Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter
Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter
Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter
http://ebooksecure.com/product/ebook-pdf-differential-
equations-4th-edition/
https://ebooksecure.com/download/linear-algebra-and-partial-
differential-equations-ebook-pdf/
http://ebooksecure.com/product/ebook-pdf-elementary-differential-
equations-10th-edition/
https://ebooksecure.com/download/schaums-outline-of-differential-
equations-ebook-pdf/
First Course in differential equations (11ed) /
Differential Equations and Boundary Value Problems
(9ed) Solutions manual 9th Edition - eBook PDF
https://ebooksecure.com/download/first-course-in-differential-
equations-11ed-differential-equations-and-boundary-value-
problems-9ed-solutions-manual-ebook-pdf/
http://ebooksecure.com/product/ebook-pdf-fundamentals-of-
differential-equations-9th-edition/
http://ebooksecure.com/product/fundamentals-of-differential-
equations-8th-edition-ebook-pdf/
http://ebooksecure.com/product/ebook-pdf-differential-equations-
and-linear-algebra-3rd-edition/
http://ebooksecure.com/product/ebook-pdf-differential-equations-
and-linear-algebra-4th-edition-2/
Advanced Differential
Equations
Youssef N. Raffoul
Professor of Mathematics
University of Dayton
Dayton, OH, United States
ISBN: 978-0-323-99280-0
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
CHAPTER 1 Preliminaries and Banach spaces . . . . . . . . . . . . . . . . 1
1.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Escape velocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Applications to epidemics . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Metrics and Banach spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.5 Variation of parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5.1 RC circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.6 Special differential equations . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
CHAPTER 2 Existence and uniqueness . . . . . . . . . . . . . . . . . . . . . . . 27
2.1 Existence and uniqueness of solutions . . . . . . . . . . . . . . . . . . 27
2.2 Existence on Banach spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.3 Existence theorem for linear equations . . . . . . . . . . . . . . . . . . 44
2.4 Continuation of solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.5 Dependence on initial conditions . . . . . . . . . . . . . . . . . . . . . . 48
2.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
CHAPTER 3 Systems of ordinary differential equations . . . . . . . . 57
3.1 Existence and uniqueness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2 x = A(t)x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.2.1 Fundamental matrix . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.2.2 x = Ax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.2.3 Exponential matrix eAt . . . . . . . . . . . . . . . . . . . . . . . . 82
3.3 x = A(t)x + g(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
CHAPTER 4 Stability of linear systems . . . . . . . . . . . . . . . . . . . . . . . 103
4.1 Definitions and examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.2 x = A(t)x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.3 Floquet theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4.3.1 Mathieu’s equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
4.3.2 Applications to Mathieu’s equation . . . . . . . . . . . . . . . 125
4.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
CHAPTER 5 Qualitative analysis of linear systems . . . . . . . . . . . . 129
5.1 Preliminary theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.2 Near-constant systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.3 Perturbed linear systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
vii
Student Resources
For the Partial Solutions Manual, visit the companion site:
https://www.elsevier.com/books-and-journals/book-companion/9780323992800
xi
strong nonlinearities not treatable by other methods. More recently, the concept of
the Lyapunov exponent related to Lyapunov’s first method of discussing stability has
received wide interest in connection with chaos theory.
Chapter 1 deals with various introductory topics, including variation of parame-
ters formula, metric spaces, and Banach spaces.
In Chapter 2 we introduce Gronwall’s inequality that we make use of to prove the
uniqueness of solutions. We introduce theorems on the existence and uniqueness of
solutions, their dependence on initial data, and their continuation on maximal inter-
val.
In Chapter 3 we introduce systems of differential equations. We briefly discuss
how the existence and uniqueness theorems of Chapter 2 are extended to suit systems.
Then we develop the notion of the fundamental matrix as a solution and utilize it to
write solutions of non-homogeneous systems so that they can be analyzed.
Stability theory is the central part of this book. Chapters 4–8 are totally devoted
to stability. In Chapter 4 we are mainly concerned with the stability of linear systems
via the variation of parameters. The chapter also includes a nice section of Floquet
theory with its application to Mathieu’s equation.
Chapters 5 and 6 are deeply devoted to the study of the stability of linear systems,
near-linear systems, perturbed systems, autonomous systems in the plane, and stabil-
ity by linearization. Chapter 5 is ended with the study of Hamiltonians and gradient
systems. We begin Chapter 6 by looking at stability diagrams in scalar equations and
move into the study of bifurcations as it naturally arises while looking at stability.
Bifurcation occurs when the dynamics abruptly change as certain parameters move
across certain values. We end Chapter 6 by considering stable and unstable man-
ifolds, which then delves into the Hartman–Grobman theorem. The theorem says
that the behavior of a dynamical system in the domain near a hyperbolic equilibrium
point is qualitatively the same as the behavior of its linearization near this equilibrium
point, where hyperbolicity means that no eigenvalue of the linearization has a zero
real part.
Chapter 7 delves deeply into the stability of general systems using Lyapunov
functions. We prove general theorems regarding the stability of autonomous and non-
autonomous systems by assuming the existence of such Lyapunov function. We touch
on the notion of ω-limit set and its correlation to Lyapunov functions. The chapter is
concluded with a detailed discussion on exponential stability.
Chapter 8 is solely devoted to the study of delay differential equations. It con-
tains recent development in the research of delay differential equations. We begin
the chapter by pointing out how basic results from ordinary differential equations
are easily extended to delay differential equations. We introduce the method of steps
and show how to piece together a solution. The transition of moving from ordinary
differential equations to delay differential equations was made simple through the
extension of Lyapunov functions to Lyapunov functionals. Then we move on to
a whole new concept, fixed point theory. The use of fixed point theory alleviates
some of the difficulties that arise from the use of Lyapunov functionals when study-
ing stability. Later on, we apply fixed point theory to the study of stability and the
existence of positive periodic solutions of neutral differential equations and neutral
Volterra integro-differential equations, respectively. We end the chapter with the use
of Lyapunov functionals to obtain necessary conditions for the exponential stability
of Volterra integro-differential equations with finite delay.
Chapter 9 deals with current research concerning the use of a new variation of
parameters formula. The objective is to introduce a new method for inverting first-
order ordinary differential equations with time-delay terms to obtain a new variation
of parameters formula that we use to study the stability, boundedness, and periodicity
of general equations in ordinary and delay differential equations.
A combination of Chapters 1–3, 6, and 7 can be used to deliver a course on non-
linear systems for engineers.
The author has not attempted to give the historical origin of the theory, except
in very rare cases. This resulted in the situation that not every reference listed in
References is mentioned in the text or the body of the book.
Exercises play an essential learning tool of the course and accompany each chap-
ter. They range from routine calculations to solving more difficult problems to open-
ended ones. Students must read the relevant material before attempting to do the
exercises.
I am indebted to Dr. Mohamed Aburakhis, who fully developed all the codes
for all the figures in the book. I like to thank the hundreds of graduate students
at the University of Dayton whom the author taught for the last 20 years and who
helped the polishing and refining of the lecture notes, most of which have become
this book.
A heartfelt appreciation to Jeff Hemmelgarn from the University of Dayton for
carefully reading the whole book and pointing out many typos.
Youssef N. Raffoul
University of Dayton
Dayton, Ohio
June 2021
1.1 Preliminaries
Let R denote the set of real numbers, let I be an interval in R, and consider a function
x : I → R. We say the function x has a derivative at a point t ∗ ∈ I if the limit
x(t) − x(t ∗ )
lim
t∈I, t→t ∗ t − t∗
x(t) − x(t ∗ )
x (t ∗ ) = lim ,
t∈I, t→t ∗ t − t∗
between the independent variable t and the dependent variable x, and its derivatives
dx d 2 x d 3 x d nx
, 2 , 3 , ..., n .
dt dt dt dt
Loosely speaking, by a solution of (1.1.1) on an interval I , we mean a function x(t) =
ϕ(t) such that
F t, ϕ(t), ϕ (t), . . . , ϕ (n) (t)
is defined for all t ∈ I and
F t, ϕ(t), ϕ (t), . . . , ϕ (n) (t) = 0
for all t ∈ I . If we require, for some initial time t0 ∈ R, a solution x(t) to satisfy the
initial conditions
dx d 2x d n−1 x
x(t0 ) = a0 , (t0 ) = a1 , (t 0 ) = a 2 , . . . , (t0 ) = an−1 , (1.1.2)
dt dt 2 dt n−1
for constants ai , i = 0, 1, 2, ..., n−1, then (1.1.1) along with (1.1.2) is called an initial
value problem (IVP).
