Vous êtes sur la page 1sur 4

Example of the Principle of

Minimum Total Potential Energy


CEE 201L. Uncertainty, Design, and Optimization
Department of Civil and Environmental Engineering
Duke University
Henri P. Gavin
Spring, 2015

Consider the bending of a simply-supported beam, with simple supports at x = 0 and x = L


carrying a concentrated transverse point load F at x = xo , (0 < xo < L).
y,v(x),F
v(x) v(xo )
11111
00000 111111
000000
x 0.1 sin(2 π x/L)
00000
11111
00000
11111 000000
111111
000000
111111
00000
11111 F 000000
111111
-1.0 sin(π x/L)
x=0 x=xo x=L -1 sin(π x/L) +0.1 sin(2 π x/L)

It is essential that any function approximating the transverse displacements of the beam,
v(x), adhere to the constraints of the supports (essential boundary conditions): v(0) = 0 and
v(L) = 0. The assumed displacement function
πx 2πx
v(x; a1 , a2 ) = a1 sin + a2 sin
L L
is admissible with respect to displacements at the supports. The semicolon notation indicates
that v is a function of x and is parameterized by coefficients a1 and a2 . Given this assumed
equation for the transverse displacements, the coefficients a1 and a2 determine the displace-
ments for any value of x. The curvature of deformation (assuming small deformation), is
π2 πx 4π 2 2πx
v 00 (x; a1 , a2 ) = − 2
a 1 sin − 2 a2 sin .
L L L L

The internal bending strain energy is given by


1Z L
U= EI (v 00 (x; a1 , a2 ))2 dx ,
2 0
the potential energy function of external forces is given by

V = −F v(xo ; a1 , a2 ) ,

(F and v(xo ; a1 , a2 ) are collocated) and the total potential energy function is given by

Π=U +V .

The principle of minimum total potential energy states that Π must be minimized with
respect to the as-of-yet unknowns in the problem, a1 and a2 .
∂U ∂V
min Π(a1 , a2 ) ⇔ δΠ(a1 , a2 ) = 0 ⇔ δai + δai = 0
a1 ,a2 ∂ai ∂ai

1
2 CEE 201L. – Uncertainty, Design, and Optimization – Duke University – Spring 2015 – H.P. Gavin

∂U ∂ 1Z L
= EI (v 00 (x; a1 , a2 ))2 dx
∂ai ∂ai 2 0
1 Z L
∂ 00
= EI (v (x; a1 , a2 ))2 dx
2 0 ∂ai
1 Z L
∂ 00
= EI 2 v 00 (x; a1 , a2 ) v (x; a1 , a2 ) dx
2 0 ∂ai
∂V ∂
= −F v(xo ; a1 , a2 ))
∂ai ∂ai

These equations need to be evaluated separately for ai = a1 and for ai = a2 .


For ai = a1 ,
! !
Z L
π2 πx 4π 2 2πx π2 πx πxo
0 = δΠ1 = EI − 2 a1 sin − 2 a2 sin − 2 sin dx − F sin
0 L L L L L L L
4 Z L 4
π πx 4π Z L
2πx πx πxo
= EI 4 a1 sin2 dx + EI 4 a2 sin sin dx − F sin
L 0 L L 0 L L L
π4 L πxo
EI 4 a1 = F sin
L 2 L
1 2L3 πxo
a1 = 4
F sin
EI π L

For ai = a2 ,
! !
Z L
π2 πx 4π 2 2πx 4π 2 2πx 2πxo
0 = δΠ2 = EI − 2 a1 sin − 2 a2 sin − 2 sin dx − F sin
0 L L L L L L L
4 4
4π Z L
πx 2πx 16π Z L
2 2πx 2πxo
= EI 4 a1 sin sin dx + EI a 2 sin dx − F sin
L 0 L L L4 0 L L
16π 4 L 2πxo
EI 4
a2 = F sin
L 2 L
1 2L3 2πxo
a2 = F sin
EI (2π)4 L

By extension, if the function for v(x) had two more terms, with sin(3πx/L) and sin(4πx/L),
the coefficients would be
1 2L3 3πxo
a3 = 4
F sin
EI (3π) L
3
1 2L 4πxo
a4 = F sin
EI (4π)4 L

Note here that expressions for each coefficient ai can be written separately, independent of the
other coefficients, because of the orthogonality of sine functions. This series of sine functions
is called a Fourier series. In general finding values for n coefficients involves solving a set of
n linear equations with n unknowns, (a linear matrix equation).

