PHY 412: Statistical Physics
24-25 Semester II
Assignment – 3
Problems in black are for submission; Blue ones are for practice. Due date: March 6th in Class
1. Random walks with a constant drift term will have a net correlation between steps. However, it can be
reduced to random walks without drift by shifting to the moving reference frame. Suppose we have a
random walk with steps independently drawn from a uniform density 𝜌(𝑙) on [0,1), but with a nonzero
mean 〈𝑙〉 = 𝑙 ̅ ≠ 0.
(a) Argue that the sums 𝑆𝑁′ = ∑𝑁 ̅
𝑛=1(𝑙𝑛 − 𝑙 ) describe random walks with zero mean in a moving
reference frame.
(b) Argue that the variance of these random walks (the squared standard deviation) is the same as the
variance 〈(𝑆𝑁 − ̅̅̅
𝑆𝑁 )2 〉 of the original random walks.
2. Consider a colloidal particle of radius 0.1 μm, density 2 g cm-3 moving in water (density 1 g cm-3,
viscosity 1 mPa.s). Answer the following questions:
(a) Calculate the sedimentation velocity 𝑉𝑠 of the particle.
(b) Calculate the diffusion constant 𝐷 of the particle.
(c) Calculate the average distance 𝑉𝑠 𝑡 that the particle moves in time 𝑡, and its root-MSD √𝐷𝑡 for 𝑡 =
1 ms, 1 s, 1 hour, 1 day. Discuss when do we need to consider Brownian motion.
3. Consider the electrical circuit given at the right; the circuit is made of a resistor (resistance R) and a
capacitor (capacitance C). Charge Q stored in the capacitor is 0 on average, but can be positive or
negative due to the thermal fluctuations of charge carriers. Answer the following questions:
(a) Write down the equation describing the fluctuating charge Q(t), under the
influence of a randomly fluctuating voltage 𝜓𝑟 (𝑡). Comparing the
equation with the Langevin equation, show that the fluctuating voltage
satisfies the fluctuation-dissipation relation (also called as Nyquist
theorem in electrical engineering context), 〈𝜓𝑟 (𝑡)𝜓𝑟 (0)〉 = 2𝑅𝑘𝐵 𝑇𝛿(𝑡).
(b) Calculate the time correlation functions 〈𝑄(𝑡)𝑄(0)〉 and 〈𝐼(𝑡)𝐼(0)〉,
where 𝐼(𝑡) = 𝑄̇(𝑡) is the current flowing in the resistor.
𝑑𝑥 𝑘 𝑥
4. The Langevin equation for a Brownian particle in a harmonic potential =− 𝑥 + 𝑣r (𝑡) = − +
𝑑𝑡 𝛾 𝜏
𝑡
𝑣r (𝑡), gives a steady-state solution for 𝑥(𝑡) as 𝑥(𝑡) = ∫0 𝑑𝑡1 𝑒 −(𝑡−𝑡1 )/𝜏 𝑣r (𝑡1 ). Answer the following
questions:
(a) Calculate 〈𝑥(𝑡)2 〉 from the solution 𝑥 (𝑡) using the time-correlation of random-velocity (𝑣r (𝑡) =
𝐹𝑟 (𝑡)
𝛾
, 𝐹𝑟 (𝑡) is the random force) 〈𝑣𝑟 (𝑡)𝑣𝑟 (𝑡 ′ )〉 = 2𝐷𝛿(𝑡 − 𝑡 ′ ).
(b) Calculate the time correlation function 〈𝑥(𝑡)𝑥(𝑡 ′ )〉 for 𝑡 > 𝑡 ′ , and 𝑡 < 𝑡 ′ .
𝐷
(c) Show that 〈𝑥̇ (𝑡)𝑥̇ (0)〉 = 2𝐷𝛿(𝑡) − 𝜏 𝑒 −|𝑡|/𝜏 starting from the Langevin equation given above.
5. Consider two Brownian particles connected by a harmonic spring, as shown in the figure at right. The
𝑑𝑥1
Langevin equations for their positions 𝑥1 and 𝑥2 are written as: 𝜁 =
𝑑𝑡
𝑑𝑥2
−𝑘(𝑥1 − 𝑥2 ) + 𝐹𝑟1 (𝑡), and 𝜁 𝑑𝑡
= −𝑘 (𝑥2 − 𝑥1 ) + 𝐹𝑟2 (𝑡). Answer
the following questions:
(𝑒)
(a) Suppose that a constant force 𝐹1 is applied to the particle 1 for t > 0. Calculate the average
displacements of the particles 〈𝑥1 (𝑡) − 𝑥1 (0)〉𝐹(𝑒) and 〈𝑥2 (𝑡) − 𝑥2 (0)〉𝐹(𝑒) in this situation.
1 1
2
(b) Discuss the behavior of 〈(𝑥1 (𝑡) − 𝑥1 (0)) 〉 at short time (𝑡 ≪ 𝜁/𝑘) and at long time (𝑡 ≫ 𝜁/𝑘).
6. Consider a Brownian particle in one dimension along x-axis. There is a source of propulsion that drives
the particle at a constant speed along the positive x direction. Hence the velocity of the particle at any
time t can be given by, v(t) = vr (t) + vx, where vr (t) is the random velocity associated with the random
force that the particle experiences, and vx is the time-independent propulsion speed. For this system
(a) Calculate and comment on the velocity autocorrelation function < v(t)v(0)>
(b) From the velocity autocorrelation, calculate the force autocorrelation < ξ(t)ξ(0)>
(c) Comment on the validity of fluctuation-dissipation theorem (FDT) from (ii)
(d) Calculate the displacement autocorrelation <(x(t1)-x(0))(x(t2)-x(0))>
(e) From displacement autocorrelation calculate and comment on MSD <(x(t) - x(0))2>
