Statistical Mechanics and Out-of-Equilibrium Systems!

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

Statistical Mechanics and Out-of-Equilibrium Systems!

Luca Giuggioli! Monday 7th Oct - 14:00-16:00 ! Wednesday 9th Oct - 11:15-13:15! Friday 11th Oct - 11:00-12:00!

Langevin and probability description

Suggested reading material


N.G. Van Kampen, Stochastic processes in physics and chemistry , Elsevier 3rd ed. (2007) C.M. Van Vliet, Equilibrium and non-equilibrium statistical mechanics , World Scientific (2008) Coffey, Kalmykov, Waldron, The Langevin Equation , World Scientific 2nd ed. (2005)

N. Pottier, Nonequilibrium statistical physics , Oxford Univ. Press (2010) N. Zwanzig, Nonequilibrium statistical mechanics , Oxford Univ. Press (2001) S. Redner, A guide to first-passage processes , Cambridge Univ. Press 2nd ed. (2007)

Generalities
Equilibrium statistical mechanics is based on the idea of a statistical ensemble, e.g. the thermodynamic properties of a gas can be found by calculating the partition function of a statistical ensemble. Non-equilibrium statistical mechanics is based on the same idea of a statistical ensemble, with the fundamental difference being that there are many non-equilibrium states. Time correlation functions in non-equilibrium systems play the same role as partition function.

I would distinguish two types of out-of-equilibrium systems: 1) those that will relax towards equilibrium 2) those that are driven away from equilibrium reaching a non-equilibrium steady state in the long time limit

time correlation functions


Time correlation functions occur whenever we analyze the statistics of time-dependent quantities

A correlation function is the time-averaged product of two fluctuations around a mean value at different times

C(t1 , t2 ) = T

Z T 1
0

(t1 + s)A (t2 + s) ds A t = t2 t1

If the process is stationary

Langevin equation for a Brownian particle

h[v(t )

D hv(t )inoise ] inoise = 1 g


2

2g t

Velocity correlation function

When ensemble averages are equal to time averages we say that the system is ergodic

Details of the calculations

Noise average

Z t2
0

ds e

g (t2 s)

Z t1
0

ds e

g (t1

s0 )

hF (s)F (s )i = 2De

g (t2 +t1 )

R t1

2g s ds e R0 t2 2g s ds e 0

t2 > t1 t1 > t2

Fluctuation-dissipation relation
From the kinetic theory of gas at equilibrium the average square of a particle speed is equal to

This tells us that there is a relation between the strength of the friction or dissipation and the strength of the noise or random fluctuations

Non-Markovian Langevin equation

Brownian motion of a harmonic oscillator

By elimination of the momentum variable from the equations

Non-Markovian version of the fluctuation-dissipation theorem

Generalized master equation


How do we combine oscillations with decay?

Generalized master equation

Generalized master equation


If we add space how do we represent deterministic and random behaviour? Through the so-called generalized master equation consider the railway track problem where we have both coherent (drift) and incoherent dynamic (scattering)

railway track is a way of representing a particle that drifts (move ballistically) to the right or left until it gets scattered at random times, e.g. by other particles or some imperfection. When that happens it reverse its direction of motion The rate at which direction reversal events occur is Q and the speed at which the particle moves is v

Exercise

P! (x, t ) P! (x, t ) +v = Q [P ( x , t ) t x

P (x, t )]

P (x, t ) t

P (x, t ) ! v = Q [P ( x , t ) x
P(x, t ) = P! (x, t ) + P (x, t )
of the form, i.e. nd the function ! f (t )

P (x, t )]

convert the two equations into a single generalized master equation for !

P(x, t ) = t

knowing that !

Z t
0

ds f (t

s)

2 P(x, s)

x2

P! (x, 0) P (x, 0) = =0 x x

You might also like