Structural Optimization Formulation Dr. Rabi M.

Najem

(Color online) The initial

configuration of the truss

(Color online) The truss after 5, 15, 40, 60, 100, 140, 200 and 260 iterations. The thick-ness of
each bar in the plot is directly proportional to its cross-sectional area

1
Structural Optimization Formulation Dr. Rabi M. Najem
Example 1 Design a uniform column of tubular section, with hinge joints at both ends, (Fig. 1.1) to carry a
compressive load P = 2500 kgf for minimum cost. The column is made up of a material that has a yield stress
(σy ) of 500 kgf/cm2, modulus of elasticity (E) of 0.85 × 106 kgf/cm2, and weight density (ρ) of 0.0025 kgf/
cm3. The length of the column is 250 cm. The stress induced in the column should be less than the buckling
stress as well as the yield stress. The mean diameter of the column is restricted to lie between 2 and 14 cm,
and columns with thicknesses outside the range 0.2 to 0.8 cm are not available in the market. The cost of the
column includes material and construction costs and can be taken as 5W + 2d, where W is the weight in
kilograms force and d is the mean diameter of the column in centimeters.
SOLUTION The design variables are the mean diameter (d) and tube thickness (t):
x1 d
X= = (E1 )
x2 t
The objective function to be minimized is given by
f (X) = 5W + 2d = 5ρlπ dt + 2d = 9.82x1 x2 + 2x1 (E2 )

Fig. 1.1 Tubular column under

compression. The behavior constraints can be expressed as

stress induced ≤ yield stress

stress induced ≤ buckling stress

The induced stress is given by

P 2500
induced stress = σi = = (E3 )
π dt πx1 x2
The buckling stress for a pin-connected column is given by

Euler buckling load π 2 EI 1

buckling stress = σb = = (E4 )
cross-sectional area l 2 π dt
where

I = second moment of area of the cross section of the column

π 4
= (d − di4 )
64 o
π 2 π
= (do + di2 )(do + di )(do − di ) = [(d + t)2 + (d − t)2 ]
64 64
× [(d + t) + (d − t)][(d + t) − (d − t)]
π π
= dt (d 2 + t 2 ) = x1 x2 (x12 + x22 ) (E5 )
8 8

2
Structural Optimization Formulation Dr. Rabi M. Najem
Thus the behavior constraints can be restated as
2500
g1 (X) = − 500 ≤ 0 (E6 )
πx1 x2
2500 π 2 (0.85 × 106 )(x12 + x22 )
g2 (X) = − ≤0 (E7 )
πx1 x2 8(250)2
The side constraints are given by

2 ≤ d ≤ 14
0.2 ≤ t ≤ 0.8

which can be expressed in standard form as

g3 (X) = −x1 + 2.0 ≤ 0 (E8 )

g4 (X) = x1 − 14.0 ≤ 0 (E9 )
g5 (X) = −x2 + 0.2 ≤ 0 (E10 )
g6 (X) = x2 − 0.8 ≤ 0 (E11 )

Since there are only two design variables, the problem can be solved graphically as shown below.

First, the constraint surfaces are to be plotted in a two-dimensional design space where the two axes
represent the two design variables x1 and x2. To plot the first constraint surface, we have

2500
g1 (X) = − 500 ≤ 0
πx1 x2
that is,
x1 x2 ≥ 1.593

Thus the curve x1x2 = 1.593 represents the constraint surface g1(X) = 0. This curve can be plotted by
finding several points on the curve. The points on the curve can be found by giving a series of values to
x1 and finding the corresponding values of x2 that satisfy the relation x1x2 = 1.593:

x1 2.0 4.0 6.0 8.0 10.0 12.0 14.0

x2 0.7965 0.3983 0.2655 0.1990 0.1593 0.1328 0.1140

These points are plotted and a curve P1Q1 passing through all these points is drawn as shown in Fig. 1.2,
and the infeasible region, represented by g1(X) > 0 or x1x2 < 1.593, is shown by hatched lines.†
Similarly, the second constraint g2(X) ≤ 0 can be expressed as x1x2(x2 + x2) ≥ 47.3 and the points lying
on the constraint
1 2 surface g2(X) = 0 can
be obtained as follows for x1 x2 (x12 + x22 ) = 47.3:

