The Principle of Minimum Potential Energy

MPE

Def : The Potential Energy of a structural System is defined as the Sum of

the Strain Energy (SE) and the Work Potential (WP),

Π=SE+WP

-The Strain Energy is the elastic energy stored in deformed structure.

It is computed by integrating the strain energy density
(i.e., strain energy per unit volume) over the entire volume of the
structure;

The strain energy density is given by : U=1/2 (Strain)x(Stress)

-The Work Potential is the negative of the work done by the external
forces acting on the structure 1
 1
𝑆𝐸 = 𝜀 𝜎 𝑑𝑉
2
𝑉

𝑁 𝑚

𝑊𝑃 = − 𝜌 𝑏 𝑢 𝑑𝑉 − 𝑇 𝑢 𝑑𝑆 − 𝑃𝑖 𝑢𝑖 − 𝑀𝑗 𝜃𝑗
𝑉 𝑆 𝑖=1 𝑗=1

The total Potential energy is then :

1 𝑁 𝑚
Π= 𝑉 2
𝜀 𝜎 𝑑𝑉 − 𝑉
𝜌 𝑏 𝑢 𝑑𝑉 − 𝑆
𝑇 𝑢 𝑑𝑆 − 𝑖=1 𝑃𝑖 𝑢𝑖 − 𝑗=1 𝑀𝑗 𝜃𝑗

1 𝜕𝑢𝑖 𝜕𝑢𝑗
Recognizing that (from constitutive law) : 𝜎 = 𝐶 𝜀 Where 𝜀𝑖𝑗 = +
2 𝜕𝑥𝑗 𝜕𝑥𝑖

The rest of the procedur is summarized by the following two steps :

- Evaluate the displacement field 𝒖 by interpolation functions of the

element’s nodal displacements;
- Derive 𝜫 in respect to each nodal displacement to obtain the 2
element stiffness matrix and the element external load vector
 Beam Element

Beam : Long, slender structural member that is generally subjected to transversal loading that
produces significant bending effects.
- Long horizontal members used in buildings and bridges;
- Shaft supported in bearings;
- Airplane and automobile structures (frames);
- etc,..

Beams are a critical area of interest because they include typically many things of interest that
we need as designers to know :
- Shear force function 𝑇(𝑥);
- Bending moment function 𝑀(𝑥);
- Deflection function 𝒗(𝒙).

To determine deflections, stresses and strains of beams, two common theory are generally used
1- Euler-Bernoulli theory;
2- Timoshenko theory.

The main difference between these two theories reside in the fact that shear
deformation in not considered in Euler-Bernoulli’s theory unlike in Timoshenko's
theory
3
Euler-Bernoulli beam assumptions:
- The deformations are small (when designing, we want deflections to be small);
- The beam is made of a linear elastic isotropic material;
- Effects of Poisson’s ratio are ignored;
- Plane sections remain plane (after deformation)

𝒚 𝒗
Beam deflection

The analysis of a Euler-Bernoulli beam can be broken down into three steps :
1. Determine the deflection function from the applied loading :
𝑑4 𝑣
𝑞=𝐸𝐼 Solve for 𝒗
𝑑𝑥 4
2. Determine the shear and bending moment function from the deflection function :
𝑑2 𝑣 𝑑3 𝑣
𝑀=𝐸𝐼 & 𝑇 = 𝐸𝐼 Functions of 𝒙
𝑑𝑥 2 𝑑𝑥 3

3. Determine the stresses components from the moment and shear functions :
𝑀𝑦 𝑇𝑄
𝜎𝑥𝑥 = − 𝐼 & 𝜎𝑥𝑦 = 𝐼 𝑏 Functions of 𝒙 & 𝒚 4
 Beam Element Stiffness Matrix
(From MPE : minimum potential energy principle )
𝒚
Let us consider now a beam element of length 𝑳 with two nodes :
𝑣1 , 𝒇𝟏 𝑣2 , 𝒇𝟐

Ѳ1 , 𝑴𝟏 Ѳ2 , 𝑴𝟐
𝟏 𝟐 𝒙
𝑳

- 𝑣1 , 𝑣2 represent the local transverse nodal displacements;

- 𝜃1 , 𝜃2 represent the nodal rotations;
- 𝑓1 , 𝑓2 represent the local transverse nodal forces;
- 𝑀1 , 𝑀2 represent the nodal moments.

