Adv Eng Math Lecture Notes 3 v4
Adv Eng Math Lecture Notes 3 v4
kr)
Objectives
• First-order ODEs: Equations including only the first derivative y ′ ( This equations may contain y and any
given functions of x. )
• Separable ODEs
• Exact ODEs
• Linear ODEs
1
Chapter 1
First-Order ODEs
1.1.1 Modeling
• Ordinary Differential Equation (ODE): An equation that contains one or several derivatives of an
– Examples:
y ′ = cos x, y ′′ + 9y = e−2x , y y ′′ − 3
2 (y ′ )2 = 0.
• Partial Differential Equation (PDE): An equation involving partial derivatives of an unknown function
2
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
– Example:
∂2u ∂2u
+ = 0.
∂x2 ∂y 2
Order: The highest derivative of the unknown function appearing in the equation.
• Examples:
3
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
First-Order ODE. Equations that contain only the first derivative y ′ , and may also contain y or any functions
of x.
Concept of Solution
A function
y = h(x)
is a solution of the ODE (4) on an open interval a < x < b if h(x) is defined and differentiable there, and if
substituting y = h and y ′ = h′ into the ODE makes it an identity. The graph of h is called a solution curve.
Here, a < x < b excludes the endpoints a and b. This notation also covers infinite intervals such as
Solution: A function (or set of functions) that makes the equation hold true.
• Singular Solution (see Problem 16): An additional solution that cannot be obtained from the general
solution.
(y ′ )2 − x y ′ + y = 0.
4
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
• General solution: y = c x − c2 .
• Particular solution: y = 2x − 4.
x2
• Singular solution: y = 4 .
An initial value problem is an ODE together with a specified value of the unknown function at a particular
y(x) = c e3x .
Example 5. Suppose we have an initial 0.5 g of a radioactive substance. Find the amount y(t) present at any
time t.
• Physical Information. Experiments show that a radioactive substance decays at a rate proportional to
dy
the amount present, i.e. dt = −k y.
• y(0) = 0.5.
5
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
dy
= −k y =⇒ initial condition: y(0) = 0.5.
dt
y(t) = c e−kt .
dy
= −0.5 k e−kt = −k y and y(0) = 0.5.
dt
lim y(t) = 0.
t→∞
Definition: A direction field is a graphical representation that includes pairs of grid points (x, y) and small line
segments through those points with slopes matching y ′ = f (x, y). Each line segment corresponds to the tangent
• You need not solve the ODE explicitly to get a qualitative understanding.
• The method shows the entire family of solutions and their typical behaviors.
6
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Euler’s method yields approximate solution values at equidistant x-values, starting from an initial value y(x0 ) =
y0 .
x1 = x0 + h, y1 = y0 + h f (x0 , y0 ),
x2 = x0 + 2h, y2 = y1 + h f (x1 , y1 ),
• Typically we plot the approximate values xn , yn to visualize the numerical solution’s trajectory.
Figure 1.1: Illustration of Euler’s Method: error between true solution and approximate steps.
7
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Definition: A differential equation is called separable if all the terms involving y can be collected (multiplied)
with the derivative, and all the terms involving x can be placed on the other side of the equality. Symbolically,
Example 1. Solve y′ = 1 + y2 .
Z Z
dy 1
y′ = 1 + y2 =⇒ = dx =⇒ dy = 1 dx + C.
1 + y2 1 + y2
Hence
Mixing problems occur frequently in chemical engineering. We solve the basic model involving a single tank.
Suppose the tank in Fig. 11 contains 1000 gal of water, initially with 100 lb of salt dissolved. Brine enters at 10
gal/min, each gallon containing 5 lb of salt. The mixture is kept uniform by stirring, and drains at 10 gal/min.
• Step 1: Setting up a model. Let y(t) be the amount of salt (lb) in the tank at time t. By the “balance
law,”
dy y
y
= (10 gal/min × 5 lb/gal) − 10 gal/min × 1000 lb/gal =⇒ y ′ = 50 − .
dt | {z } | {z } 100
50 lb/min inflow y
100 lb/min outflow
8
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
y dy 1
y ′ = 50 −
=⇒ =− y − 5000 .
