SOLUTIONS OF SELECTED PROBLEMS OF EXERCISE 2

ANS. 2.1

Method I: Using variational approach:

 L EI  d 2 w 2 kwL 2 
       2  dx     dx 
P  dw 
L 2


 0 2  dx  2  dx  2 
0 
Using properties of variational operator:
 d 2 w  d 2 w   dw  d  w
    EI  
L L

2  2 
dx   P dx  kwL  wL
0
0  dx  dx  dx  dx
On integrating by parts:

   EIw  w 0    EIw    w  dx  Pw  w 0    Pw   wdx +kwL  wL

d d d d
L L
L L

dx 0
dx dx 0
dx
 d2 
 w 0   EIw   w 0    2  EIw    Pw   wdx  Pw  w 0 +kwL wL
d d d
L
=EIw
L L L

0  dx 
dx dx dx
Putting   0 ,
 d2 
EIw  w 0   EIw   w 0    2  EIw    Pw   wdx  Pw  w 0 +kwL  wL  0
d d d
L
L L L

0  dx 
dx dx dx

For arbitrary w which vanishes on the boundaries alongwith its first derivative,
 d2 
0  dx 2  EIw  dx  Pw  wdx=0
d
L

From the above expression, if w is arbitrary, the following relation must hold good:

2 
EIw    Pw   0
d2 d
dx dx
which is the governing differential equation.
In view of the above equation, the following relation also must hold good:

 EIw  w  Pw w   EIw   w  kwL  0

 d d 
L

 dx dx 0

or (EIw '') L  wL'  ( EIw)0  w0'  ( Pw   EIw   kw) L  wL  ( Pw   EIw )0  w0  0
d d
dx dx

d w
As w and  w  are independent and arbitrary at both the boundaries, at any boundary for
dx
non-zero variation of slope and deflection, the associated bracketed term becomes zero. This
provides the following boundary conditions:

 EIw )0  w0  0
d
At x=0: (i) w is prescribed (essential boundary condition) or ( Pw 
dx
(natural boundary condition).
(ii) w’ is prescribed (essential b.c.) or EIw  0 (natural b.c.)

1
  EIw   kw)0  w0  0 (natural b. c.).
d
At x=L: (i) w is prescribed (essential b.c.) or ( Pw 
dx
(ii) w’ is prescribed (essential b.c.) or EIw  0 (natural b.c.)

Second method:

The governing functional:

EI  d 2 w 
    2  dx   
P  dw  kwL 2
L 2 L 2

 dx 
0
2  dx  0
2  dx  2

dw
where essential boundary conditions are: wx 0  0, 0
dx x 0
If F is the integrand of the functional, then governing differential equation is:
F d F d 2  F 
   0 (1)
y dx y dx 2  y  

and the boundary conditions are

 F d F 
   y  0 at x= a and x=b (2)
 y dx y 
F
 y  0 at x= a and x=b (3)
y

The functional can be written as

 EI  d 2 w 2 P  dw  2 kw2 
    2     
L

   x - L dx
0
2  dx  2  dx  2 


where   x - L  is the Dirac delta function

EI  d 2 w  P  dw 
  x - L
kw2
2

Therefore, F      
2  dx 2  2  d x  2
Putting F in (1), we get:
 d2 w 
kw  x - L  
d  dw  d 2
0  P   EI 2   0
dx  dx  dx 2  dx 
Thus, the governing differential equation is
 d2 w 
kw  x - L  
d  dw  d 2
P   EI 2   0
dx  dx  dx 2  dx 
To get boundary conditions, F is put in (2)