Following the notation of (1.1.1), a first-order differential equation takes the form
dx
F t, x, = 0.
dt
dx
Hence, if we assume that we can solve for , then we have
dt
x (t) = f (t, x)
and called the indefinite integral of the function x. When we calculate the indefi-
nite integral t 2 dt, we in fact solve the first-order differential equation x (t) = t 2 .
The family of its solutions is given by t 3 /3 + c, where c is any constant. Thus we
may write the solution as x(t) = t 3 /3 + c, which is a one-parameter family of so-
lutions, the same as the family of all the antiderivatives of t 3 . Now, if we impose
an initial condition on the differential equation, say x(t0 ) = x0 , for some initial time
t0 and real number x0 , then the constant c is uniquely determined by the relation
x0 = t03 /3 + c. In this case the differential equation has the unique solution given by
x(t) = t 3 /3 + x0 − t03 /3. However, without imposing the condition x(t0 ) = x0 , the
differential equation would have infinitely many solutions given by x(t) = t 3 /3 + c.
Differential equations play an important role in modeling the behavior of physical
systems such as falling bodies, vibration of a mass on a spring, and swinging pen-
dulum. To illustrate the need for the theoretical study of differential equations and in
particular nonlinear ones, we examine a few examples.
Consider the first-order differential equation
x (t)
= h(t).
g(x(t))
Separating the variables and then integrating both sides from t0 to t give
t x (s)ds t
= h(s)ds.
t0 g(x(s)) t0
dG 1
If for some function G, we have = , then the above expression implies that
dx g
t
G(x(t)) − G(x(t0 )) = h(s)ds
t0
or
t
x(t) = G−1 G(x0 ) + h(s)ds ,
t0
provided that the inverse of G exists. Note that the right side of the above expres-
sion depends on the initial time t0 and the initial value x0 . Therefore to emphasize
the dependence of solutions on the initial data, we may write a solution x(t) in
the form x(t) = x(t, t0 , x0 ).
FIGURE 1.1
This example shows infinitely many solutions.
In the next example, we illustrate the existence of more than one solution. Consider
the differential equation
3
x (t) = x 1/3 (t), x(0) = 0, t ∈ R.
2
It is clear that x(t) = 0 is a solution. Hence we may consider a solution x1 (t) = 0 and
let
0 for t ≤ 0,
x2 (t) =
t 3/2 for t > 0,
which is also a continuous and differentiable solution. Likewise, for t1 > 0, we have
0 for t ≤ t1 ,
x3 (t) =
(t − t1 )3/2 for t > t1 .
Continuing in this way, we see that the differential equation has infinitely many solu-
tions. Similarly, if x is a solution, then −x is also a solution (see Fig. 1.1).
In the next example, we show that solutions may escape (become unbounded) in
finite time. The differential equation
1
We can easily see that the solution is only valid for t < t0 + and becomes un-
2x02
bounded (escapes) as t approaches t0 + 1
from the left.
2x02
The most important application in engineering problems is Newton’s law
d 2 x(t) dx(t)
m = F [t, x(t), ]
dt 2 dt
for the position x(t) of a particle with mass m acted on by a force F , which may be
a function to time t, the position x(t), and the velocity dx(t)
dt . For example, if the force
is only due to gravity, then we have the second-order differential equation
d 2 x(t)
m = −mg,
dt 2
which has the solution
1
x(t) = − gt 2 + c1 t + c2 ,
2
where c1 and c2 are constants that can be uniquely determined by specifying the
position and velocity of the particle at some instant of time.
In the next example, we consider the problem of leaky bucket.
Example 1.1. We have a bucket with no water flowing into it and having a hole in the
bottom. If Q is the volumetric flow rate, then Qin − Qout = Qstored . Since Qin = 0,
we obtain Qstored = −Qout . Let h(t) be the height of the water in the bucket at time t,
and assume that the initial height at time t = 0 is h0 . Let v(t) be the velocity of
the leaked water (flow velocity). The volumetric flow rate Qstored can be calculated
by multiplying the velocity by the area of the bucket, that is, Qstored = Abucket dh(t)dt .
Similarly, Qout is the flow velocity multiplied by the area of the hole Ahole . It follows
that Qout = Ahole v(t).√For fluids of height h(t), the velocity of water coming out at
the bottom is v(t) = 2gh(t). By rearranging the terms, we arrive at the first-order
differential equation in h given by
dh(t) √
= −k h, h(0) = h0 , (1.1.3)
dt
Ahole
where k = 2g > 0. By separating the variables in (1.1.3) and then integrat-
Abucket
ing, we arrive at the solution
kt 2
h(t) = h0 − . (1.1.4)
2
Note that the solution
√
h(t) given by (1.1.4) decreases from the initial height h0 . More-
over, at time t ∗ = 2 kh0 , the water is completely drained (by gravity), h(t ∗ ) = 0, and
√
the bucket will remain empty or the height will remain zero after t ∗ = 2 h0
k . There-
fore we may write the solution as
⎧
⎪ √
⎨ √h − kt 2 , 0 ≤ t ≤ 2 h0 ,
0
h(t, 0, h0 ) = 2
√
k
⎪
⎩ 0, 2 h0
t> k .
Mm
F = −G ,
(R + x)2
where G is the proportionality constant in the universal gravitational law. The mi-
nus sign means the force on the mass m points in the direction of decreasing x. By
2
Newton’s second law, force F = m ddt x2 (mass times acceleration). When x = 0, that
is, at the Earth’s surface, the gravitational force equals −mg, where g is the gravita-
tional constant. Therefore
GM
= g, g ≈ 9.8 m/s2 .
R2
We have the second-order differential equation
d 2x M
= −G
dt 2 (R + x)2
g
=− , (1.2.1)
(1 + Rx )2
where the radius of the Earth is known to be R ≈ 6350 km. We transform Eq. (1.2.1)
2
into a first-order differential equation in terms of the velocity v by noting that ddt x2 =
dt . If we write v(t) = v(x(t)), that is, considering the velocity of the mass m as a
dv
function of its distance above the Earth. Then using the chain rule, we have
dv dv dx
=
dt dx dt
dv
=v ,
dx
since v = dx
dt . As a consequence, (1.2.1) becomes the first-order differential equation
dv g 1
=− .
dx (1 + Rx )2 v
Suppose the mass is shot vertically from the Earth’s surface with initial velocity
v(x = 0) = v0 . Separating the variables and integrating both sides, we obtain
dx
vdv = −g ,
(1 + Rx )2
which gives
v2 gR 2
= +c
2 R+x
for some constant c. Using the initial velocity condition, we find
v02
c = −gR + .
2
Substituting c back into the solution and simplifying, we arrive at the solution
gRx
v 2 = v02 − .
R+x
The escape velocity is defined as the minimum initial velocity v0 , such that the mass
can escape to infinity. Therefore v0 = vescape when v → 0 as x → ∞. Taking the
limit, we have
gRx
0 = v02 − lim = v02 − 2gR,
x→∞ R + x
or
2
vescape = 2gR.
√
With R ≈ 6350 km and g = 127,008 km/h2 , we get vescape = 2gR ≈ 40,000 km/h.
In contrast, the muzzle velocity of a modern high-performance rifle is 4300 km/h,
which is not enough for a bullet shot into the sky to escape from Earth’s gravity.
Now we formally attempt to qualitatively analyze differential equations.
subsequently found wide use in epidemiology and ecology. To describe the law of
mass action, we assume that m substances s1 , s2 , . . . , sm together form a product
with concentration p. Then the law of mass action states that dp dt is proportional to
the product of the m concentrations si , i = 1, . . . , m, that is,
dp
= ks1 s2 . . . sm .
dt
Suppose we have a homogeneous population of fixed size divided into two groups.
Those who have the disease are called infective, and those who do not have the dis-
ease are called susceptible. Let S = S(t) be the susceptible portion of the population,
and let I = I (t) be the infective portion. Then by assumption, we may normalize the
population and have S + I = 1. We further assume that the dynamics of this epidemic
satisfies the law of mass action. Hence, for positive constant λ, we have the nonlinear
differential equation
I (t) = λSI. (1.3.1)
Let I (0) = I0 , 0 < I (0) < 1, be a given initial condition. By substituting S = 1 − I
into (1.3.1) it follows that
If we can solve (1.3.2) for I (t), then S(t) can be found from the relation I + S = 1.
We separate the variables in (1.3.2) and obtain
dI
= λdt.
I (1 − I )
Using partial fraction on the left side of the equation and then integrating both sides
yield
Ln|I | − Ln|1 − I | = λt + c,
or for positive constant c1 , we have
c1 eλt
I (t) = .
