Optimization: Course Contents
Optimization: Course Contents
Course Contents
6.1 Definition
6.2 Conventional versus
Optimum Design Process
6.3 Engineering Applications
of Optimization
6.4 Statement of an
Optimization Problem
6.5 Terms used in
Optimization
6.6 Classification of
Optimization Problem
6.7 Optimum Design with
Lagrange Multipliers
6.8 Johnson’s Method of
Optimum Design
6.1 Definition
− Optimization is the act of obtaining the best result under given circumstances.
− In design, construction, and maintenance of any engineering system, engineers have
to take many technological and managerial decisions at several stages. The ultimate
goal of all such decisions is either to minimize the effort required or to maximize the
desired benefit.
− Optimization can be defined as the process of finding the conditions that give the
maximum or minimum value of a function.
1. The optimum design method has block 0, where the problem is formulated as one of
optimization. An objective function is defined that measures the merits of different
designs.
2. Both methods require data to describe the system in block 1.
3. Both methods require an initial design estimate in block 2.
4. Both methods require analysis of the system in block 3.
5. In block 4, the conventional design method checks to ensure that the performance
criteria are met, whereas the optimum design method checks for satisfaction of all of
the constraints for the problem formulated in block 0.
6. In block 5, stopping criteria for the two methods are checked, and the iteration is
stopped if the specified stopping criteria are met.
7. In block 6, the conventional design method updates the design based on the designer’s
experience and intuition and other information gathered from one or more trial
designs; the optimum design method uses optimization concepts and procedures to
update the current design.
Fig. 6.2 Comparison of (a) conventional design method and (b) optimum design method.
− The foregoing distinction between the two design approaches indicates that the
conventional design process is less formal. An objective function that measures a
design’s merit is not identified. Trend information is usually not calculated; nor is it
used in block 6 to make design decisions for system improvement. In contrast, the
optimization process is more formal, using trend information to make design changes.
x1
x
Find X = 2 which minimizes f(X)
:
xn
Subject to the constraints
gj (X) ≤ 0, j = 1, 2,….., m
lj (X) ≤ 0, j = 1, 2,….., p (6.1)
where X is an n-dimensional vector called the design vector,
f (X) is termed the objective function, and
gj (X) and lj (X) are known as inequality and equality constraints, respectively.
The problem stated in Eq. (6.1) is called a constrained optimization problem.
Some optimization problems do not involve any constraints and can be stated as
x1
x
Find X = 2 which minimizes f(X)
:
xn
Such problems are called unconstrained optimization problems.
and N2. Since these constraints depend on the performance of the gear pair, they are
called behavior constraints.
− The values of T1 and T2 cannot be any real numbers but can only be integers. Further,
there can be upper and lower bounds on T1 and T2 due to manufacturing limitations.
Since these constraints depend on the physical limitations, they are called side
constraints.
IV. Constraint Surface
− For illustration, consider an optimization problem with only inequality constraints gj
(X) ≤ 0. The set of values of X that satisfy the equation gj (X) = 0 forms a
hypersurface in the design space and is called a constraint surface.
− The constraint surface divides the design space into two regions:
− one in which gj (X) < 0 and the other in which gj(X)>0.
− Thus the points lying on the hypersurface will satisfy the constraint gj (X) critically,
whereas the points lying in the region where gj (X)>0 are infeasible or unacceptable,
and the points lying in the region where gj (X) < 0 are feasible or acceptable. The
collection of all the constraint surfaces gj (X) = 0, j = 1, 2, . . . ,m, which separates the
acceptable region is called the composite constraint surface.
− Here the design variables are functions of the length parameter t. This type of
problem, where each design variable is a function of one or more parameters, is
known as a trajectory or dynamic optimization problem.
n
Subject to g j ( X ) = ∑ g ij ( xi ) ≤ b j , j = 1, 2,……., m
i =1
Where bj is a constant
(h) Classification Based on the Number of Objective Functions
Depending on the number of objective functions to be minimized, optimization problems can
be classified as single and multiobjective programming problems.
Multiobjective Programming Problem. A multiobjective programming problem can be stated
as follows:
Find X which minimizes f1(X), f2(X), . . . ,fk(X)
subject to
gj (X) ≤ 0, j = 1, 2, . . . ,m
wheref1, f2, . . . , fk denote the objective functions to be minimized simultaneously.
