MSc-PRACTICAL EXPERIMENTS

MSc-1- FOURIER SERIES

MSc-2-BOHR'S THEORY OF ATOMIC STRUCTURE
MSc-3-NUMERICAL SOLUTION OF SCHRODINGER WAVE EQUATION
MSc-4-LORENTZ FORCE
MSc-5-EULER'S METHOD OF SOLVING ODE
MSc-6-NEWTON RAPHSON METHOD
MSc-7-NUMERICAL INTEGRATION-3-8 SIMPSON RULE
MSc-8-RANGE-KUTTA METHOD -ODEMSc-9-NUMERICAL SOLUTION OF SECOND ORDER DIFFERENTIAL
EQUATION
MSc-10-HARMONIC OSCILLATOR
MSc-11-LEARNING MOLECULAR SYMMETRY-H2O-NH4
MSc-12-STUDY MATERIAL OF HYDROGEN ATOM AND MOLECULE
MSc-13-GROUND STATE OF HELIUM ATOM -VARIATION METHOD
MSc-14-ELECTRON MOTION THROUGH RECTANGULAR POTENTIAL
BARRIER
MSc-15-NUMERICAL SOLUTION FOR THE RADIAL EQUATION OF THE
HYDROGEN ATOM
MSc-16-NUMERICAL SOLUTION OF THE RADIAL EQUATION FOR 3D
OSCILLATOR
MSc-17-XY3- PLANAR- H2CO-FORMALDEHYDE.
MSc-18-ETELON

1. FOURIER SERIES
Goal of this exercise:
To give students the opportunity to efficiently and effectively practice using a Fourier
Series Expansion. This exercise serves as a foundation for future work in quantum
mechanics where linear combinations of atomic orbitals and expressions of solutions to
the Schrodinger Equation are written as expansions using sets of functions that are
complete orthonormal sets.
FOURIER SERIES
The general equations for a fourier series are shown here. Notice the use of a and b for
the coefficients for the cos and sin terms respectively. Here F(x) is the function that is to
be fit by the Fourier series. In the sections that follow you will be lead through a series of
exercises that will put into practice the Fourier series method for both odd and even
functions.
F(x) some function of x

1
a 0 a n cos n x b n sin n x
2
n1
n1

f (x)
an

bn

2
L

F(x) cos n x dx

F(x) sin n x dx

an

2
F (x)cos(n x)dx
0

bn 0

2
bn F (x)sin(n x)dx
0
an 0
regular functions, each continuous component of the function must be determined These
expressions for a and b assume that the integrals exist and that F(x) is a continuous
function. For piecewise separately.

MATHCAD WORKSHEET WORK

Let us start with a simple Fourier expansion for a periodic function, a simple step
function. When you work with this document do not repeat the exercises using the
same variable names. This will confuse Mathcad and give you a headache trying to
debug your Mathcad code.
-x < 0

-1

F (x)

Defining the step function.

x <,=

N can equal 0 up to any large integer.

x 3.1

Notice the relationship between N and i

j 0 N

f ( x )

Notice the way the range of x is defined.

1
2 j 1

sin [ ( 2 j 1 ) x ]

Values of N are entered below just above the graph. The

entry point for this integer was placed there so as to
make exploration and graphical insights possible
without excessive scrolling of the screen view. N is a
Mathcad Global variable in this document. Depending
on how you have your worksheet options set a change
in N will immediately result in a new figure or you may
need to press F9 to do do the calculations that will
show a new figure. This can also be accomplished by
pressing the = sign in the first tool bar of the window.

Vary N, the number of terms in the expansion and describe the consequences as N is increased
to a large number. Record your observations in your notebook. Note how the Fourier series
behaves relative to the step function it is designed to match.
Concisely express the function that is being expressed by the series expansion shown above.
This means to write out explicitly the first few terms in the series for the fitting function.