2
 dw d  d 2 w  
P   EI 2    w  0
 dx dx  dx  

dw d  d 2 w 
P   EI 2   0 or w is prescribed at x=0 and x=L.
dx dx  dx 
.
dw d   2 w 
At x=0 ,w=0 and P   EI  at x=L which is the first natural boundary condition
dx dx  x 2 
at x=L.
To get another set of boundary conditions, F is put in (3)

d2 w
 EI 2  w  0
dx

d2 w
Therefore either  EI  w  0 or w is prescribed at x=0 and x=L. Here,
dx 2

d2 w
 w =0 at x=0 and EI  w  0 at x=L.
dx 2

dw d  d 2 w 
First natural bc: P   EI 2  at x=L
dx dx  dx 
d2 w
Second natural bc: EI 0 at x=L
dx 2

NOTE: In the second method, we have considered spring force just infinitesimally inside the
rod, therefore kw  x - L  term had come in the differential equation. However, if we put limits
of integration 0 to L ,so that kw  x - L  does not come in governing differential equation ,then
it would come in the boundary condition.

ANS 2.2
Consider the differential equation
L( )  f  0
where L is the differential operator,  is a scalar function defined over the domain D and f is a
known scalar function.
Multiplying the equation by a variation of  and integrating it over the domain

  L(   f  dD  0

D (1)
d2 r
Here L= ,  =T, f =- .
dx 2
kA

3
Putting the value of L,  , f in equation (1), we get

 d 2T r 
0  dx 2  kA  Tdx  0
1

By using integration by parts

  ( T )
kA 0
dT d dT r
1 1 1
T dx   Tdx  0
dx 0 0
dx dx

 1  dT 2 
    
r dT dT
1 1 1

  dx   Tdx   T   T 0
0 
 2  dx  
 kA 0
dx 0 dx

Using the Boundary conditions,

T  0,
Essential boundary condition: T= 0 at x= 0
Natural boundary condition:
dT T1
 at x=1
dx 2

 1  dT 2 rT  T1 T1
 
1

  dx  0
 2  dx  kA 
0 
2

Then, the variational form of the given differential equation is

 1  dT 2 rT 
I   
T12
1

  dx  0
0 
 2  dx  kA 
 4

4
ANS. 2.3

    2  2, x  50  dx
1 
L
Functional
0

  Fdx
L

1
where F   '2  50  '  x
2
for extremum condition   0 or Euler-Lagranges equation should satisfy
F d  F 
  0
 dx   ' 
  1 '2  d    1 '2 
  2   50   dx   '    50    0
2 

50   '   0
d
dx
50   ''  0
 ''  50  0
 ''  50
int egrating with respect to x
d
 50 x   =const. of integration
dx
again int egrating with respect to x
 =  25 x 2  x  
with the given boundry condition
x  0  0 =0
and x  L  0 =25L
on substituting the value of A and B in  , we get
 =  25 x 2  25 Lx

ANS. 2.5

To obtain variational form, multiply the above governing equation by  T ,

  2 T  2T 
 I    2  TdA  A  T TdA  0
2
(1)
A  x y 
2

The above expression can be written as

5
 I   .(T  T )dA   T .( T )dA   .(T  T )dA   T . (T )dA =0
A A A A

Applying the divergence theorem to the first term and expanding the second term, we get

  T .nˆ  T ds    x  x  y  y dA
 T T T T 
I 
s A
As the temperature is prescribed at the boundaries, the boundary integral term vanishes and we
get

 
 T T T T     T 2  T  2 
I       dA      dA
x x y y  2   x   y  
A A 

Thus, the variational form is


1   T   T  
2 2
I     dA
2   x   y  
A 

ANS. 2.6

As the first step, multiply the above governing equation by  u ,

 d 2u 
 I  
1
 u  1 udx
0 
dx 2

Integrating equation by parts,

d  u  du
 dx    u 2 dx    udx
du 1
1 1 1 1
 I  u
dx 0 0 dx dx 0
2 0

In view of the boundary conditions, the first term of the above expression becomes 0, then the
expression becomes,