100 dt 100
Rewrite:
dy 1
=− dt.
y − 5000 100
1 t
ln|y − 5000| = − t + C∗ =⇒ y − 5000 = c e− 100 ,
100
∗
where c = eC . Using y(0) = 100 gives
t
100 − 5000 = c e0 =⇒ c = −4900, so y(t) = 5000 − 4900 e− 100 .
Certain first-order ODEs that are not obviously separable can become separable through a suitable change of
y
y′ = f can be reduced to separable form by the substitution y = ux
x
y du dx
⇒ y′ = f ⇒ u′ x + u = f (u) ⇒ =
x f (u) − u x
y
y = ux ⇒ u= , y ′ = (ux)′ = u′ x + u
x
′ 2 2 1
′ y x ′ 1 1
2xyy = y − x ⇒ y = − ⇒ ux+u= u−
2 x y 2 u
9
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
2u 1
⇒ du = − dx
u2 +1 x
c y 2 c
⇒ u2 + 1 = ⇒ +1= ⇒ x2 + y 2 = cx
x x x
M (x, y) dx + N (x, y) dy = 0
is called exact if the left-hand side equals the total differential du of some function u(x, y); in other words,
∂u ∂u
du = dx + dy = M (x, y) dx + N (x, y) dy.
∂x ∂y
∂u ∂u
Hence M = ∂x , N= ∂y .
• If M (x, y) = ∂u
∂x , we can integrate M (x, y) w.r.t. x:
Z
u(x, y) = M (x, y) dx + k(y),
∂u
where k(y) is an arbitrary function of y. Then we use N = ∂y to determine k(y).
Z
u(x, y) = N (x, y) dy + l(x).
10
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Example 1.
Solve
cos(x + y) dx + 3y 2 + 2y + cos(x + y) dy = 0.
∂M
M (x, y) = cos(x + y), = − sin(x + y).
∂y
∂N
N (x, y) = 3y 2 + 2y + cos(x + y), = − sin(x + y).
∂x
∂M ∂N
Since ∂y = ∂x , it is exact.
Z Z
u(x, y) = M (x, y) dx + k(y) = cos(x + y) dx + k(y) = sin(x + y) + k(y).
Then ∂u
∂y = cos(x + y) + k ′ (y). But we must match N (x, y), so
u(x, y) = sin(x + y) + y 3 + y 2 .
u(x, y) = sin(x + y) + y 3 + y 2 = c.
∂u ∂u
dy = cos(x + y) dx + cos(x + y) + 3y 2 + 2y dy = 0.
du = dx +
∂x ∂y
Some ODEs can be made exact by multiplying by a suitable nonzero function F (x, y), called an integrating factor.
11
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Example 4. Consider −y dx + x dy = 0.
∂ ∂
(−y) = −1, (x) = 1.
∂y ∂x
1
That equation is not exact. But multiplying by x2 yields
y 1
− 2
dx + dy = 0,
x x
∂ ∂
FP = FQ .
∂y ∂x
Sometimes we guess that F depends only on x or only on y. If that fails, we look for more advanced forms.
F P dx + F Q dy = 0.
Golden Rule: If you cannot solve your problem, try to solve a simpler one.
Case 1) F = F (x)
1 dF 1 ∂P ∂Q
= R(x) = − .
F dx Q ∂y ∂x
12
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Therefore,
Z Z 1 ∂P ∂Q
F (x) = exp R(x) dx = exp − dx .
Q ∂y ∂x
Case 2) F ∗ = F ∗ (y)
1 dF ∗ ∗ 1 ∂Q ∂P
= R (y) = − .
F ∗ dy P ∂x ∂y
Hence,
Z Z 1 ∂Q ∂P
∗ ∗
F (y) = exp R (y) dy = exp − dy .
P ∂x ∂y
Example.
Let
Then,
∂P
= ex+y + ey + yey = ex+y + ey (1 + y)
∂y
∂Q
= ey
∂x
∂P ∂Q
Since ∂y ̸= ∂x , the equation is not exact.
13
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Simplify:
Let
Check exactness:
∂ P̃ ∂ Q̃
= 1, =1 ⇒ Equation is exact.
