0% found this document useful (0 votes)
110 views209 pages

Lectures 1 - ME 340

This document contains lecture notes for a course on dynamics of mechanical systems. It covers topics such as linear coordinate transformations, complex numbers, and matrices. Linear coordinate transformations can be represented by matrices and complex numbers are a type of linear transformation. The document provides examples and exercises related to these topics.

Uploaded by

Lodu Cent
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)
110 views209 pages

Lectures 1 - ME 340

This document contains lecture notes for a course on dynamics of mechanical systems. It covers topics such as linear coordinate transformations, complex numbers, and matrices. Linear coordinate transformations can be represented by matrices and complex numbers are a type of linear transformation. The document provides examples and exercises related to these topics.

Uploaded by

Lodu Cent
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/ 209

Lecture notes, ME 340

2014-2015 Harry Dankowicz


This document contains transcripts of the online lectures, as well as
in-class worksheets for ME 340, Dynamics of Mechanical Systems.
All rights reserved.

lecture notes, me 340

Lesson 0 Mathematical preliminaries


TAKE-AWAYS:
* A linear coordinate transformation transforms a point
with coordinates ( x, y) to a point with coordinates ( x 0 , y0 ) =
( ax + by, cx + dy) for some numbers a, b, c, and d, and can be
represented in terms of matrix multiplication by
!
!
!
x0
a b
x
=

.
y0
c d
y
* Complex numbers are linear transformations represented by
matrices of the form
!
r cos q
r sin q
,
r sin q r cos q
by the corresponding rectangular form r cos q + jr sin q or by
the polar form r ejj , where the imaginary unit j satisfies the
condition that j j = 1. A complex number may be described
in terms of its real and imaginary parts r cos q and r sin q, or
in terms of its magnitude r and phase j.
* An eigenvalue l of the matrix
a
c

b
d

is a real or complex number l that satisfies the characteristic


equation
a u
b
=0
c
d u
in the unknown variable u. The corresponding eigenvector is
obtained by solving a system of linear equations.

lecture notes, me 340

Online lecture: linear transformations


In a two-dimensional coordinate system, a linear coordinate
transformation1 is a rule that transforms the point with coordinates
( x, y) to the point with coordinates

( x 0 , y0 ) = ax + by, cx + dy ,

Bold faced characters are used for


terminology that you should memorize.

(1)

given some numbers a, b, c, and d.


Example 1. The numbers a = 0, b = 1, c = 1, and d = 3 define a coordinate transformation that transforms the point (5, 1) to the point (1, 2)
and, in general, the point ( x, y) to the point ( x 0 , y0 ) = (y, x + 3y).
x0

Figure 1: The linear coordinate transformation x 0 = y, y0 = x + 3y transforms


straight lines to straight lines.

y0 = 3

( x, y) = (5, 1)

2y = 3

( x 0 , y0 ) = (1, 2)

Points ( x, y) on the straight line x 2y = 3 are transformed to points


on the straight line x 0 y0 = 3.

( x 0 , y0 )

In terms of the column matrices


!
x
and
y

x0
y0

the transformation (1) can be written in matrix form as


!
!
!
!
x0
a b
x
ax + by
=

:=
.
y0
c d
y
cx + dy

(2)

(3)

We say that the coefficient matrix


a
c

b
d

represents the coordinate transformation.

(4)

lecture notes, me 340

Application of the transformation represented by the matrix in (4)


followed by the transformation represented by the matrix
!
e f
(5)
g h
is equivalent to a transformation represented by the matrix product
!
!
!
e f
a b
ea + f c eb + f d

:=
.
(6)
g h
c d
ga + hc gb + hd

Figure 2: Successive application of


two linear coordinate transformations
is represented by the product of the
corresponding matrices.

( x 0 , y0 ) = ( ax + by, cx + dy)

( x 00 , y00 ) = (ex 0 + f y0 , gx 0 + hy0 )

( x, y)

Example 2. The origin ( x, y) = (0, 0) is transformed to itself by every


linear coordinate transformation, since
!
!
!
a b
0
0

=
.
(7)
c d
0
0
If ad
with

bc = 0, other points may be transformed to the origin. For example,


2
4

1
2

(8)

2x are transformed to the origin.

all points on the straight line y =


The expression
a
c

b
d

:= ad

bc

(9)

in Example 2 is the determinant of (4). There exist points away from


the origin that are transformed to the origin only if the determinant
equals 0.

lecture notes, me 340

If the determinant is nonzero, the inverse of the matrix


!
a b
c d

(10)

equals
a
c

b
d

d/( ad bc)
c/( ad bc)

:=

b/( ad bc)
a/( ad bc)

(11)

The corresponding transformation reverses the changes to the coordinates, since the product
!
!
!
d/( ad bc)
b/( ad bc)
a b
1 0

=
(12)
c/( ad bc) a/( ad bc)
c d
0 1
equals the identity matrix, for which ( x 0 , y0 ) = ( x, y).

In-class worksheet: rotations


A linear coordinate transformation represented by a matrix of the
form
!
cos j
sin j
(13)
sin j
cos j
is a rotation by the angle j. In particular,

( x20

x10 )2 + (y20

= ( x2 cos j

y10 )2
y2 sin j

x1 cos j + y1 sin j)2

+ ( x2 sin j + y2 cos j
= ( x2

x1 )2 + ( y2

x1 sin j

y1 cos j)2

y1 )2 ,

(14)

i.e., distances between points are unchanged by the transformation.


Points on the straight line
x sin q

y cos q = 0

(15)

are transformed to points on the straight line


x 0 sin(q + j)

y0 cos(q + j) = 0,

(16)

since
cos j sin(q + j)

sin j cos(q + j) = sin q,

(17)

sin j sin(q + j) + cos j cos(q + j) = cos q.

(18)

Since the origin is transformed to the origin, the distance from each
such point to the origin is unchanged by the transformation.

lecture notes, me 340

Figure 3: Distances between points


and to the origin are unchanged by a
rotation.

Since the determinant of (13) equals cos2 j + sin2 j = 1, the inverse


is the matrix
!
!
cos j
sin j
cos( j)
sin( j)
=
,
(19)
sin j cos j
sin( j)
cos( j)
which corresponds to a rotation by the angle

j.

Exercises
1. Determine the point ( x, y) that is transformed to ( x 0 , y0 ) = (1, 5)
under the linear coordinate transformation

( x 0 , y0 ) = (4x

y, x + 3y).

2. Determine the coordinate transformation corresponding to


!
1 0
.
1 1
3. In the same diagram, graph all points that satisfy the straight-line
condition 4x 3y = 2, as well as the points they transform to
under the linear coordinate transformation
x 0 = x + 2y, y0 = 3x

y.

4. Find a linear coordinate transformation that transforms the straight


line y = 2x 1 to the straight line y0 = x 0 + 2.
5. Determine the inverse of the matrix
0
1

1
0

lecture notes, me 340

6. Find the coordinates ( x 0 , y0 ) such that


!
!
x0
4
1
=

y0
3 0

1
2

7. Find the coordinates ( x, y) such that


!
!
0
0
1
=

1
2 1

x
y

8. Choose the constant a such that the determinant of the matrix


!
1 a
2 3
equals 0. Find a straight line of points that are all transformed to
the origin by the corresponding transformation.
9. Compute the matrix product
!
1
1 1
2

0 2
0

0
1
3

1
0

1
2

1
2

and comment on the value of its n-th power.

Solutions
1. With ( x 0 , y0 ) = (1, 5) it follows that
1 = 4x

y, 5 = x + 3y,

which can be solved to yield ( x, y) = (8/13, 19/13).


2. The matrix

1
1

corresponds to the transformation


!
!
x0
1 0
=

y0
1 1

0
1

x
y

x
x+y

3. The equations
x 0 = x + 2y, y0 = 3x

may be solved for x and y to yield


x=

1 0
1
( x + 2y0 ), y = (3x 0
7
7

y 0 ),

lecture notes, me 340

and substitution in the straight-line condition 4x


yields
4 0
3
( x + 2y0 )
(3x 0 y0 ) = 2
7
7
or, in other words,
5x 0 + 11y0 = 14.

3y = 2 then

The two lines are graphed below.

5x 0 + 11y0 = 14
4x

3y = 2

( x 0 , y0 ) = ( 5, 1)
( x, y) = ( 1, 2)

4. If ( x 0 , y0 ) are chosen to that y0 = x 0 + 2, then the transformation


( x 0 , y0 ) = ( ax + by, cx + dy) implies that cx + dy = ax + by + 2 or
(b d)y = (c a) x 2. This describes the straight line y = 2x 1
provided that 2(b d) = c a and (b d) = 2, for example,
a = 0, b = 2, c = 4, d = 0.
5. It follows directly from the definition that
0
1

1
0

4
3

1
0

0
1

1
0

2
3

6. Multiplication yields
1
2

i.e., the point (1, 2) is transformed to the point (2, 3).


7. The equality
0
1

0
2

1
1

is equivalent to the system of equations


0=

y, 1 = 2x + y

x
y

lecture notes, me 340

or, in other words, x = 1/2 and y = 0. Alternatively,


x
y

0
2

8. The determinant

equals 0 if a =

implies that 2x

1
2

a
3

0
1

1/2
0

= 3 + 2a

3/2. In this case,


!
1
3/2

2
3

1
2

x
y

0
0

3y = 0.

9. The product equals


!
1
1 1
2

0 2
0
Since

1
1

1
0

1
2

0
1
3
1
2

1
0

1
0

1
2

1
2

1
2

1
0

1
12
1
3

0
0
1

it follows that
1
2

1
12
1
3

!n

1
0

1
2

1
2n

0
1
3n

1
0

1
2

1
2

Online lecture: complex arithmetic


Complex numbers are linear coordinate transformations.
Example 3. The matrix
a
0

0
a

(20)

corresponds to the linear transformation ( x 0 , y0 ) = (ax, ay). Instead of (20),


this transformation is expressed in terms of the number a:
!
!
!
x0
x
ax
= a
:=
.
(21)
y0
y
ay
When a = 1, the transformation is the identity.

lecture notes, me 340

Example 4. The matrix


0
1

1
0

(22)

corresponds to the linear transformation ( x 0 , y0 ) = ( y, x ). Instead of (22),


this transformation is expressed in terms of the imaginary unit j:
!
!
!
x0
x
y
= j
:=
.
(23)
y0
y
x
The transformation ( x 0 , y0 ) = ( by, bx ) can then be expressed in terms of
either of the products j b or b j (or, equivalently, jb or bj).
For numbers a and b, it follows by definition that a b = ab, since
application of the corresponding transformations in succession yields
x
y

ab

= a

bx
by

abx
aby

x
y

. (25)

x
y

(26)

. (27)

x
y

= (ab)

Similarly, a jb = ja b = j(ab), since


!
!
!
x
by
aby
a jb
= a
=
= j(ab)
y
bx
abx
and
ja b

x
y

= ja

bx
by

Finally, ja jb = ab, since


!
!
x
by
ja jb
= ja
=
y
bx

aby
abx

abx
aby

= j(ab)

x
y

= ( ab)

(24)

In particular, j j = 1, since application of the corresponding transformation twice in succession is equivalent to the transformation
( x 00 , y00 ) = ( y0 , x 0 ) = ( x, y) represented by the matrix product
!
!
!
0
1
0
1
1 0

=
.
(28)
1 0
1 0
0
1
Example 5. If matrix addition is defined so that
!
!
1
2
1 0
0
=
+
2 1
0 1
2
then this corresponds to the linear transformation
!
!
!
x0
1 0
x
0
=

+
0
y
0 1
y
2

2
0

2
0

(29)

x
y

(30)

10

lecture notes, me 340

or, in other words,


!
x0
= 1
y0

x
y

x
y

+ j2

= (1 + j2)

x
y

11

(31)

The application of the transformation twice in succession is equivalent to the


transformation represented by the matrix
!
!
!
1
2
1
2
3
4

=
(32)
2 1
2 1
4
3
or, equivalently,

(1 + j2) (1 + j2) = 1 1 + 1 j2 + j2 1 + j2 j2 =
since j2 j2 =

3 + j4,

(33)

4.

A linear transformation represented by a matrix of the form


!
!
!
a 0
0
b
a
b
+
=
b a
0 a
b 0

(34)

is a complex number with real part a and imaginary part b. The


rectangular form a + jb reduces to the cases in Examples 3 and 4
when a = 0 or b = 0. We define a jb := a + ( 1) jb = a + j( b).
Example 6. The complex number corresponding to the matrix
!
r cos j
r sin j
r sin j r cos j

(35)

with r > 0 rotates every straight line through the origin by j about the
origin, and stretches the result by r.
Figure 4: The linear coordinate transformation in (35) rotates straight lines
through the origin by j and stretches
the result by r.

rr cos(q + j), rr sin(q + j)

(r cos q, r sin q )

j
q

rr

lecture notes, me 340

Indeed, the point ( x, y) = (r cos q, r sin q ) is transformed to the point


!
!
!
x0
r cos j
r sin j
r cos q
=

y0
r sin j r cos j
r sin q
!
rr (cos j cos q sin j sin q )
=
rr (sin j cos q + cos j sin q )
!
rr cos( j + q )
=
,
(36)
rr sin( j + q )
where the last equality follows from the trigonometric addition formulae.
The positive factor r and the angle j are the magnitude and phase
of the complex number (35). The rectangular form
r cos j + jr sin j

(37)

is expressed in terms of the real and imaginary parts, r cos j and


r sin j, respectively. The sum of the squares of the real and imaginary
parts equals the square of the magnitude:
r2 cos2 j + r2 sin2 j = r2 .

(38)

Example 7. Application of the transformation represented by 1 + j2 followed by the transformation represented by 1 j2 is equivalent to a single
transformation represented by

(1

j2) (1 + j2) = 1 1 + 1 j2

j2 1

j2 j2 = 5.

(39)

It follows that application of the transformation represented by 1 + j2 followed by the transformation represented by 15 j 25 is equivalent to a single
transformation represented by

1
2
j
(1 + j2) = 1,
(40)
5
5
i.e., the identity transformation.
For an arbitrary complex number represented by a matrix
!
a
b
b a

(41)

or the rectangular form a + jb, its conjugate a + jb corresponds to the


matrix transpose
a
b

b
a

!T

a
b

b
a

(42)

12

lecture notes, me 340

i.e., a + jb = a jb. The product of the complex number with its


conjugate equals the square of the magnitude |a + jb|:
jb) = a2 + b2 .

(a + jb) a + jb = (a + jb) (a

(43)

The inverse 1/(a + jb) corresponds to the matrix inverse


a
b
i.e.,

b
a

a/(a2 + b2 )
b/(a2 + b2 )

b/(a2 + b2 )
a/(a2 + b2 )

1
a jb
a + jb
a + jb
= 2
=
=
.
a + jb
a + b2
|a + jb|2
a + jb (a + jb)

(44)

(45)

In particular, (a + jb) 1/(a + jb) = 1, i.e., the identity transformation.

Exercises
1. Compute
3
1

j2

2. Show that the transformation corresponding to the imaginary unit


j rotates every straight line through the origin counterclockwise by
90 .
3. Show that

(a + jb) (g + jd) = ab

bd + j(ad + bg)

by multiplying the matrices


a
b

b
a

g
d

d
g

4. Show that

(a + jb) (g + jd) = (g + jd) (a + jb)


by multiplying the matrices
a
b

b
a

g
d

d
g

5. Show that

(a + jb) (g + jd) = (a + jb) (g + jd)


6. Evaluate the product

(2

j3) (3 + j2).

13

lecture notes, me 340

7. Evaluate the product


j (1

j).

8. Show that the complex number 1 j2 satisfies the quadratic equation


u2 2u + 5 = 0.
9. Compute the conjugate of the complex number j3.
10. Compute the magnitude of the complex number
11. Compute the inverse of the complex number 3

1 + j2.
j.

12. Compute the ratio


1 j
: = (1
j+1

j)

1
.
j+1

13. Determine the magnitude and phase of the complex number


!
1
2
.
2
1
14. Find the rectangular form of the complex number with magnitude
3 and phase p/2.
15. Find the point ( x 0 , y0 ) transformed from ( x, y) = (1, 1) by the
complex number with magnitude 1/2 and phase p/3.
16. Compute the difference
1
j

1 j
2 + j2

1+j
.
2 j2

Solutions
1. By definition
j2

3
1

2
6

2. Suppose that ( x, y) = (r cos q, r sin q ) for some q and arbitrary r


and let ( x 0 , y0 ) = ( y, x ). Then

( x 0 , y0 ) = ( r sin q, r cos q ) = r cos(q + p/2), r sin(q + p/2) ,


i.e., a 90 counterclockwise rotation, for all r.

14

lecture notes, me 340

3. Here,
a
b

b
a

g
d

d
g

a
b

b
a

g
d

d
g

g
d

d
g

a
b

b
a

ag bd
ad + bg

ad
ag

bg
bd

ag bd
ad + bg

ad
ag

bg
bd

ag bd
ad + bg

ad
ag

bg
bd

4. Here,

and
.

5. Here,

(a + jb) (g + jd) = ag

bd

j(ad + bg),

(a + jb) (g + jd) = (a

jb) (g

whereas

= ag

jd)

j(ad + bg).

bd

6. Using the distributive property,

(2

j3) (3 + j2) = 2 3 + 2 j2

j3 3

j3 j2 = 12

j5.

7. Using the distributive property


j (1

j) = j 1

j j = j + 1.

8. By substitution

(1

j2) (1

9. Here, j3 =
10. Here, |

j2)

2(1

j2) + 5 = 1

2 + j4 + 5 = 0.

j3.

1 + j2|2 = ( 1 + j2) ( 1

11. Here, 1/(3

j4

j2) = 17, i.e., |

1 + j2| =

17.

j) = (3 + j)/10.

12. Here,

(1

j)

1
= (1
j+1

j)

j
2

j.

13. Here, r cos q = 1 and r sin q = 2. Squaring these expressions


p
and adding together gives r2 = 5 ) r = 5. The unique angle
p
q 2 [ p, p ) then equals p arccos(1/ 5).

15

lecture notes, me 340

14. With r = 3 and q =

p/2, the rectangular form equals


r cos q + jr sin q =

j3.

p
15. Since cos p/3 = 1/2 and sin p/3 = 3/2,
!
!
!
1
p
1
p
x0
cos
sin
1
2
3
2
3
=

=
1
p
1
p
y0
1
2 sin 3
2 cos 3

4p
1+ 3
4

16. The difference between a complex number and its complex conjugate equals j2 multiplied by its imaginary part. Since
1 j
1
= (1
2 + j2
8
the imaginary part is

j)(2

j2) =

1
j ,
2

1/2, and the expression equals

1
1
j2
= 1.
j
2

Online lecture: eigenvectors and eigenvalues


For some linear coordinate transformations, there are straight lines
through the origin that are transformed to themselves.
Example 8. For the linear transformation defined by the coefficient matrix
!
2
1
,
(46)
1 2
points with coordinates ( x, y) are transformed to points with coordinates
!
!
!
!
x0
2
1
x
2x y
=

=
.
(47)
y0
1 2
y
x + 2y
In particular, points ( x, y) = (g, g) for any number g are transformed
to points ( x 0 , y0 ) = (g, g). It follows that the straight line y = x is
transformed to the straight line y0 = x 0 .
Similarly, points ( x, y) = (g, g) for any number g are transformed
to points ( x 0 , y0 ) = (3g, 3g). It follows that the straight line y = x is
transformed to the straight line y0 = x 0 .
A number l is a real eigenvalue of
a
c

b
d

(48)

16

lecture notes, me 340

if u = l satisfies the determinant condition


a

= (a

u)(d

bc = 0,

u)

(49)

known as the characteristic equation. In this case, there exists a


nonzero column matrix
!
g
(50)
h
such that
a
c

b
d

g
h

g
h

0
l

l
0

g
h

= l

(51)

The column matrix (50) is a real eigenvector of (48). Points ( x, y) on


the straight line hx gy = 0 are transformed to points ( x 0 , y0 ) =
( ax + by, cx + dy) on the same straight line hx 0 gy0 = 0.
Example 9. A real eigenvector of the matrix
!
2
1
1 2
is a nonzero column matrix
g
h
such that

2
1

1
2

or, in other words,


2

l
1

1
2 l

g
h

(52)

(53)

g
h

= l

g
h

0
0

(55)

l)

(56)

(54)

But this can only be possible if the determinant


2
equals 0, i.e., if

l
1

1
2 l
l2

= (2

l)(2

4l + 3 = 0,

(57)

which is satisfied by the eigenvalues l = 1 and l = 3.


For l = 1, (55) is equivalent to the equations
g

h = 0,

g + h = 0 ) h = g,

which implies that every nonzero column matrix of the form


!
g
g

(58)

(59)

17

lecture notes, me 340

is an eigenvector corresponding to the eigenvalue 1.


Similarly, for l = 3, (55) is equivalent to the equations
g

h = 0,

h=0)h=

(60)

g,

which implies that every nonzero column matrix of the form


!
g
g

(61)

is an eigenvector corresponding to the eigenvalue 3.


A complex number l R + jl I is a complex eigenvalue of
!
a b
c d

(62)

if u = l R + jl I satisfies the characteristic equation


a

u
c

b
d

= (a

u)(d

u)

bc = 0.

(63)

In this case, there exists a nonzero matrix


!
gR g I
hR h I
such that
a
c

b
d

gR
hR

gI
hI

gR
hR

gI
hI

The expression
gR + jg I
hR + jh I

(64)

lR
lI

lI
lR

(65)

(66)

is the corresponding complex eigenvector.


Example 10. The characteristic equation for the complex number
!
a
b
b a
is

(a

u )2 + b2 = 0

This is satisfied by u = a + jb, since ( jb) ( jb) =


agR
ag I

(67)

(68)
b2 . The equations

g I b,

(69)

bh I = gR b + g I a

(70)

bhR = gR a

h I b,

(71)

bg I + ah I = hR b + h I a

(72)

bgR + ahR = hR a

18

lecture notes, me 340

imply that hR = g I and h I = gR . An example complex eigenvector is


obtained with gR = 1 and g I = 0:
!
1
.
(73)
j
A second eigenvalue and corresponding eigenvector is obtained by substituting j for j.
A matrix of the form

a
c

b
d

(74)

either has a single real eigenvalue, two real eigenvalues, or a pair of


complex eigenvalues that are conjugate to each other.

In-class worksheet: a matrix decomposition


For a linear transformation represented by the matrix
!
a b
,
c d

(75)

there exist angles j and y, and scaling factors r x and ry , such that the
transformation may be represented by the product
!
!
!
cos j
sin j
rx 0
cos y
sin y

,
(76)
sin j
cos j
0 ry
sin y
cos y
i.e., as a rotation by y, followed by a scaling of each of the coordinates by r x and ry , respectively, followed by a rotation by j. Here,
r x < 0 and ry < 0 correspond to reflection about the y and x axis,
respectively.
Example 11. Consider the linear coordinate transformation represented by
the matrix
!
2
1
p
.
(77)
0
3
This equals the product (76) provided that
r x cos j cos y

ry sin j sin y =

r x cos j sin y

ry sin j cos y = 1,

r x sin j cos y + ry cos j sin y = 0,


r x sin j sin y + ry cos j cos y =
Substitution of
rx =

2,

(78)
(79)

ry
rx
, ry =
sin j sin y
sin j sin y

(80)
3.

(81)

(82)

19

lecture notes, me 340

then yields
r x cot j cot y
r x cot j

2,

ry =

(83)

ry cot y = 1,

r x cot y + ry cos j = 0,
r x + ry cot j cot y =

(84)

(85)
3.

(86)

Multiplication of (83) by ry and of (86) by r x , followed by addition results


in
p
r2x r2y = 3r x 2ry .
(87)
Moreover, (84) and (85) imply that
cot j =

r2x

rx

r2y

, cot y =

ry
r2x

r2y

(88)

Finally, multiplication of (83) by r x and of (86) by ry , followed by addition results in

p
r x ry
r x ry
2r x
3ry = r2y r2x cot j cot y = 2
= p
(89)
2
r x ry
3r x 2ry
or, in other words,
p
2 3r2x + 8r x ry

p
p
ry
2 3r2y = 0 ) r x = 3ry or p
3

(90)

Substitution back into (87) then gives


2r2y = ry or

2 2
r =
3 y

ry .

(91)

Since r x and ry cannot both equal 0, we obtain two possible solutions:


!
!
p
p
3 1
3 3
(r x , ry ) =
,
and (r x , ry ) =
,
.
(92)
2 2
2 2

p
In the former case, cot j =
3 and cot y = 1, which imply that j =
p/6 + np and y = p/4 + mp, for arbitrary integers n and m and,
p
p
consequently, r x = ( 1)m+n 6 and ry = ( 1)m+n 2. In the latter
p
case, cot j = 1/ 3 and cot y = 1, which imply that j = p/3 + np
and y = p/4 + mp, for arbitrary integers n and m and, consequently,
p
p
r x = ( 1)m+n 2 and ry = ( 1)m+n 6.
The angles j and y, as well as the scaling factors r x and ry may be
obtained from the real eigenvalues and eigenvectors of
!
!
!
!
a b
a b
a c
a c
.
(93)

and

c d
b d
b d
c d

20

lecture notes, me 340

In both cases, the eigenvalues equal r2x and ry2 . The corresponding
eigenvectors are multiples of
!
!
cos j
sin j
,
(94)
sin j
cos j
and

cos y
sin y

sin y
cos y

(95)

respectively.
Example 12. In the case of the linear transformation represented by
!
2
1
p
,
0
3

(96)

the eigenvalues of
2
0

1
p

2
1

0
p

5
p

p !
3
3

are l1 = 6 and l2 = 2 with corresponding eigenvectors


!
p
3h
h
and
g
p
3g

(97)

(98)

(99)

respectively. Here,
cos j =

3h, sin j = h ) j =

for some integer n.


Similarly, the eigenvalues of
!
2
0
p

1
3

2
0

1
p

p
+ np
6

4
2

are l1 = 6 and l2 = 2 with corresponding eigenvectors


!
h
h
and
g
g

2
4

(100)

(101)

(102)

(103)

respectively. Here,
sin y = g, cos y = g ) y =

p
+ mp
4

(104)

21

lecture notes, me 340

22

for some integer m.


We select m and n, and then choose r x and ry from the positive and
negative square roots of l1 and l2 , respectively, so that (76) holds. For
p
p
example, with m = n = 0, the values r x =
6 and ry =
2 imply that
the coordinate transformation is equivalent to a rotation by p/4, followed by
p
a scaling and sign change of the x coordinate by 6 and of the y coordinate
p
by 2, followed by a rotation by p/6.
Figure 5: The coordinate transformation
can be reduced to a rotation, followed
by scaling of each of the coordinates,
followed by a rotation.

Exercises
1. Find the characteristic equation for the matrix
!
0 3
.
1
1
2. Compute all eigenvalues of the matrix
!
2 3/2
.
0
1
3. Verify that
1
0

lecture notes, me 340

is a real eigenvector of
1
0

5/2
4

4. Find a complex eigenvector of the matrix


!
5
2
.
5
1
5. Show that the matrix
1
1

1
3

only has one real eigenvalue and no complex eigenvalues. What is


the corresponding eigenvector?
6. Given a matrix

G
F

D
Y

whose entries are arbitrary expressions, a (formal) eigenvector is a


nonzero column matrix of expressions
!
s
t
such that
G
F

D
Y

s
t

equals
l

s
t

:=

:=

Gs+Dt
Fs+Yt
ls
lt

for some expression l, known as the corresponding (formal)


eigenvalue, that is a formal solution to the characteristic equation
G

l
F

:= (G

l) (Y

l)

F D = 0.

Assume that l R + jl I is a complex eigenvalue of


!
a b
c d
with corresponding complex eigenvector
!
gR + jg I
.
hR + jh I
Show that (64) is a formal eigenvector of (62) with formal eigenvalue l R + jl I .

23

lecture notes, me 340

7. Find a straight line y = kx that is transformed to itself under the


transformation given by the matrix
!
10
4
.
24
10

Solutions
1. The characteristic equation equals
0

3
1 u

u
1

= 0 ) u2 + u

2. The characteristic equation equals u2

2 = 0. Since
p
b b2 4ac
,
2a

au2 + bu + c = 0 ) u =
it follows that the eigenvalues equal
p
1 9
= 2 and
2
3. The claim follows since
!
1 5/2

0
4

1
0

1.

1
0

3 = 0.

1
0

4. The characteristic equation equals u2 4u + 5 = 0. The eigenvalues


equal
p
4
4
= 2 j.
2
With l R = 2 and l I = 1,
!
!
!
!
5
2
gR g I
gR g I
2 1

5
1
hR h I
hR h I
1 2
implies that
3gR + g I
5gR

2hR = 0,

gR + 3g I

3hR + h I = 0, 5g I

hR

2h I = 0,
3h I = 0

or, in other words,


hR =

3
1
gR + g I , h I =
2
2

1
3
gR + g I .
2
2

The choice gR = 3 and g I = 1 gives the complex eigenvector


!
3+j
.
5

24

lecture notes, me 340

5. The characteristic equation is u2


only eigenvalue is l = 2. Then
g + h = 2g,

2)2 = 0, so the

4u + 4 = (u

g + 3h = 2h ) h = g

implies that a corresponding eigenvector is


!
g
g
for any g 6= 0.
6. By definition, l R + jl I is a solution to the characteristic equation.
Moreover,
!
!
a b
gR + jg I

c d
hR + jh I
equals
a
c

b
d

gR
hR

b
d

gR + jg I
hR + jh I

a
c

+j

Similarly,

(l R + jl I )

gI
hI

equals
gR
hR

gI
hI

lR
lI

+j

gR
hR

gI
hI

lI
lR

But these must be equal according to (65).


7. The characteristic equation is u2 4 = 0, so there are two real
eigenvalues l = 2 and 2. An eigenvector
!
g
h
with g 6= 0 corresponding to the first eigenvalue must satisfy the
equations
10g

4h = 2g, 24g

10h = 2h ) h = 2g.

In particular, any point ( x, y) on the straight line y = hx/g = 2x is


transformed to
!
!
!
!
x0
10
4
x
2x
=

=
,
y0
24
10
2x
4x
i.e., points on the same straight line y0 = 2x 0 .

25

lecture notes, me 340

26

Online lecture: complex exponentials


Suppose that r and q are two numbers and define the complex exponential
er+jq := er cos q + jer sin q = er ejq .
(105)
This is clearly a complex number with magnitude er and phase q.
The complex number ejq may be represented graphically by a point
on the unit circle in the complex plane with coordinates (cos q, sin q ).
In particular, ejp/2 = j, ejp = 1, ej3p/2 = j, and ej2p = 1. Moreover,
ejq ejj = ej(q + j) .

(106)
Figure 6: Eulers formula ejq =
cos q + j sin q implies that the complex
exponential ejq is a point on the unit circle in the complex plane with coordinates
given by the real and imaginary parts.

