Go to the U of M home page
School of Physics & Astronomy
School of Physics and Astronomy Wiki

User Tools


classes:2009:fall:phys4101.001:lec_notes_1104

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
classes:2009:fall:phys4101.001:lec_notes_1104 [2009/11/05 13:28] jbarthelclasses:2009:fall:phys4101.001:lec_notes_1104 [2009/11/07 21:44] (current) yk
Line 1: Line 1:
-===== Nov 04 (Wed)  =====+===== Nov 04 (Wed) Laplacian in Spherical Coordinate, Legendre Polynomals  =====
 ** Responsible party: Captain America, Cthulhu Food **  ** Responsible party: Captain America, Cthulhu Food ** 
  
Line 19: Line 19:
 \\ \\
  
-** Mathematical method to get Laplacian**]+ 
 +** Mathematical method to get Laplacian** 
 + 
 +The Laplacian in Cartesian coordinates is: 
 + 
 +<math>\bigtriangledown^2=\frac{\partial^2\Psi}{\partial x^2}+\frac{\partial^2\Psi}{\partial y^2}+\frac{\partial^2\Psi}{\partial z^2}</math> 
 + 
 +We start the derivation of the Laplacian in spherical coordinates with the kinetic energy operator: 
 + 
 +<math>\ <KE> = <\Psi|\frac{p^2}{2m}|\Psi> = \frac{1}{2m}*<\Psi|{p^2}|\Psi> </math> 
 + 
 +For the moment we can disregard <math>\ \frac{1}{2m} </math> because it is a multiplicative constant. 
 + 
 +In 1-D: 
 + 
 +<math>\ <\Psi|p^2|\Psi> = <p\Psi|p\Psi> = \int (\frac{\hbar}{i} \frac{\partial\Psi}{\partial\Psi})^*(\frac{\hbar}{i} \frac{\partial\Psi}{\partial\Psi})dx </math> 
 + 
 +In 3-D: 
 + 
 +<math> <KE> </math> is proportional to <math> \int (\bigtriangledown \Psi)^*(\bigtriangledown \Psi) dV </math> 
 + 
 +Where: 
 + 
 +<math> \bigtriangledown=\frac{\partial\Psi}{\partial x}+\frac{\partial\Psi}{\partial y}+\frac{\partial\Psi}{\partial z}</math> (yk says) Rather, <math> \bigtriangledown\Psi=\begin{bmatrix} 
 +\frac{\partial\Psi}{\partial x}\\  
 +\frac{\partial\Psi}{\partial y}\\   
 +\frac{\partial\Psi}{\partial z}\end{bmatrix}</math> 
 + 
 +and 
 + 
 +<math> dV = r^2 sin\theta dr d\theta d\phi </math> 
 + 
 +The above integral can also be written if we introduce a few tensors: 
 + 
 +The first one, <math>g_{ij}=\begin{bmatrix} 
 +1 & 0 & 0\\ 
 +0 & r^2 & 0\\ 
 +0 & 0 & r^2\sin\theta\end{bmatrix}  </math> is a tensor, when it is bracketed by row and column vectors <math>\begin{bmatrix}dr & d\theta & d\phi\end{bmatrix}</math> and <math>\begin{bmatrix}dr \\ d\theta \\ d\phi\end{bmatrix}</math> it is form proper 3D distance, <math> ds^2 = dr^2 + r^2 d\theta^2 + (r\sin\theta)^2 d\phi^2</math>
 + 
 +The 2nd tensor is similar to the first, but it will allow <math>\begin{bmatrix}\frac{d}{dr} & \frac{d}{d\theta} & \frac{d}{d\phi}\end{bmatrix}</math> and <math>\begin{bmatrix}\frac{d}{dr} \\ \frac{d}{d\theta} \\ \frac{d}{d\phi}\end{bmatrix}</math> to form a proper 3D dot product of two //gradients//, even though some of the differentiation do not have proper units (like <math>d\theta</math>).  This 2nd tensor takes the form of <math>g^{ij}=\begin{bmatrix} 
 +1 & 0 & 0\\ 
 +0 & \frac{1}{r^2} & 0\\ 
 +0 & 0 & \frac{1}{r^2\sin\theta}\end{bmatrix}  </math>
 + 
 +These two tensors are both expressed with <math>g</math>, but one has indeces as subscripts, and the other, superscripts, in the convention we are using.  In fact, these two matrices are inverse of each other. 
 + 
 +In addition, a scaler quantity, <math>g</math> is introduced, which is the determinant of the tensor <math>g_{ij}</math> Using these notations, the expectation value for the kinetic energy in the spherical coordinates can be written as,  
 + 
 +<math> \int (\bigtriangledown_i \Psi)^* g^{ij} (\bigtriangledown_j \Psi) g^{(\frac{1}{2})} dr d\theta d\phi </math> 
 + 
 +where 
 + 
 +<math> g^{(\frac{1}{2})} = r^2 sin(\theta) </math> 
 + 
 +<math> i,j = r,\theta,\phi </math>
 + 
 +This expression can be equated to that for the normal Cartesian version of the same calculation, <math> \int (\bigtriangledown \Psi)^*(\bigtriangledown \Psi) dV = \int (\bigtriangledown \Psi)^*(\bigtriangledown \Psi) g^{\frac{1}{2}} dr d\theta d\phi</math>
 + 
 +//i.e.// 
 +<math> \int (\bigtriangledown_i \Psi)^* g^{ij} (\bigtriangledown_j \Psi) g^{(\frac{1}{2})} dr d\theta d\phi = \int (\bigtriangledown \Psi)^*(\bigtriangledown \Psi) dV</math>
 + 
 +Both sides of this equation can be modified (using integral by parts) to move the differentiation operator on the left to be operating on the "right" function.  //i.e.// 
 + 
 +<math> \int (\Psi)^* \bigtriangledown_i [g^{ij} g^{(\frac{1}{2})} (\bigtriangledown_j \Psi) dr d\theta d\phi = \int (\Psi)^*\bigtriangledown(\bigtriangledown \Psi) dV = \int (\Psi)^*(\bigtriangledown^2 \Psi) dV</math>
 + 
 +At this point, if we replace <math>dV</math> with <math>g^{\frac{1}{2}} dr d\theta d\phi</math>, we get: 
 + 
 +<math> \int (\Psi)^* \bigtriangledown_i [g^{ij} g^{(\frac{1}{2})} (\bigtriangledown_j \Psi) dr d\theta d\phi = \int (\Psi)^*(\bigtriangledown^2 \Psi) g^{\frac{1}{2}} dr d\theta d\phi</math>
 + 
 +Since this equation between two integrals hold for any wave functions, the integrand must be equal, too.  Further, we can regard the equation for the operator in front of <math>\Psi</math> From this, we can claim that 
 + 
 +<math> \bigtriangledown_i g^{ij} g^{\frac{1}{2}} \bigtriangledown_j  = \bigtriangledown^2 g^{\frac{1}{2}} </math> or  
 + 
 +<math> \bigtriangledown^2 = \frac{1}{g^{\frac{1}{2}}}\bigtriangledown_i g^{ij} g^{\frac{1}{2}} \bigtriangledown_j </math>
  