CC BY-NC-ND H.P. Gavin


Example of the Principle of Minimum Total Potential Energy 3

1.6

1.4

1.2
v(x) / ( FL3/(π4 EI) )

0.8

0.6

0.4
point load at xo/L = 0.70
0.2 a1 = 1.6180, a2 = -0.1189, a3 = 0.0076, a4 = 0.0046
Π1 = -7.0205, Π2 = -7.5753, Π3 = -7.5803, Π4 = -7.5987
0
0 0.2 0.4 0.6 0.8 1
x/L

0.6

0.5
v(x) / ( FL3/(π4 EI) )

0.4

0.3

0.2

point load at xo/L = 0.90

0.1 a1 = 0.6180, a2 = -0.0735, a3 = 0.0200, a4 = -0.0074

Π1 = -1.1691, Π2 = -1.4283, Π3 = -1.5221, Π4 = -1.5609

0
0 0.2 0.4 0.6 0.8 1
x/L

CC BY-NC-ND H.P. Gavin


4 CEE 201L. – Uncertainty, Design, and Optimization – Duke University – Spring 2015 – H.P. Gavin

1 % example o f an u n c o n s t r a i n e d minimum t o t a l p o t e n t i a l e n e r g y problem . . .


2 % s i m p l y s u p p o r t e d beam o f l e n g t h L w i t h a p o i n t l o a d a t x=x0
3 % assume v ( x ) = a1 s i n ( p i x /L) + a2 s i n (2 p i x /L)
4
5 L = 10; % l e n g t h o f t h e beam , m
6 EI =1 e0 ; % b e n d i n g r i g i d i t y , kN mˆ2
7 F = 1 % l o a d v a l u e , (F>0 : up−ward ) , kN
8
9 x0 = 7.0; % l o c a t i o n of the point load , m
10
11
12 % coefficieints ...
13
14 a1 = 1/ EI * 2* L ˆ3/( pi )ˆ4 * F * s i n ( pi * x0 / L )
15 a2 = 1/ EI * 2* L ˆ3/(2* pi )ˆ4 * F * s i n (2* pi * x0 / L )
16 a3 = 1/ EI * 2* L ˆ3/(3* pi )ˆ4 * F * s i n (3* pi * x0 / L )
17 a4 = 1/ EI * 2* L ˆ3/(4* pi )ˆ4 * F * s i n (4* pi * x0 / L )
18
19
20 dx = L /100; % x−a x i s i n c r e m e n t f o r p l o t t i n g
21 x = [0: dx : L ]; % x−a x i s f o r p l o t t i n g
22
23
24 % four l e v e l s of approximation f o r v ( x ) . . .
25
26 v1 = a1 * s i n ( pi * x / L );
27 v2 = a1 * s i n ( pi * x / L ) + a2 * s i n (2* pi * x / L );
28 v3 = a1 * s i n ( pi * x / L ) + a2 * s i n (2* pi * x / L ) + a3 * s i n (3* pi * x / L );
29 v4 = a1 * s i n ( pi * x / L ) + a2 * s i n (2* pi * x / L ) + a3 * s i n (3* pi * x / L ) + a4 * s i n (4* pi * x / L );
30
31
32 % compute t h e t o t a l p o t e n t i a l e n e r g y PI = U + V f o r each o f t h e a p p r o x i m a t i o n s . . .
33
34 i0 = x0 / dx ; % index of ”x” where the point load i s applied
35 PI1 = (1/2) * EI * trapz (( d i f f ( d i f f ( v1 )/ dx )/ dx ).ˆ2)* dx - F * v1 ( i0 )
36 PI2 = (1/2) * EI * trapz (( d i f f ( d i f f ( v2 )/ dx )/ dx ).ˆ2)* dx - F * v2 ( i0 )
37 PI3 = (1/2) * EI * trapz (( d i f f ( d i f f ( v3 )/ dx )/ dx ).ˆ2)* dx - F * v3 ( i0 )
38 PI4 = (1/2) * EI * trapz (( d i f f ( d i f f ( v4 )/ dx )/ dx ).ˆ2)* dx - F * v4 ( i0 )
39
40 % plotting ...
41
42 FL3EI = F * L ˆ3/ pi ˆ4/ EI ; % a normalization factor
43
44 txt1 = s p r i n t f ( ’ point load at x_o / L = %4.2 f ’ , x0 / L );
45 txt2 = s p r i n t f ( ’ a_1 = %6.4 f , a_2 = %6.4 f , a_3 = %6.4 f , a_4 = %6.4 f ’ , ...
46 a1 / FL3EI , a2 / FL3EI , a3 / FL3EI , a4 / FL3EI );
47 txt3 = s p r i n t f ( ’ \\ Pi_1 = %7.4 f , \\ Pi_2 = %7.4 f , \\ Pi_3 = %7.4 f , \\ Pi_4 = %7.4 f ’ , ...
48 PI1 , PI2 , PI3 , PI4 );
49
50 f i g u r e (1)
51 clf
52 plot ( x /L , v1 / FL3EI , x /L , v2 / FL3EI , x /L , v3 / FL3EI , x /L , v4 / FL3EI )
53 xlabel ( ’x / L ’)
54 ylabel ( ’v ( x ) / ( FL ˆ3/(\ pi ˆ4 EI ) ) ’)
55 text (0.15 ,0.30 , txt1 )
56 text (0.15 ,0.20 , txt2 )
57 text (0.15 ,0.10 , txt3 )
58 axis ’ tight ’

CC BY-NC-ND H.P. Gavin

Vous aimerez peut-être aussi