The potential energy 𝚷 of the beam element is given by :

𝐿 2 𝐿
1 𝑑2 𝑣
Π= 𝐸𝐼 𝑑𝑥 − 𝑞 𝑥 𝑣 𝑑𝑥 − 𝑝𝑖 𝑣𝑖 − 𝑀𝑗 𝜃𝑗
2 𝑑𝑥 2
0 0 𝑖 𝑗

Where 𝑞(𝑥) is the distributed load per unit length, 𝑝𝑖 is the point load @ point 𝑖, 𝑀𝑗 is the
𝑑𝑣
moment applied @ point 𝑗, 𝜃𝑗 = 𝑑𝑥 𝑗
is the slope at point 𝑗. 5
- To evaluate the potential energy Π we need an expression for the displacement function 𝑣 𝑥 ,
which is the only unknown in this expression;
- Since the element has four dof (𝑣1 , 𝜃1 , 𝑣2 , 𝜃2 ) choose a cubic polynomial :
𝑣 𝑥 = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥 2 + 𝑎3 𝑥 3

In order to solve for these four coefficients (𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 ) Express 𝑣 𝑥 with its
𝑣 0 = 𝑣1 𝑎0 = 𝑣1
𝑑𝑣 𝑎1 = 𝜃1
= 𝜃1
𝑑𝑥 0 −3 1
corresponding nodal dof :
𝑣 𝐿 = 𝑣2 𝑎2 = 𝐿2
𝑣1 − 𝑣2 − 𝐿
2𝜃1 + 𝜃2
𝑑𝑣 2
= 𝜃2 𝑎3 = 𝐿3 𝑣1 − 𝑣2 + 𝜃1 + 𝜃2
𝑑𝑥 𝐿
By introducing the shape functions, 𝑣(𝑥) is written as follow :
𝑣 = 𝑁1 𝑥 𝑣1 + 𝑁2 𝑥 𝜃1 + 𝑁3 𝑥 𝑣2 + 𝑁4 𝑥 𝜃2

3𝑥 2 2𝑥 3
𝑁1 = 1 − 2 + 3
𝐿 𝐿
2
2𝑥 𝑥3
𝑁2 = 𝑥 − + 2
𝐿 𝐿
3𝑥 2 2𝑥 3
𝑁3 = 2 − 3
𝐿 𝐿
𝑥2 𝑥3
𝑁4 = − + 2
𝐿 𝐿 6
In a matrix form we have :
𝑣1
𝜃1
𝑣 = 𝑁1 𝑁2 𝑁3 𝑁4 𝑣2 or in the matrix compact form 𝑣 = 𝑁 𝑑
𝜃2

𝑑𝑣 𝑑2 𝑣
We need also & in order to substitute in the Π expression :
𝑑𝑥 𝑑𝑥 2
𝑣1
𝑑𝑣 𝜃1 𝑑𝑣
= 𝑁1,𝑥 𝑁 2,𝑥 𝑁3,𝑥 𝑁4,𝑥 𝑣2 , Equivalently = 𝑁,𝑥 𝑑
𝑑𝑥 𝑑𝑥
𝜃2

𝑣1
𝑑2 𝑣 𝜃1 𝑑2 𝑣
= 𝑁1,𝑥𝑥 𝑁2,𝑥𝑥 𝑁3,𝑥𝑥 𝑁4,𝑥𝑥 𝑣2 , Equivalently = 𝑁,𝑥𝑥 𝑑
𝑑𝑥 2 𝑑𝑥 2
𝜃2

