2  br  c  0
• Quadratic formula (or factoring) yields two solutions, r1 & r2:
 b  b 2  4ac
r
2a
• When b2 – 4ac = 0, r1 = r2 = -b/2a, since method only gives
one solution:
y1 (t )  ce b t / 2 a
 Second Solution: Multiplying Factor v(t)

• We know that
y1 (t ) a solution  y2 (t )  cy1 (t ) a solution
• Since y1 and y2 are linearly dependent, we generalize this
approach and multiply by a function v, and determine
conditions for which y2 is a solution:
y1 (t )  eb t / 2a a solution  try y2 (t )  v(t )eb t / 2a
• Then
y2 (t )  v(t )e b t / 2 a
b
y2 (t )  v(t )e b t / 2 a  v(t )e b t / 2 a
2a
b b b2
y2(t )  v(t )e b t / 2 a
 v(t )e b t / 2 a
 v(t )e b t / 2 a
 2 v(t )e b t / 2 a
2a 2a 4a
 ay  by  cy  0
Finding Multiplying Factor v(t)

• Substituting derivatives into ODE, we seek a formula for v:

  b b2   b  
a v(t )  v(t )  2 v(t )  b v(t )  v(t )  cv(t )  0
b t / 2 a
e
  a 4a   2a  
b2 b2
av(t )  bv(t )  v(t )  bv(t )  v(t )  cv(t )  0
4a 2a
 b2 b2 
av(t )     c v(t )  0
 4a 2a 
 b 2 2b 2 4ac    b 2 4ac 
av(t )     v(t )  0  av(t )    v(t )  0
 4a 4a 4a   4a 4a 
 b 2  4ac 
av(t )   v(t )  0
 4a 
v(t )  0  v(t )  k3t  k 4
 General Solution

• To find our general solution, we have:

y (t )  k1e  bt / 2 a  k 2 v(t )e  bt / 2 a
 k1e bt / 2 a  k3t  k 4 e bt / 2 a
 c1e bt / 2 a  c2tebt / 2 a

• Thus the general solution for repeated roots is

y(t )  c1ebt / 2a  c2tebt / 2a

 Wronskian

• The general solution is

y(t )  c1ebt / 2a  c2tebt / 2a
• Thus every solution is a linear combination of
y1 (t )  ebt / 2a , y2 (t )  tebt / 2a
• The Wronskian of the two solutions is
e bt / 2 a tebt / 2 a
W ( y1 , y2 )(t )  b bt / 2 a  bt  bt / 2 a
 e 1  e
2a  2a 
 bt   bt 
 e bt / a 1    e bt / a  
 2a   2a 
 e bt / a  0 for all t

• Thus y1 and y2 form a fundamental solution set for equation.

 Example 1 (1 of 2)

• Consider the initial value problem

y  4 y  4 y  0
• Assuming exponential soln leads to characteristic equation:
y(t )  ert  r 2  4r  4  0  (r  2) 2  0  r  2
• So one solution is y1(t )  e 2t and a second solution is found:
y2 (t )  v(t )e 2t
y2 (t )  v(t )e 2 t  2v(t )e 2 t
y2(t )  v(t )e 2 t  4v(t )e 2 t  4v(t )e 2 t
• Substituting these into the differential equation and
simplifying yields v" (t )  0, v' (t )  k 1, v(t )  k 1t  k 2
where c1 and c 2 are arbitrary constants.
 Example 1 (2 of 2)

• Letting k 1  1 and k 2  0, v(t )  t and y 2(t )  te  2t

• So the general solution is
y(t )  c1e 2t  c2te2t
• Note that both y1 and y 2 tend to 0 as t   regardless of the
values of c1 and c2
• Using initial conditions yt

y (0)  1 and y ' (0)  3

2.0

y(t )  (1  5t )e 2t
c1  1 1.5

  c1  1, c2  5
 2c1  c2  3 1.0

• Therefore the solution to the 0.5

IVP is y(t )  e2t  5te2t 0.0 0.5 1.0 1.5 2.0 2.5 3.0
t
 Example 2 (1 of 2)

• Consider the initial value problem

y  y  0.25 y  0, y0  2, y0  1/ 3
• Assuming exponential soln leads to characteristic equation:
y(t )  ert  r 2  r  0.25  0  (r  1 / 2) 2  0  r  1 / 2
• Thus the general solution is
y(t )  c1et / 2  c2tet / 2 yt
4

y(t )  et / 2 (2  2/3 t )
• Using the initial conditions: 3

c1  2 2

 2
1 1  c  2, c  
 c2 
1
1 2
c1 3
2 3
 1 2 3 4
t

2 t/2
• Thus
1

y (t )  2e t/2
 te
3 2
 Example 2 (2 of 2)

• Suppose that the initial slope in the previous problem was

increased
y0  2, y0  2
• The solution of this modified problem is

y(t )  2et / 2  tet / 26y t

• Notice that the coefficient of the second
term is now positive. This makes a big red : y(t )  et / 2 (2  t )
4
difference in the graph, since the blue : y(t )  et / 2 (2  2/3 t )

exponential function is raised to a

2
positive power:
  1/2  0
t
1 2 3 4

2
 Reduction of Order

• The method used so far in this section also works for

equations with nonconstant coefficients:
y  p(t ) y  q(t ) y  0
• That is, given that y1 is solution, try y2 = v(t)y1:
y2 (t )  v(t ) y1 (t )
y2 (t )  v(t ) y1 (t )  v(t ) y1 (t )
y2(t )  v(t ) y1 (t )  2v(t ) y1 (t )  v(t ) y1(t )
• Substituting these into ODE and collecting terms,
y1v  2 y1  py1 v   y1  py1  qy1 v  0
• Since y1 is a solution to the differential equation, this last
equation reduces to a first order equation in v :
y1v  2 y1  py1 v  0
Example 3: Reduction of Order (1 of 3)

• Given the variable coefficient equation and solution y1,

2t 2 y  3ty  y  0, t  0; y1 (t )  t 1 ,
use reduction of order method to find a second solution:
y2 (t )  v(t ) t 1
y2 (t )  v(t ) t 1  v(t ) t  2
y2(t )  v(t ) t 1  2v(t ) t  2  2v(t ) t 3
• Substituting these into the ODE and collecting terms,
  
2t 2 vt 1  2vt 2  2vt 3  3t vt 1  vt 2  vt 1  0 
 2vt  4v  4vt 1  3v  3vt 1  vt 1  0
 2tv  v  0
 2tu  u  0, where u (t )  v(t )
 Example 3: Finding v(t) (2 of 3)

• To solve
2tu  u  0, u (t )  v(t )
for u, we can use the separation of variables method:
du du 1
2t
dt
u  0   u  2t dt  ln u  1/2 ln t  C


 u t  u  ct 1/ 2 , since t  0.
1/ 2 C
e

• Thus
v  ct1/2

and hence
v(t )  2/3 c t 3/ 2  k
 Example 3: General Solution (3 of 3)

• Since v(t )  2/3 c t 3/ 2  k

y2 (t )  2/3 c t 3/ 2  k t 1  2/3 c t1/ 2  k t 1
• Recall y1 (t )  t 1
• So we can neglect the second term of y2 to obtain
y2 (t )  t1/ 2
• The Wronskian of y1 (t ) and y2 (t ) can be computed
W ( y1 , y 2)(t )  3/2 t  3 / 2  0, t  0
• Hence the general solution to the differential equation is
y(t )  c1t 1  c2t1/ 2