1 + c1 eλt
Applying I (0) = I0 gives the solution
I0 eλt
I (t) = . (1.3.3)
1 − I0 + I0 eλt
Now for 0 < I (0) < 1, the solution given by (1.3.3) is increasing with time as ex-
pected. Moreover, using L’Hospital’s rule, we have
I0 eλt
lim I (t) = lim = 1.
t→∞ t→∞ 1 − I0 + I0 eλt
Hence the infection will grow, and everyone in the population will eventually get
infected.
It is easy to see that the function f (x) = x 2 is not Lipschitz on R. This is due to the
fact that for any x and y in R, we have that f (x) − f (y) = |x 2 − y 2 | = |x + y||x − y|,
and so there is no constant K such that |x 2 − y 2 | ≤ K|x − y|. Definition 1.4.6 implies
that f is globally Lipschitz since the constant K is uniform for all x and y in R.
Proof. Suppose {fn } is equicontinuous on [a, b]. Let ε > 0. For each x ∈ K, let
δx > 0 be such that |y − x| < δx =⇒ |fn (x) − fn (y)| < ε/2 for all n ∈ N. The
collection {B(x, δx /2) : x ∈ [a, b]} is an open cover of [a, b], so it has a finite
subcover {B(xi , δxi /2) : i = 1, . . . , k}. Let δ = min{δxi /2 : i = 1, . . . , k}. Then, if
x, y ∈ [a, b] with |x − y| < δ, then there is some i with x ∈ B(xi , δxi /2). Since
|x − y| < δ ≤ δxi /2, we have |xi − y| ≤ |xi − x| + |x − y| < δxi /2 + δxi /2 = δxi .
Hence |xi − y| < δxi and |xi − x| < δxi . So, for any n ∈ N, we have |fn (x) − fn (y)| ≤
|fn (x) − fn (xi )| + |fn (xi ) − fn (y)| < ε/2 + ε/2 = ε. So {fn } is uniformly equicon-
tinuous.
The next theorem gives us the main method of proving compactness in the spaces
we are interested in.
Proof. Since {fn (t)} is equicontinuous on [a, b], by Lemma (1.1) {fn (t)} is uni-
formly equicontinuous. Let t1 , t2 , . . . be a listing of the rational numbers in [a, b]
(note that the set of rational numbers is countable, so this enumeration is possi-
ble). The sequence {fn (t1 )}∞n=1 is a bounded sequence of real numbers (since {fn }
is uniformly bounded), so it has a subsequence {fnk (t1 )} converging to a num-
ber, which we denote φ(t1 ). It will be more convenient to represent this subse-
quence without subsubscripts, so we write fk1 for fnk and switch the index from
k to n. So the subsequence is written as {fn1 (t1 )}∞ n=1 . Now, the sequence {fn (t2 )} is
1
bounded, so it has a convergent subsequence, say {fn (t2 )}, with limit φ(t2 ). We con-
2
tinue in this way obtaining a sequence of sequences {fnm (t)}∞ n=1 (one sequence for
each m), each of which is a subsequence of the previous one. Furthermore, we have
fnm (tm ) → φ(tm ) as n → ∞ for each m ∈ N. Now, consider the “diagonal” functions
defined Fk (t) = fkk (t). Since fnm (tm ) → φ(tm ), it follows that Fr (tm ) → φ(tm ) as
r → ∞ for each m ∈ N (in other words, the sequence {Fr (t)} converges pointwise at
each tm ). We now show that {Fk (t)} converges uniformly on [a, b] by showing that
it is Cauchy in the uniform norm. Let ε > 0, and let δ > 0 be as in the definition of
uniformly equicontinuous for {fn (t)} applied with ε/3. Divide [a, b] into p intervals,
where p > b−a δ . Let ξj be a rational number in the j th interval for j = 1, . . . , p. Re-
call that {Fr (t)} converges at each of the points ξj , since they are rational numbers.
So, for each j , there is Mj ∈ N such that |Fr (ξj )−Fs (ξj )| < ε/3 whenever r, s > Mj .
Let M = max{Mj : j = 1, . . . , p}. If t ∈ [a, b], then it is in one of the p intervals, say
the j th. So |t − ξj | < δ, and thus |frr (t) − frr (ξj )| = |Fr (t) − Fr (ξj )| < ε/3 for ev-
ery r. Also, if r, s > M, then |Fr (ξj ) − Fs (ξj )| < ε/3 (since M is the maximum of
the Mi ). So for r, s > M, we have
|Fr (t) − Fs (t)| = |Fr (t) − Fr (ξj ) + Fr (ξj ) − Fs (ξj ) + Fs (ξj ) − Fs (t)|
≤ |Fr (t) − Fr (ξj )| + |Fr (ξj ) − Fs (ξj )| + |Fs (ξj ) − Fs (t)|
ε ε ε
≤ + + = ε.
3 3 3
By the Cauchy criterion for convergence, the sequence {Fr (t)} converges uniformly
on [a, b]. Since each Fr (t) is continuous, the limit function φ(t) is also continu-
ous.
The next criterion, known as the Weierstrass M-test plays an important role in
showing the existence of solutions.
We remark that the Weierstrass M-test can be easily generalized if the domain of
the sequence of functions is a subset of Banach space endowed with an appropriate
norm.
Here is an example of the Weierstrass M-test.
∞ ∞
1 1
the series converges, by the Weierstrass M-test the series con-
n 2 t + n2
2
n=1 n=1
verges uniformly on R. Moreover, as each term of the series is continuous and the
convergence is uniform, the sum function is also continuous. (As the uniform limit
of continuous functions is continuous.)
Here is another example with a simple twist to it.
Example 1.3. We prove that the series
∞ 2
n + x4
n4 + x 2
n=1
n2 + x 4 n2 + x 4 1 c4
≤ ≤ + := Mn .
n4 + x 2 n4 n2 n4
On the other hand, the series
∞
∞
∞
1 1
Mn = 2
+ c 4
n n4
n=1 n=1 n=1
2. Scalar multiplication
a. · is a function from F × V to V . Outputs are denoted α · x or αx.
b. for all α, β ∈ F and x ∈ V , α(βx) = (αβ)x.
c. for all x ∈ V , 1 · x = x.
d. for all α, β ∈ F and x ∈ V , (α + β)x = αx + βx.
e. for all α ∈ F and x, y ∈ V , α(x + y) = αx + αy.
Commonly, the real numbers and complex numbers are fields in the above defini-
tion. For our purposes, we only consider the field of real numbers F = R.
Definition 1.4.8. (Normed spaces) A vector space (V , +, ·) is a normed space if for
each x ∈ V , there is a nonnegative real number x, called the norm of x, such that
for all x, y ∈ V and α ∈ R,
1. x = 0 if and only if x = 0,
2. αx = |α|x,
3. x + y ≤ x + y.
Remark 1.3. A norm on a vector space always defines a metric ρ(x, y) = x − y on
the vector space. Given a metric ρ defined on a vector space, it is tempting to define
v = ρ(v, 0). But this is not always a norm.
Definition 1.4.9. A Banach space is a complete normed vector space, that is, a vector
space (X, +, ·) with norm · for which the metric ρ(x, y) = x − y is complete.
Example 1.4. The space (Rn , +, ·) over the field R is a vector space (with the usual
vector addition + and scalar multiplication ·), and there are many suitable norms for
it. For example, if x = (x1 , x2 , . . . , xn ), then
1. x = max |xi |,
1≤i≤n
n
2. x = xi2 ,
i=1
n
3. x = |xi |,
i=1
n 1/p
4. xp = |xi |p ,p≥1
i=1
are all suitable norms. Norm 2 is the Euclidean norm: the norm of a vector is its
Euclidean distance to the zero vector, and the metric defined from this norm is the
usual Euclidean metric. Norm 3 generates the “taxi-cab” metric on R2 , and Norm 4
is the l p norm.
Throughout the book, it should cause no confusion to use | · | instead of || · || to
denote a particular norm.
Remark 1.4. Consider the vector space (Rn , +, ·) as a metric space with its metric
defined by ρ(x, y) = x − y, where · is any of the norms in Example 1.4. The
completeness of this metric space comes directly from the completeness of R, and
hence (Rn , · ) is a Banach space.
Remark 1.5. In the Euclidean space Rn , compactness is equivalent to closedness
and boundedness (Heine–Borel theorem). In fact, the metrics generated from any of
the norms in Example 1.4 are equivalent in the sense that they generate the same
topologies. Moreover, compactness is equivalent to closedness and boundedness in
each of those metrics.
Example 1.5. Let C([a, b], Rn ) denote the space of all continuous functions
f : [a, b] → Rn .