x1 2 4 6 8 10 12 14
x2 2.41 0.716 0.219 0.0926 0.0473 0.0274 0.0172

These points are plotted as curve P2Q2, the feasible region is identified, and the infea-sible region is
shown by hatched lines as in Fig. 1.2. The plotting of side constraints is very simple since they represent
straight lines. After plotting all the six constraints, the feasible region can be seen to be given by the
bounded area ABCDEA.
Next, the contours of the objective function are to be plotted before finding the optimum point. For this,
we plot the curves given by
f (X) = 9.82x1 x2 + 2x1 = c = constant

for a series of values of c. By giving different values to c, the contours of f can be plotted with the
help of the following points.
For 9.82x1 x2 + 2x1 = 50.0:
x2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
x1 16.77 12.62 10.10 8.44 7.24 6.33 5.64 5.07
3
Structural Optimization Formulation Dr. Rabi M. Najem
For 9.82x1 x2 + 2x1 = 40.0:

x2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

x1 13.40 10.10 8.08 6.75 5.79 5.06 4.51 4.05

For 9.82x1 x2 + 2x1 = 31.58 (passing through the corner point C):

x2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

x1 10.57 7.96 6.38 5.33 4.57 4.00 3.56 3.20

For 9.82x1 x2 + 2x1 = 26.53 (passing through the corner point B):

x2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

x1 8.88 6.69 5.36 4.48 3.84 3.36 2.99 2.69

For 9.82x1 x2 + 2x1 = 20.0:

x2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

x1 6.70 5.05 4.04 3.38 2.90 2.53 2.26 2.02

These contours are shown in Fig. 1.2 and it can be seen that the objective function cannot be
reduced below a value of 26.53 (corresponding to point B) without violating some of the constraints.
Thus the optimum solution is given by point B with d∗ =
x1∗ = 5.44 cm and t ∗ = x2∗ = 0.293 cm with fmin = 26.53.

Fig. 2 Graphical optimization of Example 1.1.

Fig. 1.2 G raphical optimization of Example

4
Structural Optimization Formulation Dr. Rabi M. Najem

Example 2 Four identical helical springs are used to support a milling machine weighing 5000 lb. Formulate the
problem of finding the wire diameter (d), coil diameter (D), and the number of turns (N) of each spring (Fig. 2.1) for
minimum weight by limiting the deflection to 0.1 in. and the shear stress to 10,000 psi in the spring. In addition, the
natural frequency of vibration of the spring is to be greater than 100 Hz. The stiffness of the spring (k), the shear
stress in the spring (τ ), and the natural frequency of vibration of the spring (fn) are given by

d 4G
k=
8D 3 N
8F D
τ = Ks
πd 3
  √
1 kg 1 d 4G g Gg d
fn = = 3 2
= √
2 w 2 8D N ρ(πd /4)πDN 2 2ρπD 2 N

where G is the shear modulus, F the compressive load on the

spring, w the weight of the spring, ρ the weight density of the
spring, and Ks the shear stress correction factor. Assume that
the material is spring steel with G = 12 × 106 psi and ρ = 0.3
lb/in3, and the shear stress correction factor is Ks ≈ 1.05.

Figure 2.1 Helical spring.