Sample Problem:
Expand the function F(x)=x in the interval - to by a Fourier series.

Solution: This function is an odd function. Only sine terms will show up in the
expansion.
1. First we write the function and set the interval for the expansion.
F ( x ) sin ( x )
x 3.0
n 0 M
1
B n

F ( x ) sin ( n x ) dx

2. Then we choose an upper limit for the number of terms in the

expansion (the value for M). Explorations on the effect of changing the
upper limit are easily implemented by just changing the value for M.
M is a global variable and is placed below just above the graph.
3. Next we have the integral for Bn. In this integral note the presence
of the function F(x). Any odd function can be used here. n is the
index used to identify the integrals in the expansion for the fitting
process. Using M and n makes possible a more general exploration
of the series in the area below.
Ask Mathcad to show several of the B n coefficients in the Fourier
expansion for the function given here. Comment on their magnitude
and the significance of the magnitude as n increases.
Write out the expression that would correspond to the coeficients for
the cosine terms in a Fourier series expansion of this function. Show
that several of these An coefficient values are in fact zero.

f ( x )

B n sin ( n x )

n 1

M 5

4. We wrote f(x) here instead of F(x) so as to prevent Mathcad from

overwriting our original function. We will compare the fourier series fit,
f(x), of the function to the function, F(x), itself as an exercise. Now
write out by hand f(x) as the sum of several terms.
Notice M is an upper limit in the sum. Varying M allows rapid and
efficient exploration of the variation of the number of terms in a Fourier
expansion. Graphical presentation is a real time saver in
understanding the consequences of changing the number of terms in
the series.

2.

BOHR'S THEORY OF ATOMIC STRUCTURE

Postulates of Bohr:
1. An electron in atom can revolve in certain stable orbits without the emission of
radiant energy.The angular momentum of the electron should in the integral multiples of
(nh/2)
2. An electron may make a transition from one of its specified non-radiating
orbits to another of lower energy equal to the energy difference between initial and
final states.
In stationary states the energy of the atom is assumed to be constant. Thus the atom
does not radiate em waves while in any one of these stationary states.
By the first postulate the angular momentum of the electron is
me v n rn

nh

---------(1)

By the second postulate the emission of energy is:

h

---------(2)

Ei Ef

where is the frequency of the energy emitted : Ei and Ef are the energies of initial and final
states. Applying the Coulomb's law of force between the electron and proton:
k

1
4 0

2
rn

rn

----------(3)

Equating the RHS of the equation (3) to the cetripetal force

1
4 0

rn

me v n .rn

Solving (1) and (4) for r n and vn we get,

----------(4)

Potential Energy of the electron is:

PE( n )

k e

r( n ) e

n 1 2 50

----------------(8)

Total Energy of the electron is

E( n ) KE( n ) PE( n )

---------------------(9)

According to Bohr's second postulate, a hydrogen atom radiates energy when the electron
jumps from one orbit to another of lower energy. The energy radiated is given by the condition:

Ei Ef

h
Substituting the values of Ei and Ef using equation(9)

1 k e
KE( n )
2 r( n )

PE( n )

k e

r( n )

E( n ) KE( n ) PE( n )

E( 1)

2
nf h

h
2
2
nf
ni

E( 1)

E( 1)

2
ni h

Theoretical Spectrums of the hydrogen atom

LYMAN SERIES
( n )

( n )

h 2
2
1
n

E( 1)

c
( n)

n 2 3 20

10

BALMER SERIES

PASCHEN SERIES

BRACKET SERIES

FUND SERIES

3. NEWTON-RAPHSON METHOD
When a fluid flows in a steady fashion through a round pipe, the pressure drop
due to wall friction is given by the empirical formula:
P

2 L

0.5 f V

where:
P= pressure drop

= fluid density

f = Darcy friction factor

L = pipe length

V = fluid velocity

D = pipe diameter

Several empirical formulas exist for the friction coefficient as a function of the
DV

Reynolds number, Re, which is given by

where is the fluid viscosity.