=
1

ejq
sin q
q

cos q

<

Example 13. For the complex number 2 ejp/3 = 2 cos(p/3) + j2 sin(p/3)


with magnitude 2 and phase p/3, the inverse is given by

The product

1
2 cos(p/3) j2 sin(p/3)
e jp/3
=
=
.
4
2
2 ejp/3
2 ejp/3

jp/3

=1

(107)

(108)

is the identity transformation.


If r (t) and q (t) are some functions of t,

d r(t)+jq (t)
dr (t)
dq (t)
e
=
+j
er(t)+jq (t) .
dt
dt
dt

(109)

Example 14. With r (t) = 0 and q (t) = wt, we obtain


1
ejwt + e
2

jwt

= cos wt.

(110)

lecture notes, me 340

Differentiation on both sides with respect to t yields


jw
ejwt e jwt =
2
which is easily verified by substitution.

w sin wt,

(111)

Since

d (a+jb)t
e
= (a + jb) e(a+jb)t
dt
for constants a and b, it follows that
Z t

e(a+jb)t dt =

(112)

e(a+jb)t
.
a + jb

(113)

Example 15. The definite integral


Z p/2w jwt
e + e jwt
0

dt =

ejwt

e
j2w

jwt p/2w

=
0

1
,
w

(114)

which also results from integrating cos wt over the interval [0, p/2w ].

Exercises
1. Write the complex exponential e2+jp/3 in rectangular form.
2. Determine the magnitude of the complex exponential
3. Write the complex number 3

ejp/4 .

j2 in polar form r ejq .

4. Compute the conjugate of the complex number 3 e


5. Compute the ratio of the complex numbers 3

j3p/4 .

j and ejp/3 /2.

6. Compute the magnitude of the ratio


2 e jp/2
.
ej2p/7
7. Find the polar form of the inverse of the complex number

3 + j.

8. Compute the derivative with respect to t of the complex exponen2


tial et jt .
9. Evaluate the product
e

jpt/2

d jpt/2
e
.
dt

10. Verify that u(t) = C ejwt satisfies the differential equation


d2 u(t)
+ w 2 u(t) = 0
dt2

for any constant C.


11. Graph the real part of the function e(

1/2+j)t .

27

lecture notes, me 340

Solutions
1. Here, r = e2 and q = p/3. The rectangular form is
p
p
p
1 2
3 2
2
2
e cos + je sin = e + j
e .
3
3
2
2
2. Since

1 = ejp ,
p

ej 4 =

1 ej 4 = ejp ej 4 = ej

5p
4

whose magnitude equals 1.


3. With r cos q = 3 and r sin q = 2, we find r =
p
and q = arcsin (2/ 13). The polar form is

13 e

j arcsin p2

13

32 + 22 =

13

4. Since cos( q ) = cos q and sin( q ) = sin q, the conjugate of


ejq = cos q + j sin q is e jq = cos q j sin q. It follows that
3e
5. Since
3

j=

j 3p
4

3p

= 3 ej 4 .
j arcsin p1

10 e

10

it follows that
3
j p3

e /2

10 e

6. Since

j arcsin p1

10

2e

j p3

p
= 2 10 e

r ejq
r
= e j( q
jj
h
he

j)

the magnitude of the ratio

2e jp/2
ej2p/7
is 2.
7. Since

the inverse equals

8. Here,

3 + j = 2 ej 6 ,
1
e
2

d t
e
dt

jt2

= (1

j p6

j2t) et

jt2

p +arcsin p1
3
10

28

lecture notes, me 340

9. Here,
j pt
2

d j pt
e 2 =e
dt

j pt
2

p j pt
e 2 = e
2

or, in other words,

10. Since

j pt
2

p
p j p j pt
p
e 2 e 2 = ej 2
2
2

p
j .
2
d
C ejwt = C jw ejwt ,
dt

it follows that
d2
C ejwt = C (jw ) (jw ) ejwt =
dt2
11. The real part of the function e(
shown below.

1/2+j)t

Cw 2 ejwt =
is e

t/2 cos t

w 2 u ( t ).

whose graph is

1
e

t/2 cos t

2p

29

lecture notes, me 340

Lesson 1 Dynamic systems and modeling


TAKE-AWAYS:
* A dynamic system consists of components with inputs and
outputs. Examples are amplifiers, delay lines, and integrators.
* Harmonic signals are given by
A cos(wt

q)

in terms of an amplitude A, an angular frequency w, and a


phase shift q. An expression of the form a cos wt + b sin wt
corresponds to a harmonic signal with
p
A = a2 + b2
and

cos q = p

a
a2

+ b2

, sin q = p

b
a2

+ b2

* Physical systems using electronic, mechanical, or fluid components may be modeled as dynamic systems, and designed
to output approximately the integral or derivative of a harmonic input.

30

lecture notes, me 340

31

Online lecture: components and harmonic signals


In these notes, a dynamic(al) system is a collection of physical
components that modify or generate signals. A signal is a function
of time, denoted by t. The notation i (t) identifies the signal i as a
time-dependent function. We sometimes omit the (t), when possible
time-dependence can be inferred from the context.
disturbance
r (t) +

e(t)

controller

u(t)

plant

o (t)

y(t)

Figure 7: A block representation of


a dynamic system in the presence of
external disturbances, in which the
output o (t) of the plant is controlled by
a signal u(t), generated by a controller,
whose input e(t) is the difference
between a reference signal r (t) and the
output y(t) of a sensor.

sensor

Each component of a system may have one or several inputs,


which receive the (input) signals i1 (t), i2 (t), . . ., and one or several
outputs, which return the (output) signals o1 (t), o2 (t), . . ..
i1 ( t )
o1 ( t )
i2 ( t )

Figure 8: A block representation of a


component of a dynamic system with
three inputs and two outputs.

o2 ( t )
i3 ( t )

We refer to single-input-single-output components or systems


using the acronym SISO. We use the acronym MIMO for multiinput-multi-output components or systems.
Example 16. An amplifier is a SISO component that increases the signal
strength by a given positive factor, known as the (amplifier) gain.
i(t)

o (t)

If the gain is denoted by K, then the value of the output o at time t equals
the value of the input i at time t multiplied by K. If the input signal i (t)
received by an amplifier equals e t sin t and the gain equals 5 then the
output signal o (t) returned by the component equals 5i (t) = 5e t sin t.

Figure 9: A block representation of an


amplifier with gain K.

lecture notes, me 340

32

Example 17. A delay line is a SISO component that delays a signal in


time by a given positive amount, known as the (time) delay.
i(t)

Figure 10: A block representation of a


delay line with delay D.

o (t)

If the delay is denoted by D, then the value of the output o at time t


equals the value of the input i at time t D. If i (t) = e t sin t and D = 1,
then o (t) = i (t 1) = e (t 1) sin(t 1).
Example 18. An integrator is a SISO component that integrates a signal
with respect to time, given some initial value.
i(t)

c+

Rt

Figure 11: A block representation of an


integrator with initial value c.

o (t)

If the initial value is denoted by c, then the value of the output o at time t
equals c plus the integral of the input i over the interval [0, t]. If i (t) = te t
and c = 0, then o (t) = 1 (1 + t)e t .
A signal of the form
A cos (wt

(115)

q)

with A > 0 is said to be harmonic with amplitude A, angular frequency w, and phase shift q. A harmonic signal with angular frequency2 w is periodic with period T = 2p/w.

The angular frequency equals the


change in phase per unit time. The frequency f = 1/T = w/2p equals the
number of periods per unit time.
Figure 12: The graph of a harmonic
signal obtained from the height above
the horizontal axis of a point moving
counterclockwise with angular speed w
on a circle of radius A centered at the
origin.

A
A cos(wt

wt0
q
t0

q)
t

Example 19. The signal


f (t) := a cos wt + b sin wt

(116)

lecture notes, me 340

33

is a sum of two harmonic signals with the same angular frequency (and
therefore the same period) and phase shifts 0 and p/2, respectively, since

p
sin wt = cos wt
.
(117)
2
Every harmonic signal A cos (wt
From A cos(wt
that

q ) can be written in the form (116):

q ) = A cos wt cos q + A sin wt sin q, it follows


a = A cos q and b = A sin q.

Square each side, add, and take the square root to obtain the amplitude:
p
A = a2 + b2 .
Let the phase shift q be the unique angle in [ p, p ) that satisfies
the equations
cos q = p

a
a2

+ b2

and sin q = p

b
a2

+ b2

It follows that a signal of the form (116) can be written as a harmonic signal
with period 2p/w and with amplitude and phase shift given in terms of the
individual amplitudes a and b.
Example 20. The derivative of the signal
f (t) := a cos wt + b sin wt
equals

But,

and

d f (t)
=
dt

wa sin wt + wb cos wt.

p
sin wt = cos wt +
= cos w t
2

p
cos wt = sin wt +
= sin w t
2

We conclude that

d f (t)
= w a cos w t
dt

(118)

T
4

(119)

+ b sin w t

T
4
T
4

T
4

(120)

(121)

(122)

or, in other words,


d f (t)
= wf
dt

T
4

This equals the original signal f (t) amplified by w and advanced3 by a


quarter period.

(123)
3

A signal that is delayed by a negative


quantity is said to be advanced by the
absolute value of the quantity.

lecture notes, me 340

Example 21. The definite integral of the signal


f (t) := a cos wt + b sin wt

(124)

over the interval [0, t] equals


Z t
0

t
b
cos wt
w
0
b
b
cos wt + .
w
w

a
sin wt
w
a
= sin wt
w

f (t ) dt =

But,

sin wt = cos wt

and

T
4

cos wt = sin wt
We conclude that
Z t
0

1
f (t ) dt = f
w

T
4

(125)

(126)

T
4

(127)

b
.
w

(128)

This equals the original signal f (t) amplified by 1/w, delayed by a quarter
of a period, and added to the constant signal b/w.

In-class worksheet: an electrical circuit


Consider the electrical circuit shown below, consisting of a capacitor
and a resistor in series.
Vab

Vbc
C

!i

Figure 13: An electrical circuit.

Vac

The voltage drop Vab across the capacitor in the direction of the
current i equals Q/C, where Q is the charge accumulated on the capacitor and C is the capacitance. Similarly, the voltage drop Vbc across
the resistor in the direction of the current equals Ri, where R is the
resistance.

34

lecture notes, me 340

Using Kirchhoffs laws4 , we conclude that

(129)

Vac (t) = Vab (t) + Vbc (t)


equals the total voltage drop at time t, and
Q ( t ) = Q (0) +

Z t
0

i (t ) dt

(130)

equals the charge accumulated on the capacitor at time t. Differentiation on both sides of this relationship shows that
dQ
( t ) = i ( t ).
dt

(131)

Suppose that Vac is a given harmonic signal A cos wt and assume


that Vab (t) = B cos(wt q ) for some unknown amplitude B and
phase shift q. Since Vab = Q/C, it follows that
Q(t) = CB cos(wt

q ),

(132)

which in turn implies that


i (t) =

dQ(t)
=
dt

wCB sin(wt

(133)

q)

and
wRCB sin(wt

Vbc (t) = Ri (t) =

q ).

(134)

Substitution in (129) then yields


A cos wt = B cos(wt

q)

wRCB sin(wt

q ).

(135)

wRC cos q )

(136)

This implies that


A = B(cos q + wRC sin q ), 0 = B(sin q
or, in other words,
B= p

and
cos q = p

(137)

1 + (wRC )2

1 + (wRC )2

sin q = p

wRC
1 + (wRC )2

(138)

If we design the circuit in Fig. 13, such that RC 1/w, then


B A and q 0. In this case,
Vbc (t) =

wRCB sin(wt

or, in other words,

q)

wRCA sin wt

(139)

dVac
( t );
(140)
dt
the output voltage across the resistor is approximately proportional
to the derivative of the input voltage across the circuit.
Vbc (t) RC

35

Kirchhoffs 1st law: The voltage


change between two nodes in a circuit
is the same independently of the path
between the points. Kirchhoffs 2nd
law: Charge cannot accumulate at
nodes in a circuit.

lecture notes, me 340

If, instead, we design the circuit in Fig. 13, such that RC


then wRCB A and q p/2. In this case,
Vab (t) = B cos(wt

q)

A
sin wt
wRC

1/w,

(141)

or, in other words,


Vab (t)

1
RC

Z t
0

Vac (t ) dt;

(142)

the output voltage across the capacitor is approximately proportional


to the integral of the input voltage across the circuit.

The time constant of the electric circuit equals the product RC


of the resistance and the capacitance. When the time constant
is much smaller than the period of the harmonic input signal
(RC 1/w), the circuit acts as a differentiator. When the
time constant is much greater than the period of the harmonic
input signal (RC
1/w), the circuit acts as an integrator.

In-class worksheet: a mechanical suspension


Consider the mechanical suspension shown below, consisting of a
spring and a damper in parallel.
Free-body diagram:

~Fe

Figure 14: A mechanical suspension.

~Fe

i
k

~Fs =

kxi

~Fd =

dvi

Let the unit vector i be parallel to the line of action of the suspension. Denote by x the displacement of the massless plate along i
relative to its position when the spring is undeformed. It follows that
x ( t ) = x (0) +

Z t
0

v(t ) dt,

(143)

36

lecture notes, me 340

37

where v(t) is the velocity of the plate along i at time t. Differentiation


on both sides of this relationship shows that
dx
( t ) = v ( t ).
dt

(144)

The force ~Fs from the spring on the massless plate equals kxi,
where k is the stiffness. Similarly, the force ~Fd from the damper on
the massless plate equals dvi, where d is the damping coefficient.
Using Newtons 2nd law5 , and the fact that the plate is massless, we
conclude that
~Fe (t) = ~Fs (t) ~Fd (t)
(145)
equals the external force applied to the plate at time t.
Now suppose that ~Fe (t) = A cos(wt)i and assume that ~Fs (t) =
B cos(wt q )i for some unknown amplitude B and phase shift q.
Since ~Fs = kxi, it follows that
B
cos(wt
k

x (t) =

q ),

(146)

which in turn implies that


v(t) =

dx (t)
wB
=
sin(wt
dt
k

and

wdB
sin(wt
k

dv(t)i =

~Fd (t) =

(147)

q)

q )i.

(148)

Substitution in (145) then yields


A cos(wt) =

B cos(wt

q) +

This implies that

d
A = B cos q + w sin q , 0 =
k
or, in other words,
B= p

and
cos q =

1
1 + (wd/k)2

wdB
sin(wt
k

B sin q

q ).

d
w cos q
k

(150)

(151)

1 + (wd/k)2

sin q =

(149)

wd/k
1 + (wd/k )2

(152)

If we design the suspension in Fig. 14, such that d/k 1/w, then
B A and q p. In this case,

~Fd (t) =

wdB
sin(wt
k

wdA
q )i
sin(wt)i
k

(153)

Newtons 2nd law: The net force


acting on an object equals its mass
multiplied by its acceleration. For a
massless object, it follows that the net
force must vanish.

lecture notes, me 340

or, in other words,

d d~Fe
( t );
(154)
k dt
the force applied to the damper is proportional to the rate of change
of the external force applied to the plate.
If, instead, we design the suspension in Fig. 14, such that d/k
1/w, then wdB/k A and q p/2. In this case,

~Fd (t)

q )i

~Fs (t) = B cos(wt

kA
sin(wt)i
wd

(155)

or, in other words,

~Fs (t)

k
d

Z t
0

~Fe (t ) dt;

(156)

the force applied to the spring is approximately proportional to the


integral of the external force applied to the plate.

The time constant of the mechanical suspension equals the


ratio d/k of the damping coefficient and the spring stiffness.
When the time constant is much smaller than the period of
the harmonic input signal (d/k 1/w), the suspension acts
as a differentiator. When the time constant is much greater
than the period of the harmonic input signal (d/k
1/w), the
suspension acts as an integrator.

In-class worksheet: a water reservoir


Consider the liquid level system shown below, consisting of a water
reservoir and an inlet valve in series.
Figure 15: A water reservoir.

pa
g

pa

h
C

Dp

The pressure drop Pv across the valve equals rRq, where R is the
valve resistance, q is the fluid flow rate into the tank, and r is the liquid

38

lecture notes, me 340

density. Similarly, the pressure drop Pt between the bottom and top
of the liquid in the tank equals rgh, where g is the acceleration of
gravity and h is the head of the liquid.
By conservation of mass, it follows that
rq = rC

dh
,
dt

(157)

where C is the cross-sectional area of the tank. Moreover,


Dp = Pv + Pt

(158)

equals the pressure increase across the pump.


Now suppose that Dp(t) = p0 + A cos wt is given and assume that
Pt (t) = p0 + B cos(wt q ) for some unknown amplitude B < p0 and
phase shift q. Since Pt = rgh, it follows that
h(t) =

p0
B
+
cos(wt
rg rg

q ),

(159)

which in turn implies that


rq(t) = rC

dh(t)
=
dt

and
Pv (t) = rRq(t) =

wCB
sin(wt
g

q)

(160)

wRCB
sin(wt
g

q ).

(161)

Substitution in (158) then yields


p0 + A cos(wt) = p0 + B cos(wt

q)

wRCB
sin(wt
g

This implies that

RC
A = B cos q + w
sin q , 0 = B sin q
g
or, in other words,
B= p

and
cos q = p

1
1 + (wRC/g)2

RC
w
cos q
g

q ).

(162)

(163)

(164)

1 + (wRC/g)2

sin q = p

wRC/g
1 + (wRC/g)2

(165)

If we design the reservor in Fig. 15, such that RC/g 1/w, then
B A and q 0. In this case,
Pv (t) =

wRCB
sin(wt
g

or, in other words,


Pv (t)

q)

wRCA
sin wt
g

RC dDp
( t );
g dt

(166)

(167)

39

lecture notes, me 340

the pressure drop across the valve is proportional to the rate of


change of the pressure increase across the pump.
If, instead, we design the reservoir in Fig. 15, such that RC/g
1/w, then wRCB/g A and q p/2. In this case,
Pt (t)

p0 = B cos(wt

q)

gA
sin wt
wRC

(168)

p0 ) dt;

(169)

or, in other words,


Pt (t)

g
RC

p0

Z t
0

(Dp(t )

the alternating part of the pressure drop in the tank is approximately


proportional to the integral of the alternating part of the pressure
increase across the pump.

The time constant of the water reservoir equals the ratio RC/g
of the product of the valve resistance and tank cross section
with the acceleration of gravity. When the time constant is
much smaller than the period of the harmonic input signal
(RC/g 1/w), the reservoir acts as a differentiator. When
the time constant is much greater than the period of the harmonic input signal (RC/g
1/w), the reservoir acts as an
integrator.

Exercises
1. Find a unique angle q 2 [ p, p ), such that
1
cos q = p , sin q =
2

1
p .
2

2. Find a unique angle q 2 [ p, p ), such that


cos q =

1
p , sin q =
2

1
p .
2

3. How many angles q 2 [ p, p ) satisfy the condition that tan q = 1?


4. Determine A, w, and q such that
cos 2t

sin 2t = A cos(wt

q ).

5. Determine a, b, and w such that


3 sin(4t

p/3) = a cos wt + b sin wt.

40

lecture notes, me 340

6. Determine B, w, and j such that


3 cos(t

p/4) + 4 sin(t

p/4) = B sin(wt + j).

7. Determine B and q such that


cos t = B cos(t

2B sin(t

q)

q ).

8. Graph the signal f (t) := 3 cos pt sin pt, its derivative, and its
definite integral over the interval [0, t] in the same diagram.
9. Suppose that the input signal to each of the components below
is given by i (t) = e t for t
0 and 0 otherwise. In each case,
determine the output signal o (t) for t 0, and graph i (t) and o (t).
i(t)

i(t)

i(t)

o (t)

0.6

o (t)

0.6

0.6 +

Rt

o (t)

Solutions
p
1. Since cos is an even function, cos q = 1/ 2 implies that q is either
p/4 or p/4. Only the latter satisfies the condition that sin q =
p
1/ 2.
p
2. Since cos is an even function, cos q = 1/ 2 implies that q is
either 3p/4 or 3p/4. Only the latter satisfies the condition that
p
sin q = 1/ 2.
3. Since tan is a periodic function with period p, tan q = 1 if q is
either p/4 or 3p/4.
4. Since A cos(wt
w = 2 and

q ) = A cos q cos wt + A sin q sin wt, it follows that


A cos q = 1, A sin q =

1.

Squaring and adding the two equations yields A2 = 2 ) A =


The angle q lies in the fourth quadrant, i.e., q = p/4.

2.

41

lecture notes, me 340

5. Since cos p/3 = 1/2 and sin p/3 =

3/2,

p
3
3 3
3 sin(4t p/3) = sin 4t
cos 4t.
2
2
p
It follows that w = 4, a = 3 3/2, and b = 3/2.
6. Since
3 cos(t
4 sin(t

3
3
p/4) = p cos t + p sin t,
2
2
4
4
p/4) = p sin t p cos t,
2
2

and
B sin(wt + j) = B sin wt cos j + B cos wt sin j,
it follows that w = 1 and
1
7
p = B sin j, p = B cos j.
2
2
Squaring and adding gives 25 = B2 ) B = 5. The angle j lies in
p
the fourth quadrant and equals arcsin(1/5 2).
7. Since
cos t = cos(t

q + q ) = cos(t

q ) cos q

sin(t

q ) sin q,

it follows that
cos q = B, sin q = 2B,
p
i.e., B = 1/ 5 and q = arcsin(2/ 5).

8. Here,

and

d f (t)
=
dt
Z t
0

f (t ) dt =

3p sin pt

p cos pt

3
1
sin pt + cos pt
p
p

1
.
p

d f (t)/dt
5
p

Rt
0

9. For the amplifier, o (t) = 0.6e

for t

f (t ) dt

0 and 0 otherwise.

42

lecture notes, me 340

1
i(t)
o (t)
1

For the delay line, o (t) = e

(t 0.6)

for t

0.6 and 0 otherwise.

1
i(t)

o (t)

1
For the integrator, o (t) = 1.6

for t

0 and 0 otherwise.

o (t)

1
i(t)

43

lecture notes, me 340

Lesson 2 Block diagrams and system representations


TAKE-AWAYS:
* A block diagram is a wiring of components, using summing
and splitting junctions, to represent the input-output relationship of a dynamic system.
* The initial-value problem
do
( t ) = i ( t ), o (0) = c
dt
corresponds to the relationship between the input i (t) and
output o (t) of an integrator with initial value c.
* The initial-value problem
do
(t) + ao (t) = i (t), o (0) = c
dt
corresponds to the relationship between the input i (t) and
output o (t) of an integrator with initial value c and negative
feedback with gain a.
* The initial-value problem
d2 o
do
do
+ a + bo (t) = i (t), o (0) = d,
(0) = c
2
dt
dt
dt
corresponds to the relationship between the input i (t) and
output o (t) of a double integrator with initial values c and d
and negative feedback with gains a and b.

44

lecture notes, me 340

45

Online lecture: initial-value problems


Amplifiers, line delays, and integrators are examples of
system components with block representations shown in Lesson 1. In
addition, we consider multi-input-single-output summing junctions
and single-input-multiple-output splitting junctions.
i2 ( t )

Figure 16: Example block representations of summing and splitting


junctions.

i(t)

i1 ( t ) +

i(t)
i1 ( t )

i2 ( t )

i3 ( t )

i(t)

i(t)

i3 ( t )

By suitable wiring, a collection of components and junctions may


represent a composite relationship between inputs and outputs that
could not be captured by the simple components considered thus far.
Such a wiring diagram of components is called a block diagram. The
order in which components are wired together matters.
Example 22. Consider the following block diagram.
i(t)

x (t)

c+

Here,
x (t) = Ki (t),

o (t) = c +

Together these imply that


o (t) = c + K

Z t
0

Rt

Figure 17: An amplifier followed by an


integrator.

o (t)

Z t
0

x (t ) dt.

i (t ) dt.

If we switch the order of the components, then

Z t
o (t) = K c +
i (t ) dt ,
0

which equals the earlier result if the initial value c = 0.

(170)

(171)

(172)

lecture notes, me 340

Example 23. Consider the following block diagram.


i(t)

Here,
x (t) = c +

x (t)

Rt

c+

o (t)

o (t) = x (t

D ).

Figure 18: Integrator followed by a


delay line.

Z t
0

i (t ) dt,

(173)

Together these imply that, for t > D,


o (t) = c +

Z t D
0

i (t ) dt.

(174)

If we switch the order of the components, then


o (t) = c +
where s = t

Z t
0

D) dt = c +

i(t

Z t D
D

i (s) ds,

(175)

D. This equals the previous result if i (t) = 0 for t < 0.

The input-output relationship for an integrator


o (t) = c +

Z t
0

i (t ) dt

(176)

is equivalent to the initial-value problem


do
( t ) = i ( t ).
dt

o (0) = c,

(177)

More complicated initial-value problems may be obtained by adding


connections and additional components to the basic integrator.
Example 24. Consider the following block diagram.
i(t) +

x (t)

Rt

c+

Figure 19: Integrator with feedback.

o (t)

Here,
x (t) = i(t)

ao (t),

o (t) = c +

Z t
0

x (t ) dt.

(178)

Together these imply that


o (t) = c +

Z t
0

i(t )

ao (t ) dt,

(179)

46

lecture notes, me 340

which is equivalent to the initial-value problem


do
(t) + ao (t) = i (t);
dt

o (0) = c,

(180)

the general form of a first-order, scalar, constant-coefficient, linear, nonhomogeneous differential equation.
Example 25. Consider the following block diagram
Figure 20: A double integrator with
feedback.

i(t) +

x (t)

y(t)

ay(t)

bo (t) and

c+

Rt

d+

Rt
0

o (t)

Here, x (t) = i (t)


y(t) = c +

Z t
0

x (t ) dt,

o (t) = d +

Z t
0

y(t ) dt.

(181)

Together these imply that y(0) = c, o (0) = d, and


dy
=i
dt

ay

bo,

do
=y
dt

(182)

or, in other words,


d2 o
do
+ a + bo = i,
dt
dt2

o (0) = d,

do
(0) = c;
dt

(183)

the general form of a second-order, scalar, constant-coefficient, linear, nonhomogeneous differential equation.

In-class worksheet: a mechanical suspension


Consider the single-layered mechanical suspension shown on the
next page. Let the unit vector i be parallel to the line of action of the
suspension. Denote by x the displacement of the massless plate along
i relative to its position when the spring is undeformed. It follows
that
dx
x (t) :=
( t ) = v ( t ).
(184)
dt

47

lecture notes, me 340

Free-body diagram:
f (t)i

Figure 21: A single-layered suspension.

f (t)i

i
d

~Fs =

kxi

~Fd =

dvi

where v(t) is the velocity of the plate along i at time t.


The force ~Fs from the spring on the massless plate equals kxi,
where k is the stiffness. Similarly, the force ~Fd from the damper on
the massless plate equals dvi, where d is the damping coefficient.
Using Newtons 2nd law, and the fact that the plate is massless, we
conclude that
f (t) = kx (t) + d x (t).
(185)
We can convert this ordinary differential equation to the integral
equation
Z t
f (t ) kx (t )
x ( t ) = x (0) +
dt
(186)
d
0
A corresponding single-integrator block diagram is shown below.
f (t) +

1
d

x (0) +

Rt
0

x (t)

48

Figure 22: An equivalent block diagram.

lecture notes, me 340

49

Consider, instead, the multi-layered mechanical suspension shown


in the figure below.
f (t)i

Figure 23: A multi-layered suspension.

x2

k2
x1

d2

d1

k1

Let the unit vector i again be parallel to the line of action of the
suspension. Denote by x1 and x2 the displacements of the lower and
upper massless plates, respectively, along i relative to their position
in the configuration where both springs are undeformed. It follows
that
x 1 (t) :=

dx1
( t ) = v1 ( t ),
dt

x 2 (t) :=

dx2
( t ) = v2 ( t ),
dt

(187)

where v1 (t) and v2 (t) are the velocities of the two plates along i at
time t.
Consider the free-body diagrams shown on the next page. If we
apply Newtons 2nd law to the lower plate, we obtain the equation of
force balance
x1 (t)) = k1 x1 (t) + d1 x 1 (t),

k 2 ( x2 ( t )

(188)

where k1 and k2 are the corresponding spring stiffnesses and d1 is the


damping coefficient. Similarly, if we apply Newtons 2nd law to the
upper plate, we obtain the equation of force balance
f ( t ) = k 2 x2 ( t )

x1 (t) + d2 x 2 (t),

(189)

where d2 is the other damping coefficient.


We can convert these ordinary differential equations to the integral
equations
x1 ( t ) = x1 (0) +

Z t
k 2 x2 ( t )
0

x1 ( t )
d1

k 1 x1 ( t )

dt

(190)

lecture notes, me 340

f (t)i

Figure 24: Free body diagrams for the


two plates.

Free-body diagrams:

k 2 ( x2

x1 )i
x1 )i

k 2 ( x2

k1 x1 i

d2 v2 i

d1 v1 i

and
x2 ( t ) = x2 (0) +

Z t
f (t )
0

k 2 x2 ( t )
d2

x1 ( t )

dt

(191)

A corresponding two-integrator block diagram is shown below.


f (t) +

1
d2

x2 (0) +

Rt
0

Figure 25: An equivalent block diagram.

x2 ( t )

+
k2

1
d1

x1 (0) +

Rt
0

x1 ( t )

k1

Now suppose that the upper plate has a nonzero mass m. Application of Newtons 2nd law then results in
m x2 (t) = f (t)

k 2 x2 ( t )

where
x2 (t) :=

x1 ( t )

d2 x2
( t ).
dt2

50

d2 x 2 (t)

(192)

(193)

lecture notes, me 340

If we let y(t) := x 2 (t) and y (t) = x2 (t), then this is equivalent to the
first-order equations
x 2 (t) = y(t),

my (t) = f (t)

k 2 x2 ( t )

x1 ( t )

(194)

d2 y ( t )

and the corresponding integral equations


x2 ( t ) = x2 (0) +

Z t
0

y(t ) dt

(195)

and
y ( t ) = y (0) +

Z t
f (t )

k 2 x2 ( t )

x1 ( t )

d2 y ( t )

dt

(196)

A corresponding three-integrator block diagram is shown below.


Figure 26: An equivalent block diagram.

d2

f (t) +

1
m

y (0) +

Rt
0

y(t)

x2 (0) +

Rt
0

x2 ( t )

+
k2

1
d1

x1 (0) +

Rt

x1 ( t )

k1

Exercises
1. Determine i (t) and o (t) such that
i(t)

2 sin 2t

2. Determine i (t) and o (t) such that

1+

Rt
0

o (t)

51

lecture notes, me 340

i(t)

2+

2 cos 3t

Rt

o (t)

3. Determine a value for c, such that


i(t)

c+

x (t)

Rt
0

4. Suppose that u(t) = 1 for t


and o (t), such that
u(t)

4e

t sin 2t

0 and 0 otherwise. Determine x (t)

x (t)

Rt

o (t)

5. Suppose that u(t) = 1 for t


and o (t), such that
u(t)

0 and 0 otherwise. Determine x (t)

x (t)

o (t)

Rt
0

6. Draw a block diagram representing the initial-value problem


o (0) = 3,

do
( t ) + o ( t ) = i ( t ).
dt

7. Draw a block diagram representing the initial-value problem


o (0) = 0, 2

do
(t) = i(t)
dt

o ( t ).