1. C([a, b], Rn ) is a vector space over R.
2. If f = max |f (t)|, where | · | is a norm on Rn , then (C([a, b], Rn ), · ) is
a≤t≤b
a Banach space.
3. Let M and K be two positive constants and define
Then L is compact.
Proof (of part 3). Let {fn } be any sequence in L. The functions are uniformly
bounded by M and have the same Lipschitz constant, K. So the sequence is uni-
formly equicontinuous. By the Ascoli–Arzelà theorem there is a subsequence {fnk }
that converges uniformly to a continuous function f : [a, b] → Rn . We now show that
f ∈ L. Well, |fn (t)| ≤ M for each t ∈ [a, b], so |f (t)| ≤ M for each t ∈ [a, b] and
hence f ≤ M. Now fix u, v ∈ [a, b] and ε > 0. Since {fnk } converges uniformly
to f , there is N ∈ N such that |fnk (t) − f (t)| < ε/2 for all t ∈ [a, b] and k ≥ N . So
fixing any k ≥ N , we have
|f (u) − f (v)| = |f (u) − fnk (u) + fnk (u) − fnk (v) + fnk (v) − f (v)|
≤ |f (u) − fnk (u)| + |fnk (u) − fnk (v)| + |fnk (v) − f (v)|
< ε/2 + K|u − v| + ε/2 = K|u − v| + ε.
Since ε > 0 was arbitrary, |f (u) − f (v)| ≤ K|u − v|. Hence f ∈ L. We have demon-
strated that {fn } has a subsequence converging to an element of L. Hence L is
compact.
Example 1.6. Consider R as a vector space over R and define the metric d(x, y) =
|x − y|
. For each x ∈ R, we define x = d(x, 0). Explain why · is not a
1 + |x − y|
norm on R.
Example 1.7. Let φ : [a, b] → Rn be continuous, and let S be the set of continuous
functions f : [a, c] → Rn with c > b and f (t) = φ(t) for a ≤ t ≤ b. Define ρ(f, g) =
f − g = sup |f (t) − g(t)| for f, g ∈ S. Then (S, ρ) is a complete metric space
a≤t≤c<
but not a Banach space since f + g is not in S.
Example 1.8. Let (S, ρ) be the space of continuous bounded functions f : (−∞, 0] →
R with ρ(f, g) = f − g = sup |f (t) − g(t)|. Then:
−∞<t≤0
where
ρn (f, g) = max |f (s) − g(s)|,
−n≤s≤0
Continuing this way, we arrive at Fn = fnn , which has a subsequence of {fn } con-
verging uniformly on compact subsets of (−∞, 0] to a function f ∈ L. This proves
that L is compact.
|f (t)|
|f |g := sup
−∞<t≤0 |g(t)|
exists. Then:
1. (S, | · |g ) is a Banach space.
2. Let M and K be given positive constants. Then the set
Definition 1.4.10. Let (E, ρ) be a metric space, and let D : E → E. The operator or
mapping D is a contraction if there exists α ∈ (0, 1) such that
ρ D(x), D(y) ≤ αρ(x, y).
The next theorem is known by the name of Caccioppoli theorem, or the Banach
contraction mapping principle [9]. A proof can be found in many places, such as
Burton [15] or Smart [58].
ρ(yn , ym ) = ρ(D n y0 , D m y0 )
≤ αρ(D n−1 y0 , D m−1 y0 )
..
.
≤ α n ρ(y0 , ym−1 )
≤ α n ρ(y0 , y1 ) + ρ(y1 , y2 ) + . . . + ρ(ym−n−1 , ym−n )
≤ α n ρ(y0 , y1 ) + αρ(y0 , y1 ) + . . . + α m−n−1 ρ(y0 , y1 )
≤ α n ρ(y0 , y1 ) 1 + α + . . . + α m−n−1
1
≤ α n ρ(y0 , y1 ) .
1−α
ρ(yn , ym ) → 0 as n → ∞.
This shows that the sequence {yn } is Cauchy. Since (E, ρ) is a complete metric space,
{yn } has a limit, say y in E. Since the mapping D is continuous, we have that
and y is a fixed point. It remains to show that y is unique. Let x, y ∈ E be such that
D(x) = x and D(y) = y. Then
0 ≤ ρ(x, y) = ρ D(x), D(y) ≤ αρ(x, y),
Theorem 1.4.4. (Contraction mapping principle, Banach fixed point theorem) Let
(E, ρ) be a complete metric space, and let P : E → E be such that P m is a con-
traction for some fixed positive integer m. Then there is a unique x ∈ E such that
P (x) = x.
t
− t0 a(u)du
Multiplying both sides of (1.5.1) by the integrating factor e and observ-
ing that
d t
− a(u)du
t
− a(u)du
t
− a(u)du
x(t)e t0 = x (t)e t0 − a(t)x(t)e t0 ,
dt
we arrive at
d
− t a(u)du
− t a(u)du
x(t)e t0 = g(t, x(t))e t0 .
dt
Integration of the above expression from t0 to t and using x(t0 ) = x0 yield
t
− t a(u)du − s a(u)du
x(t)e t0 = x0 + g(s, x(s))e t0 ds,
t0
It can be easily shown that if x(t) satisfies (1.5.2), then it satisfies (1.5.1). Expression
(1.5.2) is known as the variation of parameters formula. Note that (1.5.2) is a func-
tional equation in x since the integrand is a function of x. If we replace the function
g with a function h ∈ C([0, ∞), R), then (1.5.2) takes the special form
t t t
t0 a(u)du
x(t) = x0 e + h(s)e s a(u)du ds, t ≥ t0 ≥ 0. (1.5.3)
t0
Another special form of (1.5.2) is that if the function a(t) is constant for all t ≥ t0
and g is replaced with h as before, then from (1.5.3), we have that
t
x(t) = x0 e a(t−t0 )
+ ea(t−s) h(s)ds, t ≥ t0 ≥ 0. (1.5.4)
t0
i
(a)
(b) R
+
η C
−
FIGURE 1.2
RC circuit.
1.5.1 RC circuit
Fig. 1.2 shows a resistor R and capacitor C connected in a series. The battery con-
nected to this circuit by a switch provides an electromotive force, or emf force η.
Initially, there is no charge on the capacitor. When the switch is flipped to (a), the
battery connects, and the capacitor charges. Similarly, when the switch is flipped
to (b), the battery disconnects, and the capacitor discharges with energy dissipated
in the resistor. Our aim is to determine the voltage drop across the capacitor during
charging and discharging.
We start by observing that the voltage drops across a capacitor VC and resistor VR
are given by
q
VC = , VR = iR,
C
respectively, where C is the capacitance and R is the resistance. The current i and the
charge q are related by the relation
dq
i= .
dt
dVC dq 1 i 1 VR
= . = = ,
dt dt C C C R
Kirchhoff’s voltage law states that the emf η in any closed loop is equal to the sum
of the voltage drops in that loop. When the switch is thrown to (a), this gives
VR + VC = η.
Thus the voltage starts at zero and rises slowly but exponentially to η with character-
istic time scale given by RC. In other words,
VC (t) → η as t → ∞.
Now suppose the switch is thrown to (b). Then by Kirchhoff’s voltage law
VR + VC = 0,
VC (t) = ηe−t/RC ,
and the voltage decays exponentially to zero with the characteristic time scale given
by RC.
x = f (t, x, x ), (1.6.1)
known about how to solve (1.6.1) except in certain particular cases. Two interesting
and useful cases occur when (1.6.1) has one of the following forms:
F (t, x , x ) = 0 (1.6.2)
or
F (x, x , x ) = 0. (1.6.3)
x
For (1.6.2), the transformation = u transforms the original second-order differ-
ential equation into a first-order differential equation F (t, u, u ) = 0, which can be
solved. To see this, we consider the nonlinear second-order differential equation
tx + 2x + t = 1, t > 0.
c1 t t2
x(t) = − + − + c2
t 2 6
for some constant c2 .
Equations of the form (1.6.3) can be reduced to first-order equations by the substi-
tution x = u. Eq. (1.6.3) is now replaced by the equivalent system of two differential
equations
x = u, F (x, u, u ) = 0.
By using x as the independent variable we have
du du dx du
x = = = u.
dt dx dt dx
Hence (1.6.3) reduces to the first-order differential equation
du
F (x, u, u ) = 0.
dx
As an application, we consider the differential equation
Then, under the mentioned substitution, we have the new first-order differential equa-
tion
du
u + u2 x + ux = 0
dx
or
du
+ xu = −x.
dx
By the variation of parameters, we arrive at the solution
x2
u = −1 + ce 2 .
x(t) = −t.
