DEPARTMENT OF PHYSICS

INDIAN INSTITUTE OF TECHNOLOGY MADRAS

PH1010 Physics I Tutorial 3 - Solution
February 28, 2025

1) The Damped Harmonic Oscillator Equation

We start with the damped harmonic oscillator (H.O.) equation:

ẍ + 2γ ẋ + ω02 x = 0,
b b
where 2γ = m, i.e. γ = 2m .

Trial solution. Let us try a solution of the form

x(t) = A eαt ,

as done in class. Substituting x(t) = A eαt into the differential equation gives the characteristic
equation
α2 + 2γα + ω02 = 0.
Solving for α yields the roots q
α = −γ ± γ 2 − ω02 .

General solution. Hence, the general solution can be written as

 √ 2 2 √ 2 2 
x(t) = e−γt A e γ −ω0 t + B e− γ −ω0 t ,

Now, the decay is dominated by

 √ 
− γ− γ 2 −ω 2 t
Ae ⇒ Rate of decay is
s !
p ω 2
γ − γ 2 − ω2 = γ 1 − 1 − 2
γ
ω2
 
≈γ 1−1+ 2
2γ
2
ω
= therefore, Rate decreases in the overdamped regime as 1/γ
2γ

1
2) Critical Damping
There are many ways of showing that in the case of critical damping, one obtains

x(t) = (A + B t) e−γt

Let us look at an alternate method:

d
Let D be the operator: D ≡.
dt
Now when ω = γ, which is the case for critical damping, the governing equation can be written as

(D + γ)(D + γ) x(t) = 0

Let us call
(D + γ) x(t) = u(t)

⇒ (D + γ)u(t) = 0
Immediately giving the solution
u(t) = Ae−γt
Now consider  
d d
x(t)eγt = eγt + γ x(t) = eγt u(t)

dt dt
d
x(t)eγt = Beγt e−γt

dt
Integrating once again we obtain the general solution

x(t) = (A + B t) e−γt .

Another way of approaching the same problem: for γ < ω,

p
Ω = ω2 − γ 2 .

Then the solutions

e−γt cos(Ωt) and e−γt sin(Ωt)
are two independent solutions.
sin(Ωt)
This in turn, implies that e−γt cos(Ωt) and e−γt Ω are independent solutions.
In the critically damped case, we have ω = γ, so that
p
Ω = ω 2 − γ 2 → 0.

In this limit, the first solution is

e−γt ,
while the second solution can be found by taking

sin(Ωt)
lim e−γt = t e−γt .
Ω→0 Ω
Hence, the two independent solutions are

e−γt and t e−γt .

Thus, the general solution may be written as the linear combination

x(t) = A e−γt + B t e−γt .

2
3)
(a) Critically Damped Case
For critical damping, one may write

x(t) = e−γt A + B t .

If x(t) must vanish at some t = t∗ , then

∗ A
x(t∗ ) = e−γt (A + B t∗ ) = 0 =⇒ A + B t∗ = 0 =⇒ t∗ = − .
B
No other zeros are possible in this form.

(b) Overdamped Case

For overdamping, the solution can be written
 
x(t) = e−γt A eΩt + B e−Ωt ,

where p
Ω = γ 2 − ω2 .
If we require x(t∗ ) = 0 for some t∗ , then
∗ ∗ ∗ ∗
A eΩt + B e−Ωt = 0 =⇒ A eΩt = − B e−Ωt ,

which implies
∗ B 1  B
e2 Ωt = − =⇒ t∗ = ln − .
A 2Ω A
Hence, one finds a specific t∗ (provided the argument of the logarithm is positive, etc.).