For flow in the turbulent regime between completely smooth pipe and wholly
rough pipe surfaces Colebrook developed an empirical equation for the friction
factor

Several empirical formulas exist for the friction coefficient as a function of the
DV
is given by
where is the fluid viscosity. For flow in the turbulent reg

smooth pipe and wholly rough pipe surfaces Colebrook developed an empiric
Develop a procedure to factor
determine f for specified values of /D and Re. Use the
.16
_D_ratio 2.51
approximation proposed by Genereaux (1939) , f 10.16 Re
an
2 log , to determine
1
1
3.7

initial approximation for f. The equation is transcendental in anture therefore an

2
2

f use the Newton Raphson

Re f
analytical solution is not possible. Here we will
procedure. A graphical description of the procedure is given below.

SOLUTION - find the friction factor using the Newton Raphson

procedure
Define values for /D and the Reynolds Number

_D_ratio
_D_ratio .001
_D_ratio
2.51

F ( f ) 2 log

1
3.7

Re ( n) f

n 6
1

Re ( n) 10

The Colebrook White equation

The derivative w.r.t. f of the Colebrook White equation

4. EULER'S METHOD FOR SOLVING ORDINARY

LINEAR DIFFERENTIAL EQUATIONS
It is often impossible to find elementary solutions to differential equations of the
form y' = f(x,y). In applications, though, it is usually sufficient to find a
numerical approximation to the function satisfying the governing differential
equation. We will develop a simple numerical technique for the estimation of
such solutions.
Let:

y'

d
dx

( x y)

f ( x y)

x 1

Note that a derivative "function" can be denoted in the same way as the original
function from which it came.
( x y)
f' ( x y)

x 1

In order to obtain the required relationship between variables wecneed

1
a starting point, commonly called an initial condition: y ( 0) c 1 ,
.
This is the initial condition for the desired realtionship between x and y.

Euler's Method for solving Ordinary

Differential Equations of the form y' = f(x,y)
y

yi+1 = yi + f`(x,y)*h
f'(x,y)

k1

in words: "new y value = old y value

*slope*h"

h
x

xi

xi+1

These are the estimates of points on the solution of the differential equation

f'(x,y) is the slope of the tangent line to the solution function at x,y
We will use this function to rate the accuracy of our estimates. Each method will rely on
iteratively
estimating values at N+1 evenly spaced points on [0,L]. Here, we take
L 10 domain over which the solution function is to be approximated
N 250 number of intervals into which the domain is to divided, the dependent variable
is then approximated at each point
1

h L N

size of each interval

i 0 N counter to keep track of interval values approximated

LINEAR APPROXIMATION (EULER'S METHOD)

Define sequences recursively by

y0 1 initial value for y when x = 0

x i i h= (counter)(interval size) = value of independent variable

yi 1 yi h f' x i yi in words new value = (old value)*(slope)* x

5. RUNGE KUTTA METHOD

A Second Order Runge Kutta Method
The simplest method works for differential equations of the form y' = f(x,y): Note: we
are operating on a function which specifies the derivative of another function. Our
solution will be the functional relationship between x and y We must also have an intial
condition, namely:

y' x0
y0
Here's how it works. Let's try an example in which . . . y' f'( x y )

initial condition x 0 and y 1

f'( x y ) y

O.K. we are looking for the numerical approximation of a function, f(x,y), whose derivative at any point is equal to y

Select an increment h:
h .2
In the general Runge-Kutta method, you must define a function as follows:
( x y h ) f' x

h
2

1
2

h f'( x y )

can be different in different variations on the method. Here we evaluate

the function at x+h/2 and y+(f'(x,y)*h/2). Since the function is the
derivative, the number returned each time is a slope.

This represents the slope evaluated at x+h/2 and y + (( y/2)*h).