8. Draw a block diagram representing the initial-value problem


o (0) =

1,

do
d2 o
do
(0) = 2,
(t) + (t) + 4o (t) = i (t).
2
dt
dt
dt

9. Formulate an initial-value problem corresponding to the block


diagram
2
i(t) +

Rt
0

y(t) +

1+

Rt
0

o (t)

52

lecture notes, me 340

10. Suppose that u(t) = 1 for t


x1 ( t ) =

1
3t + e
9

0 and 0 otherwise. Verify that

3t

1 , x2 ( t ) =

1
6t
18

3t

+1

satisfy the block diagram


u(t) +

Rt

1
2

x2 ( t )

+
2

Rt

x1 ( t )

Solutions
1. Here,
2 sin 2t = 2i (t), o (t) =

1+

It follows that i (t) = sin 2t and o (t) =

Z t
0

2 sin 2t dt.

cos 2t.

2. Here,
2 cos 3t =

2+

Z t
0

i (t ) dt, o (t) = 2( 2 cos 3t).

It follows that i (t) = 6 sin 3t and o (t) =


3. Here,
x (t) = c +

Z t
0

i (t ) dt, 4e

4 cos 3t.

sin 2t = Kx (t)

which implies that x (0) = c = 0.


4. For 0 t 1
For t

x (t) =

1,
x (t) =

Z t
0

Z t
0

dt = t, o (t) = 0.

dt = t, o (t) = x (t

1) = t

1.

53

lecture notes, me 340

5. For 0 t 1,
x (t) = 0, o (t) =
For t

1,
x (t) = 1, o (t) =

Z t
0

Z t
0

x (t ) dt = 0.

x (t ) dt =

Z t
1

dt = t

1.

6. Here,
do
(t) = i(t)
dt

o ( t ), o (0) = 3 ) o ( t ) = 3 +

Z t

o (t ) dt

i(t )

A corresponding block diagram is shown below.


i(t) +

3+

Rt

o (t)

7. Here,
do
i(t) o (t)
(t) =
, o (0) = 0 ) o ( t )
dt
2

Z t
i (t )

o (t )
2

dt.

A corresponding block diagram is shown below.


i(t) +

Rt

1
2

o (t)

8. Here, y(t) = do (t)/dt implies that


o (t) =
and
dy
(t) = i(t)
dt

4o (t)

1+

Z t
0

y(t ) dt

y(t) ) y(t) = 2 +

Z t
0

i(t )

4o (t )

y(t ) dt

Rt

A corresponding block diagram is shown below.

i(t) +

2+

Rt
0

y(t)

1+

o (t)

54

lecture notes, me 340

9. Here,
dy
(t) + 3y(t) = i (t), y(0) = 0
dt
do
(t) + 2o (t) = y(t), o (0) = 1
dt
This implies that
d2 o
do
dy
(t) + 2 (t) =
(t) = i(t)
dt
dt
dt2
and

3y(t) = i (t)

do
(0) = y (0)
dt

do
(t)
dt

6o (t)

2o (0) = 2

or, in other words,


d2 o
do
do
(t) + 5 (t) + 6o (t) = i (t), o (0) = 0,
(0) = 2
2
dt
dt
dt
10. Since x1 (0) = x2 (0) = 0, the given functions are compatible with
the initial values of the integrators. Moreover,
x2 ( t )

x1 ( t ) =

1
6

1
e
6

3t

Consequently,
Z t
0

2 x2 ( t )

x1 (t ) dt =

Z t
1
0

1
e
3

3t

dt = x1 (t)

and
Z t
1
0

2 x2 ( t )

x1 ( t )

dt =

Z t
1
0

1
+ e
3 6

3t

dt = x2 (t)

55

lecture notes, me 340

Lesson 3: First-order responses and system properties


TAKE-AWAYS:
* The solution to the initial-value problem
dy
(t) + ay(t) = f (t),
dt
is given by
y(t) = ce
where

at

f (#) g (#) ( t ) : =

y (0) = c

f (#) e
Z t
0

a#

f (t ) g(t

( t ),

t ) dt

is the convolution of the functions f and g.


* The free response of a dynamic system is the output given
zero input. For the initial-value problem above, the free response equals ce at .
* The unit step response of a dynamic system is the output
when the input equals 1 for t 0 and 0 otherwise, denoted by
u(t), and with c = 0. For the initial-value problem above, the
unit step response equals (1 e at )/a.
* The unit impulse response of dynamic system is the limit as
D ! 0 of the output when the input equals 1/D for 0 t D
and 0 otherwise, i.e., u(t) u(t D) /D, and with c = 0. For
the initial-value problem above, the unit step response equals
the free response with initial value c = 1, i.e., e at .
* The unit impulse d(t) is a symbolic notation representing
the limit as D ! 0 of the signal u(t) u(t D) /D. In
particular,

d (#) e

a#

=e

at

56

lecture notes, me 340

57

Online lecture: convolutions


The block diagram in Example 24 corresponds to the initialvalue problem
dy
(t) + ay(t) = f (t),
dt

y (0) = c

(197)

where we have written y for the output and f for the input. The
solution to this initial-value problem is given by
at

y(t) = ce
If we let6

a#

f (#) e

Z t
0

(t) :=

f (t )e

a(t t )

dt.

(198)
6

Z t

f (t )e

a(t t )

dt

(199)

denote the convolution of f with the exponentially decaying signal


e at , then

y(t) = ce at + f (#) e a# (t).


(200)
To verify that this is the correct solution, we substitute back into
the initial-value problem:
Since e0 = 1 and

f (#) e

a#

(0) =

Z 0

f (t )eat dt = 0,

it follows that y(0) = c. This verifies the initial condition.


To verify the differential equation, recall that
d
dt

Z t
0

g(t, t) dt =

It follows that

d
f (#) e
dt

a#

Z t
g(t, t)

(t) =

dt + g(t, t).

a f (#) e

a#

(t) + f (t)

Finally,
dy
(t) =
dt

ace

at

a f (#) e

a#

(t) + f (t) =

ay(t) + f (t).

When a = 0, there is no feedback, and the output signal equals the


output of an integrator with initial value c, i.e.,
y(t) = c +

Z t
0

f (t ) dt.

(201)

In the notation f (#), the # symbol is


used to represent the argument of the
function f , in order to avoid duplicate
use of t.

lecture notes, me 340

Now suppose that a > 0.


Example 26. If f (t) = 0 for t
0, i.e., in the absence of any input, the
solution in (198) becomes the free response
y(t) = ce

at

(202)
Figure 27: The free response.

y(t)
c

ce

at

c/e

1/a

The output decays to a fraction 1/e of its original value after the characteristic time 1/a, known as the time constant of the system.
The unit step u(t) is a signal that equals 1 for t
wise.

0 and 0 other-

Figure 28: The unit step signal.

u(t)

Example 27. If f (t) = u(t) and c = 0, the solution7 in (198) becomes

Z t
t
1
y(t) = u(#) e a# (t) :=
e a(t t ) dt = e a(t t )
(203)
a
0
0
or, in other words,

1
1 at
e .
(204)
a a
The output converges to a steady-state value yss = 1/a in such a way
that y(t) yss decays to 0 exactly as the free response of a first-order system
with time constant 1/a.
y(t) =

This is the unit step response.

58

lecture notes, me 340

The delayed unit step signal u(t D) equals 1 for t


D and 0
otherwise. The signal
u(t) u(t D)
(205)
D
then equals 0 for t < 0 and t > D and 1/D for 0 t D. This is a
pulse with amplitude 1/D and pulse width D.
Figure 29: A pulse signal with amplitude 1/D and width D.

u(t) u(t D)
D

1/D

Example 28. For f (t) = D1 u(t)


in (198) for t D becomes
y(t) =

1
D

Z D
0

a(t t )

D) and with c = 0, the solution

u(t

dt =

at Z D

eat dt =

e at at
e
aD

(206)
0

or, in other words,

eaD 1
e
aD
Using lHospitals rule, we recall that
y(t) =

lim

x !0

ex

1
x

at

= 1.

(207)

(208)

In the limit as the pulse width D ! 0, the output thus converges to the free
response with initial condition equal to 1. We call this the unit impulse
response. In particular, we have

u (#) u (# D )
at
a#
e
= lim
e
( t ).
(209)
D
D !0
We represent the right-hand side symbolically by the convolution

d(#) e a# (t).

(210)

59

lecture notes, me 340

60

The unit impulse d(t) acts in convolution with the exponentially decaying signal e at as multiplication by 1.

Figure 30: The unit impulse represents


a limit of pulse signals.

d(t)
D!0

)
t

In-class worksheet: a mechanical suspension


Consider the single-layered mechanical suspension and the corresponding block diagram of an integrator with feedback shown in the
figure below.
f (t)i

Figure 31: A single-layered suspension


and the corresponding block diagram.

i
d

f (t) +

1
d

x (0) +

Rt
0

x (t)

lecture notes, me 340

If x (0) = 0, it follows that

Z t
f (#)
f (t )
x (t) =
e k#/d (t) =
e
d
d
0

k (t t )/d

dt.

(211)

Suppose that f (t) = 1/D for 0 t D and 0 otherwise.


f (t)
1/D

D
Then, for 0 t D,
x (t) =

Z t
1
0

dD

k(t t )/d

or, in other words,


x (t) =
Similarly, for t

dt =

1
1
kD

1
e
kD
kt/d

k (t t )/d

t =t

(212)
t =0

(213)

D,
x (t) = x (D)e

k (t D)/d

(214)

But,

1
1 e kD/d ,
(215)
kD
which converges to 1/d as D ! 0. By substitution into (214) It follows that the unit impulse response equals
x (D) =

x (t) =
Suppose that f (t) = 2(D

1
e
d

kt/d

(216)

t)/D2 for 0 t D and 0 otherwise.

f (t)
2/D

61

lecture notes, me 340

Then, for 0 t D,
x (t) =

Z t
2( D
0

t)

dD2

k (t t )/d

dt = 2d

1 + k (D t )/d
e
k 2 D2

k (t t )/d

t =t
t =0

(217)

or, in other words,


x (t) = 2d
As before, for t

1 + k(D t)/d
k 2 D2

2d

1 + kD/d
e
k 2 D2

kt/d

(218)

D,
x (t) = x (D)e

k (t D)/d

(219)

But,

2d
kD/d
1
1
+
kD/d
e
(220)
k 2 D2
which also converges to 1/d as D ! 0 by lHospitals rule. It follows that the response to the given input f (t) converges to the unit
impulse response as D ! 0.
x (D) =

In-class worksheet: linearity, causality, and time-invariance


A linear SISO dynamic system satisfies the superposition principle:
If the input signal i1 (t) results in the output o1 (t), and the input signal i2 (t) results in the output o2 (t), then the input signal ai1 (t) + i2 (t) results in the output ao1 (t) + o2 (t).
Example 29. For a zero-initial-value integrator
Z t
0

(a f (t ) + g(t )) dt = a

Z t
0

f (t ) dt +

Z t
0

What happens if the initial value is


different from zero?

g(t ) dt.

(221)

A system consisting of a single zero-initial-value integrator is therefore


linear. Similarly, for a delay line
ai1 (t

D ) + i2 ( t

D) = ai1 (#) + i2 (#) (t

D)

(222)

A system consisting of a single delay line is therefore linear. It is straightforward to show that the same conclusion applies to a system consisting of a
single amplifier.
Block diagrams consisting of linear components, summing, and
splitting junctions are linear.
Example 30. For a zero-initial-value integrator with feedback, linearity
follows from a study of the two block diagrams below. It also follows from
the explicit solution

Z t
a#
o ( t ) = i (#) e
(t) :=
i (t )e a(t t ) dt
(223)
0

since the convolution integral is linear in i.

62

lecture notes, me 340

i1 ( t ), i2 ( t ) +

Rt
0

Figure 32: An integrator with feedback


and zero initial value is linear.

o1 ( t ), o2 ( t )

a
ai1 (t) + i2 (t) +

Rt

ao1 (t) + o2 (t)

A time-invariant SISO dynamic system satisfies the time-shift


principle:
If the input signal i (t)u(t) results in the output o (t)u(t),
then the input signal i (t D)u(t D) results in the output
o ( t D ) u ( t D ).
This is clearly true of amplifiers, delay lines, and zero-initial-value
integrators. Block diagrams consisting of time-invariant components,
summing, and splitting junctions are time invariant.
Example 31. Replace f (t) by f (t)u(t) in the integral in (198) and assume
that c = 0. It follows that

Z t
y(t) = f (#)u(#) e a# (t) :=
f (t )u(t )e a(t t ) dt.
(224)
0

Replace, instead, f (t) by f (t D)u(t D) in the integral in (198) and


assume again that c = 0. Then, for t < D, we obtain y(t) = 0. For t D,
y(t) =

Z t
D

f (t

D)u(t

D)e

a(t t )

dt

(225)

or, in other words,


y(t) =

Z t D
0

f (s)u(s)e

a(t D s)

where t = D + s and dt = ds. But this equals

f (#)u(#) e a# (t D),

ds,

63

(226)

(227)

lecture notes, me 340

i.e., the output signal obtained with f (t)u(t) as input, but delayed by D.
Thus, since

f (# D)u(# D) e a# (t) = f (#)u(#) e a# (t D)u(t D),

(228)

a zero-initial-value integrator with feedback is time invariant.


A causal SISO dynamical system satisfies the causality principle:
If the input signals i1 (t) and i2 (t) are identical for t D, then
the outputs o1 (t) and o2 (t) are identical for t D.
This is clearly true for amplifiers and delay lines. It is also true for a
zero-initial-value integrator with feedback, since the output is given
by the convolution

Z t
i (#) e a# (t) =
i (t )e a(t t ) dt,
(229)
0

which only depends on values of i (t ) for t t.


Block diagrams consisting of causal components, summing, and
splitting junctions are causal.

Exercises
1. Use a variable substitution to show that

f (#) g (#) ( t ) = g (#) f (#) ( t )


2. Suppose that f (0) = 0 and let F (t) = d f (t)/dt. Show that

u (#) F (#) ( t ) = f ( t )
3. Compute the convolution

a#

b#

( t ).

when a 6= b.
4. Use the previous result to compute the convolution

sin 2# sin 3# (t).

64

lecture notes, me 340

5. Describe the convolution

for t

a#

sin w# e

(t)

1/a.

6. Find the solution for t

0 to the initial-value problem

dy
(t) + 2y(t) = t,
dt
7. Find the solution for t

y (0) =

1.

0 to the initial-value problem

dy
(t) + y(t) = e t ,
dt

y(0) = 1.

8. Find the free response, the unit step response, and the unit impulse response of the dynamic system shown below.
i(t) +

2+

Rt

o (t)

1
2

9. Find the free response, the unit step response, and the unit impulse response of the dynamic system shown below.
i(t) +

1+

Rt

o (t)

1
2

10. Find the response of the dynamic system below in the limit as
b ! and compare this to the unit impulse response. Graph the
input and output for several representative values of b.
be

bt

Rt

o (t)

65

lecture notes, me 340

Solutions
1. Here,

f (#) g (#) ( t ) =

or, with the substitution s = t


Z 0

f (t

2. Here,

or, since u(t

s ) g(s ) ds =

u (#) F (#) ( t ) =

t ) dt

f (t ) g(t

t,

Z t

g (#) f (#) ( t )
t ) dt

F (t )u(t

t ) = 1 for 0 t t,
Z t

F (t ) dt = f (t)

3. Here,

Z t

a#

b#

or, in other words,

(t) =

f (0) = f ( t ).

Z t

eat e b(t

t =t

e(a b)t + bt
a b

=
t =0

eat
a

t)

dt

e bt
.
b

since a 6= b.
4. Since
sin wt =

ejwt

it follows that

j2#
e
sin 2# sin 3# (t) =

e
j2
e
j2

jwt

j2#

ej3#

e
j2

But,

ej3t
ej2# ej3# (t) =

ej2# e

j3#

(t) =

ej2t

j3#

( t ).

ej2t
j
e
j5

j3t

ej3t e j2t
ej3# (t) =
j5

j2t
e
e j3t
e j2# e j3# (t) =
j
e

j2#

By linearity, (sin 2# sin 3#) (t) equals

1 ej3t ej2t
ej2t e j3t
ej3t e
4
j
j5
j5
2
3
=
sin 3t + sin 2t
5
5

j2t

j2t

e
j

j3t

66

lecture notes, me 340

5. Since

ejwt

sin wt =
it follows that

sin w# e

a#

1 jw#
e e
j2

(t) =

jwt

e
j2

a#

1
e
j2

(t)

jw#

a#

But,

For t

ejw# e

jw#

a#

a#

(t) =
(t) =

ejwt e at
a + jw
jwt

e
jw

1/a, e at 0 and

jwt
1
e
a#
sin w# e
(t)
j2 a + jw

at

jwt

e
a

jw

which equals the imaginary part of


ejwt
a + jw
i.e.,

a sin wt w cos wt
a2 + w 2

6. Using integration by parts,

#e

2#

(t) =

or, in other words,

Z t
0

#e

te

2#

2( t t )

it follows that

8. For t

y(t) =

t
2

Z t

(t) =

1
4

t
2

t
2

1 1
+ e
4 4

1
4

3
e
4

(t) =

It follows that

7. Since

dt =

2t

2t

(t t )

dt = te t ,

y ( t ) = (1 + t ) e t .

0,
o (t) = 2e

t/2

+ i (#) e

2( t t )

#/2

( t ).

t =t
t =0

(t)

67

lecture notes, me 340

For the free response, i (t) = 0 for all t, in which case


o (t) = 2e

t/2

For the unit step response, i (t) = u(t) and we ignore the contribution from the initial condition, in which case

#/2
o ( t ) = u (#) e
(t) = 2(1 e t/2 ).
For the unit impulse response, i (t) = d(t) and we ignore the
contribution from the initial condition, in which case

#/2
o ( t ) = d (#) e
(t) = e t/2 .
9. Here,

Z t

1+

o (t) =

or, in other words,

2 i(t )

1
o (t )
2

do
(t) + o (t) = 2i (t), o (0) =
dt
It follows that, for t

dt

0,

o (t) =

+ 2i (#) e

(t)

For the free response, i (t) = 0 for all t, in which case


o (t) =

e t.

For the unit step response, i (t) = u(t) and we ignore the contribution from the initial condition, in which case

#
o (t) = 2u(#) e
( t ) = 2(1 e t ).
For the unit impulse response, i (t) = d(t) and we ignore the
contribution from the initial condition, in which case

#
o (t) = 2d(#) e
(t) = 2e t .
10. For t

0 and b 6= 2,

o (t) = be

b#

2#

(t) = b

bt

e
b

2t

In the limit as b ! , o (t) converges to e 2t , i.e., the unit impulse response. The diagrams below show the input and output,
respectively, for three different values of b.

68

lecture notes, me 340

b!

be

bt

2t

o (t) = b e

bt

e
2 b

2t

69

lecture notes, me 340

Lesson 4 Laplace transforms and transfer functions


TAKE-AWAYS:
at
* If | f (t)| Me for all t
Laplace transform

0 and some M > 0, then the

L f (#) ( s ) : =

st

f (t)e

dt

is defined for all complex numbers s with real part greater


than a. The Laplace transform is a linear operation.
* For any complex number a,
h i
L ea# (s) =

1
s

For any real number D > 0,

L f (#

Moreover,

L
and

D) (s) = e

D ) u (#
Z

#
0

f (t ) dt (s) =

sD

L f (#) u (#) ( s ).

1
L f (#) ( s )
s

f (##) g(##) (#) (s) = L f (#) (s) L g(#) (s)

known as the convolution theorem.

* In a convolution dynamic system with

L o (#) ( s ) = L i (#) ( s ) H ( s ),

the transfer function H (s) is the Laplace transform of the unit


impulse response h(t) and

o ( t ) = i (#) h (#) ( t ).

70

lecture notes, me 340

Online lecture: time and frequency domains


A piece of magic: If we replace every signal f (t) by its Laplace
transform
Z

L f (#) ( s ) : =
f (t)e st dt,
(230)
0

then the action of amplifiers, delay lines, zero-initial-value integrators, and zero-initial-value integrators with feedback in the time
domain correspond to multiplication with an appropriate function of
the complex number s in the frequency domain.
Example 32. The zero-initial-value integrator with negative feedback
may be represented in a block diagram in the time domain or the frequency
domain, as shown below.
Time domain:
i(t)

Rt

o (t)

Figure 33: A representation of a zeroinitial-value integrator with negative


feedback in the time and frequency
domains.

a
Frequency domain:

L i (#) ( s ) +

1
s

L o (#) ( s )

In the time domain,


o (t) =
In the frequency domain

Z t
0

i(t )

ao (t ) dt.

1
L o (#) ( s ) =
L i (#) ( s )
s

or, in other words,

L o (#) ( s ) =

(231)

a L o (#) ( s )

1
L i (#) ( s )
s+a

(232)

(233)

We obtain the output signal o (t), for a given input i (t), by returning to the
time domain on both sides of this equation. It should come as no surprise
that, in this case, the answer is in the form of a convolution.

71

lecture notes, me 340

For piecewise continuous functions f (t) such that | f (t)| Me at for

all t
0 and some M > 0, the Laplace transform L f (#) (s) is finite
for all complex numbers s with real part greater than a. In particular,

L f (#) ( s ) = L f (#) u (#) ( s ).

(234)

Example 33. The unit step f (t) = u(t) satisfies the inequality | f (t)|
Me at for all t
0 for M = 1 and a = 0. Its Laplace transform is finite
provided that the real part of s is positive and equals
Z

L u (#) ( s ) =

st

dt =

1
e
s

st

1
= .
s

(235)

The unit step is the unique signal whose Laplace transform equals 1/s.
Example 34. For every real or complex number a, the exponential function
f (t) = eat satisfies the inequality | f (t)| Me at for all t
0 for M = 1
and a equal to the real part of a. Its Laplace transform is finite provided that
the real part of s is greater than the real part of a, in which case
Z
h i
L ea# (s) =

e(a

s)t

dt =

1
a

e(a

s)t

=
0

1
s

(236)

The exponential function eat is the unique signal whose Laplace transform
equals 1/(s a).
If the input signal to the zero-initial-value integrator with negative
feedback in Example 32 is the unit step, i.e., i (t) = u(t), then

1
1
1 1
1
L o (#) ( s ) =
L i (#) ( s ) =
=
(237)
s+a
s(s + a)
a s
s+a
or, by linearity,

1
1 e at
(238)
a
0. This analysis is an example of the general observation:
o (t) =

for t

For dynamic systems consisting of amplifiers and integrators,


as well as summing and splitting junctions, calculus in the
time domain becomes algebra in the frequency domain provided that a few key facts are recalled about the Laplace transform, e.g.,
h i
1
L ea# (s) =
s a
for all complex numbers a. In the special case that a = 0, we
obtain

1
L u (#) ( s ) = .
s

72

lecture notes, me 340

Example 35. For a unit pulse,

Z
u (#) u (# D )
1 D
L
(s) =
e
D
D 0
By lHospitals rule,
1

lim

D !0

e
Ds

Ds

st

dt =

e
Ds

Ds

(239)

=1

(240)

We represent this observation by the definition L d(#) (s) := 1.

If f (t) is piecewise continuous and | f (t)| Me at for all t


some M > 0, then the integral
F (t) :=

Z t
0

73

f (t ) dt

0 and

(241)

is continuous and differentiable with F (0) = 0, F 0 (t) = f (t), and


| F (t)| M0 e at for all t 0, the same a, and some M0 > 0.
The Laplace transform L [ F (#)] (s) is finite, provided that the real
part of s is greater than a, and is given by
Z
0

F (t)e

st

1
F (t)e
s

dt =

st

+
0

1
s

Z
0

F0 (t)e

st

dt

(242)

or, in other words,

#
0

f (t ) dt (s) =

1
L f (#) ( s ).
s

(243)

As illustrated in the figure below, the action of a zero-initial-value


integrator is equivalent to multiplication by 1/s.
Frequency domain:

Time domain:
i(t)

o (t)

Rt
0

L i (#) ( s )

1
s

L o (#) ( s )

Example 36. Note that


Z t
0

f (t ) dt =

Z t
0

u(t

t ) f (t ) dt =

u (#) f (#) ( t ).

Since L u(#) (s) = 1/s, it follows from (243) that

L u(##) f (##) (#) (s) = L u(#) (s) L f (#) (s)

(244)

(245)

i.e., convolution with the unit step in the time domain is equivalent to multiplication by the Laplace transform of the unit step in the frequency domain.

Figure 34: An integrator represented in


the time domain and in the frequency
domain

lecture notes, me 340

We generalize the result of Example 36


by considering the Laplace

transform of the convolution

f (#) g(#) (t), i.e.,

f (##) g(##) (#) (s) =

Z
0

st

Z Z t
0

t
0

st

t ) g(t ) dt

f (t

dt

t ) g(t ) dt dt

f (t

(246)

t=t

Figure 35: Integration over vertical or


horizontal strips.

t=t
dt
t

dt

By changing order of integration, as suggested in the figure above,


we get
(
)
Z Z
0
t
=
t
+
t
e st f (t t ) g(t ) dt dt =
dt = dt0
0
t

=
=

Z Z
0

Z
0

st0

s(t +t0 )

f (t0 ) dt0

f (t0 ) g(t ) dt dt0


Z
0

st

g(t ) dt.

(247)

By the convolution theorem,

L f (##) g(##) (#) (s) = L f (#) (s) L g(#) (s).

the response of a dynamic system consisting of amplifiers,


zero-initial-value integrators, summing junctions, and splitting junctions, can be written in terms of a convolution with
the input.
Example 37. Recall that for a zero-initial-value integrator with feedback the
output o (t) is given by the convolution i (#) e a# (t). It follows that

L o (#) ( s ) = L i (#) ( s )

1
.
( a)

(248)

As illustrated in the figure below, the action of a zero-initial value integrator


with feedback is thus equivalent to multiplication by 1/(s + a).

74

lecture notes, me 340

Time domain:
i(t)

Rt

Figure 36: A representation of a zeroinitial-value integrator with negative


feedback in the time and frequency
domains.

o (t)

75

a
Frequency domain:

L i (#) ( s )

L o (#) ( s )

1
s+a

In-class worksheet: transfer functions


The action of a SISO dynamical system consisting of amplifiers, zeroinitial-value integrators, splitting and summing junctions is equivalent to multiplication by some function H (s) known as the transfer
function. Such a dynamical system is called a convolution, since

L o (#) ( s ) = L i (#) ( s ) H ( s )
(249)
implies that

o (t) =

i (#) h (#) ( t )

(250)

for some function h(t), where H (s) = L h(#) (s).


i(t)

o (t)

h(t)

or

L i (#) ( s )

H (s)

L o (#) ( s )

Suppose that h(t) is differentiable. Then, Taylors theorem implies


that there exists a function g, such that
h(t

t ) = h(t) +

dh
(t)t + g(t
dt

t )t

(251)

and limt !0 g(t t ) = 0. For t > D, it follows that

u (#) u (# D )
h (#) ( t )
D

Z
1 D
dh
=
h(t) +
(t)t + g(t t )t dt
D 0
dt

= h(t) +

dh D
1
(t) +
dt
2
D

Z D
0

g(t

t )t dt

(252)

Figure 37: A convolution system represented in the time domain by convolution with the unit impulse response
h(t) or in the frequency domain by
multiplication with the transfer function
H ( s ).

lecture notes, me 340

or, in other words,


lim

D !0

u (#)

u (#
D

D)

h (#) ( t ) = h ( t )

(253)

It follows that h(t) is the unit impulse response of the system, and
H (s) is its Laplace transform.
Example 38. The transfer function of a zero-initial-value integrator is

1/s. But L u(#) (s) = 1/s. It follows that the unit impulse response of a
zero-initial-value integrator is the unit step u(t).
Similarly, the transfer function of a zero-initial-value integrator with

feedback is 1/(s + a). But L e a# (s) = 1/(s + a). It follows that the
unit impulse response of a zero-initial-value integrator with feedback is the
exponential function e at .
Example 39. Now consider the following equivalent block diagrams with

I (s) := L i (#) (s), X (s) = L x (#) (s), and O(s) := L o (#) (s).
b

i(t) +

x (t)

Rt
0

Rt

o (t)

I (s) +

X (s)

1
s

1
s

O(s)

It follows that
O(s) =

1
X ( s ),
s

X (s) =

1
I (s)
s

aX (s)

bO(s)

(254)

Figure 38: Alternative block diagram


representations of a double integrator with feedback show that it is a
convolution.

76

lecture notes, me 340

77

from which we obtain


s2 O(s) + asO(s) + bO(s) = I (s)
or, in other words,
O(s) =

s2

(255)

1
I ( s ).
+ as + b

(256)

As suggested in the figure below, the system is thus equivalent to a convolution with transfer function
H (s) =

1
s2 + as + b

I (s)

(257)

Figure 39: The equivalent transfer


function representation of a double
zero-initial-value integrator with
feedback.

O(s)

s2 +as+ b

In Lesson 5, we use convolution integrals to compute the corresponding


impulse response.
Example 40. Now consider the following block diagram.
Figure 40: A zero-initial-value integrator with feedback and feedforward is a
convolution.

I (s) +

1
s

O(s)

Let I (s) := L i (#) (s) and O(s) := L o (#) (s). It follows that
O(s) = bI (s) +

1
I (s)
s

aO(s)

(258)

from which we obtain


O(s) =

b + 1s
bs + 1
I (s) =
I (s)
1 + as
s+a

(259)

As suggested in the figure below, the system is thus equivalent to a convolution with transfer function
H (s) =

bs + 1
1 ab
= b+
.
s+a
s+a

(260)

lecture notes, me 340

I (s)

Since

bs+1
s+a

H (s) = L bd(#) + (1

78

Figure 41: The equivalent transfer


function representation of an integrator
with feedback and feedforward.