∂y ∂x
Z
∂u
= ex + y ⇒ u(x, y) = (ex + y) dx = ex + xy + k(y)
∂x
Now compute:
∂u !
= x + k ′ (y) = x − e−y ⇒ k ′ (y) = −e−y ⇒ k(y) = e−y
∂y
So,
u(x, y) = ex + xy + e−y
Final Answer:
ex + xy + e−y = 1 + e
14
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Linear ODE: An ODE that is linear in both the unknown function and its derivatives. For example,
whereas
y ′ + p(x) y = r(x).
y ′ + p(x) y = 0
Z
′ dy R
y = −p(x) y =⇒ = −p(x) dx =⇒ ln |y| = − p(x) dx + C =⇒ y = C e− p(x) dx
.
y
15
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
y ′ + p(x) y = r(x).
R
p(x) dx
µ(x) = e .
d h i
µ(x) y ′ + µ(x) p(x) y = µ(x) r(x) =⇒ µ(x) y = µ(x) r(x).
dx
Z R hZ R i
− p(x) dx p(x) dx
µ(x) y = µ(x) r(x) dx + C =⇒ y=e e r(x) dx + C .
Example. 0
Z
x
p = −1, r=e , h= p dx = −x
Z Z Z
−h h x −x x x
⇒ y=e e r dx + c = e e e dx + c = e 1 dx + c = ex (x + c)
⇒ y = ex x + cex
16
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Example 1.
Solution. Here,
Z Z
h= p dx = tan x dx = ln | sec x|
Z
y(x) = cos x 2 sin x dx + c = cos x (−2 cos x + c) = c cos x − 2 cos2 x
y(0) = 1 ⇒ 1 = c · 1 − 2 · 12 ⇒ c = 3
y = 3 cos x − 2 cos2 x
Here, 3 cos x is the response to the initial data, and −2 cos2 x is the response to the input sin 2x.
17
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
We set
1−a
u(x) = [y(x)]
Example 4: Logistic Equation Solve the following Bernoulli equation, known as the logistic equation (or
Verhulst equation):
y ′ = Ay − By 2
Z Z Z
p = A, r=B ⇒ h= p dx = Ax & u = e−h eh r dx + c = e−Ax eAx B dx + c
−Ax B Ax B −Ax
=e e +c = + ce
A A
1 1
y= =
u B
+ ce−Ax
A
18
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
The logistic equation is important in modeling population growth of plants, animals, or humans over time.
• If B = 0, the equation y ′ = A y yields exponential growth y(t) = c eAt , known as Malthus’s law.
Orthogonal Trajectory: A family of curves that intersect a given family of curves at right angles.
• Step 1. Find the ODE y ′ = f (x, y) whose general solution represents the original family of curves.
y = c x2 .
• Differentiate: y ′ = 2 c x.
• But c = y
So y ′ = 2 y 2y
x2 . x2 x= x .
• Solve that ODE to get the family of orthogonal curves (e.g. by separating variables). y 2 + 21 x2 = c∗
• No solution,
19
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
Examples.
x3
• y ′ = x2 , y(0) = 1. Precisely one solution, y = 1 + 3 .
• Some ODEs can admit infinitely many solutions for a given initial condition (not typical for well-behaved
ODEs).
Problem of Existence: Under what conditions does an IVP have at least one solution?
Problem of Uniqueness: Under what conditions does an IVP have at most one solution?
|x − x0 | < a, |y − y0 | < b.
Assume |f (x, y)| ≤ K for all (x, y) in that region, for some constant K. Then there exists at least one
(This statement can vary slightly in different textbooks, but the idea is the same: if f is bounded, the
20
Advanced Engineering Mathematics (Spring 2025)-SNU CEE Hyoseob Noh (hyoddubi1@[Link])
∂f
Let f and its partial derivative ∂y be continuous in the rectangle
|x − x0 | < a, |y − y0 | < b.
∂f
|f (x, y)| ≤ K and ∂y (x, y) ≤M
has precisely one solution y(x). By the existence theorem, it exists at least for
b
|x − x0 | < min a, ,
K
21
References
22