SOLUTION The design vector is given by

   
x1   d 
X = x2 = D
x3 N
   

and the objective function by

πd 2 (E1 )
f (X) = weight = πDNρ
4
The constraints can be expressed as
F 8FD3 N
deflection = = ≤ 0.1
k d 4G
that is,
d 4G (E2 )
g1 (X) = >1
80FD3 N
8FD
shear stress = Ks ≤ 10,000
πd 3
that is,
1250πd 3 (E3 )
g2 (X) = >1
Ks FD
√
Gg d
natural frequency = √ ≥ 100
2 2ρπ D 2 N
that is,
√
Gg d (E4 )
g3 (X) = √ >1
200 2ρπD 2 N

5
Structural Optimization Formulation Dr. Rabi M. Najem

Since the equality sign is not included (along with the inequality symbol, >) in the constraints of Eqs. (E2)
to (E4), the design variables are to be restricted to positive values as

d > 0, D > 0, N >0 (E5 )

By substituting the known data, F = weight of the milling machine/4 = 1250 lb, ρ = 0.3 lb/in3, G = 12 × 106
psi, and Ks = 1.05, Eqs. (E1) to (E4) become
2
f (X) = 14π2(0.3)d2DN = 0.7402x 1 x2 x3 (E6)
d 4 (12 × 106 )
g1 (X) = = 120x14 x2−3 x3−1 > 1 (E7 )
80(1250)D 3 N
1250πd 3
g2 (X) = = 2.992x13 x2−1 > 1 (E8 )
1.05(1250)D
√
Gg d
g3 (X) = √ = 139.8388x1 x2−2 x3−1 > 1 (E9 )
200 2ρπD 2 N

6
Structural Optimization Formulation Dr. Rabi M. Najem
Example 3 A scaffolding system consists of three beams and six ropes as shown in Fig. 3.1. Each of the top
ropes A and B can carry a load of W1, each of the middle ropes C and D can carry a load of W2, and each of the
bottom ropes E and F can carry a load of W3. If the loads acting on beams 1, 2, and 3 are x1, x2, and x3,
respectively, as shown in Fig. 3, formulate the problem of finding the maximum

load (x1 + x2 + x3) that can be supported

by the system. Assume that the weights of
the beams 1, 2, and 3 are w1, w2, and w3,
respectively, and the weights of the ropes
are negligible.

Figure 3.1 Scaffolding system with three beams.

SOLUTION Assuming that the weights of the beams act through their respective middle points, the
equations of equilibrium for vertical forces and moments for each of the three beams can be written as

For beam 3:
TE + TF = x3 + w3
x3 (3l) + w3 (2l) − TF (4l) = 0
For beam 2:
TC + TD − TE = x2 + w2
x2 (l) + w2 (l) + TE (l) − TD (2l) = 0
For beam 1:
TA + TB − TC − TD − TF = x1 + w1
x1 (3l) + w1 ( 92 l) − TB (9l) + TC (2l) + TD (4l) + TF (7l) = 0
where Ti denotes the tension in rope i. The solution of these equations gives
TF = 34 x3 + 12 w3
TE = 14 x3 + 12 w3
TD = 12 x2 + 18 x3 + 12 w2 + 14 w3
TC = 12 x2 + 18 x3 + 12 w2 + 14 w3
TB = 13 x1 + 13 x2 + 23 x3 + 12 w1 + 31 w2 + 59 w3
TA = 23 x1 + 23 x2 + 13 x3 + 12 w1 + 32 w2 + 49 w3
The optimization problem can be formulated by choosing the design vector as

 
x1 
X = x2
x3
 

Since the objective is to maximize the total load

f (X) = −(x1 + x2 + x3 ) (E1 )

The constraints on the forces in the ropes can be stated as

TA ≤ W 1 (E2 )
TB ≤ W 1 (E3 )
TC ≤ W 2 (E4 )
7
Structural Optimization Formulation Dr. Rabi M. Najem

TD ≤ W 2 (E5 )
TE ≤ W 3 (E6 )
TF ≤ W 3 (E7 )

Finally, the nonnegativity requirement of the design variables can be expressed as

x1 ≥ 0
x2 ≥ 0
x3 ≥ 0 (E8 )

8
Structural Optimization Formulation Dr. Rabi M. Najem

Example 4 Formulate the problem of designing a minimum-cost rectangular under-reinforced concrete beam
that can carry a bending moment M with a probability of at least 0.95. The costs of concrete, steel, and
formwork are given by Cc = $200/m3, Cs =$5000/m3, and Cf = $40/m2 of surface area. The bending moment
M is a probabilistic quantity and varies between 1 × 105 and 2 × 105 N-m with a uniform probability. The
strengths of concrete and steel are also uniformly distributed probabilistic quantities whose lower and upper
limits are given by

fc = 25 and 35 MPa
fs = 500 and 550 MPa

Assume that the area of the reinforcing steel and the cross-sectional dimensions of the beam are deterministic
quantities.