O(s)

a#

ab)e

( s ),

(261)

it follows that the unit impulse response is given by an impulse scaled by b


added to an exponentially decaying signal with initial value 1 ab and time
constant a 1 .
Consider, again, the multilayered mechanical suspension shown
in the figure below, and its representation by the double integrator
block diagram shown in Figure 43 on the next page.
f (t)i

Figure 42: A multilayered suspension.

x2

k2
x1

d2

d1

k1

Let F (s) := L f (#) (s), X1 (s) := L x1 (#) (s), and X2 (s) :=

L x2 (#) (s). Figure 44 below shows an alternative frequency-domain


block-diagram representation of the suspension in the case of zero
initial values. In this case, it follows that

X1 ( s ) =

k 2 X2 ( s )

and

X1 ( s )
d1 s

k 1 X1 ( s )

k 2 X2 ( s ) X1 ( s )
d2 s
After solving for X1 (s) and X2 (s), we obtain
X2 ( s ) =

X1 ( s ) =

d1 d2

s2

F (s)

k2 F (s)
= H1 (s) F (s)
+ ( d2 k 1 + d1 k 2 + d2 k 2 ) s + k 1 k 2

(262)

(263)

(264)

lecture notes, me 340

79

and
X2 ( s ) =

( d2 s + k 1 + k 2 ) F ( s )
= H2 (s) F (s)
d1 d2 s2 + ( d2 k 1 + d1 k 2 + d2 k 2 ) s + k 1 k 2

(265)

in terms of the transfer functions H1 (s) and H2 (s).

f (t) +

1
d2

x2 (0) +

Rt
0

x2 ( t )

Figure 43: An equivalent block diagram


in the time domain.

+
k2

1
d1

x1 (0) +

Rt
0

x1 ( t )

k1

F (s) +

1
d2

1
s

X2 ( s )

+
k2

1
d1

1
s

k1

X1 ( s )

Figure 44: An equivalent block diagram


in the frequency domain.

lecture notes, me 340

80

A frequency-domain representation of this single-input-two-output


dynamical system is shown below.

F (s)

H1 (s)
H2 (s)

Figure 45: The equivalent transfer function representation of the multilayered


suspension.

X1 ( s )
X2 ( s )

In-class worksheet Liquid-level systems


Let y (t) denote the liquid level at time t of a tank of constant crosssectional area C. Suppose that the fluid flow rate into the tank is qi (t)
and that the fluid flow rate out of the tank is qo (t). By the principle
of conservation of mass
d
(266)
(rCy) = rqi rqo ,
dt
since an increase in the amount of mass inside the tank must be
accompanied by a larger flow rate of mass per unit time into the tank
than out of the tank.
Figure 46: A single-tank liquid-level
system.

qi

pa

qo

pa

g
C

Ro

Assume that the mass flow rate out of the tank through a valve is
proportional to the pressure drop across the valve:
rqo =

Dp
,
Ro

(267)

where Ro is the resistance of the output valve. Then, if the pressure


drop equals the difference between the ambient atmospheric pressure
p a and the pressure at the bottom of the tank p a + rgy, it follows that
Dp = rgy and thus
d
rg
y.
(268)
(rCy) = rqi
dt
Ro

lecture notes, me 340

81

This can be written as


dy
1
1
( t ) + y ( t ) = q i ( t ),
dt
T
C
where the time constant
T=

(269)

CRo
.
g

(270)

If y(0) = y0 , then
y ( t ) = y0 e

t/T

1
q (#) e
C i

#/T

( t ).

(271)

If we consider qi (t) as the input and y(t) as the output, then for
y(0) = 0, the system is a convolution with transfer function
H (s) =

1 T
.
C sT + 1

(272)

An equivalent block diagram is shown below.

L q i (#) ( s )

L y (#) ( s )

1 T
C sT +1

Figure 47: The equivalent transfer function representation of the relationship


between the input fluid flow rate and
the liquid level.

Example 41. The unit impulse response of the single-tank liquid level
system considered above is the signal whose Laplace transform equals the
transfer function, namely
y(t) =

1
e
C

t/T

(273)

which also equals the free response (in the absence of input) given an initial
condition y(0) = 1/C.
Example 42. The unit step response of the single-tank liquid level system
considered above is the signal whose Laplace transform equals

11 T
T 1
T
=
(274)
s C sT + 1
C s
sT + 1
or, in other words,

T
1 e t/T
(275)
C
for t 0. In particular, y (t) ! T/C = Ro /g as t ! corresponding to
a steady-state level of the liquid in the tank, as the amount of liquid flowing
into the tank equals that flowing out of the tank.
y(t) =

We replace the open inlet with a pump that raises the ambient
pressure by dp(t), and an inlet valve with resistance Ri connected to
the bottom of the tank, as shown in the figure below.

lecture notes, me 340

pa
g

pa

qi

dp

Figure 48: A single-tank liquid-level


system with a pump.

Ri

In this case,

Ro

d
(rCy) = rqi
dt

where
rqi =

dp

and
rqo =
so that

82

rqo ,

qo

(276)

rgy
Ri

(277)

rgy
,
Ro

(278)

dy
1
1
(t) + y(t) =
dp(t),
dt
T
rCRi

(279)

where the time constant


T=

CRi Ro
.
g ( Ri + R o )

(280)

If y(0) = h0 , then
y ( t ) = y0 e

t/T

1
+
dp(#) e
rCRi

#/T

( t ).

(281)

If we consider dp(t) as the input and y(t) as the output, then for
y(0) = 0, the system is a convolution with transfer function
H (s) =

1
T
.
rCRi sT + 1

(282)

An equivalent block diagram is shown below.

L dp(#) (s)

1
T
rCRi sT +1

L y (#) ( s )

Example 43. We consider the two-tank liquid-level system shown on the


next page. If y1 (0) = 0, then the analysis of the single-tank liquid-level system in Fig. 48 shows that the relationship between the pressure differential

Figure 49: The equivalent transfer function representation of the relationship


between the pressure differential across
the pump and the liquid level.

lecture notes, me 340

Figure 50: A coupled liquid-level


system of two tanks.

pa
g

pa

qi

y1

dp

C1

R1

R2

C2
pa

pa

qo

y2

R3

across the pump and the liquid height in the first tank is a convolution with
transfer function
1
T1
H1 (s) =
,
(283)
rC1 R1 sT1 + 1
where
T1 =

C1 R1 R2
g ( R1 + R2 )

(284)

Similarly, if y2 (0) = 0, then the analysis of the single-tank liquid-level


system in Fig. 46 shows that the relationship between the fluid flow rate into
the second tank and the liquid height in the second tank is a convolution
with transfer function
H2 (s) =
where
T2 =

1
T2
,
C2 sT2 + 1
C2 R3
.
g

(285)

(286)

Finally, Eq. (278) implies that the relationship between the liquid height in
the first tank and the fluid flow rate into the second tank is a convolution
with transfer function
g
H1 2 (s) =
.
(287)
R2
The coupled liquid-level system may now be represented by the block
diagram below.

83

lecture notes, me 340

L dp(#) (s)

L y1 (#) ( s ) L q i (#) ( s )
T1
1
rC1 R1 sT1 +1

g
R2

T2
1
C2 sT2 +1

L y2 (#) ( s )

The input-output relationship between the pressure differential across the


pump and the liquid height in the second tank is therefore a convolution
with transfer function
1
T1
g 1
T2
rC1 R1 sT1 + 1 R2 C2 sT2 + 1

(288)

g
T1 T2
rC1 R1 C2 R2 (sT1 + 1)(sT2 + 1)

(289)

H (s) =
or, in other words,
H (s) =

This is the transfer function of a sequential wiring of two zero-initial-value


integrators with negative feedback followed by (or preceded by) amplification.

Exercises
1. Consider the following block diagram in the time domain.
i(t) +

Rt

1
2

o (t)

Draw a corresponding diagram in the frequency domain and

express L o (#) (s) in terms of L i (#) (s).

2. Consider the following block diagram in the time domain.


2
i(t) +

Rt
0

y(t) +

Rt
0

o (t)

84

Figure 51: The equivalent transfer function representation of the relationship


between the pressure differential across
the pump, the liquid level in each of the
tanks, and the flow rate into the second
tank.

lecture notes, me 340

Draw a corresponding diagram in the frequency domain and

express L o (#) (s) in terms of L i (#) (s).

3. Consider the following block diagram in the time domain.


i(t) +

Rt

1
2

o (t)

+
2

Rt
0

y(t)

Draw a corresponding diagram in the frequency domain and

express L o (#) (s) in terms of L i (#) (s).

4. Suppose that a 6= b and compute the Laplace transform

ea#
a

e b#
(s)
b

5. Compute the Laplace transforms

L sin w# (s), L cos w# (s)

Verify that your result agrees with (243).


6. Use (243) to show that


n!
L #n ( s ) = n +1 ,
s

provided that the real part of s is greater than 0.


7. Show that

dn
L #n f (# ) ( s ) = ( 1 ) n n L f (# ) ( s )
ds
and use the result to show that

n!
L #n ( s ) = n +1 ,
s

provided that the real part of s is greater than 0.


8. Compute the Laplace transform
h i
L #e# (s)

85

lecture notes, me 340

9. Show that

L f (#

D) (s) = e

D ) u (#

sD

L f (#) u (#) ( s ).

and illustrate the result using block diagrams in the time and
frequency domains.
10. Suppose that

L o (#) ( s ) =

1
L i (#) ( s )
s ( s + 2)

Find an expression for o (t) in terms of a convolution with i (t).


11. Suppose that

s+1
L o (#) ( s ) = 2
L i (#) ( s )
s +1
Draw a corresponding frequency-domain block diagram.

12. Suppose that

1 + sb
L o (#) ( s ) =
L i (#) ( s )
s+a

Find an expression for o (t) in terms of a convolution with i (t).


13. Find the unit impulse response corresponding to the transfer function
s 1
2
1
H (s) =
=
(s 2)(s 3)
s 3 s 2
14. Suppose that the unit impulse response of a convolution dynamical system equals tet . Find the Laplace transform of the unit step
response.

Solutions
1. A corresponding frequency-domain block diagram is given below.

L i (#) ( s ) +

Here,

1
2

1
s

1
L o (#) ( s ) =
L i (#) ( s )
2s

or, in other words,

L o (#) ( s ) =

L o (#) ( s )

1
L i (#) ( s ).
1 + 2s

L o (#) ( s )

86

lecture notes, me 340

2. A corresponding frequency-domain block diagram is given below.

L i (#) ( s ) +

1
s

L y (#) ( s )

1
s

L o (#) ( s )

3
Here,

1
L o (#) ( s ) =
L y (#) ( s )
s

1
L y (#) ( s ) =
L i (#) ( s )
s

or, in other words,

L o (#) ( s ) =

2L o (#) ( s )

3L y (#) ( s )

1
L i (#) ( s )
(s + 2)(s + 3)

3. A corresponding frequency-domain block diagram is shown below.

L i (#) ( s ) +
L o (#) ( s )
1
1

s
2

1
s

L y (#) ( s )

Here,

1
L o (#) ( s ) =
L i (#) ( s ) 2 L o (#) ( s )
2s

2
L y (#) ( s ) =
L o (#) ( s ) L y (#) ( s )
s

or, in other words,

L o (#) ( s ) =

s+2
L i (#) ( s )
2s(s + 3)

L y (#) ( s )

87

lecture notes, me 340



4. Since L ea# (s) = 1/(s

ea#
a

a ),

e b#
(s) =
b

5. Since sin wt = (ejwt

L sin w# (s) =

1
s jw

L cos w# (s) =

1
s+jw

1
a)(s

(s

j2

Similarly, since cos wt = (ejwt + e

Since,

1
s b

b)

jwt ) /j2,

1
s a

s2

w
+ w2

jwt ) /2,

1
s jw

1
s+jw

s
s2 + w 2

w
L sin w# (s) = L cos w# (s)
s
this agrees with (243).

6. Recall that L 1 (s) = 1/s, provided that the real part of s is greater
than 0. With f (t) = ntn 1 , it follows from (243) that
i

n h
L #n ( s ) = L #n 1 ( s )
s
and, consequently, that

7. Here,


nn 1 h n
L #n ( s ) =
L #
s s

( 1) n

(s) = =

dn
n d
L
f
(
#
)
(
s
)
=
(
1
)
dsn
dsn

= ( 1) n

Z t
0

n!
s n +1

f (t)e

st

dt

( t)n f (t)e

st

dt

or in other words,

dn
L #n f (# ) ( s ) = ( 1 ) n n L f (# ) ( s ).
ds

In particular, if f (t) = u(t), then


dn 1
n!
L #n ( s ) = ( 1 ) n n = n +1 .
ds s
s

8. From the previous result,

h i
L #e# (s) =

d 1
1
=
.
ds s 1
( s 1)2

88

lecture notes, me 340

9. The claim follows from the observation that

L f (#

D) (s) =

D ) u (#

sD

=e

D)u(t

f (t

Z
0

st

D)e
st

f (t )u(t )e

dt

dt,

where t = t D. As illustrated below, the action of a delay line


with delay D is thus equivalent to multiplication by the exponential e sD .
Frequency domain:

Time domain:
i(t)

10. Since

1
1
=
s ( s + 2)
2

it follows that

L i (#) ( s )

o (t)

1
s

1
s+2

1
= L 1
2

L o (#) ( s )

Ds

2#

( s ),

1
2#
o (t) =
1 e
i (#) ( t ).
2

11. Let O(s) = L o (#) (s) and I (s) = L i (#) (s) so that

(s2 + 1)O(s) = (s + 1) I (s) , O(s) =

1
I (s)
s2

O(s) +

1
I ( s ).
s

Then,
Y (s) =

1
I (s)
s

O(s) ) O(s) =

1
Y ( s ) + O ( s ) + Y ( s ).
s

A corresponding block diagram is shown below.

I (s) +

1
s

1
s

Y (s)

12. Since
h
1 + sb
1 ab
= b+
= L bd(#) + (1
s+a
s+a

it follows that

o (t) = bi (t) + (1

ab) e

a#

ab)e

a#

i (#) ( t ).

( s ),

O(s)

89

lecture notes, me 340

13. Since

2
s

1
3

= L 2e3#

the unit impulse response equals 2e3t

e2# ](s),

e2t .

14. Since the corresponding transfer function equals



L #e# (s) =

1
1)2

(s

the Laplace transform of the unit step response is given by


1
s(s

1)2

90

lecture notes, me 340

Lesson 5 Inverse Laplace transforms


TAKE-AWAYS:
* The transfer function
H (s) =

1
=
(s
s2 + as + b

1
g1 )(s

g2 )

has poles at s = g1 and s = g2 . The corresponding dynamic


system is overdamped if a2 > 4b, in which case g1 and g2 are
real and unequal; critically damped if a2 = 4b, in which case
g1 and g2 are real and equal; and underdamped if a2 < 4b, in
which case g1 and g2 are conjugate complex numbers.
* If h(t) is the unit impulse response corresponding to the
transfer function H (s), then sH (s) is the Laplace transform
of the distributional derivative Dh(t)/dt of h(t)u(t). If h(t) is
continuous and differentiable for all t > 0, then for t 0
Dh
dh
(t) =
(t) + d(t) lim h(t).
dt
dt
t #0
For example,

D at
e
= ae at + d(t).
dt
If limt#0 h(t) = 0, then sH (s) is the Laplace transform of the
derivative of h(t).
* If the function H (s) is a ratio of two polynomials in s, then
the corresponding inverse Laplace transform can be computed
using convolution integrals and distributional derivatives.

91

lecture notes, me 340

92

Online lecture: second-order responses


The transfer function for a double zero-initial-value integrator
with negative feedback is given by
H (s) =

1
1
1
=
s g1 s g2
s2 + as + b

(290)

(291)

where
g1,2 :=

a2
4

and a, b > 0. We say that the system has poles8 at s = g1 and s = g2 ,


respectively.
Recall that
h
i
1
L e a# (s) =
(292)
s+a
It follows that
h
i
h
i
H (s) = L eg1 # (s) L eg2 # (s)
(293)

The poles are points in the complex


plane, where the denominator equals 0.

By the convolution theorem, the unit impulse response h(t) is then


given by

Z t
Z t
g1 #
g2 #
e e
(t) =
eg1 t eg2 (t t ) dt = eg2 t
e(g1 g2 )t dt
(294)
0

or, in other words,


h(t) =

eg1 t
g1

eg2 t
g2

(295)

provided that g1 6= g2 , and


h(t) = tegt

(296)

if g1 = g2 = g.
Example 44. A double zero-initial-value integrator with negative feedback
is overdamped if a2 > 4b. In this case, g1 and g2 are real, negative, and
distinct.
h(t) = e

Figure 52: The unit impulse response


of an overdamped double zero-initialvalue integrator with negative feedback.

( a2 /4 b t)
at/2 sinhp
a2 /4 b

The unit impulse response shows an initial increase from zero and a
subsequent decrease back to zero without any overshoot.

lecture notes, me 340

93

Example 45. A double zero-initial-value integrator with negative feedback


is underdamped if a2 < 4b. In this case, g1 and g2 are complex conjugate
numbers with negative real part given by a/2 and imaginary part given
p
by b a2 /4. Substitution then yields
p
a2 /4 t)
at/2 sin( b
p
h(t) = e
(297)
b a2 /4
where we recall that sin x = (ejx

h(t) = e

jx ) /j2.

Figure 53: The unit impulse response


of an underdamped double zero-initialvalue integrator with negative feedback.

2
at/2 sin(p b a /4 t)
2
b a /4

The unit impulse response is an exponentially decaying harmonic


signal with initial amplitude
p

a2 /4
p
time constant 2/a, angular frequency b a2 /4, and no phase shift.
b

(298)

Example 46. A double zero-initial-value integrator with negative feedback


is critically damped if a2 = 4b. In this case, g1 = g2 = a/2. The unit
impulse response is given by
h(t) = te

h(t) = te

at/2

(299)

Figure 54: The unit impulse response


of a critically damped double zeroinitial-value integrator with negative
feedback.

at/2

The unit step response of a critically damped double zero-initial-value

lecture notes, me 340

94

Figure 55: The unit step response of a


critically damped double zero-initialvalue integrator with negative feedback.

4/a2

h (#) u (#) ( t )
t

integrator with negative feedback is then given by

Z t
h (#) u (#) ( t ) =
te at/2 u(t
0

=
or, in other words,

h (#) u (#) ( t ) =

2e

at/2 ( at

+ 2)

a2
4

2e

t ) dt
t

(300)
0

at/2 ( at + 2)

(301)

a2

whose Laplace transform equals H (s)/s.

In-class worksheet: distributional derivatives


Suppose that h(t) is the impulse response of a convolution corresponding to the transfer function H (s). Then, sH (s) is the Laplace
transform of the distributional derivative9 Dh(t)/dt of h(t)u(t).
Example 47. The function h(t) = e at u(t) is the unit impulse response
of a convolution corresponding to the transfer function H (s) = 1/(s + a).
But,
h
i
s
a
sH (s) =
=1
= L d(#) ae a# (s)
(302)
s+a
s+a
We conclude that, for t 0,
D
e
dt

at

= d(t)

ae

at

(303)

This equals the regular derivative of h(t) plus a unit impulse at t = 0


corresponding to the discontinuous jump from 0 to e a0 = 1. The special
case when a = 0 in (303) implies that, for t 0,
D
u ( t ) = d ( t ).
dt

(304)

Since sin wt = (ejwt e jwt )/j2, substitution of a = jw on the


right-hand side of (303) implies that
d(t) + jw ejwt

d(t) + jw e
j2

jwt

=w

ejwt + e
2

jwt

(305)

For piecewise-continuous functions,


the distributional derivative equals
the regular derivative at points of
continuity plus unit impulses shifted to
points of discontinuity and scaled by
the size of the discontinuities.

lecture notes, me 340

0,

or, in other words, for t

D
sin wt = w cos wt
dt

(306)

i.e., the regular derivative of sin wt for all t. Since L sin w# (s) =
w/(s2 + w 2 ), it follows that the Laplace transform of cos wt equals
s
w
s
= 2
w s2 + w 2
s + w2

(307)

In contrast,
d(t) + jw ejwt + d(t)
2

jw e

jwt

= d(t)

ejwt

e
j2

jwt

(308)

0,

or, in other words, for t

D
cos wt = d(t) w sin wt,
(309)
dt
i.e., the regular derivative of cos wt plus a unit impulse at t = 0 corresponding to the discontinuous jump from 0 to cos(0) = 1.
Example 48. The Laplace transform of the convolution

a#
a#
d(#) ae
e
(t) = e at ate
equals

It follows that, for t

at

(310)

s
1
s
=
s+as+a
( s + a )2

(311)

D
te
dt

(312)

0,

i.e., the regular derivative of te

at

=e

at

ate

at

at .

Example 49. Consider the transfer function


H (s) =

s2 + s 2
s2 + s 2
=
s3 + 5s2 + 7s + 3
( s + 1)2 ( s + 3)

(313)

corresponding to the block diagram on the next page. We obtain the corresponding Inverse Laplace Transform by several successive convolution
integrals, as follows

1
1
#
#
L
(
t
)
=
e

e
(t) = te t .
(314)
(# + 1)2

1
(
t
)
=
#e
(# + 1)2 (# + 3)
Z t

3#

te t e 3(t
1 3t
=
e
+ 2te
4

(t)
t)

dt
e

(315)

95

lecture notes, me 340

Figure 56: A template block diagram


for rational transfer functions.

+
1
I (s) +

1
s

1
s

1
s

+ O(s)

+
+

+
3

whose first and second distributional derivatives equal


d(t)

3e

3t

2te

+ 2e
4

d(t) + e

3e

3t

2te
4

+ 3e

(316)

and
3d(t) + 9e

3t

+ 2te

2e

+ 3d(t)

3e

9e

3t

+ 2te
4

5e

(317)

The corresponding unit impulse response then equals


1
9e
4

3t

+ 2te
1
2
e
4

t
3t

5e

+ 2te

1
3e 3t 2te t + 3e t
4

e t = e 3t te t

(318)

as is also easily verified from the definition of the Laplace transform.

Example 50. Consider a block diagram representing a convolution with an


integrator in the feedback loop.
Here,

1
1
X ( s ) = O ( s ),
O(s) =
I (s) 3O(s) 2X (s)
(319)
s
s
from which we obtain
s2 O(s) + 3sO(s) + 2O(s) = sI (s).

(320)

The system is thus equivalent to a convolution with transfer function


H (s) =

s
s2 + 3s + 2

(321)

96

lecture notes, me 340

X (s)

I (s) +

97

Figure 57: A convolution with an


integrator in the feedback loop.

1
s

1
s

O(s)

We obtain the corresponding unit impulse response by computing the distributional derivative of the unit impulse response corresponding to the
transfer function 1/ s2 + 3s + 2 .

In-class worksheet: initial-value problems


For t

0, the output of an integrator with initial value c is given by

Z t
o (t) = c +
i (t ) dt = cu(t) + i (#) u(#) (t)
(322)
0

With O(s) = L o (#) (s) and I (s) = L i (#) (s),


O(s) =

c
I (s)
c + I (s)
+
=
.
s
s
s

(323)

When I (s) = 0, the output equals the free response. When c = 0


and I (s) = 1, the output equals the unit impulse response. The
free response is always a multiple of the unit impulse response and
equals the unit impulse response when c = 1.
Figure 58: An integrator with initial
value represented in the time domain
and frequency domains.

Frequency domain:

Time domain:

c
i(t)

c+

Rt

o (t)

I (s) +

1
s

O(s)

Example 51. The figure on the next page shows two representations of a
nonzero-initial-value integrator with negative feedback. In this case,
O(s) =

1
c + I (s)
s

aO(s)

(324)

lecture notes, me 340

Time domain:
i(t)

c+

Rt

Figure 59: A nonzero-initial-value


integrator with negative feedback
represented in the time and frequency
domains.

o (t)

a
Frequency domain:
c
I (s)

1
s

O(s)

or, in other words,


O(s) =

c + I (s)
.
s+a

(325)

When I (s) = 0, the output equals the free response. When c = 0 and
I (s) = 1, the output equals the unit impulse response. The free response is
always a multiple of the unit impulse response and equals the unit impulse
response when c = 1.
Example 52. The figure on the next page shows two representations of a
double nonzero-initial-value integrator with negative feedback. In this case,
O(s) =

1
1
d + X (s) , X (s) =
c + I (s)
s
s

or, in other words,


O(s) =

aX (s)

ds + c + da + I (s)
s2 + as + b

bO(s)

(326)

(327)

When I (s) = 0, the output equals the free response. When c = 0 and
I (s) = 1, the output equals the unit impulse response. The free response is a
linear combination of the unit impulse response of the transfer function
1
s2 + as + b

98

(328)

and its distributional derivative, and equals the unit impulse response when
c = 1 and d = 0.

lecture notes, me 340

Figure 60: A double nonzero-initialvalue integrator with negative feedback


represented in the time and frequency
domains.

Time domain:
b

i(t) +

Rt

c+

x (t)

d+

Rt

o (t)

Frequency domain:
b
c
I (s) +

99

1
s

X (s)

+
+

1
s

Exercises
1. Consider the mechanical suspension shown below.
f (t)i
xm
2 kg
i
3 N/m

0.1 Ns/m

O(s)

lecture notes, me 340

Use a free-body diagram and Newtons 2nd law to derive a differential equation governing x (t) and determine whether the system
is overdamped, underdamped, or critically damped. Graph the
corresponding unit impulse response.
2. Explain why the unit impulse response of an overdamped system
is a nonnegative function of time.
3. Use a frequency-domain block diagram representation of the
initial-value problem
do
(t) = e
dt

cos 2pt, o (0) =

to find the Laplace transform of its solution.


4. Use a frequency-domain block diagram representation of the
initial-value problem
do
(t) + 2o (t) = te
dt

t/2

, o (0) = 1

to find the Laplace transform of its solution.


5. Use a frequency-domain block diagram representation of the
initial-value problem
d2 o
do
(t) + 3 + 2o (t) = u(t)
2
dt
dt

u(t

1), o (0) = 1,

do
(0) = 3
dt

to find the Laplace transform of its solution.


6. Solve the initial-value problem
d2 o
do
do
(t) + 2 + o (t) = 0, o (0) = 3,
(0) =
dt
dt
dt2

1.

7. Determine the initial values for which the free response of a double integrator with negative feedback equals the unit impulse
response.
8. Find the inverse Laplace transform of the function
4s3

4s2 + 4s + 5
4s2 + 4s + 5
=
.
2
+ 12s + 9s + 2
(2s + 1)2 (s + 2)

9. Find the inverse Laplace transform of the function


1 + sb
.
s+a

100

lecture notes, me 340

Solutions
1. Consider the following free-body diagram, where x = dx/dt.
f (t)i N

3xi N

0.1x i N

Using Newtons 2nd law, we obtain the differential equation


2

d2 x
dx
(t) + 0.1 (t) + 3x (t) = f (t)
dt
dt2

and the equivalent representation as a double integrator with


negative feedback shown below.
3
2

F (s)

1
2

1
s

1
s

X (s)

5
100

It follows that

F (s)
1
,
2
2 s + 0.05s + 1.5
which is underdamped. The graph of the corresponding unit
impulse response
p
1 t/40 sin( 2399/40t)
p
h(t) = e
2
2399/40
X (s) =

is shown in the diagram below.


0.5

t
30

0.5

60

90

101

lecture notes, me 340

2. For an overdamped system with distinct poles at g1 and g2 , the


unit impulse response equals
h(t) =

eg1 t
g1

eg2 t
.
g2

In particular, h(0) = 0, h0 (0) = 1, and h(t) = 0 , e(g1


which is only true for t = 0.
3. Since

it follows that

h
L e

i
cos 2p# (s) =

g2 )t

s+1
,
(s + 1)2 + 4p 2

1
s +1
(s+1)2 +4p 2

1
s

O(s)

or, in other words,


s2 + s + 4p 2
.
s3 + 2s2 + s(1 + 4p 2 )

O(s) =
4. Since,

h
L #e

it follows that

#/2

(s) =

4
,
(2s + 1)2

1
4
(2s+1)2

1
s

O(s)

2
or, in other words,
O(s) =
5. Since

it follows that

L u (#)

4s2 + 4s + 5
.
4s3 + 12s2 + 9s + 2

u (#

1
1) ( s ) =

e
s

= 1,

102

lecture notes, me 340

2
3
1 e
s

X (s)

1
s

1
s

O(s)

3
or, in other words,
O(s) =

s2 + 6s + 1 e
s3 + 3s2 + 2s

6. The initial-value problem is equivalent to the frequency-domain


block diagram shown below.

1
1
0

+
+

3
X (s)

1
s

1
s

O(s)

2
It follows that
O(s) =
Since

the response equals

L #e

o (t) = 5te

3s + 5
.
( s + 1)2

(s) =

1
,
( s + 1)2

D
te
dt

+3

= 2te

+ 3e t .

7. For a double nonzero-initial-value integrator with negative feedback,


ds + c + da + I (s)
O(s) =
.
s2 + as + b

103

lecture notes, me 340

The free response (when I (s) = 0) equals the unit impulse response (when c = d = 0 and I (s) = 1), provided that
ds + c + da = 1
for all s, i.e., c = 1 and d = 0.
8. Since

L
it follows that

1
L

1
1
(t) = te
2
4
(2# + 1)

1
(t) =
(2# + 1)2 (# + 2)

1
#e
4

1
e
9

2t

t/2

#/2

2#

3t 2
e
18

(t)

t/2

It we denote this function by h(t), then

4#3

4#2 + 4# + 5
D
D2
(t) = 5h(t) + 4 h(t) + 4 2 h(t)
2
dt
+ 12# + 9# + 2
dt
13 2t 6t 4 t/2
= e +
e
.
9
9

Since h(0) = dh
dt (0) = 0, both distributional derivatives are simply
regular derivatives.
9. Since

L
it follows that

1 + sb
L 1
(t) = e
s+a

1
(t) = e
#+a

at

+b

D
e
dt

at

at

= bd(t) + (1

ab)e

at

In this case, the distributional derivative differs from the regular


derivative, since limt#0 e at = 1.

104

lecture notes, me 340

Lesson 6 - Modal analysis


TAKE-AWAYS:
* If there exist initial values such that the free response of a
dynamic system is harmonic, then the corresponding angular
frequency is a natural frequency of the system.
* When they exist, natural frequencies correspond to eigenvalues of a matrix, whose elements are combinations of system
gains. Each corresponding eigenvector describes the mode
shape associated with a natural frequency.
* Each natural frequency is associated with a modal oscillation, in which representative outputs oscillate harmonically
with the same angular frequency, and with amplitudes and
phase shifts related by the corresponding mode shape.
* For mechanical systems with n masses that are coupled to
each other and to ground by springs, there are n natural frequencies and n distinct mode shapes. Arbitrary initial values
result in a free response that is a sum of modal oscillations.
Such a sum is only periodic if the corresponding natural frequencies are rationally related.

105

lecture notes, me 340

Online lecture: natural frequencies


If the free response of a dynamic system is harmonic for some
collection of initial values, then the corresponding angular frequency
is a natural (or modal) frequency of the system.
Example 53. Consider the double integrator with feedback shown below.

Figure 61: An undamped double


integrator with feedback.

4
c
I (s) +

+
+

Here,
X (s) =

X (s)

1
s

1
c + I (s)
s

+
+

4O(s) , O(s) =

or, in other words,


O(s) =

1
s

O(s)

1
d + X (s)
s

ds + c + I (s)
.
s2 + 4

(329)

(330)

Since

L sin(w#) (s) =

s2

it follows that

w
s
, L cos(w#) (s) = 2
,
2
+w
s + w2

c
I (s)
O(s) = dL cos 2# (s) + L sin 2# (s) + 2
.
2
s +4

(331)

(332)

When I (s) = 0, the free response equals the harmonic signal


o (t) = d cos 2t +

c
sin 2t
2

(333)

with angular frequency 2.