4)
Resonance Amplitude.
A typical driven-damped harmonic oscillator amplitude near resonance is proportional to
1
α(ωd ) ∝ q .
2 2


ω 2 − ωd + 4 γ 2 ωd2

If we define the function

f (ωd ) = (ω 2 − ωd2 )2 + 4 γ 2 ωd2 ,
then the resonant frequency(s) can be found by differentiating f (ωd ) with respect to ωd and setting
the derivative to zero:
d p
f (ωd ) = 0 =⇒ ωd = ω 2 − 2 γ 2 .
dωd
Hence the response amplitude α(ωd ) attains a maximum at this value of ωd (assuming ω 2 > 2 γ 2 , so
that the square root is real).

5) Simple Harmonic Motion: Time and Spatial Averages

Consider a particle of mass m on a spring with spring
q constant k, executing a simple harmonic motion
k
(SHM) of amplitude A and angular frequency ω = m. The displacement is

x(t) = A sin(ωt).

3
 • Potential energy:
1 2 1
2 1
U (t) = kx (t) = k A sin(ωt) = k A2 sin2 (ωt).
2 2 2

• Kinetic energy:
1 1
2 1
T (t) = mẋ2 (t) = m Aω cos(ωt) = m A2 ω 2 cos2 (ωt).
2 2 2

(a) Time Averages

2π
Over one full period T = ω , the time average of a quantity f (t) is
Z T
1
⟨f ⟩T = f (t) dt.
T 0

Average Potential Energy. Since

1
sin2 (ωt) averages to over one full period,
2
we have
T
k A2
Z
1 1 1 1
⟨U ⟩T = k A2 sin2 (ωt) dt = k A2 · = .
T 0 2 2 2 4
2
Using k = mω , this becomes
m ω 2 A2
⟨U ⟩T = .
4

Average Kinetic Energy. Similarly, since cos2 (ωt) also averages to 1/2 over one period,
T
m ω 2 A2
Z
1 1 1 1
⟨T ⟩T = m A2 ω 2 cos2 (ωt) dt = m A2 ω 2 · = .
T 0 2 2 2 4

Hence,
m ω 2 A2
⟨T ⟩T = ⟨U ⟩T = .
4
Because the total mechanical energy is E = T (t) + U (t) = 12 m ω 2 A2 , it follows that

E
⟨T ⟩T = ⟨U ⟩T = .
2

(b) Spatial Averages

Consider a spatial average over x from 0 to A. For instance,
A A A
x3 k A2 mω 2 A2
Z Z 
1 1 2 k 2 k
⟨U ⟩x = k x dx = x dx = = = .
A 0 2 2A 0 2A 3 0 6 6

Similarly,
Z A
1 1
m ω 2 A2 − x2 dx,


⟨T ⟩x =
A 0 2
2k A2 2mω 2 A2
= =
6 6
1
noting that for a turning point x = A, the kinetic energy T = 0 and T (x) = E − U (x) = 2 m ω 2 A2 −
1 2 2
2 mω x .
A straightforward integral or geometric argument shows ⟨T ⟩x = 2 ⟨U ⟩x .

4
Graphical Interpretation
A convenient way to see these relationships is to plot
1 1
m ω 2 x2 m ω 2 A2 − x2


U (x) = and T (x) = E − U (x) =
2 2
versus x from −A to +A. The total energy is E = 21 m ω 2 A2 . One can show by computing areas under
these curves that the (signed) area under T (x) from −A to +A is twice the area under U (x) from −A
to +A. This is yet another way of seeing the ratio ⟨T ⟩x : ⟨U ⟩x = 2 : 1.

6) Forced Damped Oscillation

(a) Steady-State Solution
Consider the driven, damped oscillator
F0
ẍ + 2γ ẋ + ω 2 x = cos(ωd t). (1)
m
In the long-time limit t → ∞, the steady-state (particular) solution is of the form

F0

xp (t) = cos ωd t − ϕ , (2)
mR
where
2 γ ωd
q
2
R = ω 2 − ωd2 + 4 γ 2 ωd2 , tan ϕ = . (3)
ω 2 − ωd2
Hence, the amplitude of the forced motion is
F0
A = .
mR
The velocity in a steady state is
F0 ωd

ẋp (t) = − sin ωd t − ϕ = v (4)
mR

Instantaneous Power. The driving force is F (t) = F0 cos(ωd t). Then, the instantaneous power
input is
P (t) = F (t) ẋp (t) = F0 cos(ωd t) ẋp (t).