So we can find a set of points on the curve as follows:
1. Set a counter to determine the number of points to find N 500 i 0 N
2. Define the x values as : x x i h. Note that x o and yo are already defined as 0 and 1.
i
0
3. Now use: y

i 1

y h x y h to define the y values. Note that the relationship is recursive in that

i

each new y value is then used in ( x y h)

f' x

h
2

h
2

f'( x y ) to compute the next y value.

In words : y i+1 = y i + ( x * (average between y i and y i+1 ))

d
y
dx

y has an analytical solution: y dep xind e

xind

also written as x

ln ( y )

xind 0 .2 100
Note that the derivative of y
=y

xind

returns e

xind

demonstrating that it is in deed the solution to the given ODE, y'

If we plot the analytical solution we see that it exhibits pronounced curvature once x approaches 9 to 10. This type of
behavior is difficult to reproduce with a numerical approximation algorithm. We will verify this fact by evaluating and
plotting both the analytical solution and the numerical solution on the same set of axes.

6. NUMERICAL INTEGRATION 3/8-SIMPSON'S RULE

Stream cross sectional areas (A) are required for a number of tasks in water resource
engineering, including flood forecasting and reservoir design. Unless electronic sounding
devices are available to obtain continuous profiles of the channel bottom, the engineer
must rely discrete depth measurements to compute A. An example of a typical stream
cross section is shown below. The data represents places where a boat was anchored and
depth measurements taken. Estimate the cross sectional area of the channel.

Data below scaled from drawing

origin default is 0, reset to 1 in the Math / built-in variables submenu
iInMathcad
1 12
Disatance
6.5

0 depth

13.00
19.5

26

32.5

5.5

52.5

59

65.5

72

46.5

39

Data

ft

3.5

2.8

1.4

78.5

n 12

number of points

h 6.5 ft

distance between sequential data

points

disti Data i 1
depthi Data i 2

label each column separately

compute the flow area using Simpson's rule

or

i 1 2 n 1

j 2 3 12

depthi depth j
i

Simpson's rule with summations multiplied

through

h ft

Simpson's 1/3 rule - each segment requires 3 points

i 1 3 n 1

Area 2

j 2 4 n

k 3 5 n 1

( width) ( average_height)

2h

segments

Area 2

segments

f ( 1) 4 f ( 2) f ( 3)
6

Simpson's 3/8 rule - each segment requires 4 points

If the average flow velocity in the stream is 6 fps what is the flow rate?

7.

LORENTZ FORCE

We study the motion of a particle of charge q and mass m in a magnetic field B and in an
electric field E. The force acting on the particle is given by the Lorentz formula: F = qE
+ qvxB where v is the velocity vector. For small enough velocities, v << c, we can use
Newtonian mechanics to describe the motion of the charged particle: ma = F . Lorentz
formula and Newton's second law give: dv/dt = (q/m)E + (q/m)vxB.
First we input the mass and charge of the particle, proton for example.
19
27
q 1.6 10
m 1.67 10

Next we input the magnetic field B and the electric field E. In this computation we will
study the case of no electric field E =
0.0
0
B

E 0

0.0001

The Larmor (cyclotron) angular velocity and period are T:

We will compute the position and velocity of the particle 1000 times during a period.
t

T
n 0 5000

1000

t n t
n

The initial velocity is v0 and the initial position is r0.

106

v
0

10

r 0
0

We integrate numerically the equation of motion a = F/m by using a = dv/dt = (vn+1 vn)/t and v = (rn+1 - rn)/t. This implementation is called the Euler method and its
accuracy depends on the smallness of the time interval: t << T.

n 1

n 1

v t
n

q v B q E

m
m n

r t v
n

To graph the trajectory, let us use the 3d scatter plot available in Mathcad.
vx
x

vy v
n
n
vz
n

y r
n
n
zn

8.