In Example 53, every free response is harmonic, and the only
natural frequency is 2.
Example 54. Consider the mechanical suspension shown in the figure
on the next page. Using Newtons 2nd law, we obtain the corresponding
initial-value problem
m

d2 x
dx
dx
(t) + b (t) + kx (t) = 0, x (0) = x0 ,
(0) = v0
2
dt
dt
dt

(334)

106

lecture notes, me 340

Free-body diagram:
x

Figure 62: A mechanical suspension.

i
b

~Fs =

kxi ~Fd =

b dx
dt i

where m is the plates mass, b is the damping coefficient, k is the spring


stiffness, and x0 and v0 denote the initial displacement and velocity, respectively, of the plate. The corresponding integral equations
x ( t ) = x0 +

Z t
0

y(t ) dt, y(t) = v0

1
m

Z t
0

by(t ) + kx (t ) dt

(335)

bY (s) + kX (s)
s

(336)

translate into
X (s) =

v0
x0 + Y ( s )
, Y (s) =
s

1
m

or, in other words,


X (s) =

b
m x0
,
b
k
ms + m

x0 s + v0 +
s2 +

Y (s) =

v0 s
s2 +

k
m x0
b
k
ms + m

107

(337)

The free response is harmonic if and only if b = 0, i.e., in the absence


of damping. In this case, the natural frequency of the mechanical system is
p
k/m, and
r
r
r
k
m
k
x (t) = x0 cos
t + v0
sin
t,
(338)
m
k
m
r
r
r
k
k
k
y(t) = v0 cos
t x0
sin
t.
(339)
m
m
m
For b 1, i.e., in the presence of weak damping, the free response is approxp
imately harmonic with angular frequency k/m, but with an amplitude
that decays exponentially in time.

lecture notes, me 340

In-class worksheet: mode shapes


Consider the coupled, identical, double integrators with feedback,
shown below. Here,
ds + c + I1 (s)
s2 + 2
f s + e + I2 (s)
O2 (s) =
s2 + 2
O1 (s) =

or, in formal matrix form,


!
s2 + 2
1

1
s2 + 2

O1 (s)
O2 (s)

O2 (s)

(340)

O1 (s)

ds + c + I1 (s)
f s + e + I2 (s)

(341)

(342)

Figure 63: Two coupled undamped


double integrators with feedback.

2
c
I1 (s) +

+
+

I2 (s)

+
+

d
1
s

1
s

+
+

+
+

1
s

1
s

O1 (s)

O2 (s)

e
2

When I1 (s) = I2 (s) = 0, the outputs o1 (t) and o2 (t) are harmonic
with angular frequency w, provided that
O1 (s) =

o11 s + o12
o s + o22
, O2 (s) = 212
2
2
s +w
s + w2

(343)

for some coefficients o11 , o12 , o21 , and o22 . Substitution into (342),
multiplication on both sides by s2 + w 2 , and the requirement that
coefficients of equal powers of s be identical on both sides of the
equation, shows that o11 = d, o12 = c, o21 = f , and o22 = e. In this
case,
!
!
1
O1 (s)
ds + c
= 2

(344)
s + w2
O2 (s)
fs + e

108

lecture notes, me 340

must be a (formal) eigenvector of


s2 + 2
1

1
2
s +2

(345)

corresponding to the (formal) eigenvalue s2 + w 2 , since


!
!
!
1
s2 + 2
1
ds + c
ds + c
2

=
.
s + w2
1
s2 + 2
fs + e
fs + e

(346)

But an eigenvalue l of (345) is a solution u = l of the characteristic


equation
s2 + 2
1

1
s2 + 2

= ( s2 + 3

u)(s2 + 1

u) = 0,

(347)

i.e., l = s2 + 3 or l = s2 + 1. For the corresponding eigenvectors

T
, we find
s t

and

s2 + 2

s2 + 2

( s2 + 3) s + t = 0 ) t = s

( s2 + 1) s + t = 0 ) t =

s,

(348)

(349)

respectively.
We conclude that the system of two coupled double integrators
p
with feedback have two natural frequencies: w = 3 and w = 1. If
the initial values are chosen so that f = d and e = c, associated with
the first eigenvector, then
O1 (s) = O2 (s) =

ds + c
,
s2 + 3

(350)

and the free response equals


o1 (t) = o2 (t) = d cos

p
c
3t + p sin 3t
3

(351)

In this case, both outputs have identical amplitude and phase shifts.
Both outputs reach their maximal and minimal values simultaneously.
If, instead, the initial values are chosen so that f = d and e = c,
associated with the second eigenvector, then
O1 (s) =

O2 (s) =

ds + c
,
s2 + 1

(352)

and the free response equals


o1 ( t ) =

o2 (t) = d cos t + c sin t

(353)

109

lecture notes, me 340

In this case, the two outputs have identical amplitude, but phase
shifts that differ by p. When one of the outputs reaches its maximal
value, the other reaches its minimal value.
For arbitrary e and f ,
!
!
!
(d+ f )s+(c+e)
(d f )s+(c e)
ds + c
2
2
=
+
,
(354)
(d+ f )s+(c+e)
( f d)s+(e c)
fs + e
2
2
where the first column matrix is an eigenvector of
!
s2 + 2
1
1
s2 + 2

(355)

corresponding to the eigenvalue s2 + 3, and the second is an eigenvector corresponding to the eigenvalue s2 + 1. It follows by linearity
that the free response of this dynamical system is a combination of
p
a harmonic signal with angular frequency 3 and a harmonic signal with angular frequency 1. Such a combination is typically not
p
periodic, since 3 is irrational.
Each natural frequency is associated with a mode shape, corresponding to the relationship between the amplitudes and
phase shifts of the harmonic signals associated with each output. In the example, the mode shapes are 1 : 1 and 1 : 1,
corresponding to an in-phase and an out-of-phase mode, respectively.

In-class worksheet: coupled oscillators


Consider the layered mechanical suspension, shown below.
Using Newtons 2nd law, we obtain the corresponding initial-value
problem
m1

d2 x1
( t ) + k 1 x1 ( t ) =
dt2

K x1 ( t )

x2 (t) , x1 (0) = d,

dx1
(0) = c
dt
(356)

m2

d2 x2
( t ) + k 2 x2 ( t ) =
dt2

K x2 ( t )

x1 ( t ) , x2 (0) = f ,

dx1
(0) = e
dt
(357)

and the equivalent frequency-domain representation


ds + c + K X2 (s) X1 (s) /m1
,
s2 + k1 /m1
f s + e + K X1 (s) X2 (s) /m2
X2 ( s ) =
s2 + k2 /m2
X1 ( s ) =

(358)
(359)

110

lecture notes, me 340

x2

Figure 64: A multi-layered mechanical


suspension.

m2

K
x1

m1

k2

k1

or, in formal matrix form,


s2 +

k1 +K
m1
K
m2

K
m1
s2 + k2m+2K

X1 ( s )
X2 ( s )

ds + c
fs + e

(360)

The response is a modal oscillation, in which the displacements


x1 (t) and x2 (t) of the two plates are harmonic with angular frequency w, provided that
X1 ( s ) =

x11 s + x12
x s + x22
, X2 (t) = 212
s2 + w 2
s + w2

(361)

for some coefficients x11 , x12 , x21 , and x22 . Substitution into (360),
multiplication on both sides by s2 + w 2 , and the requirement that
coefficients of equal powers of s be identical on both sides of the
equation, shows that x11 = d, x12 = c, x21 = f , and x22 = e. In this
case,
!
!
1
X1 ( s )
ds + c
= 2

(362)
s + w2
X2 ( s )
fs + e
must be a (formal) eigenvector of
s2 +

k1 +K
m1
K
m2

K
m1
s2 + k2m+2K

(363)

corresponding to the (formal) eigenvalue s2 + w 2 . But an eigenvalue


l of this matrix is a solution u = l of the characteristic equation
s2 + (k1 + K )/m1
K/m2

K/m1
s2 + (k2 + K )/m2

= 0,

111

(364)

lecture notes, me 340

It follows that v = l

s2 is a solution of the characteristic equation

(k1 + K )/m1
K/m2

K/m1
(k2 + K )/m2

= 0,

(365)

This is of the form v2

tv + d = 0, whose solutions are


r
t
t2
v=
d
2
4
Here, t is the sum of the diagonal elements of the matrix
!
(k1 + K )/m1
K/m1
K/m2
(k2 + K )/m2

(366)

(367)

and d is its determinant. The solutions to the characteristic equation


are real and positive provided that t > 0 and 0 < d < t 2 /4. It follows
from inspection that t and d are both positive. Moreover,

t2
k1 + K
k2 + K 2
4K2
d=
+
(368)
4
m1
m2
m1 m2
which is always positive.
It follows that l s2 is real and positive for both eigenvalues
of (363). If we denote these values by w12 and w22 , we conclude that
the mechanical system always has two natural frequencies. Since
the corresponding eigenvectors are also real, it follows that for each
modal oscillation, x1 (t) and x2 (t) are harmonic signals with different
amplitudes but identical phase shifts, if both components of the
eigenvector have the same sign (in-phase), or harmonic signals with
different amplitudes and phase shifts that differ by p (out-of-phase),
if the components of the eigenvector are of opposite sign.
For arbitrary e and f , the column matrix
!
ds + c
(369)
fs + e
may be expressed as a sum of an eigenvector corresponding to w1
and an eigenvector corresponding to w2 . It follows by linearity that
the free response of the system is a sum of modal oscillations with
angular frequencies w1 and w2 . Such a combination is periodic only
if w1 and w2 are rationally related, i.e., if there are integers m and n
such that mw1 = nw2 .

Exercises
1. Suppose that

5s + 6 kX (s)
.
s2 + 5
Determine k such that x (t) is harmonic with angular frequency 3.
X (s) =

112

lecture notes, me 340

2. Find the natural frequency of the dynamical system shown below.


9
d
1
s

X (s)

1
s

O(s)

c
and express x (t) in terms of the initial values c and d.
3. Show that the dynamical system shown below has no natural
frequencies.
c

1
s

X (s)

1
s

O(s)

9
4. Determine x0 and v0 such that the solution to the initial-value
problem
m

d2 x
dx
(t) + kx (t) = 0, x (0) = x0 ,
(0) = v0
dt
dt2

is harmonic with amplitude 2 and phase shift p/4.


5. Consider the block diagram below.

b
c
I (s) +

+
+

d
1
s

X (s)

+
+

1
s

O(s)

Let w denote its natural frequency. Show that the transfer function
of the corresponding system with zero initial values has poles at
jw.

113

lecture notes, me 340

6. Find the (formal) eigenvalues and corresponding eigenvectors of


the matrix
!
s2 + 3
2
.
1
s2 + 1
7. Find the relationship between the eigenvalues of the matrix
!
s2 + a
b
.
g
s2 + d
and the roots of its determinant as a function of s. What happens
in the special case that an eigenvalue equals s2 + w 2 for some real
number w?
8. The outputs of a pair of coupled, identical, double integrators with
feedback satisfy the equations
ds + c aO2 (s)
,
s2 + b
f s + e aO1 (s)
O2 (s) =
s2 + b
O1 (s) =

in terms of the gains a and b and initial values c, d, e, and f . Find


a condition on a and b that guarantees the existence of two natural
frequencies.
9. Consider a mechanical system, in which two displacements x1 (t)
and x2 (t) satisfy the coupled differential equations
d2 x1
(t) + 4x1 (t)
dt2
d2 x2
(t) + 12x2 (t)
dt2

3x2 (t) = 0
3x1 (t) = 0

Graph a representative free response corresponding to each of the


two natural frequencies.
10. For the mechanical system shown below, there exist two values
of k such that the response to f 1 (t) = d(t) and f 2 (t) = kd(t),
applied to the lower and upper plate, respectively, with zero initial
displacements and velocities equals a modal oscillation in the
absence of inputs but with nonzero initial displacements and
velocities. Find the corresponding values of k.

114

lecture notes, me 340

x2

2 kg

4 N/m
x1

2 kg

4 N/m

i
10 N/m

11. Suppose that the free response of the mechanical system shown
below to some choice of initial conditions is harmonic in both
outputs with period approximately equal to 1.66 s. Sketch the configuration of the system when the first mass reaches its maximal
displacement in either direction, as well as when the first mass
is halfway between its two maximal displacements, and indicate
the elapsed time since the initial time at which both displacement
equal 0.
x2

4.1 kg

5.2 N/m
x1

0.5 kg

12.6 N/m

i
1.3 N/m

12. Describe the response of the two-degree-of-freedom mechanical


system
m1 x1 (t) =

k 1 x1 ( t ) + K ( x2 ( t )

x1 (t)) + ad(t)

m2 x2 (t) =

k 2 x2 ( t )

x1 (t)) + bd(t)

K ( x2 ( t )

115

lecture notes, me 340

with initial conditions x1 (0) = d, x 1 (0) = c, x2 (0) = f , x 2 (0) = e,


and specialize to the case that m1 = m2 = 2, k1 = 10, k2 = K = 4
and
(a) a = 1, b = 3, c = f = 0, d = e = 1,
(b) a = b = c = e = 0, d = f = 1,
(c) a = b = d = f = 0, c = e =

2.

Solutions
1. If x (t) is harmonic with angular frequency 3 then
X (s) =

ds + c
s2 + 9

for some constants d and c. Here,


X (s) =
2. Here,
X (s) =

1
c
s

s2

5s + 6
) k = 4.
+5+k

9O(s) , O(s) =

or, in other words,

1
d + X (s)
s

cs 9d
.
s2 + 9
3d sin 3t and the natural frequency

X (s) =
It follows that x (t) = c cos 3t
is 3.
3. Here,
X (s) =

1
c
s

9X (s) , O(s) =

1
d + X (s)
s

or, in other words,


X (s) =

c
ds + c + 9d
, O(s) =
s+9
s2 + 9s

No choice of c and d will result in a combination of s/(s2 + w 2 )


and w/(s2 + w 2 ).
4. Here,

r
r
k
m
k
x (t) = x0 cos
t + v0
sin
t
m
k
m
q
is harmonic with amplitude x02 + v20 m/k and phase shift q,
where
p
x0
v0 m/k
cos q = q
, sin q = q
.
x02 + v20 m/k
x02 + v20 m/k
p
p
It follows that x0 = 2 and v0 = 2k/m.

116

lecture notes, me 340

5. Here, w =

b. Furthermore,
O(s) =

ds + c + I (s)
s2 + b

from which we obtain the transfer function


H (s) =

s2

1
+b

in the case of zero initial values. Its poles are found at s = j


6. The characteristic equation equals

( s2 + 3

u)(s2 + 1

and its solutions are s2 + 2

T
satisfy
s t
s2 + 3

( s2 + 2 +

s2 + 3

( s2 + 2

u)

b.

2=0

3. The corresponding eigenvectors

3) s

2t = 0 ) t =

3) s

2t = 0 ) t =

p
2

and

p
1+ 3
s,
2

respectively.
7. If l is an eigenvalue then u = l satisfies the characteristic equation
s2 + a
g
This implies that v = l

u
s2

b
+d

= 0.

s2 satisfies the characteristic equation


v

a
g

b
d

= 0,

which in turn means that any square root of


determinant
s2 + a
b
g
s2 + d

v is a root of the

as a function of s. In the special case that v = w 2 and, consequently, l = s2 + w 2 , it follows that jw are two roots of the
determinant.
8. If they exist, the natural frequencies correspond to eigenvalues of
the matrix
!
s2 + b
a
a
s2 + b

117

lecture notes, me 340

In particular, l is an eigenvalue if u = l satisfies the characteristic


equation
s2 + b u
a
=0
2
a
s +b u
i.e., l = s2 + b + a or l = s2 + b a. These are both of the form
s2 + w 2 for a real number w, provided that b > a.
9. Here,

1
1
d + c + 3X2 (s)
s
s

1
1
X2 ( s ) =
f + e + 3X1 (s)
s
s
X1 ( s ) =

4X1 (s)

12X2 (s)

or, in matrix form,


s2 + 4
3

3
2
s + 12

X1 ( s )
X2 ( s )

ds + c
fs + e

dx2
1
where c = dx
dt (0), d = x1 (0), e = dt (0), and f = x2 (0). The
natural frequencies w1 and w2 correspond to eigenvalues of the
matrix
!
s2 + 4
3
3
s2 + 12

i.e., to solutions u = l of the characteristic equation


s2 + 4 u
3

s2

3
+ 12

= 0.

It follows that l = s2 + 3 and l = s2 + 13 and, consequently, that


p
p
w1 = 3 and w2 = 13.

T
For the corresponding eigenvectors s t
,

and

s2 + 4

s +4

s2 + 3

s + 13

3t = 0 ) s = 3t

3t = 0 ) t =

3s

respectively. We conclude that w1 corresponds to an in-phase


mode shape 3 : 1, for which
p
x1 (t) = 3 cos 3t
p
x2 (t) = cos 3t
if c = 0 and d = 1. The corresponding modal oscillation is graphed
below.

118

lecture notes, me 340

3
2

x1 (t) = 3 cos

3t

x2 (t) = cos

3t

2p

p
1

3p

2
3
Similarly, w2 corresponds to an out-of-phase mode shape 1 :
for which
p
x1 (t) = cos 13t
p
x2 (t) = 3 cos 13t

3,

if c = 0 and d = 1. The corresponding modal oscillation is graphed


below.
3
2

x2 ( t ) =

3 cos

13t

1
x1 (t) = cos
p
1

13t

2p

3p

2
3
10. Here,
ds + c + 12 F1 (s) + 2 X2 (s)
s2 + 5
1
f s + e + 2 F2 (s) + 2 X1 (s)
X2 ( s ) =
s2 + 2
X1 ( s ) =

X1 ( s )
X2 ( s )

,
.

It is clear that c = d = e = f = 0, F1 (s) = 2, and F2 (s) = 2k


is equivalent to the case that c = 1, e = k, d = f = 0, and
F1 (s) = F2 (s) = 0. The latter case results in a modal oscillation,
provided that 1 : k describes the corresponding mode shape or,

119

lecture notes, me 340

T
equivalently, that the column matrix 1 k
is an eigenvector
of
!
s2 + 7
2
2
s2 + 4
An eigenvalue l of this matrix is a solution u = l of the characteristic equation
s2 + 7 u
2
=0
2
2
s +4 u
i.e., l = s2 + 3 or l = s2 + 8. For the corresponding eigenvectors

T
,
s t

2
2
s +7
s +3
s 2t = 0 ) t = 2s
and

s +7

s +8

2t = 0 ) t =

respectively. It follows that k = 2 or k =

1
s
2

1/2.

11. A harmonic response with period close to 1.66 s corresponds to a


modal oscillation with natural frequency w 3.78 rad/s. For this
mechanism,
13.0 w 2
10.4
=0
1.27
4.34 w 2
in the appropriate units, implies that w = w! 1.74
rad/s and
w = w2 3.78 rad/s. For the eigenvector s t corresponding
to w2 , t 0.13s. It follows that the second mass reaches its maximal displacement in the negative direction when the first mass
reaches its maximal displacement in the positive direction and vice
versa. When the first mass is half-way between its maximal displacements, this is also true of the second mass. The sketch below
gives a qualitative sense of the corresponding mode shape.

x1

x2

x2

t = 0 s:

t 0.42 s:
x1

120

lecture notes, me 340

x2

x2

t 0.83 s:

t 1.25 s:

x1

x1

12. In the frequency domain,


X1 =

ds + c0 + K ( X2 X1 )/m1
f s + e0 K ( X2 X1 )/m2
, X2 =
2
s + k1 /m1
s2 + k2 /m2

or, equivalently,
s2 +

k1 +K
m1
K
m2

K
m1
s2 + k2m+2K

X1
X2

ds + c0
f s + e0

where c0 = c + a/m1 and e0 = e + b/m2 account for the change to


the initial velocities that results from the impulsive forcing.
The response is a modal oscillation with angular frequency w
provided that
ds + c0
f s + e0
X1 = 2
,
X
=
2
s + w2
s2 + w 2
which, in turn, is true if w 2 is an eigenvalue of
k1 +K
m1
K
m2

and

K
m1
k2 +K
m2

ds + c0
f s + e0

is a corresponding eigenvector.
With the given numerical values for m1 , m2 , k1 , k2 , and K, the
eigenvalues are 8 and 3 with corresponding eigenvectors
!
!
2
1
,
1
2

121

lecture notes, me 340

It follows that the response is a modal oscillation with frequency


p
8 if both d = 2 f and c0 = 2e0 , and a modal oscillation with
p
frequency 3 if both f = 2d and e0 = 2c0 . In the first case,
x1 ( t ) =

2 f cos

p
p
p
2e0
e0
p sin 8t, x2 (t) = f cos 8t + p sin 8t
8
8

8t

In the second case,


x1 (t) = d cos

p
p
p
c0
2c0
3t + p sin 3t, x2 (t) = 2d cos 3t + p sin 3t
3
3

If neither of these relations hold for the initial conditions, then the
response is a sum of the two modal oscillations, i.e.,

p
p
p
2B
D
p sin 8t + C cos 3t + p sin 3t
8
3
p
p
p
p
B
2D
x2 (t) = A cos 8t + p sin 8t + 2C cos 3t + p sin 3t
8
3
where
x1 ( t ) =

2A cos

8t

2A + C = d,

2B + D = c0 , A + 2C = f , B + 2D = e0

In case (a), it follows that


A=

2
3
1
11
,B =
,C = ,D =
5
10
5
10

In case (b), it follows that


A=

1
3
, B = 0, C = , D = 0
5
5

or, in other words,

p
p
2
3
cos 8t + cos 3t,
5
5
p
p
1
6
x2 ( t ) =
cos 8t + cos 3t.
5
5
p
In particular, for 8t = p/2 + np, x1 : x2 = 1 : 2, i.e., according to
p
the second mode shape. Similarly, for 3t = p/2 + np, x1 : x2 =
2 : 1, i.e., according to the first mode shape.
x1 ( t ) =

Finally, in case (c), it follows that


A = 0, B =

2
, C = 0, D =
5

6
5

or, in other words,

p
p
4
6
p sin 8t
p sin 3t,
5 8
5 3
p
p
2
12
p sin 3t
x2 (t) = p sin 8t
5 8
5 3
x1 ( t ) =

122

lecture notes, me 340

p
In particular, for 8t = np, x1 : x2 = 1 : 2, i.e., according to the
p
second mode shape. Similarly, for 3t = np, x1 : x2 = 2 : 1, i.e.,
according to the first mode shape.

123

lecture notes, me 340

Lesson 7 Frequency response


TAKE-AWAYS:
* The response of a stable convolution to a bounded input is
bounded.
* For large t, the steady-state response of a stable convolution
to the input ejwt equals H (jw )ejwt , where w 7! H (jw ) is the
frequency response function.
* The steady-state response to the harmonic input
A cos(wt

q)

is harmonic with amplitude A| H (jw )| and phase shift equal to


the difference between q and the phase of H (jw ).
* If the magnitude of the frequency response function has a
local maximum for some frequency w , then w is a resonance
frequency and the corresponding magnitude is the resonance
amplitude.
* If the magnitude of the frequency response function has a
local minimum for some frequency w , then w is an attenuation frequency. When this occurs in a mechanical system,
input vibrations at the frequency w are attenuated relative to
vibrations at nearby frequencies.

124

lecture notes, me 340

125

Online lecture: Bode diagrams


A signal f (t) is bounded if | f (t)| M for all t > 0 and some
M > 0.
Figure 65: A bounded signal.

M
f (t)
t

A convolution is stable if the response to a bounded input is


bounded. Suppose that |i (t)| < M. Since

|o (t)| =

Z t
0

t ) dt M

h ( t )i ( t

Z t
0

|h(t )| dt,

(370)

a convolution is stable if and only if


Z
0

|h(t )| dt < .

(371)

For a stable convolution,


Z

H ( s ) = L h (#) ( s ) =

exists for all s with real part

| H (s)|

st

h(t)e

dt

(372)

0, since in this case

Z
0

|h(t)| dt < .

(373)

Example 55. Consider the zero-initial-value integrator with negative


feedback shown below.
i(t)

Rt
0

o (t)

Figure 66: A zero-initial-value integrator with negative feedback.

lecture notes, me 340

126

This corresponds to a stable convolution with transfer function


H (s) =

1
s+a

(374)

at

(375)

1
<
a

(376)

and unit impulse response


h(t) = e
since

Z
0

provided that a > 0.

|e

at

| dt =

Example 56. Consider the sequential wiring of two identical zero-initialvalue integrators with negative feedback shown below.
Figure 67: A sequential wiring of two
identical zero-initial-value integrators
with negative feedback.

a
i(t) +

y(t) +

Rt
0

Rt

o (t)

This corresponds to a stable convolution with transfer function


H (s) =

1
( s + a )2

(377)

and unit impulse response


h(t) = te
since

Z
0

provided that a > 0.

|te

at

at

| dt =

(378)

1
<
a2

(379)

Consider the input i (t) = ejwt to a stable convolution with transfer


function H (s) and unit impulse response h(t). Since
Z t

o (t) = h(#) ejw# (t) = ejwt


h(t )e
0

jwt

dt,

(380)

it follows that

o (t) ejwt

Z
0

h(t )e

jwt

dt = H (jw )ejwt

(381)

for large t. We call this the steady-state behavior of the output and
write oss (t).

lecture notes, me 340

127

The complex-valued function w 7! H (jw ) is known as the frequency response function. Since the magnitude of ejwt equals
1, the magnitude of the steady-state output oss (t) = H (jw )ejwt
equals the magnitude of the frequency-response function.
A pair of graphs of the magnitude (in dB10 ) and phase, respectively,
of H (jw ) versus w is called a Bode diagram. Each graph is a Bode
plot.

10

In other words, represented by the


quantity 20 log10 | H (jw )|.

Example 57. For the zero-initial-value integrator with negative feedback in


Example 55 with a = 1, the frequency response function equals
H (jw ) =
with magnitude

1
jw + 1

(382)

(383)

1 + w2

and phase
arctan w.

(384)

Figure 68: A Bode diagram for a zeroinitial-value integrator with negative


feedback. A logarithmic scale is used
along the horizontal axes.

0
dB

p 1
1+ w 2

20

40
10
0

10

100

101

102

101

102

arctan w
p/4

p/2
10

10

100

A corresponding Bode diagram is shown in the figure above.


Example 58. For the sequential wiring of two identical zero-initial-value
integrators with negative feedback in Example 56 with a = 1, the frequency

lecture notes, me 340

128

response function equals


H (jw ) =

1
=
(jw + 1)2

with magnitude
p

and phase

w2

1
+ 2jw + 1

(385)

(1

arctan
if w < 1,
arctan

(386)

w 2 )2 + 4w 2
2w
1 w2

2w
w2 1

(387)

(388)

if 1 < w, and p/2 if w = 1. A corresponding Bode diagram is shown in


the figure below.
Figure 69: A Bode diagram for a sequential wiring of two identical zeroinitial-value integrator with negative
feedback. A logarithmic scale is used
along the horizontal axes.

0
dB
20
40

1
(1 w 2 )2 +4w 2

60
80
10
0

10

100

arctan 1 2ww2
arctan w2w
p
2 1

p/4

102

101
w<1
w>1

p/2
3p/4
p
10

10

100

101

102

In-class worksheet: resonance


Consider the mechanical suspension shown below. Using Newtons
2nd law, we obtain the differential equation
m

d2 x
dx
(t) + b (t) + kx (t) = f (t)
dt
dt2

(389)

lecture notes, me 340

Free-body diagram:
f (t)i

Figure 70: A mechanical suspension.

f (t)i

i
d

kxi

b dx
dt i

for the displacement x (t) of the plate, relative to the undeformed


configuration of the spring, in the direction of the unit vector i. With
zero initial displacement and velocity, the relationship between the
input f (t) and the output x (t) is described by a convolution with
transfer function
1
H (s) =
(390)
ms2 + bs + k
The corresponding unit impulse response is the convolution

g1 #
g2 #
h(t) = e e
(t)
(391)
where
g1,2 =

b2
2m

4mk

(392)

are the poles of H (s).


Provided that b > 0, this dynamical system is stable, since the unit
impulse response is a combination of exponentially decaying signals
with polynomial or trigonometric coefficients. The corresponding
frequency response function equals
H (jw ) =

mw 2

1
1
=
k1
+ jbw + k

1
+ j2zW

W2

(393)

p
where wn := k/m, z := b/(2mwn ), and W = w/wn . As shown in
the Bode plot on the next page, its magnitude
1
p
k (1

1
W2 )2

+ 4z 2 W2

(394)

decays monotonically from its maximum of 1/k at w = 0 if z 2 > 1/2.


In contrast, if z 2 < 1/2, the largest magnitude equals
1
p
2zk 1

z2

129

(395)

lecture notes, me 340

130

and is obtained when


w = w : = wn

2z 2

(396)

For small values of the damping coefficient, the resonance frequency


w is close to the natural frequency wn of the corresponding undamped system. The resonance peaks shifts to the left from the natural frequency as the damping increases.
20

Figure 71: Bode plots of the magnitude


of the frequency response function for
z 2 < 1/2 and z 2 > 1/2.

20

40
10

z 2 < 1/2

z 2 > 1/2

dB

1
(1 W2 )2 +4z 2 W2

10

101

In-class worksheet: riding comfort


Consider the quarter-car vehicle-suspension model shown below.
x

i
k

f (t)

Let the unit vector i be parallel to the line of action of the suspension. Denote by x the displacement of the vehicle body of mass m
along i relative to the undeformed configuration of the spring, when
the lower end of the suspension is located at some reference level.
Suppose that displacements along i of the lower end of the suspension relative to the reference level are described by the function f (t).

Figure 72: A quarter-var vehicle suspension model that accounts for uneven
terrain through the function f (t).

lecture notes, me 340

Using Newtons 2nd law, we conclude that

d2 x
dx
df
m 2 (t) = b
(t)
(t)
dt
dt
dt

k x (t)

f (t)

(397)

where b is the damping coefficient and k is the spring stiffness.


Let
dx
df
y(t) =
(t) and g(t) =
(t)
(398)
dt
dt
It follows that
x ( t ) = x (0) +
and
y ( t ) = y (0) +

1
m

Z t
0

y(t ) dt, f (t) = f (0) +

Z t
0

b g(t )

Z t
0

y(t ) + k f (t )

g(t ) dt,

x (t )

(399)

dt.

