This preview shows pages 1–4. Sign up to view the full content.
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: ACM 100c  Methods of Applied Mathematics Week 8 May 17, 2009 1 Laplace equation in a semiinfinite strip In this section, we consider the solution of the Laplace equation in 2D Cartesian coordi nates, but in a semiinfinite region: ∇ 2 φ ( x,y ) = 0 ≤ x ≤ L < y < + ∞ . The boundary conditions here are given by φ (0 ,y ) = g 1 ( y ) , φ ( L,y ) = g 2 ( y ) , φ ( x, 0) = f ( x ) . We have to assume that the solution behaves well as y → ∞ We can think of the solution as the sum of the solutions of two subproblems: φ = φ 1 + φ 2 where ∇ 2 φ 1 ( x,y ) = 0 with boundary conditions φ 1 ( x, 0) = 0 φ 1 (0 ,y ) = g 1 ( y ) φ 1 ( L,y ) = g 2 ( y ) , and where ∇ 2 φ 2 ( x,y ) = 0 with boundary conditions φ 2 ( x, 0) = f ( x ) φ 2 (0 ,y ) = 0 φ 2 ( L,y ) = g 2 ( y ) . We look at the solution for φ 1 and φ 2 in turn. For φ 2 , the solution is easily accomplished using finite transforms: φ 2 ( x,y ) = ∞ X n =1 A n ( y ) sin( nπx/L ) . 1 The ODE’s for the coefficients A n are d 2 A n dy 2 n 2 π 2 L 2 A n = 0 . From this, we can see that φ 2 ( x,y ) = ∞ X n =1 F n exp( nπy/L ) sin( nπx/L ) , where F n = 2 L Z L f ( x ) sin( nπx/L ) dx. In order to get φ 1 we use the Fourier sine transform since this is a useful approach when you have Dirichlet boundary conditions at one end of the semiinfinite interval. We write the solution as φ 1 ( x,y ) = Z ∞ U S ( x,k ) sin( ky ) dy, U S ( x,k ) = 2 π Z ∞ φ 2 ( x,y ) sin( dy ) dy. When we transform the equation we get the ODE d 2 U S dx 2 + 2 kφ 2 ( x, 0) π k 2 U S ( x,k ) = 0 , with boundary conditions U S (0 ,k ) = G 1 S ( k ) U S ( L,k ) = G 2 S ( k ) , where G 1 S and G 2 S are the sine transforms of the boundary conditions g 1 ( y ) and g 2 ( y ) . Note that the boundary value at y = 0 also appears as an inhomogeneous term in the ODE so in principle we really did not need to solve earlier for φ 2 . We can do it all at once with the sine transform. The solution is U S ( x,k ) = G 2 S ( k ) sinh( kL ) sinh( kx ) + G 1 S ( k ) sinh( kL ) sinh( k ( L x )) At this point the solution is obtained by performing the inverse transform: φ 1 ( x,y ) = Z ∞ U S ( x,k ) sin( ky ) dy,. At this point, the solution will be in the form of an integral. Note there is more than one approach to this problem. For example we could have written φ ( x,y ) = ψ ( x,y ) + g 1 ( y ) + ( g 2 g 1 ) x/L. 2 If we plug this into the Laplace equation for φ we get a Poisson equation which now has homogeneous boundary conditions at x = 0 and x = L . We could then expand the solution as before using sines: ψ ( x,y ) = ∞ X n =1 C n ( y ) sin( nπx/L ) and solve the ODE’s for C n . All these formulas are simply different representations of the same solution φ ( x,y ) ....
View
Full
Document
This note was uploaded on 12/12/2009 for the course ACM 95c taught by Professor Nilesa.pierce during the Spring '09 term at Caltech.
 Spring '09
 NilesA.Pierce

Click to edit the document details