 **Legendre Polynomials** **Legendre Polynomials**
Line 27: Line 100:
 where <math>\alpha</math> is separation of the r-dependent part and <math>\beta</math> is the separation of the <math>\phi</math> dependent part. where <math>\alpha</math> is separation of the r-dependent part and <math>\beta</math> is the separation of the <math>\phi</math> dependent part.
  
-Then the Differential equation form of <math>\Theta(\theta)= P_l^m (cos\theta)</math>+Then the Differential equation form of <math>\Theta(\theta)= P_l^m (cos\theta)</math> replacing <math>z=cos\theta</math> 
 + 
 +is: <math>(1-z^2)\frac{d^2U}{dz^2} - 2z \frac{dU}{dz} + \alpha U = 0</math> 
 + 
 +Then we can take 
 + 
 +<math>U(z)= \sum_{n=0}^\infty a_n z^n </math> 
 + 
 + 
 +We can then take <math>\xi=\frac{1-z}{2}</math> where <math>-1 < z < 1 </math> and therefore <math>0 < z < 1</math> 
 +Then <math>U(\xi)</math> is also a differential equation where
  
 +<math>U(\xi)=\sum_{n=0}^\infty a_n \xi^2</math>
  
 +From this we get 
  
 +<math> const. U'' + const. U' + const. U= const</math> for all <math>\xi</math>
  
  
classes/2009/fall/phys4101.001/lec_notes_1104.1257449286.txt.gz · Last modified: 2009/11/05 13:28 by jbarthel