(400)

or, in the frequency domain


X (s) =

x (0) + Y ( s )
f (0) + G ( s )
, F (s) =
,
s
s

(401)

and

y(0) + mb G (s) Y (s) + mk F (s) X (s)


Y (s) =
(402)
s

where X (s) = L x (#) (s), Y (s) = L y(#) (s), F (s) = L f (#) (s), and

G (s) = L g(#) (s). If we assume that x (0) = y(0) = f (0) = 0, then


X (s) =

bs + k
F (s)
+ bs + k

ms2

(403)

i.e., the relationship between the displacement of the lower end of the
suspension and the displacement of the plate is a convolution with
transfer function
bs + k
H (s) =
.
(404)
2
ms + bs + k
The dynamic system governed by this transfer function is stable,
since the same is true for the transfer function 1/(ms2 + bs + k). The
magnitude of the steady-state response xss (t) to the complex exponential input f (t) = ejwt is given by the magnitude of the frequency
response function, i.e.,
p
p
k 2 + b2 w 2
1 + 4z 2 W2
= p
(405)
| H (jw )| = p
2
2
2
2
(k mw ) + b w
(1 W2 )2 + 4z 2 W2
p
where wn := k/m, z := b/(2mwn ), and W = w/wn . For W 1,
1, | H (jw )| 2z/W.
| H (jw )| 1. For W
As shown in the Bode plot below, there exists a unique global
maximum at the resonance frequency
rq
wn

w = w :=
1 + 8z 2 1
(406)
2z

131

lecture notes, me 340

132

corresponding to the resonance amplitude

| H (jw )| = s

(1 + 8z 2 )1/4

p
1 + 1 + 8z 2 + 1

1+

1+8z 2

4z 2

(407)

Here, w < wn for all z. The resonance peak in the corresponding


Bode plot shifts to the left from the natural frequency as the damping
1
increases. For z 1, w wn and | H (jw )| 2z
.
Figure 73: Bode plots of the magnitude
of the frequency response function
for the quarter-car suspension model
transfer function. Resonance implies
a significant amplification of ground
oscillations into movement of the
vehicle.

0
dB

10

20
10

1+4z 2 W2

(1 W2 )2 +4z 2 W2

10

101

In-class worksheet: a washing machine


Consider a front-loaded washing machine in the spin cycle. Assume
that the drum rotates with angular frequency w and that its total
mass (including clothes) is M, out of which an imbalance of mass m
is located at a distance d from the axis of rotation. Suppose that the
rotating drum is constrained to move only in the vertical direction
and is suspended from the frame of the washing machine by a linear
spring of stiffness k and damper with damping coefficient b.
The vertical motion x (t) of the rotating drum relative to the equilibrium position is then described by the differential equation
M

d2 x
dx
(t) + b (t) + kx (t) = mdw 2 cos wt
dt
dt2

(408)

For design purposes, it may be desirable to select values of the stiffness and damping that ensure that x (t) remains below a critical
bound during steady-state operation across a range of frequencies.
The response x (t) equals the real part of the solution of the equation
d2 x
dx
M 2 (t) + b (t) + kx (t) = mdw 2 ejwt
(409)
dt
dt
Here, x (t) is the output of a double integrator with negative feedback
and gains b/M and k/M that results from the input mdw 2 ejwt /M. It

lecture notes, me 340

wt

Figure 74: A rotating imbalance.

i
b

follows that the magnitude of the steady-state response xss (t) equals

| H (jw )| = p

mdw 2

(k

Mw 2 )2 + b2 w 2

W2

md
p
M (1

W2 )2 + 4z 2 W2

(410)

p
where wn := k/M, z := b/(2Mwn ), and W = w/wn .
p
As shown in the Bode plot below, if z > 1/ 2, the magnitude
increases monotonically from 0 at w = 0 toward a limiting value of
p
md/M as w ! . In contrast, if z < 1/ 2, there exists a global
maximum at the resonance frequency
w = w := p

wn
1

(411)

2z 2

corresponding to the resonance amplitude

| H (jw )| =

md
1
p
M 2z 1

z2

(412)

The resonance peak in the corresponding Bode plot shifts to the right
from the natural frequency wn as the damping increases.

In-class worksheet: vibration absorption


Consider movements induced in the mechanical suspension shown
below by the horizontal motion of the supports.
In this case, the displacement x1 (t) and x2 (t) of each of the masses
relative to the equilibrium configuration satisfy the differential equa-

133

lecture notes, me 340

20

Figure 75: Bode plots of the magnitude


of the frequency response function for
z 2 < 1/2 and z 2 > 1/2.

z 2 < 1/2

dB

z 2 > 1/2

20

40
10

134

W2
(1 W2 )2 +4z 2 W2

10

x1 , x2

i
z/2

f (t)

101

Figure 76: A vibration absorber.

z/2

f (t)

1
k/2

1/2

1/2

k/2

1/

tions
1 dy1
(t) =
dt

k x1 ( t )

f (t)

z y1 ( t )
dy2
( t ) = x1 ( t )
dt
where

x1 ( t )

x2 ( t )
(413)

y2 ( t )

x2 ( t ) + z y1 ( t )

y2 ( t )

(414)

dx1
dx2
( t ), y2 ( t ) =
( t ).
(415)
dt
dt
In the case of zero initial displacements and velocities, this corresponds to the system of coupled integrators with feedback shown
below.
Here,
1
1
X1 (s) = Y1 (s), X2 (s) = Y2 (s)
(416)
s
s
y1 ( t ) =

lecture notes, me 340

F (s)

1
s

Y1 (s)

Figure 77: A block-diagram representation of the two-degree-of-freedom


mechanism.
1
s

X1 ( s )

+
+

and

1
s

Y1 (s) =

k X1 ( s )
s

Y2 (s) =

1
s

X1 ( s )

Y2 (s)

1
s

X2 ( s )

+ z Y1 (s) Y2 (s) ,

X2 (s) + z Y1 (s) Y2 (s) ,

X2 ( s )

F ( s ) + X1 ( s )

(417)
(418)

where X1 (s) = L x1 (#) (s), X2 (s) = L x2 (#) (s), Y1 (s) = L y1 (#) (s),

Y2 (s) = L y2 (#) (s), and F (s) = L f (#) (s). It follows that


s2 + zs + 1
kF (s) (419)
s4 + z (1 + )s3 + (1 + + k)s2 + kzs + k
zs + 1
X2 ( s ) = 4
kF (s). (420)
s + z (1 + )s3 + (1 + + k)s2 + kzs + k
X1 ( s ) =

It is possible to show that the transfer functions


s2 + zs + 1
k
+ (1 + + k)s2 + kzs + k
zs + 1
H2 (s) = 4
k
s + z (1 + )s3 + (1 + + k)s2 + kzs + k
H1 (s) =

s4

+ z (1 + ) s3

135

(421)
(422)

are both stable. It follows that the ratio of the magnitudes of the
steady-state responses x1,ss (t) and x2,ss (t) to the complex exponential
input f (t) = ejwt is given by
p
| H1 (jw )|
(1 w 2 )2 + z 2 w 2
p
=
.
(423)
t :=
| H2 (jw )|
1 + z 2 w2
As seen in the diagram on the next page, this has a global minimum
at
rq
1

w = w :=
1 + 2z 2 1.
(424)
z

lecture notes, me 340

136

In particular, for z 1, w 1 z 2 /4 and t z. In this case, and


for input frequencies near w , the design absorbs the energy from the
motion of the supports predominantly in motion of the second mass.
3

Figure 78: The ratio t between the


steady-state amplitudes for the oscillatory motion of the first and second
mass, respectively.

(1 w 2 )2 + z 2 w 2

1+ z 2 w 2

Exercises
1. Show that convolution with h(t) = u(t) is not stable.
2. Show that the convolution corresponding to the transfer function
H (s) = s/(s + 1) is stable.
3. Show that the convolution corresponding to the transfer function
H (s) = 1/(s2 + bs + 1) is stable for all b > 0.
4. Show that the convolution corresponding to the transfer function
H (s) = (s2 + 1)/(s4 + zs3 + 3s2 + zs + 1) is stable for all z > 0.
5. Suppose that oss (t) is the steady-state response of a stable convolution with transfer function H (s) to the input i (t) = ejwt . Find the
steady-state response to the input i (t) = e jwt .
6. Suppose that oss (t) is the steady-state response of a stable convolution with transfer function H (s) to the input i (t) = ejwt . Find the
steady-state response to the input i (t) = cos wt.
7. Describe the steady-state response of the stable convolution with
transfer function
s 1
H (s) = 2
s +s+4
to a general harmonic input, and specialize to the case that i (t) =
p
2 sin 3t.
8. Draw a Bode diagram for the transfer function H (s) = s/(s + 1).

lecture notes, me 340

9. Draw a Bode diagram for the transfer function H (s) = s/(s + 1)2 .
10. Show that the frequency response function corresponding to the
transfer function
1
H (s) =
ms2 + bs + k
has a resonance peak only if the dynamical system is underdamped.
11. Suppose that a convolution is described by the Bode diagram
shown below.
0
dB
20
40
60
1

100

101

102

100

101

102

Find the steady-state response to the input i (t) = cos 3t

cos 10t.

10
p
p/2
p
p/2
p
10

Solutions
1. Here,
o (t) =
where s = t

Z t
0

u ( t )i ( t

t ) dt =

Z t
0

i (s ) ds,

t. The input i (t) = u(t) is bounded, but


o (t) =

is not bounded.

Z t
0

u(s ) ds = t

137

lecture notes, me 340

2. Here,
h(t) =
and

Z t
0

h ( t )i ( t

D
e
dt

= d(t)

t ) dt = i (t)

or, in other words,


Z t
0

h ( t )i ( t

t ) dt |i (t)| +

Z t
0

Z t
0

i(t

i(t

t ) dt

t ) dt .

If i (t) is bounded, i.e., |i (t)| < M for all t, then


Z t
0

h ( t )i ( t

t ) dt M + M

Z t
0

dt 2M.

for all t.
3. Since s2 + bs + 1 = (s g1 )(s g2 ) for some complex numbers g1
and g2 corresponding to the poles of H (s), it follows that h(t) is a
convolution of eg1 t and eg2 t . In particular,
Z t
0

|h(t )| dt <

provided that the real parts of g1 and g2 are both negative.


Let g(b) represent the dependence of one of the poles on b, such
that
g2 (b) + bg(b) + 1 = 0
Then, g(0) equals either j or
substitution of b = 0 yields

j. Moreover, differentiation and

2g(0)g0 (0) + g(0) = 0 ) g0 (0) =

1
2

For b 1, it follows g(b) g(0) b/2, i.e., that the real part
is negative for small, but positive b. But g(b) is purely imaginary
only if b = 0, since
g(b) = jw )

w 2 + jbw + 1 = 0

is possible only if b = 0. It follows that the real part must be


negative for all b > 0.
4. Since s4 + zs3 + 3s2 + zs + 1 = (s g1 )(s g2 )(s g3 )(s
g4 ) for some complex numbers g1 , g2 , g3 , and g4 corresponding
to the poles of H (s), it follows that h(t) is combination of terms
obtained by various convolutions of eg1 t , eg2 t , eg3 t , and eg4 t , and
their distributional derivatives. In particular,
Z t
0

|h(t )| dt <

138

lecture notes, me 340

provided that the real parts of g1 , g2 , g3 , and g4 are all negative.


Let g(z ) represent the dependence of one of the poles of H (s) on
z, such that
g4 (z ) + zg3 (z ) + 3g2 (z ) + zg(z ) + 1 = 0
Then, g(0) equals one of the purely imaginary numbers
s
p
3 5
j
2
Differentiation and substitution of z = 0 yields
4g3 (0)g0 (0) + g3 (0) + 6g(0)g0 (0) + g(0) = 0,
or, in other words,
0

g (0) =

p
5 5
< 0.
20

g2 (0) + 1
=
4g2 (0) + 6

But g(z ) is purely imaginary only if z = 0, since


g(z ) = jw ) w 4

jzw 3

3w 2 + jzw + 1 = 0

only if z = 0. It follows that the real part of g(z ) must be negative


for all z > 0.
5. Here,

oss (t) = H (jw )ejwt

The response to the input i (t) = e


Z t
0

h(t )e

jw (t t )

jwt

dt = e

then equals

jwt

Z t
0

h(t )ejwt dt

and the steady-state response becomes


e

jwt

Z
0

h(t )ejwt dt = ejwt

Z
0

h(t )e

jwt

dt

or, in other words


H (jw )ejwt = oss (t).
6. Since

ejwt + e jwt
2
the steady-state response to i (t) = cos wt equals
cos wt =

oss (t) + oss (t)


2
i.e., the real part of oss (t) = H (jw )ejwt . If q denotes the phase of
H (jw ), then this equals

| H (jw )| cos(wt + q ).

139

lecture notes, me 340

7. The transfer function corresponds to the frequency response function


jw 1
H ( jw ) =
.
4 w 2 + jw
It follows from the previous exercise that
i (t) = A cos(wt

q ) ) oss (t) = A| H ( jw )| cos(wt

q + arg H ( jw )),

where | H ( jw )| and arg H ( jw ) denote the magnitude and phase of


H ( jw ). Here,
p
p
1+j 3
2e j2p/3
p = jp/3 = e jp/3
H ( j 3) =
2e
1+j 3
and
i (t) = 2 sin

p
3t = 2 cos( 3t

p/2)

implies that

p
oss (t) = 2 cos( 3t

p
p/2 + p/3) = 2 sin( 3t + p/3).

p
The amplitude of the steady-state response equals 2| H ( j 3)| = 2.
The absolute value of the difference in phase between the input
p
and the steady-state response equals | arg H ( j 3)| = p/3.
In general,

| H ( jw )| = p

p
(4

1 + w2
w 2 )2 + w 2

A resonance frequency w corresponds to a local maximum of


| H ( jw )|. This is also a local maximum of | H ( jw )|2 , i.e., such that
d| H ( jw )|2 /dw = 0 when evaluated at w = w . Here,
d
w 4 + 2w 2 23
| H ( jw )|2 = 2w 4
dw
(w
7w 2 + 16)2
p
p
equals 0 when w = 0 or w =
1 + 2 6 1.97.

A Bode plot of the magnitude | H ( jw )| is a graph of 20 log10 | H ( jw )|


against w that uses a logarithmic scale for w. Here,
p
20 log10 | H ( j 3)| = 0, 20 log10 | H ( j1)| 7.0,
and
20 log10 | H ( j0.1)|

12.0.

Moreover, 20 log10 | H ( jw )| = 20 provided that | H ( jw )| = 0.1,


which is true for w 10.4. Accordingly, a Bode plot of the magnitude has the form shown below.

140

lecture notes, me 340

0
dB
10

20
10

10

101

101

102

8. The frequency response function


H (jw ) =
has magnitude

jw
jw + 1

w
1 + w2

and phase

p
arctan w
2
A corresponding Bode diagram is shown below.
0
dB

p w
1+ w 2

20

40
10

10

100

p/2
p
2

arctan w

p/4

0
10

10

100

9. The frequency response function


H (jw ) =

jw
(jw + 1)2

101

102

141

lecture notes, me 340

has magnitude

w
1 + w2

and phase

p
2 arctan w
2
A corresponding Bode diagram is shown below.
0
dB

w
1+ w 2

20

40
10

10

100

101

102

p/2
p
2

p/4

2 arctan w

0
p/4
p/2
10

10

100

101

102

10. The frequency response function


H (jw ) =

1
mw 2 + jbw + k

has magnitude
p

(k

mw 2 )2 + b2 w 2

The derivative with respect to w:


2m(k

(k

mw 2 )

b2

mw 2 )2 + b2 w 2

3/2

equals zero for w = 0 or w 2 = (2mk b2 )/2m2 . It follows that


the derivative is negative for all w > 0 if 2mk < b2 , but has a
unique global maximum away from w = 0 when 2mk > b2 . A
resonance peak thus exists provided that b2 < 2mk, i.e., only in the
underdamped case (for which b2 < 4mk).

142

lecture notes, me 340

11. From the Bode diagram, the magnitude and phase of H (j3) are
approximately 0.56 and p. Similarly, the magnitude and phase of
H (j10) are approximately 0.1 and p/2. By linearity, it follows
that the steady-state response is approximately equal to
0.56 cos(3t + p )

0.1 cos(10t

or, in other words,


0.56 cos 3t

0.1 sin 10t

p/2)

143

lecture notes, me 340

Lesson 8 - Newtonian dynamics


TAKE-AWAYS:
* Newtonian dynamics is a systematic formalism for deriving
equations of motion for mechanical systems by applying Newtons 2nd law to individual particles in the system, and using
physical observations to eliminate unknown reaction forces.
* In order to equate the product of mass and acceleration relative to an inertial coordinate system with the net interaction
force on each particle, the configuration of the mechanism is
described in terms of unknown functions of time, known as
generalized coordinates.
* Infinitesimal masses and forces are common in the study of
rigid bodies, in order to ensure that the total mass and force
are finite. Equations of motion that omit internal interaction
forces are obtained by integration over each of the constituent
bodies.

144

lecture notes, me 340

145

Online lecture: the Newtonian recipe


Newtonian dynamics follows a systematic recipe for deriving
differential equations governing the motion of a mechanism. These
are called the equations of motion of the mechanism.
Step 1: Introduce an inertial (cartesian) coordinate system,
relative to which isolated particles move along straight
paths with constant speed, and let i, j, and k denote
unit vectors along the x-, y-, and z-axes.
Step 2: Formulate a position vector to each particle in the
mechanism. Express its components in terms of
unknown functions of time, known as generalized
coordinates.
The smallest number of generalized coordinates required to describe all particles is the number of (geometric) degrees of freedom
of the mechanism.
Example 59. Consider the slender rod of length ` shown below, able to
rotate about the stationary axis A that is perpendicular to the rod.

Figure 79: A slender rod constrained to


rotate about an axis through one of its
end points.

q (t)

`
i
A

Introduce a coordinate system such that the stationary end point of the
rod coincides with the origin, and A coincides with the z-axis, and assume
that this is inertial. Other than the stationary end point of the rod, all particles on the rod move along circles centered at the origin. It follows that these
particles cannot be isolated, and must be affected by forces.
Let q (t) denote the time-dependent angle of rotation of the rod relative
to the x-axis. A rotation corresponds to a positive angle, when the direction
of rotation would result in motion along the z-axis of a right-handed screw,
and is negative otherwise. The position vector to the particle at distance r

lecture notes, me 340

146

from the origin is then given by

~r = r cos q (t)i + r sin q (t) j

(425)

All particles of the mechanism may be described this way by considering all
values of r on the interval [0, `]. Since q (t) is the only generalized coordinate required to describe the positions of all the particles in the mechanism,
the mechanism has a single degree of freedom.
Step 3: Associate mass with each particle, keeping in mind
that the total mass in the mechanism must be finite.
Step 4: Compute the accelerations of each particle in the
mechanism, keeping in mind that the unit vectors i, j,
and k do not change with time.
Example 60. If the rod in Example 59 is homogeneous and has total mass
M, then each particle is associated with an infinitesimal mass dm = dr,
in terms of a constant density per unit length and an infinitesimal
length dr. In particular,
Z

rod

dm =

Z `
0

dr = ` = M,

(426)

i.e., = M/`.
Figure 80: An infinitesimal element
R of
mass dm, such that the integral rod dm
equals the total mass of the rod.

dm = dr
r

~r = r cos q (t)i + r sin q (t) j

For each particle on the rod, the acceleration is given by

~a :=

d2~r
=
dt2

r q 2 (t) cos q (t) + q(t) sin q (t) i


r q 2 (t) sin q (t)

where q (t) = dq (t)/dt and q(t) = d2 q (t)/dt2 .

q(t) cos q (t) j,

(427)

lecture notes, me 340

Step 5: For each particle, use a free body diagram to


represent interactions with its environment, including
interactions with other particles in the mechanism,
keeping in mind that the total force acting on the
mechanism must be finite.
Step 6: For each particle, use Newtons 2nd law to equate the
product of mass and acceleration with the total force
acting on the particle.
Step 7: Use physical observations to eliminate unknown
interaction forces from the system of equations
obtained from Newtons 2nd law.
Example 61. Suppose that the rod in Example 59 interacts with its environment through a uniform force field with field strength g and parallel to the y-axis, such that each particle experiences an infinitesimal force
d~Fg = g dm j. In addition, each particle interacts with other particles in
the mechanism to ensure that the rod retains its shape and mass distribution.
Some subset of particles are also interacting with the joint at the origin to
ensure that the rods motion is a rotation about the joint axis.

Figure 81: A free-body diagram including both known and unknown


interactions.

d~Fr

g dm j

d~Fg =
i
z

In the free-body diagram above, the force vector d~Fr denotes an unknown,
infinitesimal reaction force associated with internal interactions within the
rod and/or with the hinge. Newtons 2nd law now becomes
dm ~a = d~Fg + d~Fr

(428)

or, in other words,


r dm q 2 (t) cos q (t) + q(t) sin q (t) i
r dm q 2 (t) sin q (t) q(t) cos q (t) j =

g dm j + d~Fr .

(429)

Physical observations suggest that


k

rod

~r d~Fr = 0,

(430)

147

lecture notes, me 340

i.e., that the total torque of the unknown reaction forces about the hinge
axis vanishes. Then, since
k ~r i =

r sin q (t), k ~r j = r cos q (t),

(431)

the equation
Z

rod

k ~r i

Z
r dm q 2 (t) sin q (t) q(t) cos q (t) k ~r j
rod

Z
Z
~r d~Fr ,
=
g dm k ~r j + k
(432)

r dm q 2 (t) cos q (t) + q(t) sin q (t)

rod

rod

can be simplified to

M
q (t)
`
It follows that

Z `
0

r2 dr =

M
cos q (t) g
`

Z `
0

r dr.

2`
q + g cos q = 0
3
is the equation of motion of the rod.

(433)

(434)

In-class worksheet: a projectile


Consider a body that is able to translate freely, but not rotate. Let
x (t)i + y(t) j + z(t)k denote the position vector of some particle on
the body relative to an inertial coordinate system. Then, for every
particle on the body,

~r = ( x (t) + r x )i + (y(t) + ry ) j + (z(t) + rz )k

(435)

for some particle-dependent constants r x , ry , and rz .


The acceleration equals

~a = x (t)i + y (t) j + z(t)k

(436)

where x (t) = d2 x (t)/dt2 , y (t) = d2 y(t)/dt2 , and z (t) = d2 z(t)/dt2 ,


independently of the particle. Each particle may be associated with
the infinitesimal mass dm, such that
Z

body

dm = M

(437)

equals the total mass of the body.


In the free-body diagram below, d~Fr denotes unknown reaction
forces that ensure that the body retains its shape and mass distribution, and d~Fe denotes all other interaction forces. Newtons 2nd law
then becomes
dm x (t)i + y (t) j + z (t)k = d~Fr + d~Fe .

(438)

148

lecture notes, me 340

x (t)i + y(t) j + z(t)k


j

149

Figure 82: A projectile with three


degrees of freedom.

r x i + ry j + rz k

~r = ( x (t) + r x )i + (y(t) + ry ) j + (z(t) + rz )k


x
i

Figure 83: An infinitesimal mass element.

dm

~r
x

i
z

Figure 84: A free-body diagram including known and unknown forces.

d~Fr

i
z

d~Fe

lecture notes, me 340

150

Physical observations suggest that


Z

body

d~Fr = ~0

(439)

i.e., the net reaction force vanishes. Then,


Z

body

dm x (t)i + y (t) j + z (t)k =

body

d~Fr +

body

d~Fe

(440)

can be simplified to
M x (t)i + y (t) j + z (t)k = ~Fe ,
where

~Fe :=

body

(441)

d~Fe = Fe,x i + Fe,y j + Fe,z k

(442)

is the total (non-reaction) force on the body. It follows that


M x (t) = Fe,x , My (t) = Fe,y , M z (t) = Fe,z

(443)

are the equations of motion of the body.

In-class worksheet: rotor dynamics


For particles on a circular cylinder of radius R and length ` that
is able to rotate about a perpendicular axis A through its center,
let r, j, and z denote cylindrical coordinates relative to a reference
configuration of the disk, and let q (t) be a time-dependent angle of
rotation of the disk relative to the reference configuration.
j

Figure 85: A thin disk constrained to


rotate about an axis through its center.

q (t)
R

Introduce a coordinate system with origin at the center of the


cylinder and z-axis along the axis of rotation, and assume that this is
inertial. For each particle, it follows that

~r = r cos j + q (t) i + r sin j + q (t) j + zk

(444)

lecture notes, me 340

151

and

~a =

r q 2 (t) cos( j + q (t)) + q(t) sin( j + q (t)) i


r q 2 (t) sin( j + q (t)) q(t) cos( j + q (t)) j,

(445)

since r, j, and z are constant.


If the cylinder is homogeneous and has total mass M, then each
particle is associated with an infinitesimal mass dm = r dr dj dz, in
terms of a constant density per unit volume , infinitesimal lengths
dr and dz, and an infinitesimal angle dj. In particular,
Z

cylinder

dm =

Z `/2 Z 2p Z R
`/2 0

r dr dj dz = pR2 ` = M,

(446)

i.e., = M/pR2 `.
j

Figure 86: An infinitesimal mass element.

j + q (t)

dm = r dr dj dz
r

In the free-body diagram shown below, d~Fr denotes unknown


reaction forces that ensure that the disk retains its shape and mass
distribution, as well as constrain it to rotate about A, and d~Fe denotes
all other interaction forces. Newtons 2nd law then becomes

M
r dr dj dz
r q 2 (t) cos( j + q (t)) + q(t) sin( j + q (t)) i
pR2

r q (t) sin( j + q (t)) q (t) cos( j + q (t)) j

= d~Fr + d~Fe .

(447)

Physical observations suggest that


k

cylinder

r d~Fr = 0

(448)

i.e., the net reaction torque about the joint axis vanishes. Since
k (r i) =

r sin( j + q (t)), k (r j) = r cos( j + q (t)),

(449)

lecture notes, me 340

Figure 87: A free-body diagram including both known and unknown


interactions.

d~Fe

d~Fr

it follows that
M
q (t)
pR2

Z `/2 Z 2p Z R
`/2 0

r3 dr dj dz = k

cylinder

+ k
can be simplified to

where

r d~Fr

cylinder

MR2
q (t) = Te
2
Te = k

cylinder

r d~Fe

r d~Fe .

(450)

(451)
(452)

is the total (non-reaction) torque about the joint axis.

In-class worksheet: rolling without slipping


Consider the roller mechanism shown below, whose rolling-withoutslipping motion along the inclined plane is affected by a linear spring
and a uniform gravitational field.
Choose the inertial coordinate system, such that the origin coincides with the center of the disk in the reference configuration
corresponding to the undeformed length of the spring, the inclined
plane is parallel to the x-axis, and the axis of rotation of the disk is
parallel to the z-axis. The position vector to the center of the disk is
then given by
x (t)i
(453)
in terms of the generalized coordinate x (t). For a particle with position vector
r cos ji + r sin j j
(454)
in the reference configuration, in terms of polar coordinates r and j,
the position vector then equals

x (t)
x (t)

~r = x (t) + r cos j
i + r sin j
j.
(455)
R
R

152

lecture notes, me 340

153

Figure 88: A disk rolling without


slipping along an incline in a uniform
gravitational field with strength g and
suspended from a stationary support by
a spring with stiffness K.

K
R
a

The expression x (t)/R for the angle of rotation ensures that the
disk rolls without slipping along the incline. The mechanism has a
single degree of freedom.
j y

Figure 89: The angle of rotation is


proportional to the displacement of the
center of the disk.

x (t)i

k
z

x (t)/R

For an arbitrary particle, the acceleration now equals


x (t)
x (t)
x (t)
x (t)
~a = x (t) r
cos
j
sin
j
i
R
R
R
R2
2

x (t)
x (t)
x (t)
x (t)
r
sin
j
+
cos
j
j.
(456)
R
R
R
R2
where x = dx (t)/dt and x = d2 x (t)/dt2 . Each particle may be
associated with the infinitesimal mass
dm =

M
r dr dj.
pR2

(457)

lecture notes, me 340

154

In the free-body diagram below, d~Fr denotes unknown reaction


forces that ensure that the disk retains its shape and mass distribution, as well as ensures that the disk rolls without slipping along the
inclined plane. Let ~p := R j x (t)i +~r be the position vector from the
point of contact with the inclined plane and an arbitrary particle on
the disk. Physical observations suggest that
k

disk

~p d~Fr = 0,

(458)

i.e., that the net reaction torque about the point of contact with the inclined plane along a direction parallel to the axis of rolling vanishes.
Figure 90: The construction of the
vector ~p from the contact point to
the infinitesimal mass element, and a
superimposed free-body diagram.

d~Fs

~r

x (t)i

dm =

M
r dr dj
pR2

~p
d~Fr

R j
d~Fg
a

Since

x (t)
R r sin j
k
R

x (t)

~p j = r cos j
k,
R

~p i =

the contribution

d~Fg = g dm(sin ai

(459)
(460)

cos a j)

(461)

from the gravitational field to the non-reaction forces satisfies


Z

disk

~p d~Fg
Mg
pR2

Z 2p Z R
0

or, in other words,

disk

r cos j

~p d~Fg =

x (t)
R

a + R sin a r dr dj k

MgR sin a k.

(462)

(463)

lecture notes, me 340

Similarly, the contribution d~Fs from the spring to the non-reaction


forces satisfies
Z

disk

~p d~Fs = R j ( Kx (t)i) = KRx (t)k

(464)

in terms of the spring stiffness K, since this equals the net torque of
the force Kx (t)i applied at the center of the disk about the contact
point.
Finally,
Z

disk

~p dm ~a

M
=
pR3

Z 2p Z R
0

x (t)
r x (t)r cos j
R
0


x (t)
2
2
x (t) r + R + 2rR sin j
dr dj k
R
(465)
2

or, in other words,


Z

disk

~p dm ~a =

3MR x (t)
k.
2

(466)

From Newtons 2nd law


dm ~a = d~Fg + d~Fs + d~Fr ,

(467)

it follows that
3MR x (t)
=
2

MgR sin a + KRx (t)

(468)

corresponding to the equation of motion of the disk.

Exercises
1. Find the equation of motion for a small bead of mass m that is able
to translate along the straight line y = kx + in a two-dimensional
inertial coordinate system under the influence of a spring of undeformed length ` and stiffness K, connecting the bead to the origin.
y

y = kx +
m
K, `
x

155

lecture notes, me 340

2. Find the equation of motion for a small bead of mass m that is able
to translate along the parabolic curve y = bx2 in a vertical inertial
coordinate system under the influence of gravity.

y = bx2

x
3. Find the equation of motion for a small bead of mass m that is
able to translate along a vertical circular curve of radius R that
is spinning about a vertical axis through its center with angular
velocity w. Consider the influence of gravity.
w

m
R

4. Find the equation of motion for a small bead of mass m, suspended in a plane from a taut string that is wrapped around a
vertical pulley of radius R, and moving under the influence of
gravity.

R
g

5. Find the equations of motion for two small beads of mass m1 and
m2 that are able to translate along the x- and y-axis, respectively,
of an inertial coordinate system under the influence of a spring of
undeformed length ` and stiffness K, connecting the two beads.