SOLUTION The breadth b in meters, the depth d in meters, and the area of reinforcing steel As in square meters
are taken as the design variables x1, x2, and x3, respectively (Fig. 4.1). The cost of the beam per meter length is
given by

f (X) = cost of steet + cost of concrete + cost of formwork

= As Cs + (bd − As )Cc + 2(b + d)Cf (E1 )

The resisting moment of the beam section is given by [1.119]


As fs
MR = As fs d − 0.59
fc b

Figure 4.1 Cross section of a reinforced concrete beam.

and the constraint on the bending moment can be expressed as [1.120]

  
As fs
P [MR − M ≥ 0] = P As fs d − 0.59 − M ≥ 0 ≥ 0.95 (E2 )
fc b
where P [· · ·] indicates the probability of occurrence of the event [· · ·].
To ensure that the beam remains underreinforced,† the area of steel is bounded by the balanced steel area A(sb) as

As ≤ As(b) (E3 )

where
fc 600
A(b)
s = (0.542) bd
fs 600 + fs
Since the design variables cannot be negative, we have
d≥0
b≥0
As ≥ 0 (E4)

9
Structural Optimization Formulation Dr. Rabi M. Najem

Example 5 A uniform column of rectangular cross section is to be constructed for supporting a water tank of
mass M (Fig. 5.1). It is required (1) to minimize the mass of the column for economy, and (2) to maximize the
natural frequency of trans-verse vibration of the system for avoiding possible resonance due to wind. Formulate
the problem of designing the column to avoid failure due to direct compression and buckling. Assume the
permissible compressive stress to be σmax.

SOLUTION Let x1 = b and x2 = d denote the cross-sectional dimensions of the column. The mass of
the column (m) is given by

m = ρbdl = ρlx1 x2 (E1 )

where ρ is the density and l is the height of the column. The natural frequency of transverse vibration of the
water tank (ω), by treating it as a cantilever beam with a tip mass M, can be obtained as

 1/2
3EI
ω= 33
(E2 )
(M + 3
140 m)l

Figure 5.1 Water tank on a column.

where E is the Young’s modulus and I is the area moment of inertia of the column given by

1 3 (E3 )
I= 12 bd

The natural frequency of the water tank can be maximized by minimizing −ω. With the help of Eqs. (E1) and
(E3), Eq. (E2) can be rewritten as
 1/2
Ex1 x23
ω= 33
(E4 )
4l 3 (M + 140 ρlx1 x2 )

The direct compressive stress (σc) in the column due to the weight of the water tank is given by

Mg Mg
σc = = (E5 )
bd x1 x2

and the buckling stress for a fixed-free column (σb) is given by

π 2 EI 1 π 2 Ex22

σb = = (E6 )
4l 2 bd 48l 2

10
Structural Optimization Formulation Dr. Rabi M. Najem

To avoid failure of the column, the direct stress has to be restricted to be less than σmax and the buckling stress
has to be constrained to be greater than the direct compressive stress induced.

Finally, the design variables have to be constrained to be positive. Thus the multiobjective
optimization problem can be stated as follows:

x1
Find X = which minimizes
x2
f1 (X) = ρlx1 x2 (E7 )
 1/2
Ex1 x23
f2 (X) = − 33
(E8 )
4l 2 (M + 140 ρlx1 x2 )

subject to
Mg
g1 (X) = − σmax ≤ 0 (E9 )
x1 x2
Mg π 2 Ex22
g2 (X) = − ≤0 (E10 )
x1 x2 48l 2
g3 (X) = −x1 ≤ 0 (E11 )
g4 (X) = −x2 ≤ 0 (E12 )

11