1.7 Exercises
Exercise 1.1. Show that the differential equation
3
x (t) = x 1/3 (t), x(0) = 0, t ∈ R
2
has infinitely many solutions.
Exercise 1.3. Find the escape time of the solutions for the differential equation
x (t) = t, x(t0 ) = x0 = 0, t ≥ 0
Exercise 1.6. Consider the following power series L(x), which is also known as
Euler’s dilogarithm function:
∞ n
x
L(x) = .
n2
n=1
(a) Compute the domain of convergence for L(x). Be sure to give a full analysis of
the endpoints.
(b) Show that L(x) uniformly converges on its entire domain of convergence. (Use
Weierstrass M-test.)
(c) Explain why the L(x) is continuous on its domain of convergence.
Exercise 1.8. Show that the space C(I, R), where I is any compact subset of R,
endowed with the supremum norm
is a Banach space.
Exercise 1.9. Show that the space T of all continuous T -periodic functions given
by
T = {f ∈ C(R, R) : f (t + T ) = f (T ) for all t ∈ R}
with the maximum norm
||f || = max |f (t)|
t∈[0,T ]
is a Banach space.
Exercise 1.10. Show that the space C([0, 1], R) endowed with the L1 norm
1
||f ||1 = |f (t)|dt
0
and
1/p
||a||p = |ak |p .
k
Show that lp endowed with || · ||p is a Banach space.
Exercise 1.12. Prove Remark 1.1.
Exercise 1.13. Find a continuous solution satisfying
Show that
t
x(t) ≤ x(0)ekt + ek(t−s) f (s)ds, t ∈ [0, d].
0
Hint: There is a continuous and nonnegative function g(t) such that
Show that there is a constant K such that ϕ(t) ≤ eKt for all t ∈ [0, d].
Exercise 1.18. In Problems 1–5 solve the differential equation. If no initial condi-
tions are given, find the general solution.
1. t 2 x + t = 1, t > 0.
2. t 2 x + 2tx = 1, t > 0.
3. xx − 2(x )2 + 4x 2 = 0, x(1) = 1, x (1) = 2.
4. sin(x )x = sin(x), x(1) = 2, x (1) = 1.
5. tx − x = t 2 et , t > 0.
Existence and
uniqueness
2
In Chapter 1 we laid out the basics we need to use in this and the following chap-
ters. This chapter is devoted to the existence and uniqueness and the continuation of
solutions. We limit our discussions and proofs to scalar differential equations, and in
Chapter 3 we indicate how the theory can be naturally extended to vector equations.
∂f
|f (t, x)| ≤ M and | |≤K (2.1.4)
∂x
Advanced Differential Equations. https://doi.org/10.1016/B978-0-32-399280-0.00008-5
Copyright © 2023 Elsevier Inc. All rights reserved.
27
for all points (t, x) in R. Now for any two points (t, x1 ), (t, x2 ) in R, by the mean
value theorem there exists a constant η ∈ (x1 , x2 ) such that
∂f
f (t, x1 ) − f (t, x2 ) = (t, η)(x1 − x2 ),
∂x
from which it follows that
∂f
|f (t, x1 ) − f (t, x2 ≤ | (t, η)||x1 − x2 |
∂x
≤ K |x1 − x2 |. (2.1.5)
∂f
We have shown that if f and ∂y are continuous on R, then f satisfies a global Lips-
chitz condition on R.
Example 2.1. Consider f (t, x) = x 2/3 in the rectangle R = {(t, x) : |t| ≤ 1, |x| ≤ 1}.
We claim f does not satisfy the Lipschitz condition on R. Consider the pair of points
(t, x1 ) and (t, 0) in R where x1 > 0. Then
|f (t, x1 ) − f (t, 0) −1/3
| = x1 → ∞ as x1 → 0+ ,
x1 − 0
and hence there exists no K such that (2.1.5) holds.
Remark 2.1. Consider (2.1.1) on the rectangle D ⊂ R × R defined by
D = {(t, x) : |t − t0 | ≤ a, |x − x0 | ≤ b},
τ = t − t0 , u = x − x0 , and g(τ, u) = f (τ + t0 , u + x0 ).
D ∗ = {(τ, u) : −a ≤ τ ≤ a, −b ≤ u ≤ b},
and
b
h = min{a, }. (2.1.8)
M
Then the IVP (2.1.1) has a unique solution, denoted by x(t, t0 , x0 ), on the interval
|t − t0 | ≤ h and passing through (t0 , x0 ). Furthermore,
|x(t) − x0 | ≤ b for |t − t0 | ≤ h.
FIGURE 2.1
Interval of existence.
x0 (t) = x0 ,
t
x1 (t) = x0 + f (s, x0 (s))ds,
t0
..
.
t
xn (t) = x0 + f (s, xn−1 (s))ds. (2.1.9)
t0
t
Therefore f (s, x1 (s))ds can be defined for |t − t0 | ≤ h, and hence
t0
|x2 (t) − x0 | ≤ hM ≤ b.
Similarly, we may show that x3 (t), . . . , xn (t) are well defined on |t − t0 | ≤ h. Thus
t
|xn (t) − x0 | = | f (s, xn−1 (s))ds| ≤ M|t − t0 |. (2.1.10)
t0
xn (t) = x0 (t) + [x1 (t) − x0 (t)] + [x2 (t) − x1 (t)] + · · · + [xn (t) − xn−1 (t)],
or
n−1
xn (t) = x0 (t) + [xj +1 (t) − xj (t)].
j =0
n−1
lim xn (t) = x0 (t) + lim [xj +1 (t) − xj (t)]. (2.1.11)
n→∞ n→∞
j =0
Our next task is to compute |xj +1 (t) − xj (t)|. Using (2.1.9) and then (2.1.2) followed
by (2.1.10) give
t
|x2 (t) − x1 (t)| ≤ f (s, x1 (s)) − f (s, x0 (s))ds
t0
t
≤K x1 (s) − x0 (s)ds
t0
t t
≤ KM |s − t0 |ds = KM (s − t0 )ds
t0 t0
(t − t0 )2
= KM , |t − t0 | ≤ a.
2
In a similar fashion, we obtain
t
|x3 (t) − x2 (t)| ≤ f (s, x2 (s)) − f (s, x1 (s))ds
t0
t
≤K x2 (s) − x1 (s)ds
t0
t (s − t0 )2
≤K KM ds
t0 2
(t − t0 )3
= MK 2 , |t − t0 | ≤ a.
3!
Continuing this way, we arrive at
(t − t0 )j
|xj (t) − xj −1 (t)| ≤ MK j −1 . (2.1.12)
j!
To complete the induction argument, we assume that (2.1.12) holds for j and show
that it holds for j + 1. Using (2.1.12), we arrive at
t
|xj +1 (t) − xj (t)| ≤ f (s, xj (s)) − f (s, xj −1 (s))ds
t0
t
≤K xj (s) − xj −1 (s)ds
t0
t (s − t0 )j
≤K MK j −1 ds
t0 j!
(t − t0 )j +1 (t − t0 )j +1
= MK j = MK j , |t − t0 | ≤ a.
j !(j + 1) (j + 1)!
M |K(t − t0 )|j +1
n−1
lim xn (t) = x0 + lim
n→∞ K n→∞ (j + 1)!
j =0
M
n−1
(Kh)j +1
≤ x0 + lim
K n→∞ (j + 1)!
j =0
M Kh
≤ x0 + e − 1 < ∞.
K
Hence, by the Weierstrass M-test, {xn (t)} converges uniformly, say to a function x(t).
Next, we show that x(t) is a solution of (2.1.1) on D. We have
= 0 + M|t − t0 |
≤ Mh ≤ b
for t ∈ (t0 − h, t0 + h). So x(t) ∈ D. If the sequence {xn (t)} converges uniformly and
{xn (t)} is continuous on the interval |t − t0 | ≤ h, then
t t
lim f (s, xn (s))ds = lim f (s, xn (s))ds
n→∞ t t0 n→∞
0
that is,
t
x(t) = x0 + f (s, x(s))ds, |t − t0 | ≤ h. (2.1.13)
t0
The integrand f (s, x(s)) in (2.1.13) is a continuous function, and hence x(t) is dif-
ferentiable with respect to t, and its derivative is equal to f (t, x(t)). So the proof of
the existence is complete. It remains to show that the solution x(t) on D is unique.
Let y(t) be another solution of (2.1.1) with y(t0 ) = x0 . Then
t
y(t) = x0 + f (s, y(s))ds.
t0
Let
N = sup |x(t) − y(t)|.
|t−t0 |≤h
or
|x(t) − y(t)| ≤ KN |t − t0 |.