Substitution in the expression of the potential energy, give :

𝐿 𝐿
1
Π= 𝐸𝐼 𝑁,𝑥𝑥 𝑑 𝑁,𝑥𝑥 𝑑 𝑑𝑥 − 𝑞 𝑥 𝑁 𝑑 𝑑𝑥 − 𝑝𝑖 𝑁 𝑖 𝑑 𝑖− 𝑀𝑖 𝑁,𝑥 𝑑 𝑖
2 𝑖
0 0 𝑖=1,2 𝑖=1,2

7
 𝑇 𝑇 2 𝑇
Noting that : 𝑁,𝑥𝑥 𝑑 = 𝑑 𝑁,𝑥𝑥 and 𝑑 = 𝑑 𝑑
Π
𝐿 𝐿
1 𝑇 𝑇 𝑇 𝑇 𝑇 𝑇
= 𝐸𝐼 𝑑 𝑁,𝑥𝑥 𝑁,𝑥𝑥 𝑑 𝑑𝑥 − 𝑞 𝑥 𝑑 𝑁 𝑇 𝑑𝑥 − 𝑝𝑖 𝑑 𝑖 𝑁 𝑇
𝑖− 𝑀𝑖 𝑑 𝑖 𝑁,𝑥
2 𝑖
0 0 𝑖=1,2 𝑖=1,2

Differentiating w.r.t nodal displacements :

𝜕Π 𝒌𝒆
=0 , 𝑘 = 1, … , 4
𝜕𝑑𝑘

𝝏 1 𝐿 𝑇 𝐿 𝑇
𝟏 𝒔𝒕
𝑻𝒆𝒓𝒎 ∶
𝝏𝒅𝒌 2 0
𝐸 𝐼 𝑑 𝑇
𝑁,𝑥𝑥 𝑁,𝑥𝑥 𝑑 𝑑𝑥 = 0
𝐸 𝐼 𝑁,𝑥𝑥 𝑁,𝑥𝑥 𝑑𝑥 𝒅

𝐿/2
𝜕 𝐿 𝐿 𝐿2 /12 If 𝒒 = 𝑪𝒕𝒆
𝟐𝒔𝒕 𝑻𝒆𝒓𝒎 ∶ 0
𝑞 𝑥 𝑑 𝑇
𝑁 𝑑𝑥 𝑇
= 0
𝑞 𝑇
𝑥 𝑁 𝑑𝑥 = 𝑞
𝐿/2
𝜕𝑑𝑘
−𝐿2 /12
𝟑𝒓𝒅 & 𝒕𝒉𝒆 𝟒𝒕𝒉 𝑻𝒆𝒓𝒎𝒔 𝒈𝒊𝒗𝒆𝒔 𝒓𝒆𝒔𝒑𝒆𝒄𝒕𝒊𝒗𝒆𝒍𝒍𝒚:
𝑝1 0
0 𝑀1
&
𝑝2 0
8
0 𝑀2
 Beam Element Stiffness Matrix
𝐿
𝑇
𝑘𝑒 = 𝐸 𝐼 𝑁,𝑥𝑥 𝑁,𝑥𝑥 𝑑𝑥
0

−6 12𝑥 −4 6𝑥 6 12𝑥 −2 6𝑥
𝑁,𝑥𝑥 = 2 + 3 + 2 − 3 + 2
𝐿 𝐿 𝐿 𝐿 𝐿2 𝐿 𝐿 𝐿

After calculations !
12 6𝐿 −12 6𝐿
𝐸 𝐼 6𝐿 4𝐿2 −6𝐿 2𝐿2
𝑘𝑒 = 3
𝐿 −12 −6𝐿 12 −6𝐿
6𝐿 2𝐿2 −6𝐿 4𝐿2