POTENTIAL ENERGY, ∏.

The total potential energy of an elastic body , is

defined as the sum of total strain energy (U) and the
work potential (WP) .

∏ = U + WP
 For linear elastic materials , the
strain energy per unit volume in
1
the body is “ σ T ε ”
2

For elastic body total strain

energy (U) is U = 1 ∫ σ T ε dv
2
 The work potential is given by
WP = − ∫ u fdV − ∫ u Tds − ∑ u Pi
. .
T T T
V S i
i

The total potential energy for the general

elastic body is
1 T
π = ∫ σ ε dv − ∫ u fdV − ∫ u Tds − ∑ ui Pi
. .
T T T

2 V S
i
 Principal of minimum potential
energy

For conservative systems, of all the

kinematically admissible displacement
fields, those corresponding to equilibrium
extremize the total potential energy . If the
extremum condition is a minimum , the
equilibrium state is stable
Example

1
1

K2` K1` q1
2
q2
2 q3 K4
K3
3

3 4

Figure-1
Figure 1 ,shows a system of spring .

The total potential energy is given by

1 1 1 1
π = k1δ1 + k2δ 2 + k3δ 3 + k4δ 42 − F1q1 − F3q3
2 2 2

2 2 2 2
where δ1 , δ2 , δ3 , and δ4 are extensions of four spring .
since
δ1 = q 1 - q 2

δ2 = q 2

δ3 = q 3 - q 2

δ4 = - q 3
 we have
1 1 1 1
π = k1 ( q1 − q2 ) + k2 q2 + k3 ( q3 − q2 ) + k4 q32 − F1q1 − F3q3
2 2 2

2 2 2 2

where
q 1 , q 2 , and q 3 are the displacements of nodes
1 , 2 and 3 respectively
• For equilibrium of this 3- DOF system , we need
to minimize to ∏ with respect to q 1 , q 2 , and q 3
the three equations are given by

∂π =0
∂qi i = 1 ,2 ,3

which are

∂π = k1 (q 1 - q 2) - F1 = 0
∂q1
∂π
∂q2 = -k1 (q 1 - q 2) + k2 q2 - k3 (q 3 - q 2) = 0

∂π = k3 (q 3 - q 2) + k4 q3 – F3 = 0
∂q3
Equilibrium equation can be put in the form of

K q = F as follows

K1 -K1 0 q1 F1

q2 .. ….. 1 1
-K1 K1+ K2+ K3 -K3 = 0

0 -K3 K3+ K4 q3 F3
If on the other hand , we proceed to write the
equilibrium of the system by considering the
equilibrium of each separate node as shown in
figure 2
We can write
K1δ1 = F1

K2δ2 - K1δ1 - K3δ3 = 0

K3δ3 - K4δ4 = F3

Which is precisely the set of equations represented in

Eq-1
We see clearly that the set of equation “1”
is obtained in a routine manner using the
potential energy approach, without any
reference to the free body diagrams .

This make the potential energy approach

attractive for large and complex problems .
RAYLEIGH-RITZ METHOD

Rayleigh-Ritz method involves the construction of

an assumed displacement field, say

u = ∑ai Φi ( x, y, z) i = 1 to L

v = ∑aj Φj( x, y, z) j = L + 1 to M Eq-1

w = ∑ak Φk( x, y, z) k = M + 1 to N

N > M > L
The functions Φi are usually taken as
polynomials. Displacements u, v, w must
satisfy boundary conditions.