For n = 1, we have
t
x1 (t) = 2sds = t 2 ,
0
and for n = 2,
t t4
x2 (t) = 2s(1 + s 2 )ds = t 2 + .
0 2
We leave it to the reader to verify that
t 2n (t 2 )i
n
t4 t6
xn (t) = t + + + . . . ,
2
= − 1.
2 3! n! i!
i=0
n
(t 2 )i
xn (t) = −1
i!
i=0
2
and the true solution of the IVP is given by x(t) = et − 1 on the interval [0, 1]. Thus
∞
n
(t 2 )i 1
max − 1 − (et − 1) = lim
2
lim ||xn − x|| = lim = 0.
n→∞ n→∞ t∈[0,1] i! n→∞ i!
i=0 i=n+1
Therefore
n
(t 2 )i 2
lim xn (t) = lim − 1 = et − 1.
n→∞ n→∞ i!
i=0
Example 2.3. Next, we use Theorem 2.1.2 to find the interval of existence of the
unique solution of
x (t) = 1 + x 2 , x(0) = 1.
Let
D = {(t, x) : |t| ≤ a, |x − 1| ≤ b},
where a and b are constants. It is clear that
and
b b
h = min{a, } = min{a, }.
M 1 + (1 + b)2
b
So we must find the maximum of the function g(b) = . Using calculus,
1 + (1 + b)2
√ √
we see
√
that the function g has its maximum at b = 2 and is given by g( 2) =
2√
2
. Thus the interval of existence and uniqueness is
1+(1+ 2)
√ √ √
2 2 2
|t| ≤ √ , or t ∈ − √ , √ .
1 + (1 + 2)2 1 + (1 + 2) 1 + (1 + 2)2
2
Now suppose in Example 2.3 we chose D = {(t, x) : |t| ≤ 10, |x − 1| ≤ b}, that
is, a = 10. Then the interval of existence found above is much smaller or included in
(−10, 10). Recall that the interval of existence is included in D.
The next corollary is an immediate consequence of Theorem 2.1.2 and the discus-
sion leading to (2.1.5).
Corollary 2.1. Suppose D ⊂ R × R, f ∈ C(D, R), and ∂f ∂x is continuous on D. Then
for any (t0 , x0 ) ∈ D, the IVP (2.1.1) has a unique solution on an interval containing t0
in its domain.
As an example, consider
x (t) = tx 1/2 .
In woodfinishing, as in woodworking, the aim has been to have the pupil treat
the subject in a serious and workmanlike manner. In seventh grade little
woodfinishing is done. The woodworking processes need the centering of the
pupil’s attention, in the first place. Second, the simple pieces which the beginner
is able to make require no finish as a rule. In one group stain and wax is used.
This is the group in which decorative design is emphasized. In the eighth grade
the woodfinishing problem becomes important. Almost all of the pieces require a
finish.
The greatest obstacle to proper woodfinishing lies in the desire of the pupil to
take his piece home as soon as the woodwork is completed. Unless a definite
understanding is had with the class beforehand, proper woodfinishing is difficult
to obtain. Most boys are subject to reason, so that it is not at all necessary to
have woodfinishing slighted or to resort to makeshifts. The writer makes it a
practice to take plenty of time when the subject of woodfinishing comes up for
its first discussion to explain in detail the commercial methods of finishing fine
furniture, a piano for illustration, counting the different operations and coatings it
will receive and the labor and time expended upon the finish. A comparison is
then made between a finely rubbed finish and the cheap, sticky, unrubbed
finishes of cheap furniture.
Having established in the minds of the pupils the fact that woodfinishing is an
art second to none and that it requires time to do it well, there is not that
impatience that breeds sullen looks when the woodfinishing is to be begun after
the woodwork has been completed. The pupil will take the woodfinishing as a
matter of course and goes about it in a cheerful and manly spirit.
In grammar schools, woodfinishing has been made as simple as is consistent
with good work. Coming as the boys do but once a week and each finishing
application requiring over night for drying or hardening, the total time is quite
long even with the simple finish of filler, shellac, and wax. If the pupil wishes a
very dark finish, a stain which requires one or more periods must precede his
filler.
In high school, pupils come every day thus permitting the application of
rubbed varnish finishes, either shellac or copal, without unnecessary loss of
time. Here special finishing rooms are necessary.
5. Structural and Decorative Design. Among other requirements for a
course in woodwork and drawing as stated in the foreword is this: “At least a few
problems should be given which involve invention or design or both, thereby
stimulating individual initiative on the part of the pupils.” The present outlines in
woodwork and drawing have been planned with this in mind. In the seventh
grade the pupil is given little opportunity to exercise his initiative in either
woodwork or drawing. The reason for this, as has been previously stated, is a
firm belief that initiative in any subject to be of value must be based upon a fair
knowledge of the subject matter dealt with, its limitations and its possibilities. In
other words, that appreciation must precede invention or initiative.
With the limited time allowed manual training, at most one-half day each week
in the general educational scheme, a seventh grade beginner has about all he
can well manage in becoming familiar with his subject matter, with learning to
handle his tools and work his material.
But one group in the seventh grade will admit of decorative design. These
problems, Group VI, have purposely been made simple as to woodwork that the
pupil may give most of his attention to the design. In eighth grade, modifications
of outline and dimensions of any project are permitted where a fair degree of
merit is shown. Modifications of joints or fastenings are not to be made,
however, unless a pupil wishes to transfer a project from some other group into
the group in which the class is working.
In high school the pupil is expected to “work up” in his drawing class projects
original in so far as his ability will permit, subject to limitations mentioned
hereafter.
Eighth grade boys are expected to make at least one application of decorative
design to the pieces of woodwork made. The projects made by the high school
boys are, as a rule, not so well calculated to take decorative design. Their
efforts at decorative design will come later in connection with the metalwork of
the first year.
In high school the design is to be taught by special drawing teachers who
have informed themselves of the limitations of the shop methods when it comes
to applying these designs. It is for the shop instructor to specify the kind of joint
or joints that are to be used and the material, also the limitations as to
decoration. Present methods of organization in high schools hardly permit of the
teaching of shopwork and design and by the same instructor, which is the ideal
way providing, of course, that the instructor is expert in both. This is a
combination difficult to find. It is gratifying, however, to know that some schools
are insisting that their shop men become informed in design as well as
shopwork.
While these drawings are being worked up in the drafting room the pupil’s
shop periods are given over to the making of the exercise joints and mastering
the principles involved in their making. By the time these exercises are
completed, the working drawing will be completed ready for use in the shop.
The proper correlation of design and shopwork is not a problem beyond
solution, because of the direct relation of the two departments, providing there is
a strong administrative head able to secure proper esprit de corps. In the
grammar schools, however, the problem becomes less satisfactory of solution
by correlation.
The first objection lies in the fact that the regular grade teacher has both boys
and girls to teach and the problems must therefore be the same for the whole
room. The second objection lies in the fact that the problem in design has to
pass thru too many hands before it reaches the boy. If design is to be taught to
the best advantage, it must have the interest of the teacher and she must have
an intelligent understanding not only of the subject of design but of the particular
problem that is to be presented. The difficulties in the way are not
insurmountable where the drawing supervisor herself presents the problem to
the pupils. Even here, however, one frequently finds the drawing supervisor so
much more interested in the freehand drawing that her dislike for the design
makes her unfitted for such correlation work.
When, however, as is the case in cities, the drawing supervisor must reach
the pupils thru the regular teacher, correlation becomes in most every instance a
farce. The teaching of design is another imposition on an already overburdened
grade teacher. Very seldom does she understand the problem and it becomes a
distasteful subject to be got over in the easiest way possible. Department
teaching in the upper grammar grades would do much to aid in the correlation of
drawing and shop. Until this is made possible, we can hope for little in the way
of results from grammar school correlation, unless it be in a small system where
the supervisor teaches the children directly.
The whole subject of design as it relates to woodworking is a constant source
of discussion among manual training shop men. Many good teachers insist that
design has no place at all in a course in woodworking. Others admit that it ought
to have a place but feel that the results obtained do not justify the time spent
upon it. Still others approach the whole field of woodworking from the side of
design, tool processes and organized woodworking subject matter being mere
incidents to the problem in design.
Like every extreme position each of these points of view has good in it, but
there is sufficient error accompanying each to impair the validity of the
conclusions and to make the resulting applications unhappy as related to
ordinary public school conditions.
The whole subject of design as it relates to the manual training shop is one
that has demanded thought on the part of the author. It is one of those places
where teaching theory failed to bring efficiency either in the results obtained in
design or in the reaction upon the boy. He has been forced to the opinion, from
his own experience and from his observation of the efforts of others to teach
design to grammar school pupils, that the cause for dissatisfaction and
discouragement is due to our insistence upon one and only one method of
presentation—the inductive or synthetic.