5
One may show that the time-averaged power delivered by the driver over one full cycle is
−F02 ωd
⟨P (t)⟩ = ⟨cosωd t + sin (ωd t − ϕ⟩T
mR
2π/ωd
F 2 ω2 sin (2ωd t − ϕ) + sin(−ϕ)
Z
=− 0 d dt
2πmR 0 2
Z 2π/ωd
F 2 ωd sin ϕ
sin (2ωd t − ϕ) = 0 ⇒ ⟨P (t)⟩ = 0
0 2mR

F02 ωd sin ϕ
P (t) = (5)
2mR
Using geometry or algebra relating sin ϕ to γ, ω, ωd can give a more explicit expression. In terms of
F0
the amplitude A = mR , we can write

F02 ωd 2γωd
P (t) = = m γ ωd2 A2
2 m R2
again consistent with the idea that the driver must supply energy at the rate it is dissipated by
damping.

(b) Energy Lost to Damping

Over one full cycle of the motion, the total energy lost to the damping force can be written as
Z Z 2π
ωd dx
∆E = fdis dx = fdis dt. (6)
cycle 0 dt

The damping force is (by Newton’s law of viscous damping) fdis = − 2 γ m ẋ. Inserting the steady-state
solution xp (t) and integrating over one period Td = 2π/ωd yields
Z Td
2
∆E = 2 γ m ẋp (t) dt.
0
Z 2π/ωd  2
dx
= 2γm dt
0 dt
F0 F0
x= cos (ωd t − ϕ) ; ẋ = − ωd sin (ωd t − ϕ)
mR mR
2π/ωd
2γmF02 ωd2 0 sin2 (ωd t − ϕ) dt
R
=
m2 R2
2 2
2γF0 ωd 1 2π F 2 ωd π
= 2
= 2γ 0 2 = 2γA2 mπωd
mR 2 ωd mR
Because ∆E is the energy lost per cycle, the rate of energy dissipation (i.e., the average power lost)
is
∆E ∆E
= 2π = ⟨P (t)⟩ = m γ A2 ωd2 ,
Td ωd

matching the power input from the driving force in a steady state.
The rate at which work is done = the rate at which energy is dissipated.

(c) Average Power and Resonance Condition

From the above, the time-averaged power absorbed by the oscillator is often given by

F02 ωd2 γ 1
⟨P ⟩ = m γ ωd2 A2 =
2 . (7)
m ω − ωd + 4 γ 2 ωd2
2 2

The amplitude A or the power ⟨P ⟩ as functions of ωd typically show a peak (resonance) near ωd ≈ ω,
with a small shift depending on γ.

6
 Consider the function;
x
f (x) =
2 ,
ω2 −x + 4 γ2 x
where x stands for a driving frequency, and lets finds the maximum by setting f ′ (x) = 0.
1 x
f ′ (x) = 0 ⇒ 2 2



2 − 2 2 x − ω + 4γ
2
(ω − x) + 4γ x 2 2
(ω 2 − x) + 4γ 2 x

2
ω 2 − x + 4γ 2 x = x 2x − 2ω 2 + 4γ 2



2
⇒ ω 2 − x = 2x x − ω 2

⇒ x − ω 2 2x − x − ω 2 = 0




x = ±ω 2
ωd = ω
Thus, a bit of algebra reveals that, for light damping, the maximum occurs near x ≈ ω. In the
strict limit of negligible damping, this is exactly ωd = ω.

7)
Undamped Oscillator γ = 0

ẍ + ω 2 x = F (t).