156

lecture notes, me 340

y
m2

K, `
m1
x

6. Find the equation for the vertical motion of a motor of mass M,


resting on a mechanical suspension with stiffness K and damping
b, due to a rotor imbalance equivalent to a particle of mass m
moving around a vertical circle of radius r with angular frequency
w.
M

m
r

7. Find the equation for a slender ring of mass M and radius R


swinging in a vertical plane about a horizontal axis through a
point on the ring under the influence of gravity and a torque t (t)
about the suspension point.
t (t)

g
R

157

lecture notes, me 340

8. Find the equation of motion for a mechanism consisting of a small


bead of mass m1 that is suspended in a vertical plane at a distance R from a second bead of mass m2 that is able to slide along
a horizontal axis, under the influence of gravity and a spring of
undeformed length ` and stiffness K connecting the second bead
to a fixed location along the axis.
K, `

m2

m1

Solutions
1. Introduce the basis vectors i and j such that the position vector to
the bead equals ~r = x (t)i + (kx (t) + ) j and

~a = x (t)i + k x (t) j.
y
j

~r = x (t)i + (kx (t) + ) j


x

p
The distance to the origin equals k~r k = x2 (t) + (kx (t) + )2 . It
follows that the force of the spring on the bead equals

~Fs =

K (k~r k

where K is the spring stiffness.

`)

~r
k~r k

158

lecture notes, me 340

~Fr

i + k j
j

~Fs
x

In the free-body diagram above, ~Fr denotes a reaction force that


ensures that the bead remains on the straight line. Physical observations suggest that this force is perpendicular to the line, i.e.,

~Fr (i + k j) = 0
since ~r/x (t) = i + k j is tangent to the line. From Newtons 2nd
law,
m~a = ~Fs + ~Fr
or, after dot multiplication with ~r/x (t),

(1 + k 2 )m x (t) =

K (k~r k

`)

(1 + k 2 ) x (t) + k
.
k~r k

2. Introduce the basis vectors i and j such that the force of gravity on
the bead equals ~Fg = mg j and such that the position vector to the
bead equals ~r = x (t)i + bx2 (t) j for some constant b. It follows that

~a = x (t)i + 2b x 2 (t) + x (t) x (t) j.


y
j

~r = x (t)i + bx2 (t) j

In the free-body diagram below, ~Fr denotes a reaction force that


ensures that the bead remains on the parabolic curve. Physical observations suggest that this force is perpendicular to the parabolic
curve at the location of the bead, i.e.,

~Fr i + 2bx (t) j = 0


since ~r/x (t) = i + 2bx (t) j is tangent to the curve at this location.

159

lecture notes, me 340

i + 2bx (t) j

~Fr

~Fg
x

i
From Newtons 2nd law,
m~a = ~Fg + ~Fr
or, after dot multiplication with ~r/x (t),

m x (t) + 4mb2 x (t) x 2 (t) + x (t) x (t) =

2mgbx (t).

3. Let the origin of an inertial coordinate system coincide with the


center of the circle. Introduce the basis vectors i, j, and k such that
the force of gravity on the bead equals ~Fg = mgk and such that

~r = R sin q (t) cos(wt)i + R sin q (t) sin(wt) j

R cos q (t)k

and ~a can be obtained by differentiating twice with respect to t.


k

i
x

q (t)

j
y

y
wt
x

In the free-body diagram below, ~Fr denotes a reaction force that


ensures that the bead remains on the circular curve. Physical observations suggest that this force is perpendicular to the circular
curve at the location of the bead, i.e.,

~Fr R cos q (t) cos(wt)i + R cos q (t) sin(wt) j + R sin q (t)k = 0,


since
~r
= R cos q (t) cos(wt)i + R cos q (t) sin(wt) j + R sin q (t)k
q (t)
is tangent to the circle at this location.

160

lecture notes, me 340

~Fr
m

~Fg

x
From Newtons 2nd law,
m~a = ~Fg + ~Fr
or, after dot multiplication with ~r/q (t),
mR2 q(t)

mR2 w 2 cos q (t) sin q (t) =

mgR sin q (t).

4. The distance between the bead and the point where the string
leaves the disk equals r0 + Rq (t) for some constant r0 and in terms
of the angle q (t) shown in the figure below. It follows that

~r = R cos q (t) + r0 + Rq (t) sin q (t) i

+ R sin q (t)
r0 + Rq (t) cos q (t) j
and the acceleration ~a may be obtained by differentiating twice
with respect to t.

r0 + Rq (t)

R q (t)

q (t)
m
i

In the free-body diagram below, ~Fr denotes a reaction force that


ensures that the bead remains at the end of the taut string. Physical observations suggest that this force is parallel to the string,
i.e.,
~Fr cos q (t)i + sin q (t) j = 0
since this implies that it is perpendicular to the vector from the
center of the pulley to the point where the string leaves the disk.

161

lecture notes, me 340

cos q (t)i + sin q (t) j

~Fr

q (t)

mg j

i
From Newtons 2nd law,
m~a = ~Fr

mg j

or, after dot multiplication with cos q (t)i + sin q (t) j,


mRq 2 (t) + m r0 + Rq (t) q(t) =

mg sin q (t).

5. Introduce the basis vectors i and j such that ~r1 = x (t)i and ~r2 =
y(t) j and, consequently, ~a1 = x (t)i and ~a2 = y (t) j.
y
j

m2

K, `
m1
x

p
The distance between the beads equals k~r2 ~r1 k = x2 (t) + y2 (t).
It follows that the force of the spring on the first bead equals

~F1,s = K k~r2 ~r1 k ` ~r1 ~r2


k~r2 ~r1 k
and that ~F2,s =

~F1,s is the force of the spring on the second bead.


y
j

~F2,r
~F1,s

~F2,s
i

~F1,r
x

In the free-body diagrams above, ~F1,r and ~F2,r denote reaction


forces that ensure that the beads remain on the coordinate axes.

162

lecture notes, me 340

Physical observations suggest that ~F1,r i = 0 and ~F2,r j = 0. From


Newtons 2nd law,
m1~a1 = ~F1,s + ~F1,r , m2~a2 = ~F2,s + ~F2,r
or, after dot multiplication with i and j, respectively:
q

x (t)
2
2
m1 x (t) = K
x (t) + y (t) ` p
2
x ( t ) + y2 ( t )
q

y(t)
m2 y (t) = K
x 2 ( t ) + y2 ( t ) ` p
x 2 ( t ) + y2 ( t )

6. Consider an inertial coordinate system and introduce the basis


vectors i and j such that the displacement of a particle on the
motor is described by the position vector

~r M = r x i + (y(t) + ry ) j
for some constants r x and ry and such that the spring is undeformed when y(t) = 0. Suppose that the rotating imbalance is
located at the top of its circular motion at t = 0 such that its displacement may be described by the position vector

~rm = (rc,x

r sin wt)i + (y(t) + rc,y + r cos wt) j,

for some constants rc,x and rc,y . It follows that

~a M = y (t) j, ~am = rw 2 sin wt i + (y (t)

rw 2 cos wt) j
M

wt r

y
rc,x i + (rc,y + y(t)) j

163

lecture notes, me 340

In the free-body diagram below, ~Fm,r denotes a reaction force that


ensure that the rotating imbalance moves along its circular path
with the desired angular frequency. Similarly, d~FM,r denotes an
infinitesimal reaction force that ensures that the motor retains
its shape and mass distribution, that it translates only vertically
relative to the mechanical suspension and that its axis remains at
the center of the circular path of the rotating imbalance. Physical
observations suggest that
Z

motor

j d~FM,r =

j ~Fm,r

i.e., that the vertical component of the reaction force on the particle
is equal and opposite to the vertical component of the reaction
force on the motor. Finally,
Z

motor

j d~Fs =

Ky(t),

motor

j d~Fd =

by (t)

since these are the net vertical force components acting on the
motor from the spring and damper, respectively.

~Fm,r

dM

d~FM,r
d~Fs

d~Fd
x

i
From Newtons 2nd law,

m ~am = ~Fm,r , dM ~a M = d~FM,r + d~Fs + d~Fd


or, after dot multiplication by j, integration over the volume of the
motor and addition,
m(y (t)

rw 2 cos wt) + ( M

m)y (t) =

Ky(t)

by (t).

164

lecture notes, me 340

7. Introduce basis vectors i, j, and k such that

~r = R cos( j + q (t)) + sin q (t) i + R sin( j + q (t))

cos q (t) j

for some angle j 2 [0, 2p ). The acceleration ~a may be obtained by


differentiating twice with respect to t.
y
j
i
x

R
q (t)
q (t)

In the free-body diagram below, d~Fr denotes an infinitesimal reaction force that ensures that the ring retains its shape and mass
distribution and is restricted to rotary motion about the horizontal
axis. Physical observations suggest that
Z

ring

k ~r d~Fr = 0

i.e., that the net torque of the reaction forces about the axis of
rotation vanishes. With dm = M dj/2p and d~Fg = dm g j, it
further follows that
Z

ring

k ~r d~Fg =

MgR sin q (t).

Finally, by assumption
Z

ring

k ~r d~Ft = t (t).

y
j
i

~r

d~Ft

k
dm
d~Fr
d~Fg

165

lecture notes, me 340

From Newtons 2nd law,


dm ~a = d~Fr + d~Fg + d~Ft
and
or, after cross multiplication with ~r, dot multiplication with k,
integration over the entire ring,
2MR2 q(t) =

MgR sin q (t) + t (t).

8. Introduce basis vectors i and j such that

~r2 = x (t)i, ~r1 = ~r2 + R sin q (t)i

R cos q (t) j

where x (t) = 0 corresponds to the undeformed configuration of


the spring. We obtain the accelerations ~a1 and ~a2 by differentiating
twice with respect to t.
j

~r2

m2

k
R
q (t)
m1
In the free-body diagrams below, ~Fr,1 and ~Fr,2 denotes reaction
forces that ensure that the beads remain at distance R from each
other, that the motion of the first bead remains in the vertical
plane, and the motion of the second bead remains along the axis.

~Fr,2
j

Kx (t)i

m2

k
m2 g j

~Fr,1
m1
m1 g j

Physical observations suggest that


k ~r1

~r2 ~Fr,1 = 0

166

lecture notes, me 340

i.e., the torque of the reaction force on the first bead about the axis
of rotation through the second bead vanishes. Moreover,
i ~Fr,1 + ~Fr,2 = 0
i.e., the horizontal component of ~Fr,2 is equal in magnitude and
opposite in direction to the horizontal component of ~Fr,1 .
From Newtons 2nd law,
m1~a1 =

m1 g j + ~Fr,1 , m2~a2 =

m2 g j

Kx (t)i + ~Fr,2

or, by making use of the physical observations above,


m1 R2 q(t) + m1 R x (t) cos q (t) =

m1 gR sin q (t)

and
m1 Rq(t) cos q (t)

m1 Rq 2 (t) sin q (t) + (m1 + m2 ) x (t) =

Kx (t).

167

lecture notes, me 340

Lesson 9 - Lagrangian dynamics


TAKE-AWAYS:
* Lagrangian dynamics is a systematic formalism for deriving equations of motion for mechanical systems by applying
Lagranges equations to the kinetic energy and generalized
forces.
* The elimination of unknown reaction forces is achieved by
assuming that these do not contribute to the total infinitesimal
work that results from instantaneous infinitesimal changes to
the generalized coordinates, provided that these are independent.
* The left-hand side of Lagranges equations involves computing partial derivatives of the total kinetic energy with respect
to the generalized coordinates and their derivatives, as well as
differentiation with respect to time.

168

lecture notes, me 340

Online lecture: the Lagrangian recipe


Lagrangian dynamics follows an alternative systematic recipe
for deriving equations of motion of a mechanism that are identical
to those obtained using Newtonian dynamics, provided that certain
assumptions are made about the unknown reaction forces.
The first three steps of the recipe are identical to those using Newtonian dynamics.
Step 1: Introduce an inertial (cartesian) coordinate system,
relative to which isolated particles move along straight
paths with constant speed, and let i, j, and k denote
unit vectors along the x-, y-, and z-axes.
Step 2: Formulate a position vector to each particle in the
mechanism. Express its components in terms of
unknown functions of time, known as generalized
coordinates.
Step 3: Associate mass with each particle, keeping in mind
that the total mass in the mechanism must be finite.
A difference between the recipes appears in the fourth step.
Step 4: Compute the total kinetic energy of all particles in the
mechanism.
Example 62. For the rod in Example 59, the velocity of each particle is
given by
d~r
~v :=
= rq (t) sin q (t)i + rq (t) cos q (t) j.
(469)
dt
The integral
1
T :=
2

Mq 2 (t)
~v ~v dm =
2`
rod

Z `
0

r2 dr =

M `2 q 2 (t)
.
6

(470)

equals the total kinetic energy.


While the fifth step is identical to that in the Newtonian recipe,
further differences are apparent in the remaining steps of the recipe.
Step 5: For each particle, use a free body diagram to
represent interactions with its environment, including
interactions with other particles in the mechanism,
keeping in mind that the total force acting on the
mechanism must be finite.

169

lecture notes, me 340

Step 6: For each instantaneous infinitesimal change in a


generalized coordinate, compute the total
infinitesimal work performed by all forces on the
mechanism, assuming that unknown reaction forces do
not contribute to this work. Use the result to compute
the generalized force for each generalized coordinate.
Step 7: Use Lagranges equations to equate various
combinations of derivatives of the kinetic energy with
each of the generalized forces.
If the generalized coordinates and corresponding generalized
forces are denoted by qi (t) and Fi , respectively, for i = 1, . . . , n,
and if the generalized coordinates are independent, then Lagranges
equations are given by

d
T
T
= Fi
(471)
dt q i (t)
qi (t)
for i = 1, . . . , n. Here, q i denotes the derivative of qi (t) with respect
to time. Care should be taken when differentiating the kinetic energy
with respect to qi (t) and q i (t).
Example 63. For the rod in Example 59, an instantaneous infinitesimal
change dq to the value of the generalized coordinate q (t) results in an instantaneous infinitesimal change
d~r =

~r
dq = ( r sin q (t)i + r cos q (t) j) dq
q (t)

(472)

to the particles position vector.

g dm j

d~Fg =
i
z

Figure 91: A free-body diagram including both known and unknown


interactions.

d~Fr

Using the free-body diagram above, and ignoring the contributions of


the unknown reaction forces, we obtain the total infinitesimal work due to

170

lecture notes, me 340

gravity
dW =

rod

Mg cos q (t)
`

d~Fg d~r =

`
0

r dr dq

(473)

or, in other words,

Mg` cos q (t)


dq.
(474)
2
The generalized force Fq corresponding to the generalized coordinate q (t) is
the coefficient of dq in this expression for the total infinitesimal work, i.e.,
dW =

Fq =

Mg` cos q (t)


.
2

According to Lagranges equations

d
T
dt q (t)

T
= Fq
q (t)

(475)

(476)

With the expression for the kinetic energy in Example 62, we obtain

M `2 q 2 (t)
6
q (t)

M `2 q 2 (t)
d M `2 q (t)
M `2 q(t)
=
=
.
q (t)
6
dt
3
3
(477)
The equation of motion becomes
d
dt

M `2 q(t)
=
3

Mg` cos q (t)


2

(478)

or, in other words,

2`q(t)
+ g cos q (t) = 0
3
which is identical to the expression in Example 61.

(479)

In-class worksheet: a projectile


Consider a body that is able to translate freely, but not rotate. Let
x (t)i + y(t) j + z(t)k denote the position vector of some particle on
the body relative to an inertial coordinate system. Then, for every
particle on the body,

~r = ( x (t) + r x )i + (y(t) + ry ) j + (z(t) + rz )k

(480)

for some particle-dependent constants r x , ry , and rz .


The velocity of each particle equals

~v = x (t)i + y (t) j + z (t)k

(481)

171

lecture notes, me 340

x (t)i + y(t) j + z(t)k


j

Figure 92: A projectile with three


degrees of freedom.

r x i + ry j + rz k

~r = ( x (t) + r x )i + (y(t) + ry ) j + (z(t) + rz )k


x
i

where x (t) = dx (t)/dt, y (t) = dy(t)/dt, and z (t) = dz(t)/dt,


independently of the particle. Each particle may be associated with
the infinitesimal mass dm, such that
Z

body

dm = M

(482)

equals the total mass of the body. It follows that the total kinetic
energy equals
T=

1
2

body

~v ~v dm =

M 2
x (t) + y 2 (t) + z 2 (t) .
2

(483)

In the free-body diagram below, d~Fr denotes unknown reaction


forces that ensure that the body retains its shape and mass distribution, and d~Fe denotes all other interaction forces. Instantaneous
infinitesimal changes dx, dy, and dz to the generalized coordinates
x (t), y(t), and z(t) result in the instantaneous infinitesimal change
d~r =

~r
~r
~r
dx +
dy +
dz = dx i + dy j + dz k
x (t)
y(t)
z(t)

(484)

to each particles position vector. The total infinitesimal work performed by the non-reaction forces then equals
Z

Z
~
~
dW =
d Fe d~r =
d Fe d~r
(485)
body

body

or, in other words,


dW = Fe,x dx + Fe,y dy + Fe,z dz,
where

~Fe :=

body

d~Fe = Fe,x i + Fe,y j + Fe,z k

(486)
(487)

is the total (non-reaction) force on the body. The generalized forces


Fx , Fy , and Fz corresponding to the generalized coordinates x (t), y(t),

172

lecture notes, me 340

173

and z(t) are the coefficients of dx, dy, and dz, respectively, in this
expression for the total infinitesimal work, i.e.,
Fx = Fe,x , Fy = Fe,y , Fz = Fe,z

(488)
Figure 93: A free-body diagram including known and unknown forces.

d~Fr

d~Fe

According to Lagranges equations

d
T
T
= Fx ,
dt x (t)
x (t)

d
T
T
= Fy ,
dt y (t)
y(t)

d
T
T
= Fz .
dt z (t)
z(t)

But,

d
dt

M
x (t) 2

x (t)

x (t) + y (t) + z (t)

(489)
(490)
(491)

M 2
x (t) + y 2 (t) + z 2 (t) = M x (t)
2

(492)

and similarly for y(t) and z(t). The equations of motion become
M x (t) = Fe,x , My (t) = Fe,y , M z (t) = Fe,z

(493)

as expected.
The assumption that the unknown reaction forces do not contribute to the generalized forces implies that
Z

Z
~
~
d Fr d~r =
d Fr d~r = 0
(494)
body

body

for arbitrary dx, dy, and dz. This implies that


Z

body

d~Fr = ~0

(495)

which agrees with the physical observation that the net reaction force
should vanish.

lecture notes, me 340

174

In-class worksheet: rotor dynamics


For particles on a circular cylinder of radius R and length ` that
is able to rotate about a perpendicular axis A through its center,
let r, j, and z denote cylindrical coordinates relative to a reference
configuration of the disk, and let q (t) be a time-dependent angle of
rotation of the disk relative to the reference configuration.
j

Figure 94: A thin disk constrained to


rotate about an axis through its center.

q (t)
R
k

Introduce a coordinate system with origin at the center of the


cylinder and z-axis along the axis of rotation, and assume that this is
inertial. For each particle, it follows that

~r = r cos j + q (t) i + r sin j + q (t) j + zk


and

~v =

rq (t) sin( j + q (t))i + rq (t) cos( j + q (t)) j.

(496)
(497)

since r, j, and z are constant.


j

Figure 95: An infinitesimal mass element.

j + q (t)

dm = r dr dj dz
r

If the cylinder is homogeneous and has total mass M, then each


particle is associated with an infinitesimal mass dm = r dr dj dz, in
terms of a constant density per unit volume , infinitesimal lengths
dr and dz, and an infinitesimal angle dj. In particular,
Z

disk

dm =

Z `/2 Z 2p Z R
`/2 0

r dr dj dz = pR2 ` = M,

(498)

lecture notes, me 340

i.e., = M/pR2 `. It follows that the total kinetic energy equals


T :=

1
2

disk

~v ~v dm =

M q 2 (t)
2pR2 `

or, in other words,


T=

`/2 Z 2p Z R

`/2

r3 dr dj dz

MR2 q 2 (t)
.
4

(499)

(500)

In the free-body diagram shown below, d~Fr denotes unknown


reaction forces that ensure that the cylinder retains its shape and
mass distribution, as well as constrain it to rotate about A, and d~Fe
denotes all other interaction forces.
y

Figure 96: A free-body diagram including both known and unknown


interactions.

d~Fe

d~Fr

An instantaneous infinitesimal change dq to the generalized coordinate q (t) results in the instantaneous infinitesimal change
d~r =

r dq sin( j + q (t))i + r dq cos( j + q (t)) j = k ~r dq

(501)

to each particles position vector. Recall the vector identity

~a (~b ~c) = ~b (~c ~a).

(502)

The total infinitesimal work performed by the non-reaction forces


then equals
dW =

disk

d~Fe d~r =

disk

d~Fe (k ~r ) dq

(503)

or, in other words,


dW =

disk

k (~r d~Fe ) dq.

(504)

The generalized force Fq corresponding to the generalized coordinate


q (t) is the coefficient of dq in this expression for the total infinitesimal work, i.e.,
Z
Fq =
k (~r d~Fe )
(505)
disk

175

lecture notes, me 340

which equals the total (non-reaction) torque Te about the joint axis.
According to Lagranges equations

d
T
T
= Fq .
(506)

dt q (t)
q (t)
But,
d
dt

MR2 q 2 (t)
4
q (t)

MR2 q 2 (t)
MR2
=
q ( t ).
q (t)
4
2

(507)

The equation of motion becomes


MR2
q (t) = Te
2

(508)

as expected.
The assumption that the unknown reaction forces do not contribute to the generalized forces implies that
Z

~
~
~r d Fr dq = 0
d Fr d~r = k
(509)
disk

disk

for arbitrary dq. This implies that


k

disk

r d~Fr = 0

(510)

which agrees with the physical observation that the net reaction
torque about the joint axis should vanish.

In-class worksheet: rolling without slipping


Consider the roller mechanism shown below, whose rolling-withoutslipping motion along the inclined plane is affected by a linear spring
and a uniform gravitational field.
Choose the inertial coordinate system, such that the origin coincides with the center of the disk in the reference configuration
corresponding to the undeformed length of the spring, the inclined
plane is parallel to the x-axis, and the axis of rotation of the disk is
parallel to the z-axis. The position vector to the center of the disk is
then given by
x (t)i
(511)
in terms of the generalized coordinate x (t). For a particle with position vector
r cos ji + r sin j j
(512)
in the reference configuration, in terms of polar coordinates r and j,
the position vector then equals

x (t)
x (t)

~r = x (t) + r cos j
i + r sin j
j.
(513)
R
R

176

lecture notes, me 340

177

Figure 97: A disk rolling without


slipping along an incline in a uniform
gravitational field with strength g and
suspended from a stationary support by
a spring with stiffness K.

K
R
a

The expression x (t)/R for the angle of rotation ensures that the
disk rolls without slipping along the incline. The mechanism has a
single degree of freedom.
j y

Figure 98: The angle of rotation is


proportional to the displacement of the
center of the disk.

x (t)i

k
z

x (t)/R

For an arbitrary particle, the velocity now equals

r
x (t)
r

~v = x (t) 1 + sin j
i x (t) cos j
R
R
R

x (t)
j
R

(514)

where x = dx (t)/dt. Each particle may be associated with the infinitesimal mass
M
dm =
r dr dj.
(515)
pR2
from which we obtain the total kinetic energy
T=

1
2

disk

~v ~v dm =

3M 2
x (t).
4

(516)

lecture notes, me 340

178

In the free-body diagram below, d~Fr denotes unknown reaction


forces that ensure that the disk retains its shape and mass distribution, as well as ensures that the disk rolls without slipping along the
inclined plane. Let ~p := R j x (t)i +~r be the position vector from the
point of contact with the inclined plane and an arbitrary particle on
the disk.
Figure 99: The construction of the
vector ~p from the contact point to
the infinitesimal mass element, and a
superimposed free-body diagram.

d~Fs

~r

x (t)i

dm =

M
r dr dj
pR2

~p
d~Fr

R j
d~Fg
a

An instantaneous infinitesimal change dx to the generalized coordinate x (t) results in the instantaneous infinitesimal change

r
x (t)
r
x (t)
d~r = dx 1 + sin j
i dx cos j
j
(517)
R
R
R
R
or, in other words,

dx
R
to each particles position vector. Recall the vector identity
d~r =

(k ~p)

~a (~b ~c) = ~b (~c ~a).


It follows that
d~F d~r =

dx
k (~p d~F ) .
R

(518)

(519)

(520)

Since

x (t)
R r sin j
k
R

x (t)

~p j = r cos j
k,
R

~p i =

and

d~Fg = g dm(sin a i

cos a j)

(521)
(522)

(523)

lecture notes, me 340

it follows that

~p d~Fg =

Mg
r cos j
pR2

x (t)
R

a + R sin a r dr dj

the total infinitesimal work due to gravity then equals


Z

Z
dx

~
~
~p d Fg
dWg =
d Fg d~r = k
R
body
body

(524)

(525)

or, in other words,


dWg = Mg sin a dx
Similarly, the total infinitesimal work due to the spring equals
Z

dx
~p d~Fs
dWs = k
= Kx (t) dx
R
body

(526)

(527)

in terms of the spring stiffness K, since this equals the net work of the
force Kx (t)i applied at the center of the disk. The generalized force
Fx corresponding to the generalized coordinate x (t) is the coefficient
of dx in the expression for the total infinitesimal work, i.e.,
Fx = Mg sin a

Kx (t).

According to Lagranges equations

d
T
T
= Fx
dt x (t)
x (t)

(528)

(529)

or, in other words,


3M
x (t) = Mg sin a
2

Kx (t)

(530)

which agrees with the equation of motion obtained using the Newtonian recipe.
The assumption that the unknown reaction forces do not contribute to the generalized forces implies that
Z

Z
dx

~
~
~p d F
d Fr d~r = k
=0
(531)
R
body
body
for arbitrary dx. This implies that
k

body

~p d~Fr = 0,

(532)

which agrees with the physical observation that the net reaction
torque about the contact point along a direction parallel to the axis of
rotation should vanish.

179

lecture notes, me 340

Exercises
1. Find the equation of motion for a small bead of mass m that is able
to translate along the straight line y = kx + in a two-dimensional
inertial coordinate system under the influence of a spring of undeformed length ` and stiffness K, connecting the bead to the origin.
y

y = kx +
m
K, `
x

2. Find the equation of motion for a small bead of mass m that is able
to translate along the parabolic curve y = bx2 in a vertical inertial
coordinate system under the influence of gravity.
y

y = bx2

x
3. Find the equation of motion for a small bead of mass m that is
able to translate along a vertical circular curve of radius R that
is spinning about a vertical axis through its center with angular
velocity w. Consider the influence of gravity.
w

m
R

4. Find the equation of motion for a small bead of mass m, suspended in a plane from a taut string that is wrapped around a

180

lecture notes, me 340

vertical pulley of radius R, and moving under the influence of


gravity.

R
g

5. Find the equations of motion for two small beads of mass m1 and
m2 that are able to translate along the x- and y-axis, respectively,
of an inertial coordinate system under the influence of a spring of
undeformed length ` and stiffness K, connecting the two beads.
y
m2

K, `
m1
x

6. Find the equation for the vertical motion of a motor of mass M,


resting on a mechanical suspension with stiffness K and damping
b, due to a rotor imbalance equivalent to a particle of mass m
moving around a vertical circle of radius r with angular frequency
w.
M
m
r

181

lecture notes, me 340

7. Find the equation for a slender ring of mass M and radius R


swinging in a vertical plane about a horizontal axis through a
point on the ring under the influence of gravity and a torque t (t)
about the suspension point.
t (t)

g
R

8. Find the equation of motion for a mechanism consisting of a small


bead of mass m1 that is suspended in a vertical plane at a distance R from a second bead of mass m2 that is able to slide along
a horizontal axis, under the influence of gravity and a spring of
undeformed length ` and stiffness K connecting the second bead
to a fixed location along the axis.
K, `

m2

m1

Solutions
1. Introduce the basis vectors i and j such that the position vector to
the bead equals ~r = x (t)i + (kx (t) + ) j and

~v = x (t)i + k x (t) j.
It follows that
T=
is the kinetic energy.

m
m x 2 (t)
~v ~v =
(1 + k 2 )
2
2

182

lecture notes, me 340

y
j

~r = x (t)i + (kx (t) + ) j


x

p
The distance to the origin equals k~r k = x2 (t) + (kx (t) + )2 . It
follows that the force of the spring on the bead equals

~Fs =

K (k~r k

where K is the spring stiffness.

`)

~r
k~r k

~Fr

y
j

~Fs
x

In the free-body diagram above, ~Fr denotes a reaction force that


ensures that the bead remains on the straight line. An instantaneous infinitesimal change dx to the generalized coordinate x (t)
results in the instantaneous infinitesimal change
d~r = dx i + k dx j
to the beads position vector. The total infinitesimal work performed by the spring then equals
dW = ~Fs d~r =

K (k~r k

`)

(1 + k 2 ) x (t) + k
dx.
k~r k

The generalized force Fx corresponding to the generalized coordinate x (t) is the coefficient of dx in this expression for the total
infinitesimal work, i.e.,
Fx =

K (k~r k

`)

(1 + k 2 ) x (t) + k
.
k~r k

According to Lagranges equations

d
T
T
= Fx .
dt x (t)
x (t)

183

lecture notes, me 340

But,
d
dt

m x 2 (t)
(1 + k 2 )
x (t)
2

m x 2 (t)
(1 + k 2 ) = m x (t)(1 + k 2 ).
x (t)
2

The equation of motion becomes


m x (t)(1 + k 2 ) =

K (k~r k

`)

(1 + k 2 ) x (t) + k
.
k~r k

2. Introduce the basis vectors i and j such that the force of gravity on
the bead equals ~Fg = mg j and such that the position vector to the
bead equals ~r = x (t)i + bx2 (t) j for some constant b. It follows that

~v = x (t)i + 2bx (t) x (t) j


and
T=

m
m x 2 (t)
~v ~v =
(1 + 4b2 x2 (t)).
2
2
y

~r = x (t)i + bx2 (t) j

In the free-body diagram below, ~Fr denotes a reaction force that


ensures that the bead remains on the parabolic curve. An instantaneous infinitesimal change dx to the generalized coordinate x (t)
results in the instantaneous infinitesimal change
d~r = dx i + 2bx (t) dx j
to the beads position vector. The total infinitesimal work performed by the non-reaction forces then equals
dW = ~Fg d~r =

2mgbx (t) dx.

The generalized force Fx corresponding to the generalized coordinate x (t) is the coefficient of dx in this expression for the total
infinitesimal work, i.e.,
Fx =

2mgbx (t).