In judging results we must consider the results obtained from every member
of a class and the good each boy has got out of his experience. This efficiency
test most effectively excludes the exhibition of a few “accidentals” as evidence
that our method is the correct one. There is no reason why design should seek
justification on any ground other than that offered by other subjects.
Inductive or synthetic teaching of design has its place; so also has the
deductive or analytic. Happily those educators who insist on the use of one
method or the other only are becoming few. In other subjects we are finding that
the teaching results which demand the respect and approval of educators of
safe and sane judgment are obtained by the use of both methods
interchangeably. There is no formal notice when one is to be used or the other—
whichever method fits the occasion is used without apology. This is right; to do
otherwise is to sacrifice the boy or girl for the sake of the method. We are all
agreed that the child is the more important consideration. In fact, some
psychologists tell us that induction and deduction are one and the same
process, the difference being merely a matter of emphasis. It is this difference in
placing the emphasis that we seek to discuss.
Our methods in the high school have made much of the inductive. This is
right. Pupils of high school and college age are ready for this method, tho our
high school pupils often would profit by having a little less of this with more of
the deductive.
However, when it comes to grammar school teaching, the maximum of use
has to be made of the deductive or analytic method. This is acknowledged in the
academic subjects. Woodworking when taught so as to meet the efficiency test
that is applied to academic teaching also makes use of this method mostly. Our
design, however, has always been taught by the inductive or synthetic method,
no one seeming to have the temerity to make use of any other. As a result we
find the views of design in the grammar school as stated above. Those who
advocate it urge the “accidentals” as sufficient justification. Those who reject it
base their argument on the fact that results based on a few accidentals will not
satisfy the same efficiency test that is applied to other subjects.
Experience has shown, at least to the author’s satisfaction, that the deductive
or analytic method when given maximum emphasis with beginners in design is
all that is needed to bring the results up to a standard equal to that of other
subjects. It is the rational method of presenting any subject to beginners.
The terms deductive and inductive have such wide application that it may be
well to specify more particularly just what we mean. A concrete illustration will
suffice to show the distinction we seek to make between what we choose to
designate the deductive or analytic and the inductive or synthetic methods.
Suppose we wish to have a class, with little or no information about the
subject, design a booklet to meet certain specified conditions. Three distinct
stages of progress manifest themselves in what we shall call the complete
method. First, the pupils must be given information bearing upon the problem.
Second, they must be given experience in handling problems of that type. Third,
they will utilize this information and experience in designing the booklet to meet
given conditions.
The first step will be the taking of a type form and analyzing it. Either the
instructor will demonstrate or, better, each pupil may be given a booklet of type
form and required to take it apart and put it together again. Any way to give the
pupil the information in a form that will cause it “to stick.” In woodwork, it would
be done by means of the traditional shop demonstration—a wise practice, since
psychology teaches us that sight percepts are among the strongest.
Second, the pupils must acquire experience. Let them make a booklet
according to definite specifications provided them by the instructor.
The process thus far is mainly deductive or analytic. So far there has been no
invention or design, but the pupils are now prepared for it. Using the information
and experience now available, let them design a booklet to meet certain
conditions. This latter part we would call the inductive or synthetic process.
We should have two aims in our teaching of design: (1) Appreciation, (2)
Development of the creative faculty. Since all must be able to appreciate good
line and good form when they get out into life while only a few will ever become
designers in a creative sense, it is essential, as it is also rational, that attention
should be paid first to appreciation. Past efforts show how hopeless is the
problem when we strive to give to the pupils appreciation of and feeling for line
and form by demanding original forms in the very beginning. The beginner’s
efforts at creation are abortive and the appreciation that he derives is nil. By our
insistence on this method we have given to our pupils the idea that design
means making something out of nothing. He is not far wrong if we demand of
him original designs before we have given him anything tangible upon which to
work. We say tangible as distinguished from academic principles or rules of
design. If nothing tangible is given the pupil he must get it outside of his school
experience. This explains the superabundance of “wienerwurst” forms, bouquets
tied with ribbons, circles, etc., etc.
It is possible to create unknown out of what is seemingly unknown. When we
stop to analyze the process, however, we find that we have made use of
information, appreciation, and feeling that are known. Sometimes we make
ourselves believe that our pupils are creating unknown out of unknown without
these requisites. Analysis will show that our continued suggestions to him,
drawn from our own fund of known are the causes, and not the pupil’s faculty.
This method of teaching is the kind we have been used to in design. It works
pretty well with small classes and individual instruction. Try it on large classes of
beginners and it is not possible to bring results that stand for class efficiency.
And why should this particular method be insisted upon exclusively with
beginners? Why should not design, like mechanical drawing and woodwork and
other subjects be developed upon a substructure consisting of information and
appreciation secured by allowing or even insisting that the boy handle good
design until he becomes saturated with a feeling for good line and good form?
Of course, if any pupil comes to a beginning class with this information and
feeling, due allowance should and can readily be made. It is highly probable that
there would be less inclination on the part of our pupils to insist that designers
are born not made were more use made of the deductive method. When the
boys no longer see their efforts result in crudities and are enabled to acquire the
necessary feeling and information as their work proceeds, then you find a happy
and interested class that as a whole takes design as a matter of course and not
as something intended only for the few.
Whatever the method of teaching design in the regular classroom, lack of time
demands the most direct treatment of shop design. A grammar school boy is not
inclined to listen very patiently to anything that smacks of the academic. (1) Give
the boy something definite with which to work and (2) keep him working, or
“playing,” as one has fittingly designated it, until he has made a conscientious
effort to “make it a part of himself,” that is, until he succeeds in changing the
form until it no longer resembles the original but still possesses the pleasing
appearance of the original.
If he succeeds in doing this, he is well on the way to creative effort. Not all
boys are of equal ability in other lines of endeavor, neither are they in this. By
this method of attack, however, even the stupidest—usually stupid only in the
matter of design—is not without compensation for his effort. He has learned
somewhat of the principles that govern good design by hearing them explained
and seeing them illustrated in a piece of good design. He will have developed
some feeling for line and form thru having played with good line and form. He
can at the very least fit the form given him to an outline made by himself after
suggestions of good line placed upon the board. To this extent, at least, you
have benefited him, whereas, by the usual method he—and there are many like
him—would have simply sat idle in discouragement—if he were not more
mischievously occupied.
If our old art schools were to be criticised because they made too much use of
the imitative method when they strove to give to their students information and
appreciation and feeling for form and line thru copying historic ornament, it
would seem, from results obtained, both tangible and in the effect upon the
pupils, that our modern schools are open to criticism when they seek to force
originality upon immature minds before they have given these minds any
information or feeling.
Of course grammar school boys are not interested in historic ornament, at
least not in America. This is the weakness of the imitative method and helped to
bring in the movement which now seems to have swung to the opposite extreme
—it lacks vitality for young pupils. Instead of giving the boy historic fragments,
give him a form that is vitally interesting to him because he sees its immediate
application in the thing that is to be made in wood. Let him play with this form
combining imitation and modification and creation just as far as he is able.
Make the problem concrete, stating the principles you have to state in a
language the boy can understand. There will not be time to bring out every
principle that might be involved in design. There must be time to bring out those
involved in the particular problem under discussion. Balance and symmetry, for
illustration, are pretty well understood by the boy in the simple form in which he
will have occasion to use them.
Take as an illustration the bookrack, Fig. 7. To present such a problem we
would place upon the blackboard the blank forms as shown, also the decorative
form as shown.
The lesson immediately divides itself into two parts for consideration: (1) The
Construction, (2) The Decoration. Under the subject of Construction our normal
school notes would suggest the following points to be brought up: Use,
Construction, Decoration; Requirements of Utility; Limitations of Materials and
Processes; Proportions of Parts and Details; Harmony of Parts and Details;
Points of Force; Construction as Decoration. (According to Payne.) Under
Decoration: Supporting Outline; Center of Interest; Symmetry; Repetition;
Radiation; Rhythm; Contrast; Proportion in Curves; Proportion in Spaces; Unity;
Subordinate Centers of Interest; Balance.
FIG. 7. TEACHING DESIGN IN THE PROBLEM OF THE BOOK-
RACK.
CONSTRUCTION
DECORATION
OUTLINES
MOTIVE GIVEN ADAPTATION
APPLICATION—
BY A PUPIL
Taking these in their natural order, but without making much ado about the
“framework,” the shop man who has made some study of the principles involved
can call the boys’ attention to the most important points:
(1) The construction. Since the shopwork is to be carried on by class
instruction and not individually, it will be necessary to limit the joint or joints used
to those specified for the Group in which the project is to be worked out. Joints
of previous Groups may be used also. The book rack will be made in Group VII.