Particular solution for t ≤ T : We look for a particular solution of the form

xp (t) = C t.
F0 t
Then, over t ≤ T, the forcing is F (t) = T , so the equation becomes

F0 t
ẍ + ω 2 x = .
T
Since ẍp (t) = 0 for xp (t) = C t, we plug into
F0 t
ω2 C t = .
T
Hence,
F0
C = (call this result (2)).
ω2 T
General solution for t ≤ T :
x(t) = xp (t) + xh (t)
where the homogeneous solution to ẍ + ω 2 x = 0 is

x(t) = C1 cos(ωt) + C2 sin(ωt) + Ct.

Thus,
F0
x(t) = C1 cos(ωt) + C2 sin(ωt) + t (call this (1)).
ω2 T

7
Initial conditions: We assume x(0) = 0 and ẋ(0) = 0. From x(0) = 0 we get

C1 = 0

(because sin(0) = 0 and t = 0). From ẋ(0) = 0, note

F0
ẋ(t) = −ω C1 sin(ωt) + ω C2 cos(ωt) + .
ω2 T
F0
At t = 0, ẋ(0) = ω C2 sin(0) + ω2 T = 0. Hence

F0 F0
ω C2 + =0 =⇒ C2 = − .
ω2 T ω3 T
Call this result (3).
Solution for t ≤ T : Putting C1 = 0 and C2 from above, we get
 
F0 sin(ωt)
x(t) = 2 t − (call this (4)).
ω T ω

At t = T, we then have

F0 h sin(ω T ) i
x(T ) = T − (call this (5)),
ω2 T ω
F0 h i
ẋ(T ) = 2 1 − cos(ω T ) = V0 (call this (6)).
ω T


Shifted time for t ≥ T : We now look at a situation where x(T ), ẋ(T ) from (5) and (6) serve as
new initial conditions at t = T . Define a shifted time

τ = t − T.

We want to write the equation of motion for t ≥ T .

Equations of motion in the regime t ≥ T :

d2 x
+ ω02 x − F0


ω02
= 0,
dτ 2
which describes a “shifted” harmonic oscillator. Let
F0
y(τ ) = x(τ ) − .
ω02

Then
d2 y
+ ω 2 y = 0,
dτ 2
and the general solution is
y(τ ) = A sin(ω τ ) + B cos(ω τ ).
Hence
F0
x(τ ) = A sin(ω τ ) + B cos(ω τ ) +
ω2

Boundary conditions at τ = 0 (t = T ): We must match x(T ) and ẋ(T ) given by (5) and (6). From
(5):
F0 F0 h sin(ω0 T ) i F0
x(T ) = B + 2 = 2 T− =⇒ B = − 3 sin(ω0 T ).
ω0 ω0 T ω0 ω0 T
From (6):
F0 h i
ẋ(T ) = ω0 A = 1 − cos(ω0 T ) .
ω02 T

8
Hence
F0  
A = 3 1 − cos(ω0 T ) .
ω0 T

Amplitude of the resulting oscillations: The problem asks us to determine the amplitude, which
involves p
A2 + B 2 .
We can write
F0   F0
A = 3 1 − cos(ω0 T ) , B = − sin(ω0 T ).
ω0 T ω03 T
Combining terms,

F02 h
2 2
i F02 h 2
i
A2 +B 2 = 1−cos(ω0 T ) + sin (ω0 T ) = 1 − 2 cos(ω 0 T ) + cos2
(ω 0 T ) + sin (ω0 T ) .
ω06 T 2 ω06 T 2

Since cos2 θ + sin2 θ = 1, the expression in brackets simplifies to

1 − 2 cos(ω0 T ) + 1 = 2 1 − cos(ω0 T ) .

Hence
F0
p q

A2 + B 2 = 2 1 − cos(ω0 T ) .
ω03 T
We recognize 1 − cos(θ) = 2 sin2 θ


2 , so
r
p F0 ω T 
0 2 F0
2 · 2 sin2 ω0 T


A2 + B2 = 3 = 3 sin 2 .
ω0 T 2 ω0 T

Hence the amplitude of the oscillations is often written as

2 F0 ω0 T


A = sin .
ω03 T 2