NUMERICAL SOLUTION OF SCHRODINGER WAVE

EQUATION

Calculations for the One-dimensional H-Atom

Energy
operator:

1 d2
1


2 dx2
x

Momentum
operator:

1 d

i d

Integral:

The following exercises pertain to the one-dimensional hydrogen atom which has the 1s
wave function shown below: ( x) 2 xexp ( x)
1

Plot (x) and (x)2:

x 0 .01 10

Show that this wave function is normalized:

Calculate the most probable position for the electron:

Calculate the average value of the electron position:

Calculate the probability density at both the most probable and average positions of the
electron:

dx

Calculate the probability that the electron is between the nucleus and the most probable
value of the electron position:

Calculate the probability that the electron is between the nucleus and the average value of
the electron position:

Calculate the probability that the electron is beyond the most probable position:

Calculate that position from the nucleus for which the probability of finding the electron
is 0.95:

Calculate the uncertainty in position:

Calculate the average value of the electron momentum:

Calculate the uncertainty in momentum:

Demonstrate that the position-momentum uncertainty relation is obeyed:

Calculate the average value for kinetic energy:

Calculate the average value for potential energy:

Demonstrate that the wave function is not an eigenfunction of the kinetic energy operator
and comment on the significance of this result:

Electron in the hydrogen atom does not have a well-defined value for kinetic energy.
Demonstrate that the wave function is not an eigen function of the potential energy
operator and comment on the significance of this result:

Demonstrate that the wave function is an eigenfunction of the total energy operator and
comment on the significance of this result:

In spite of not having a well-defined kinetic or potential energy, the electron in the
hydrogen atom has a well-defined total energy.

What is the energy eigenvalue and how does it compare to previous calculations in this
exercise: The energy eigenvalue is -0.5 which is in agreement with <T> + <V> calculated
previously.
Calculate the overlap integral with the 12s orbital at the right:
x

2 ( x)

x( 2 x) exp
2

Interpret the result:

The orbitals are orthogonal.
Fourier transform the coordinate-space wave function into momentum space and plot the
momentum distribution function.

and |(p)|
graph:
p Plot
(x)|
5 24.99
5 2 onxthesame
0 .01
5

Simplified by user:

1 ( p)

2
( ip 1)

In momentum space:
Momentum space integral:

dp

Momentum operator:

Kinetic energy operator:

p
2

Position operator:

d
dp

Demonstrate that the 1s momentum wavefunction is normalized:

Calculate the average value of the momentum and compare it to value obtained with the
coordinate space wave function

Calculate the average value of the kinetic energy and compare it to value obtained with
the coordinate space wave function

Calculate the average value of the electron position and compare it to value obtained with
the coordinate space wave function

Calculate the uncertainty in momentum using the momentum wave function:

Demonstrate that the position-momentum uncertainty relation is satisfied:

1s

One-dimensional hydrogen atom wave functions: 1

x
1 ( x) 2 xexp ( x)
2s
2 ( x)
x( 2 x) exp
2
8

3s

3 ( x)

2
2
x
3 x 27 18 x 2 x exp
243
3

Fourier transform the 2s wavefunction into momentum space:

Demonstrate the 2s wavefunction is normalized

Demonstrate the 1s and 2s momentum wavefunctions are orthogonal

Fourier transform the 3s wavefunction into momentum space.

Demonstrate the 3s momentum wavefunction is normalized

Demonstrate that the 1s and 3s momentum wavefunctions are orthogonal.

Demonstrate that the 2s and 3s momentum wavefunctions are orthogonal.

Plot the 1s, 2s and 3s momentum wavefunctions and interpret the graph in terms of the
uncertainty principle.
p 3 2.99 3

As shown below, as the principal quantum number increases, the spatial distribution of
the electron becomes more delocalized. Therefore, according to the uncertainty principle,
the momentum distribution must become more localized. The graph above shows a more
localized momentum distribution as the principle quantum number increases.

x 0 .01 30