Some form of the groove joint is to be used, none other.
Here we call attention to the difference between the designer and the shop
man in their handling of the problem. The discussion of construction gives the
designer an opportunity to display the possibilities of his subject. He enumerates
all the joints that may be used with propriety in making such a piece as the
bookrack, and the pupils are encouraged to make use of as many varieties as
possible. He is totally oblivious of the fact that, while this is good teaching in
design, it is making the applications impossible except with individual instruction
—a method of instruction that may be used in small school systems but not in
cities.
(2) The manner of placing the members and the use to which the rack is to be
put will together determine the proportions of the members.
(3) For decoration, we might depend entirely upon the good form of the
outline and the stain and grain of the wood. With this particular piece, however,
we shall make use of a decorative form which will be outlined or incised and
colored with a dye.
(4) Since the design is to be made in wood and wood splits easily along the
grain, we must be careful in making an outline not to get sharp points. Also, in
making a decorative design we must avoid thin parts that will bring incised lines
close together. Also, we must take into account in planning the members the
facts of shrinkage or swelling and the strength of the wood. The grain on the
vertical members must extend vertically and that of the horizontal member must
extend from vertical member to vertical member. This to be illustrated by
referring to some similar construction.
(5) In striving for pleasing outlines, or decorative forms either, strive to avoid a
sameness made by using many lines or forms of the same size. “Large, medium
and small” is a key that unlocks many a puzzle as to what causes unpleasant
feelings in both outline and decoration. Long sweeping curves with short snappy
ones, rather than a series made with a compass. Make a special point of the
fact, which almost every boy overlooks, that the simple forms of outline are
invariably the more pleasing. To the beginner design means making something
unlike anything that was ever seen on the earth below or heaven above—hence
the freakish, fussy forms that are usually offered. Try telling the class you are
going to place an excellent form on the board then draw a well proportioned
oblong and watch the expressions on their faces. Yet a well proportioned oblong
with appropriate decorative form is one of the most pleasing of forms. There will
be no need to urge them to make “unique” forms. Their inexperience and their
zeal will produce a sufficient number. Rather urge, or insist that they postpone
search for “unique” form until they have more information.
Illustrate with blackboard sketches as you go along each of these points.
Keep the boys “playing” with outline forms until you have assured yourself they
have done their best. With them, pick out three of the best and place these in
permanent form for keeping—put them on another sheet of paper. Next, start
them on the decoration. The development of a decorative form will come much
harder than the outline. Here again the beginner will want to exhibit “unique”
forms—unique only in that they are founded upon his ignorance. Unless the boy
is not a beginner, it will be necessary in about twenty-four out of every twenty-
five cases to insist that he start with the form you have placed upon the board
for his use. If you were dealing with a few pupils, you might take his “original”
form and step by step get him to work it into a good form. With large classes this
is not possible, nor is it necessary. Simply insist that he place the form given him
in his outline and in so doing he will acquire enough feeling for line and form to
enable him to proceed of his own accord.
(6) Have the boy put on a supporting outline, that is, tell him to draw a line
around his outline and parallel to it. Show the class on the blackboard how this
is to be done.
(7) Put in the main mass and break it up explaining as you do so that you are
seeking to get large, medium and small forms-proportion of parts. Call attention
to the efforts made to keep the lines in harmony.
(8) Call attention to the center of interest you have created. It is unfortunate
that lack of time forbids the boy’s placing colors on these designs. Very
frequently a touch of color is used to create a center of interest, the form for this
in black and white not giving the proper significance at all. A design which in
outline seems to be fussy because of too many parts will, by a proper selection
and placing of colors, be made most pleasing. On the other hand, a design in
outline that seems agreeable may, when in color, not be agreeable because the
colors make certain parts stand out too prominently. A study of the color plate in
Projects in Beginning Woodwork and Mechanical Drawing will make this clear.
(9) If the form proposed happens to illustrate repetition, radiation, symmetry,
or if some boy develops a form that does, take time to say a word about them.
While you will not have time to “teach design” in the few lessons, a word here
and there may serve to awaken further interest on the part of some boy.
After all is said, we recognize that the time is short, that not much can be
done. On the other hand, what little can be done is worth doing and doing well;
its possible significance can not be overestimated.
6. Shop Excursions. In the grammar schools, and more especially in the
high schools, plans should be made for several excursions to near by shops in
which the pupils may get an insight into the workings of related industries. The
saw-mill, lumber yards, planing mills, furniture factories, architectural or drafting-
rooms and, in fact, anything relating to the industrial employment of men and
machinery may be visited.
That the trip may be one of profit the instructor should see to it that the pupils
are prepared for the trip by previous talks on what is to be seen and by after
talks on the meaning of what they saw.
In every case it will be necessary, or at least advisable, to have a time
arranged with the superintendent of the factory to be visited. Pupils should be
given to understand that they are being privileged and must act the part of
gentlemen, refraining from asking needless questions of the workmen or
handling the equipment. In many factories no talking to the men at all is desired.
The questions of young pupils are often impertinent and embarassing without
their intending them so to be. The better plan is, as has just been suggested, to
have the pupils prepared by preliminary talks then take them thru the shop with
eyes and ears only open, clinching the lessons of observation afterward.
Pupils should keep together in solid lines and, should any accident occur, the
instructor should see that any loss to the factory owner or workmen is “made
good.” Usually the class will voluntarily make recompense. It is safer and less
likely to cause embarassment if it is understood beforehand that all members of
the class who go will be expected to help repay the instructor for any money so
expended.
One might think the company well able to stand such loss. It is, but it is not
always the company’s loss. Even if it were, their courtesy ought not to be
abused. We have in mind a mold for an intricate piece of casting representing a
day’s labor for two men ruined by a student’s accidentally brushing against it
with his overcoat. As the men were on “piece work” it meant no loss to the
company, except delay in getting out the finished article. It did mean a loss to
the two men, who could ill afford it. The instructor quietly settled for the damage
or loss and the pupils reimbursed him upon reaching school. This probably
prevented the factory from excluding succeeding classes as undesirables. In
woodworking shops there is little chance for such accidents. Nevertheless
workmen there do not wish their tools or work handled. Each class should bear
constantly in mind, while on the shop excursions, that it is making succeeding
classes welcome or unwelcome in that shop.
7. Stock Bills. Every piece of woodwork made by a pupil consisting of
more than one member should have in addition to the working drawing a
carefully made stock bill. The reason is two-fold: It not only prevents the pupil’s
cutting out stock wrongly thru misreading the drawing, but it saves time for the
pupil. It is a practice that he will have to master later in life if he follows any of
the mechanical trades and is just as essential a part of his shopwork as is the
drawing or woodwork. Where the drawings are made by referring to plates,
experience has shown that many a boy will be able to make a good drawing
without fully interpreting its meaning. The making of the stock bill will show him
his weakness, also it will show the instructor. No boy can make out his stock bill
without being able to read his drawing. After the drawing has been made and
then its stock bill, the boy will have become so conversant with the plans of the
thing he is to make that few mistakes are made in working the wood, that is,
mistakes due to ignorance of the drawing.
STOCK BILL (Form)
Kind of
Grade ________________________ ________________________
Wood
INSTRUCTIONS
All articles in seventh grade will be made of White Pine or Yellow Poplar; those in
eighth grade of Chestnut.
Stock bills are not needed for articles composed of one piece of material only.
Finished sizes are the sizes to which the pieces are to be planed. Your drawing will
tell you these sizes.
Pieces of irregular shape are to be figured at their widest and longest dimensions.
Cutting sizes are obtained from the finished sizes by adding 1⁄4″ to the width and
1⁄2″ to the length. Cutting sizes are the sizes to which you work in sawing out the stock
preparatory to planing it.
All stock will be mill-planed on two surfaces to the correct thickness except that for
the ring toss, spool holder, game-board, and laundry register. Thickness of mill-planed
stock will be the same whether for finished sizes or cutting sizes. On rough stock, or
stock that has not been mill-planed, if the finished size is 3⁄4″ thick the cutting size will
be 1″ thick.
Sometimes it is possible to save material by combining two irregular pieces. The
finished stock sizes will indicate the number of pieces while the cutting size will indicate
the size of the single piece from which they are to be cut.
Remember that length always means “along the grain of the wood,” and that a
piece may be wider than long. Under the word “Pieces” put the number of pieces that
are of the same size.
In the elementary schools the form of stock bill used should be as simple and
explicit as is possible. The appended form is one that has proven satisfactory.
That it may be in convenient form for student use, it has been included with