184

lecture notes, me 340

~Fr

~Fg
x

i
According to Lagranges equations

d
T
T
= Fx .
dt x (t)
x (t)
But,
d
dt

m x 2 (t)
1 + 4b2 x2 (t)
x (t)
2

m x 2 (t)
1 + 4b2 x2 (t)
x (t)
2

= m x (t) 1 + 4b2 x2 (t) + 8mb2 x 2 (t) x (t)

4mb2 x (t)2 x (t)

= m x (t) + 4mb2 x (t) x 2 (t) + x (t) x (t)


The equation of motion becomes
m x (t) + 4mb2 x (t) x 2 (t) + x (t) x (t) =

2mgbx (t).

3. Let the origin of an inertial coordinate system coincide with the


center of the circle. Introduce the basis vectors i, j, and k such that
the force of gravity on the bead equals ~Fg = mgk and such that

~r = R sin q (t) cos(wt)i + R sin q (t) sin(wt) j

R cos q (t)k

and ~v can be obtained by differentiating once with respect to t.


After simplification, the kinetic energy equals
T=

mR2 2
q (t) + w 2 sin2 q (t) .
2

i
x

q (t)

j
y

y
wt
x

185

lecture notes, me 340

In the free-body diagram below, ~Fr denotes a reaction force that


ensures that the bead remains on the circular curve. An instantaneous infinitesimal change dq to the generalized coordinate q (t)
results in the instantaneous infinitesimal change

d~r = R cos q (t) cos(wt) i + cos q (t) sin(wt) j + sin q (t) k dq


to the beads position vector. The total infinitesimal work performed by the non-reaction forces then equals
dW = ~Fg d~r =

mgR sin q (t) dq.

The generalized force Fq corresponding to the generalized coordinate q (t) is the coefficient of dq in this expression for the total
infinitesimal work, i.e.,
mgR sin q (t).

Fq =
k

~Fr
m

j
y

~Fg

x
According to Lagranges equations

d
T
T
= Fq .

dt q (t)
q (t)
But,
d
dt

mR2 2
q (t) + w 2 sin2 q (t)
q (t) 2

mR2 2
q (t) + w 2 sin2 q (t)
q (t) 2
= mR2 q(t) mR2 w 2 sin q (t) cos q (t).

The equation of motion becomes


mR2 q(t)

mR2 w 2 sin q (t) cos q (t) =

mgR sin q (t).

186

lecture notes, me 340

4. The distance between the bead and the point where the string
leaves the disk equals r0 + Rq (t) for some constant r0 and in terms
of the angle q (t) shown in the figure below. It follows that

~r = R cos q (t) + r0 + Rq (t) sin q (t) i

+ R sin q (t)
r0 + Rq (t) cos q (t) j
and the velocity ~v may be obtained by differentiating once with
respect to t. After simplification, the kinetic energy equals
T=

m
2
r0 + Rq (t) q 2 (t).
2
r0 + Rq (t)

R q (t)

q (t)
m
i

In the free-body diagram below, ~Fr denotes a reaction force that


ensures that the bead remains at the end of the taut string. An instantaneous infinitesimal change dq to the generalized coordinate
q (t) results in the instantaneous infinitesimal change

d~r = r0 + Rq (t) cos q (t) i + sin q (t) j dq


to the beads position vector. The total infinitesimal work performed by the non-reaction forces then equals
dW = ~Fg d~r =

mg r0 + Rq (t) sin q (t) dq

The generalized force Fq corresponding to the generalized coordinate q (t) is the coefficient of dq in this expression for the total
infinitesimal work, i.e.,
Fq =

mg r0 + Rq (t) sin q (t).

~Fr

q (t)

mg j

187

lecture notes, me 340

According to Lagranges equations

d
T
T
= Fq .

dt q (t)
q (t)
But,
d
dt

m
2
r0 + Rq (t) q 2 (t)

q (t) 2

m
2
r0 + Rq (t) q 2 (t)
q (t) 2

2
= m r0 + Rq (t) q(t) + mR r0 + Rq (t) q 2 (t)

After division by r0 + Rq (t), the equation of motion becomes


m r0 + Rq (t) q(t) + mRq 2 (t) =

mg sin q (t).

5. Introduce the basis vectors i and j such that ~r1 = x (t)i and ~r2 =
y(t) j and, consequently, ~v1 = x (t)i and ~v2 = y (t) j. The total kinetic
energy then equals
T=

m1
m
m
m
~v ~v1 + 2 ~v2 ~v2 = 1 x 2 (t) + 2 y 2 (t).
2 1
2
2
2
y
j

m2

K, `
m1
x

p
The distance between the beads equals k~r2 ~r1 k = x2 (t) + y2 (t).
It follows that the force of the spring on the first bead equals

~F1,s = K k~r2 ~r1 k ` ~r1 ~r2


k~r2 ~r1 k
and that ~F2,s =

~F1,s is the force of the spring on the second bead.


y
j

~F2,r
~F1,s

~F2,s
i

~F1,r
x

188

lecture notes, me 340

In the free-body diagrams above, ~F1,r and ~F2,r denote reaction


forces that ensure that the beads remain on the coordinate axes.
Instantaneous infinitesimal changes dx and dy to the generalized
coordinates x (t) and y(t) result in the instantaneous infinitesimal
changes
d~r1 = dx i, d~r2 = dy j
to the beads position vectors. The total infinitesimal work performed by the non-reaction forces then equals

x (t) dx + y(t) dy
~
~
dW = F1,s d~r1 + F2,s d~r2 = K k~r2 ~r1 k `
.
k~r2 ~r1 k
The generalized forces Fx and Fy corresponding to the generalized
coordinates x (t) and y(t) are the coefficients of dx and dy in this
expression for the total infinitesimal work, i.e.,

x (t)
Fx = K k~r2 ~r1 k `
,
k~r2 ~r1 k

y(t)
Fy = K k~r2 ~r1 k `
.
k~r2 ~r1 k
According to Lagranges equations

d
T
T
= Fx ,
dt x (t)
x (t)

d
T
T
= Fy .
dt y (t)
y(t)
But,
d
dt

x (t)

d
dt

y (t)

m1 2
m2 2
x (t) +
y (t)
2
2

m1 2
m2 2
x (t) +
y (t) = m1 x (t)
x (t) 2
2

and

m1 2
m2 2
x (t) +
y (t)
2
2

m1 2
m2 2
x (t) +
y (t) = m2 y (t)
y(t) 2
2

The equations of motion become


q
m1 x (t) = K
x 2 ( t ) + y2 ( t )
m2 y (t) =

x 2 ( t ) + y2 ( t )

` p

` p

x (t)
x 2 ( t ) + y2 ( t )
y(t)
x 2 ( t ) + y2 ( t )

,
.

189

lecture notes, me 340

6. Consider an inertial coordinate system and introduce the basis


vectors i and j such that the displacement of a particle on the
motor is described by the position vector

~r M = r x i + (y(t) + ry ) j
for some constants r x and ry and such that the spring is undeformed when y(t) = 0. Suppose that the rotating imbalance is
located at the top of its circular motion at t = 0 such that its displacement may be described by the position vector

~rm = (rc,x

r sin wt)i + (y(t) + rc,y + r cos wt) j,

for some constants rc,x and rc,y . It follows that

~v M = y (t) j, ~vm =

rw cos wt i + (y (t)

rw sin wt) j

and the total kinetic energy equals


T=

m
2

y 2 (t) +

m 2
y (t) + r2 w 2
2

2rw y (t) sin wt .


M

wt r

y
rc,x i + (rc,y + y(t)) j

In the free-body diagram below, ~Fm,r denotes a reaction force that


ensure that the rotating imbalance moves along its circular path
with the desired angular frequency. Similarly, d~FM,r denotes an
infinitesimal reaction force that ensures that the motor retains its
shape and mass distribution, that it translates only vertically relative to the mechanical suspension and that its axis remains at the

190

lecture notes, me 340

center of the circular path of the rotating imbalance. An instantaneous infinitesimal change dy to the generalized coordinate y(t)
results in the instantaneous infinitesimal changes
d~r M = d~rm = dy j
to the position vectors to a particle on the motor and to the rotating imbalance. The total infinitesimal work performed by the
non-reaction forces then equals
dW =

motor

d~Fs + d~Fd d~r M =

by (t) dy

Ky(t)

The generalized force Fy corresponding to the generalized coordinate y(t) is the coefficient of dy in this expression for the total
infinitesimal work, i.e.,
Fy =

Ky(t)

by (t).

~Fm,r

dM

d~FM,r
d~Fs

d~Fd
x

i
According to Lagranges equations

d
T
T
= Fy .
dt y (t)
y(t)
After simplification, the left-hand side equals
My (t)

mrw 2 cos wt.

The equation of motion becomes


M y (t)

mrw 2 cos wt =

Ky(t)

by (t)

191

lecture notes, me 340

7. Introduce basis vectors i, j, and k such that

~r = R cos( j + q (t)) + sin q (t) i + R sin( j + q (t))

cos q (t) j

for some angle j 2 [0, 2p ). The velocity ~v may be obtained by


differentiating once with respect to t. After simplification, the
kinetic energy equals
T=

MR2 q 2 (t)
2p

since dm = M dj/2p.

Z 2p
0

sin j

1) dj = MR2 q 2 (t).

y
j
i
x

R
q (t)
q (t)

In the free-body diagram below, d~Fr denotes an infinitesimal reaction force that ensures that the ring retains its shape and mass
distribution and is restricted to rotary motion about the horizontal
axis. An instantaneous infinitesimal change dq to the generalized
coordinate q (t) results in the instantaneous infinitesimal changes

d~r = R
sin j + q (t) + cos q (t) dq i

+ R cos j + q (t) + sin q (t) dq j = k ~r dq

to the position vector to a particle on the ring. Since d~Fg = dm g j,


the contribution to the infinitesimal work due to gravity then
equals
dWg =

ring

=
=

d~Fg d~r
Z 2p

Mg
R cos j + q (t) + sin q (t) dj dq
2p
0

MgR sin q (t) dq

Similarly, by assumption, the contribution to the infinitesimal work


due to the applied torque equals
dWt =

ring

d~Ft d~r = k

ring

~r d~Ft dq = t (t) dq

192

lecture notes, me 340

The generalized force Fq corresponding to the generalized coordinate q (t) is the coefficient of dq in the expression for the total
infinitesimal work, i.e.,
MgR sin q (t) + t (t).

Fq =
y
j

i
x

d~Ft

~r
k
dm

d~Fr
d~Fg
According to Lagranges equations

d
T
T
= Fq ,

dt q (t)
q (t)
i.e.

2MR2 q(t) =

MgR sin q (t) + t (t).

8. Introduce basis vectors i and j such that

~r2 = (` + x (t))i, ~r1 = ~r2 + R sin q (t)i

R cos q (t) j

where x (t) = 0 corresponds to the undeformed configuration of


the spring. We obtain the velocities ~v1 and ~v2 by differentiating
once with respect to t. In particular, the total kinetic energy equals
T=

m1 + m2 2
m R2
x (t) + 1 q 2 (t) + m1 R cos q (t) x (t)q (t).
2
2
j

~r2

m2

k
R
q (t)
m1

193

lecture notes, me 340

In the free-body diagrams below, ~Fr,1 and ~Fr,2 denotes reaction


forces that ensure that the beads remain at distance R from each
other, that the motion of the first bead remains in the vertical
plane, and the motion of the second bead remains along the axis.

~Fr,2
j

Kx (t)i

m2

k
m2 g j

~Fr,1
m1
m1 g j

Instantaneous infinitesimal changes dx and dq to the generalized


coordinates x (t) and q (t) result in the instantaneous infinitesimal
changes
d~r2 = dx i, d~r1 = d~r2 + R cos q (t) dq i + R sin q (t) dq j
to the beads position vectors. The total infinitesimal work performed by the non-reaction forces then equals
dW =

Kx (t) dx

m1 gR sin q (t) dq.

The generalized forces Fx and Fq corresponding to the generalized


coordinates x (t) and q (t) are the coefficients of dx and dq in this
expression for the total infinitesimal work, i.e.,
Fx =

Kx (t), Fq =

m1 gR sin q (t).

According to Lagranges equations

d
T
T
= Fx ,
dt x (t)
x (t)

d
T
T
= Fq .
dt q (t)
q (t)
But,

T
x (t)
T
q (t)
T
x (t)
T
q (t)

= (m1 + m2 ) x (t) + m1 R cos q (t)q (t),


= m1 R2 q (t) + m1 R cos q (t) x (t),
= 0,
=

m1 R sin q (t) x (t)q (t),

194

lecture notes, me 340

The equations of motion become

(m1 + m2 ) x (t)
and

m1 R sin q (t)q 2 (t) + m1 R cos q (t)q(t) =

m1 R2 q(t) + m1 R cos q (t) x (t) =

m1 gR sin q (t).

Kx (t)

195

lecture notes, me 340

Lesson 10 - Linearization
TAKE-AWAYS:
* Systems that violate the principle of superposition are nonlinear. Such systems may be linearized about an operating
point by substituting first-order Taylor expansions for of all
nonlinear terms. The linearization provides an approximate
description of the dynamics for as long as the system state is
close to the operating point.
* Equilibria correspond to constant signals for all system variables. Changes in the number of stability of equilibria are bifurcations.
* The linearization about an equilibrium may predict the
stability of the equilibrium. In particular, an equilibrium is
asymptotically stable if small perturbations decay to zero as
t ! , and unstable if some perturbations (no matter how
small) grow without bound as t ! .
* For lightly damped mechanical systems with more than one
degree of freedom, the linearization about asymptotically stable equilibria can be understood in terms of modal oscillations. These describe transient small oscillations about the
equilibrium in the nonlinear problem, as well as resonance
to harmonic excitation.

196

lecture notes, me 340

Online lecture: stability of equilibria


Dynamic systems that do not satisfy the principle of superposition are nonlinear. While we are, in principle, able to understand
linear systems completely, our understanding of all but a few exceptional nonlinear systems is very limited.
Example 64. Consider a small bead of mass m that is able to translate along
a vertical circular curve of radius R shown below. Assume that the curve is
spinning about a vertical axis through its center with angular velocity w,
and that motion takes place in a uniform gravitational field with acceleration
g and in the presence of a viscous damping force proportional to the angular
speed of the bead along the curve.
w

Let q (t) denote the angle between the ray from the center of the circle
to the bead and the vertical axis, as illustrated below, and define q (t) =
dq/dt and q(t) = d2 q/dt2 . Using either Newtons 2nd law or Lagranges
equations, we obtain the equation of motion
mRq(t)

mRw 2 sin q (t) cos q (t) =


z

mg sin q (t)

(533)

m
q (t)
x

bq (t).

m
y

y
wt
x

This equation is nonlinear due to the presence of the trigonometric functions sin q (t) and cos q (t). It is not true that aq (t) is a solution if q (t) is a
solution, in violation of the principle of superposition.

197

lecture notes, me 340

We linearize a nonlinear differential equation about an operating point by replacing every nonlinear term by a first-order Taylor
expansion in small deviations from the operating point.
Example 65. To linearize the nonlinear differential equation
mRq(t)

mRw 2 sin q (t) cos q (t) =

bq (t)

mg sin q (t)

(534)

about the operating point 0, assume that q (t) := #d(t) for some function
d(t) and 0 < # 1. It follows that
q (t) = #d(t), q(t) = #d(t),

(535)

where d(t) = dd(t)/dt and d(t) = d2 d(t)/dt2 . Moreover, to first-order in


#,
cos q (t) = cos #d(t) cos 0 + ( sin 0)#d(t) = 1
(536)
and, similarly,

sin q (t) = sin #d(t) sin 0 + (cos 0)#d(t) = #d(t).


In this case, the linearized differential equation becomes

# mRd(t) mRw 2 d(t) = #


mgd(t)

bd(t)

(537)

(538)

or, in other words,


mRd(t) + bd(t) + m( g

Rw 2 )d(t) = 0,

(539)

since # 6= 0.
Analysis of the linearized problem gives insight into the original
nonlinear problem, as long as the linearization remains valid.

Example 66. If D(s) := L d(#) (s), then the linearized differential


equation
mRd(t) + bd(t) + m( g Rw 2 )d(t) = 0,
(540)
is equivalent to the frequency-domain representation
D(s) =

mRd(0)s + bd(0) + mRd(0)


mRs2 + bs + m( g Rw 2 )

(541)

The free response of this dynamic system equals a combination of the unit
impulse response h(t) corresponding to the transfer function
H (s) =

mRs2

1
+ bs + m( g

Rw 2 )

and its distributional derivative Dh(t)/dt. Here,

g1 #
g2 #
h(t) = e e
( t ),

(542)

(543)

198

lecture notes, me 340

where
g1,2 =

b2

4m2 R( g
2mR

Rw 2 )

(544)

are poles of H (s).


Suppose that q (0) and q (0) are both small. Provided that w 2 < g/R,
the system is either underdamped, overdamped, or critically damped. In each
case, q (t) remains small and q (t) ! 0 as t ! , i.e., the linearization
continues to be valid for all t. The configuration with q (t) = 0 for all t is an
asymptotically stable equilibrium.
If, instead, w 2 > g/R, then one of the poles is positive, and |q (t)| grows
exponentially fast for almost any small initial values q (0) and q (0). In
this case, the linearization fails to be valid after some short time, and the
configuration with q (t) = 0 for all t is an unstable equilibrium.

In-class worksheet: bifurcations


A nonlinear system may have more than one equilibrium configuration. Linearization about each equilibrium may determine its
stability. A change in stability or in the number of equilibria is a bifurcation.
Consider again the small bead translating along the spinning
vertical circular curve shown below. Here, an equilibrium configuration corresponds to a position of the bead on the curve that does
not change in time, even as the curve rotates about the vertical axis
through its center.
w

m
R

Equilibrium configurations for the corresponding nonlinear differential equation


mRq(t)

mRw 2 sin q (t) cos q (t) =

mg sin q (t)

bq (t)

(545)

correspond to time histories q (t) = q , q (t) = q(t) = 0 for some


constant q 2 ( p, p ]. It follows that
Rw 2 sin q cos q = g sin q

(546)

199

lecture notes, me 340

i.e., either sin q = 0 or cos q = g/Rw 2 . In the first case, q =


0 or p, independently of the value of w. In the second case, q =
arccos( g/Rw 2 ), provided that w 2 > g/R.
The figure below shows a corresponding bifurcation diagram.
Each point in the diagram represents an equilibrium configuration
for some value of w. In particular, we see that there are four equilibria when w 2 > g/R, but only two when w 2 < g/R. The bifurcation at
w 2 = g/R is known as a pitchfork bifurcation because of the shape
of the graphs.

Figure 100: A pitchfork bifurcation


occurs when w 2 = g/R.

3p
4
p
2
p
4

arccos( g/Rw 2 )
w2
g/R

p
4
p
2

To linearize the governing differential equation about an equilibrium operating point q , let q (t) = q + #d(t). It follows that
q (t) = #d(t), q(t) = #d(t)

(547)

Moreover, to first order in #,


cos q (t) = cos q + #d(t) cos q

sin q #d(t)

(548)

sin q (t) = sin q + #d(t) sin q + cos q #d(t),

(549)

sin q (t) cos q (t) sin q cos q + cos 2q #d(t).

(550)

and
The linearized differential equation becomes

mRw 2 cos q sin q + # mRd(t) mRw 2 cos 2q d(t)

= mg sin q + #
mg cos q d(t)

bd(t)

(551)

200

lecture notes, me 340

201

or, after elimination using (546),


mRd(t) + bd(t) + m( g cos q

Rw 2 cos 2q )d(t) = 0.

(552)

The equilibrium configuration at q = 0 was analyzed in the


previous example. In particular, it was found to be asymptotically
stable when w 2 < g/R and unstable, otherwise. In contrast, the
equilibrium configuration at q = p is always unstable.
If cos q = g/Rw 2 , the coefficient in front of d(t) equals
g cos q

Rw 2 cos 2q =

or, in other words,


Rw

g2
Rw 2 2 2 4
R w

g2
Rw 2

g2
R2 w 4

(553)

(554)

which is always positive when w 2 > g/R. Thus, when it exists, the
corresponding equilibrium configuration is asymptotically stable. In
the figure below, we modify the bifurcation diagram to use dashed
curve to represent unstable equilibria.

Figure 101: Solid curves represent stable equilibria. Dashed curves represent
unstable equilibria.

3p
4
p
2
p
4

arccos( g/Rw 2 )
w2
g/R

p
4
p
2

In-class worksheet: a two-degree-of-freedom oscillator


Consider a mechanism consisting of a small bead of mass m1 that
is suspended in a vertical plane at a distance R from a second bead
of mass m2 , which in turn is able to slide along a horizontal axis, as

lecture notes, me 340

shown below. Suppose that motion takes place under the influence
of a uniform gravitational field with acceleration g, a spring of undeformed length ` and stiffness K connecting the second bead to a fixed
location along the axis, and a viscous damping force acting on the
second bead.
K, `

m2

m1
Let x (t) denote displacements of the second bead from the undeformed configuration of the spring, and let q (t) denote the angular
displacement of the first bead from a position immediately beneath
the second bead. Suppose that the units of mass, length, and time are
chosen so that m1 = 1, m2 = 1/, g = , and R = 1. The equations of
motion become

(1 + 1/) x (t)
and

sin q (t)q 2 (t) + cos q (t)q(t) =


q(t) + cos q (t) x (t) =

Kx (t)

b x (t)

sin q (t),

(555)

(556)

where b is the damping coefficient, x (t) = dx (t)/dt, x (t) = d2 x (t)/dt2 ,


q (t) = dq (t)/dt, and q(t) = d2 q (t)/dt2 . These equations are nonlinear due to the presence of the trigonometric functions in q (t) and the
square of the angular velocity.
Equilibrium configurations correspond to x (t) = x , q (t) = q ,
x (t) = x (t) = q (t) = q(t) = 0. This implies that x = 0 and sin q = 0.
To linearize about an equilibrium, let
x (t) = #dx (t), q (t) = q + #dq (t).

(557)

It follows that
x (t) = #dx (t), x (t) = #dx (t), q (t) = #dq (t), q(t) = #dq (t).

(558)

Moreover, to first order in #,


cos q (t) cos q

sin q #d(t) = 1,

sin q (t) sin q + cos q #d(t) = #dq (t),


q 2 (t) 0,

(559)
(560)
(561)

202

lecture notes, me 340

203

where the upper and lower signs refer to q = 0 and q = p, respectively. The linearized differential equations become

(1 + 1/)dx (t) dq (t) = Kdx (t)


dq (t) dx (t) = dq (t).

bdx (t)

(562)
(563)

or, after solving for dx (t) and dq (t),

dx (t) = dq (t) bdx (t) Kdx (t)


dq (t) = Kdx (t) bdx (t) (1 + )dq (t)

(564)
(565)

An equivalent frequency-domain block diagram is shown below.

Figure 102: A block diagram representation of the linearization about an


equilibrium configuration.

dx (0)

dx (0)
1
s

1
s

D x (s)

+
dq (0)

1
s

+
+

1
s

Dq (s)

dq (0)
1+

Here, D x (s) := L dx (#) (s) and Dq (s) := L dq (#) (s) satisfy the
matrix equation
!
D x (s)
A(s)
= B ( s ),
(566)
Dq (s)
where
A(s) =
and
B(s) =

s2 + bs + K
(bs + K )

s2 (1 + )

(s + b)dx (0) + dx (0)


sdq (0) + dq (0) bdx (0)

(567)

(568)

lecture notes, me 340

It follows that

D x (s)
Dq (s)

=A

( s ) B ( s ),

(569)

where
A

(s) =

(s2 (1 + ))/ | A(s)|


/ | A(s)|
2
(bs + K )/ | A(s)| (s + bs + K )/ | A(s)|

(570)

and the determinant

| A(s)| = s4 + bs3 + K (1 + ) s2 bs K.

(571)

The stability of either equilibrium configuration is clearly determined by the unit impulse response h(t) corresponding to the
transfer function
1
H (s) =
(572)
| A(s)|
since dx (t) and dq (t) are linear combinations of h(t) and its distributional derivatives. These decay exponentially as t ! provided
that all poles of H (s) have negative real part, and grow unbounded
exponentially fast if any pole has positive real part.

Exercises
1. Linearize the function f ( x ) = x/(1 + x2 ) about the operating point
x = 3.
p
2. Linearize the function g( x, q ) = 1 + x2 (q sin q ) about the
operating point ( x , q ) = (1, p/4).
3. The bead in the figure below slides along the straight line y =
kx + .
y

y = kx +
m
K, `
x

The corresponding equation of motion is


m x (t)(1 + k 2 ) =

K (k~r k

`)

(1 + k 2 ) x (t) + k
,
k~r k

p
where k~r k = x2 (t) + (kx (t) + )2 . Find all equilibrium configurations, and draw a bifurcation diagram with ` along the horizontal axis for the case that = 1, k = 0.

204

lecture notes, me 340

4. Linearize the nonlinear equation in the previous exercise about


the equilibrium configuration that exists for all values of `, and
describe the dynamics of the linear system.
5. The bead in the figure below is suspended from the taut string
that is wrapped around a vertical pulley.

R
g

The corresponding equation of motion is


m r0 + Rq (t) q(t) + mRq 2 (t) =

mg sin q (t).

Find all physically realizable equilibrium configurations and describe the corresponding linearized dynamics.
6. Show that one of the roots of the polynomial
s2 + bs + K
bs + K

s2

has positive real part when b > 0. What is the implication to


the stability of the equilibrium configuration with q = p in the
two-degree-of-freedom mechanism in the second worksheet?
7. Show that all roots of the polynomial
s2 + bs + K
bs K

s2

+1+

have negative real part when b > 0. What is the implication to


the stability of the equilibrium configuration with q = 0 in the
two-degree-of-freedom mechanism in the second worksheet?
8. Describe the linearized dynamics near q = 0 in the two-degreeof-freedom mechanism in the second worksheet in the case that
= K = 1 and b = 0.

Solutions
1. Let x = 3 + #d for # 1. By Taylors theorem, to first order in #,
f (3 + #d) f (3) + # f 0 (3)d

205

lecture notes, me 340

But, f (3) = 3/10 and


f 0 (x) =
It follows that
f (x)

1 x2
(1 + x 2 )2

3
10

2
(x
25

3)

2. Let x = 1 + #dx and q = p/4 + #dq . By Taylors theorem, to first


order in #,

g
g
g(1 + #dx , p/4 + #dq ) g(1, p/4) + #
(1, p/4)dx + (1, p/4)dq
x
q
But,

p
g(1, p/4) = p
2 2

and
g
x
( x, q ) = p
(q
x
1 + x2

p
g
( x, q ) = 1 + x2 (1
q

sin q ),

It follows that
p
g( x, q ) p
2 2

1+

p
p

4 2

1
(x
2

1) +

cos q )

p
4

3. An equilibrium configuration corresponds to x (t) = x , x (t) =


x (t) = 0 for some constant x . Here,
K

x 2 + (kx + )2

if either (1 + k 2 ) x + k = 0 or
case,
x =

(1 + k 2 ) x + k
p
=0
x 2 + (kx + )2

x 2 + (kx + )2 = `. In the first


k
1 + k2

independently of the value of `. In the second case,


p
k (1 + k 2 )`2 2
x =
1 + k2
p
as long as ` > ||/ 1 + k 2 . A bifurcation diagram for the case
when = 1 and k = 0 is shown below

206

lecture notes, me 340

q
2
1

`2

`2
1

2
4. Let
x (t) =
so that

k
+ #d(t)
1 + k2

x (t) = #d(t), x (t) = #d(t).

Then, to first order in #,


q

x2 (t) + (kx (t) + )2

2
1 + k2

After simplification, the linearized differential equation becomes


s

1 + k2

md(t) + K 1
` d(t) = 0
2
p
As long as ` < ||/ 1 + k 2 , all solutions to this equation are
harmonic signals with angular frequency
v
s
u

uK
1 + k2
t
1
`
.
m
2

p
In contrast, for ` > ||/ 1 + k 2 , |d(t)| grows exponentially fast for
almost all initial conditions.
5. An equilibrium configuration corresponds to q (t) = q , q (t) =
q(t) = 0 for some constant q . Here, sin q = 0 implies that
q = 0 or p. A moments reflection shows that only the former is
a physically realizable equilibrium in which the string is taut and
able to provide a force that balances that of gravity.
To linearize the equation of motion about the equilibrium configuration, let q (t) = #d(t), such that
q (t) = #d(t), q(t) = #d(t).

207

lecture notes, me 340

To first order in #, the linearized differential equation becomes


mr0 d(t) + mgd(t) = 0
All solutions are harmonic signals with angular frequency

g/r0 .

6. As long as no two roots coincide, they depend smoothly on the


damping coefficient b. When b = 0, the roots satisfy the equation
s4 + K

s2

i.e., s = jw or s = g, where
1

w2 =
and

+ K +

( 1
2

K = 0

+ K )2 + 4K

(1 + K )2 + 4K
2
are both positive. It follows that there exists one real positive root
when b = 0. As b increases, this can become negative only if
it crosses 0, but 0 is never a root for any b. It follows that one
root has positive real part when b > 0 and the corresponding
equilibrium configuration is unstable.
2

g =

1+

K +

7. Assume that no two roots coincide. Then, if s(b) denotes the dependence of a root on b, then s(0) satisfies the equation
s(0)4 + K + 1 + s(0)2 + K = 0
i.e., s(0) = jw, where
2

w =

K + 1 +

(K + 1 + )2
2

4K

are both positive. There are purely imaginary roots only if b = 0.


Implicit differentiation of A s(b)
substitution of b = 0, yields
s 0 (0) =

= 0 with respect to b and

1 + s2 (0)
2 1 + + K + 2s2 (0)

which is purely real. After substitution of s(0) and simplification,

1 + + K
s 0 (0) =
1+ p
,
4
(1 + + K )2 4K
which is always negative, since

(1 + + K )2

4K

( 1 + + K )2 = 4

It follows that all roots have negative real part when b > 0 and the
corresponding equilibrium configuration is asymptotically stable.

208

lecture notes, me 340

8. In the frequency domain, the linearized dynamics are described by


the matrix equation
!
D x (s)
A(s)
= B ( s ),
Dq (s)
where
A(s) =
and
B(s) =

1
s2 + 2

sdx (0) + dx (0)


sdq (0) + dq (0)

s2 + 1
1

The matrix A(s) has two formal eigenvalues of the form s2 + w 2 ,


where
q
p
w = (3 5)/2

T
For the corresponding eigenvectors s t
,

p
1 5
s=
t.
2

For arbitrary initial conditions, the free response


q is apsum of
modal oscillations with angular frequencies (3 + 5)/2 and
q
p
(3
5)/2, respectively. The corresponding mode shapes are
p
p
1
5 : 2 and 1 + 5 : 2 and correspond to harmonic signals in
each degree of freedom that differ in phase by p or are identical
in phase, respectively. Such a sum of modal oscillations is typically not periodic, since the angular frequencies are not rationally
related.

209

You might also like