 I       dx    u 2 dx    udx
1  du  1
1 2 1 1

0
2  dx  0
2 0
Or

 1  du 2 1 2 
 I         u  u  dx
1

 2  dx  2 
0 

Hence the variational form is given by

6
  1  du  2 1 
I        u 2  u  dx
1

 2  dx  2 
0 

ANS. 2.7

The given differential equation is

d2  d 2v  dmz
 zz 2  
EI  py  0
dx 2  dx  dx

Taking  v as an arbitrary variational function

 d2  d 2 v  dmz 
0  dx 2  zz dx 2   dx  p y   vdx =0
l

 EI

By using integration by parts

l
d  d 2 v  d ( v)

d  d v
l

 EI zz 2   v dx  mz v |l0
2
EI
 zz 2 
dx  dx  0
dx  dx  dx
0

 ( v)dx   p y vdx  0
d
l l
mz
0
dx 0

  d  d 2v 
   EI zz / 2(
d 2v 2 dv
l
)  m  p v
y  dx  ((  zz 2   mz ) v) |x l
EI
 dx 2 dx  dx  dx 
z
0

d  d 2v  d 2 v dv d 2 v dv
((  EI zz 2   mz ) v) |x  0  EI zz 2  ( ) |x l  EI zz 2  ( ) |x  0
dx  dx  dx dx dx dx

Substituting the following boundary conditions:

Essential boundary conditions

dv
v | x  0 = v and | x  0 = z
dx
Natural boundary conditions
d 2v d  d 2v 
EI zz |  M and  zz 2   mz |x l  v y
EI
 *

dx 2 dx  dx 
x l z

7
we get,

 
 (   EI zz / 2(
d 2v 2 dv dv
l
)  mz  p y v dx  v*y vl  mz  l )  0
 dx dx  dx
2
0

The variational form is

 

d 2v 2 dv dv
l
I=  EI zz / 2( )  mz  p y v dx  v*y vl  mz  l
dx dx  dx
2
0

ANS. 2.8

Let the coordinates of the two end points be (x1, y1) and (x2,y2)

Let the shortest curve joining these two points be y = y(x) , then the length of the curve between
the two given points will be: ds= dx 2  dy 2
where dx= x2-x1 & dy=y2-y1

or, ds=( 1  (dy / dx) 2 ) dx

Now our aim is to minimize

 ( 1(dy / dx)2 )dx

b

a
Denoting, dy/dx = y’

where a and b are the x- coordinates of the two points.

The differential form of the above variational form will be given by E-L eq.

F d F
 ( )0 where, F= ( 1  y '2 )
y dx y '
F
Since = 0, therefore
y

d F
( )0
dx y '

8
or
d 
( ( 1  y '2 )  0
dx y '
Solving, we get
y '/( 1  y '2 )  const
or,
y '  const
i.e.
dy
 const
dx
This is nothing but the equation of a straight line.

ANS. 2.9

The differential time taken to traverse the distance ds is that distance divided by the velocity
ds
( )
v


 ds 
Total time =
(x 0, y0 )
  (1)
(0,0)
 v 

9
Since y is positive downward and the bead of mass is released from the point A.

v  2 gy  (2)

 ds    dx    dy 
2 2 2

or

  dy  2 
ds =  1+    dx
  dx  
  (3)

= 1+y dx 2

Substituting equations 2, 3 in equation no. 1

Total time

1  y  dx

1
x0 2

T=
2g 0
y

For shortest time we have to minimize this functional

F=
1  y  2

F = F (y, y)

By Euler-Lagrange equation

 F d  F  
 -   = 0
 y dx  y   
Multiplying by y’,

 F d  F   F  d  F  
 -     y= y -  y     0
 y dx  y   y  dx  y  
It can be verified that when F is independent of x, the above expression
reduces to

10
 d  F 
0=  F  y
dx  y 

Therefore,

F
F  y = C (constant) (eqn. *)
y
In our case,

F=
1  y 
2

and
F y
y 1  y2 
=
y

Substituting these terms in equation (*)

1  y  -
2
y 2
y 1  y2 
=C
y

Solving this equation, we get the solution.

Sol. 2.10 can be seen from any standard book of

vibration.

11