Introducing stress-strain and strain-

displacement relation Substituting
equation – 1 in to ∏ (PE)
∏= ∏ ( a1, a2, a3, ….…….. ar )

where r = no of independent unknowns

the extremum with respect to ai,( i = 1 to r)
yields the set of r equation

∂π =0 i = 1, 2, 3……… ,r
∂ai
 Example

The potential energy of for the linear 1-D

rod with body force is neglected , is

( )
l
1 ⎡ EA du 2
⎤
π =
2 ∫
0
⎢⎣ dx
d x − 2 u1 ⎥
⎦
where u1= u (x = 1)
Y

E = 1, A= 1

X
1
2

1 1
let as consider a polynomial function
u = a1 + a2x + a3x3 this must satisfy

u = 0, at x = 0
u = 0 at x = 2

thus 0 = a1
0 = a1 +2 a2+ 4a3
Figure- 2
Hence
a2 = -2a3
u = a3 (-2x + x2)
u1 = -a3
du
then , = 2a3 ( −1 + x ) and
dx

( )
l
π = ∫ ⎢ EA du dx dx − 2u1 ⎤⎥
1 ⎡ 2

20⎣ ⎦
⎛2⎞
= 2a ⎜ ⎟ + 2a3
2
3
⎝3⎠
we set ∂π =0
∂a3
Resulting in a3 = -0.75
u1 = - a3 = 0.75

the stress in bar given by

du
σ =E = 1.5 (1 − x )
dx
exact solution is obtained if piecewise
polynomial interpolation is used in the
construction of u .
 GALERKIN’S METHOD
• Galerkin’s method uses the set of governing
equations in the development of an integral form.

• It is usually presented as one of the weighted

residual methods.

• Let us consider a general representation of a

governing equation on a region “V”
Lu = P
Where , “L” as operator operating on “u”
 For the one-dimensional rid considered in
previous example
Governing equation

d
dx ( EA du
dx ) =0

We may consider L as operator, operating

on “u”
d ⎛ EA d ( ) ⎞
dx ⎜⎝ dx ⎟⎠
• The exact solution needs to satisfy “Lu=P”
at every point x .
• If we seek an approximate solution u , if
introduces an error ε ( x ), called the residual
ε (x ) =Lu − P
• The approximate methods revolve around
setting the residual relative to a weighting
function Wi ,i = 0 to n

∫ W ( Lu − P ) dV = 0
i
 The weighting function Wi are chosen from
the basis functions used for constructing

n
u = ∑ Qi Gi
i =1

• Here ,we choose the weighting function to

be linear combination of the basis function
Gi . Specifically ,consider an arbitrary
function φ
Given by
n
φ = ∑i=1
φ iG i

Where the coefficient φi are arbitrary , except

for requiring that φ satisfy boundary
conditions were u is prescribed.
 For elastic materials
∂σ x ∂τ xy ∂τ xz
∫V [( ∂x + ∂y + ∂z + f x )Φ x + ......]dV = 0

∂α ∂θ
∫V ∂x θdV = −V∫ α ∂x dV + ∫S nxαθdS

∫ σ T
ε (φ ) dV − ∫ φ T
fdV − ∫ φ T
TdS −
i
∑ P
φ T

V V S
Example
let us consider the problem of the previous
example and solve it by Galerkin’s approach.
The equilibrium equation is

d EA du =0
dx dx
• u=0 at x=0
• u=0 at x=0
Multiplying this differential equation by
Integrating by pars, we get
Figure- 2
 ( dx ) ( dx )
2
dφ 1 2

∫
0
− EA du
dx dx
+ φ EA du
0
+ φ EA du
1
=0

Where φ is zero at x = 0 and x = 2.

EA du
dx is the tension in the rod ,which
takes a jump of magnitude 2 at x = 1 , thus

2
dφ
∫
0
− EA du
dx dx
+ 2φ1 = 0
Now we use the same polynomial (basis ) for u
and φ
if u1 and φ are the value at x = 1 ,thus

u = ( 2 x − x ) u1
2

φ = ( 2x − x ) φ1
2

• Substituting these and E = 1, A = 1 in the

previous integral yields
⎡ 2
⎤
φ ⎢ − u1 ∫ (2 − 2 x ) d x + 2 ⎥ = 0
2

⎣ 0 ⎦
 ( )
φ1 − 8 3 u1 + 2 = 0
This is to be satisfied for every φ1 .
We get
u1 = 0.75