This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Dispersive Quantization of the linear Schr¨odinger equation Sudarshan Balakrishnan September 20, 2011 Abstract . The study of linear dispersive partial differential equations with piecewise constant periodic initial data leads to quantized structures at times which are rational multiples of π and fractal profiles at irrational multiples of π . We will present an overview of these results for the Schrodinger equation on the torus. § Introduction The purpose of this paper is to study the behavior of the solution to the Schr¨ odinger wave equation with specific initial conditions. First, we go over the methods to analyze the linearly dispersive wave equation established in [ ? ]. Next, we proceed to carry out the same analysis for the Schr¨ odinger wave equation and write programs in MATLAB to plot its solution. We also present an overview of the Fast Fourier Algorithim(FFT) and plot the error involved in the solution to the Schr¨ odinger wave equation. § 1 Linearly dispersive wave equation The Linearly dispersive wave equation is a specific periodic initialboundary value prob lem on the interval 0 ≤ x ≤ 2 π given by ∂u ∂t = ∂ 3 u ∂x 3 (1) With initial conditions: u ( t, 0) = f ( x ) , (2) Satisfying: u ( t, 0) = u ( t, 2 π ) , (3) ∂u ∂x ( t, 0) = ∂u ∂x ( t, 2 π ) , (4) 1 ∂ 2 u ∂x 2 ( t, 0) = ∂ 2 u ∂x 2 ( t, 2 π ) . (5) First, we take the initial data as a unit step f ( x ) = σ ( x ) = 0 , 0 < x < π ; 1 , π < x < 2 π (6) The precise values assigned at its discontinuities are not important, although choosing f ( x ) = 1 2 at x = 0 ,π, 2 π is consistent with Fourier analysis. The boundary conditions allow us to extend the initial data and solution to be 2 π periodic functions in x. In order to construct the solution, let us consider a timedependent Fourier Series u ( t,x ) ∼ ∞ X k =∞ b k ( t ) e ikx . (7) As we substitute (7) in equation (1) we get b k ( t ) will satisfy the elementary linear ordinary differential equation: db k dt = ik 3 b k ( t ) . (8) Then, we get: b k ( t ) = b k (0) e ik 3 t . (9) Resubstituting (9) in (7), we get : u ( t,x ) ∼ ∞ X k =∞ b k (0) e i ( kx k 3 t ) . (10) Observe that the solution (10) is 2 π periodic in both t and x . Moreover, since the kth summand is a function of , periodic waves of frequency k move with wave speed k 2 ie., ω ( k ) = k 2 , thereby justifying the dispersive nature of the system. As we solve for the initial value problem, we note the initial condition for the Fourier series given by: f ( x ) ∼ ∞ X k =∞ c k e ikx . (11) Where we have : c k = 1 2 π Z 2 π f ( x ) e ikx dx. (12) Now, when we equate u (0 ,x ) = f ( x ), we see that b k (0) = c k which will be the fourier coefficients of the initial data. We observe the step function (6), we notice that the fourier coefficients are given by: 2 b k (0) = i πk k odd ; 1 2 π k= 0 (13) Note, here that when k = even, we have that b k (0) equals zero. Now, when we combine this with (10), we get that the solution to the Linearly dispersive wave equation is given...
View
Full Document
 Spring '08
 wormer
 Equations, Fourier Series, wave equation, Dirac delta function

Click to edit the document details