Example 6.1:
Design a uniform column of tubular section, with hinge joints at both ends, (Fig. 6.4) 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 × 106kgf/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.
The design variables are the mean diameter (d) and tube thickness (t):
x d
X = 1=
x2 t
The objective function to be minimized is given by
f (X) = 5W + 2d
= 5 (ρv) +2d
= 5 (ρAl) + 2d
= 5 (ρlπdt) + 2d
= 9.82x1x2 + 2x1
The behavior constraints can be expressed as
stress induced ≤ yield stress
stress induced ≤ buckling stress
The induced stress is given by
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
These points are plotted and a curve P1Q1 passing through all these points is drawn as shown
in Fig. 6.7, 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(x12+ x22 ) ≥
47.3 and the points lying on the constraint surface g2(X) = 0 can be obtained as follows for
x1x2(x12+ x22 ) = 47.3:
These points are plotted as curve P2Q2, the feasible region is identified, and the infeasible
region is shown by hatched lines as in Fig. 6.7. 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.
For 9.82 x1 x2 + 2 x1 = 50
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
For 9.82 x1 x2 + 2 x1 = 40
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 9.82 x1 x2 + 2 x1 = 20
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. 6.7 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
t* = x2* = 0.293 cm
fmin = 26.53.
Example 6.2:It is required to design an oil can, shown in Figure 6.8, to hold at least 400 ml
of oil (1 ml = 1 cm3), as well as to meet other design requirements. The cans will be produced
in the billions, so it is desirable to minimize their manufacturing costs. Since cost can be
directly related to the surface area of the sheet metal used, it is reasonable to minimize the
amount of sheet metal required. Fabrication, handling, aesthetics, and shipping considerations
impose the following restrictions on the size of the can: The diameter should be no more than
8 cm and no less than3.5 cm, whereas the height should be no more than 18 cm and no less
than 8 cm.
The two design variables are defined as
D = diameter of the can, cm
H = height of the can, cm
x D
Design vector X = 1 =
x2 H
The objective function is minimization of cost of manufacturing of can
π
f ( x) = π DH + 2 D2
4
f ( x) = 3.14( x1 x2 + 0.5 x12 ) …………………….. (1)
Fig. 6.8 Can
Side constraints
8 ≥ D ≥ 3.5 cm
18 ≥ H ≥ 8 cm ……………………………….. (2)
Behavior constraints
V ≥ 400 ml (or cm3)
π
D 2 H ≥ 400 cm3
4
x12 x2 ≥ 509.3 …………………………………(3)
g(x) = x12 x2 ≥ 509.3
Get the solution from (2), (1) and (a)
x1 3.5 4.5 6 7 8
x2 41.57 25.15 14.14 10.39 8
f(x) 476.08 387.16 323.05 305.37 301.44
At x1 = 8 and x2 = 8 objective function is minimize and also satisfy the side constraints so,
D = 8 cm, H = 8 cm.
To plot the graph and to find feasible region with optimum solution, find x2 for various f(x)
by varying x1.
Let f ( x) = 3.14( x1 x2 + 0.5 x12 ) = 200
x1 1 2 3 4 5 6 7 8
x2 63.19 30.84 19.73 14 10.24 7.61 5.6 3.96
x1 1 2 3 4 5 6 7 8
x2 95.04 46.71 30.39 21.88 16.6 12.92 10.15 7.94
From the graph BCE is Feasible Region. Point B shows optimum solution which gives
minimum surface area and minimum cost.
D = x1 = 8 cm
H = x2 = 8 cm
f(x) = 301.44
Example 6.3: A company manufactures two machines, A and B. Using available resources;
either 28 A or 14 B can be manufactured daily. The sales department can sell up to 14 A
machines or 24 B machines. The shipping facility can handle no more than 16 machines per
day. The company makes a profit of $400 on each A machine and $600 on each B machine.
How many A and B machines should the company manufacture every day to maximize its
profit?
Design variable
The following two design variables are identified in the problem statement:
xl = number of A machines manufactured each day
x2 = number of B machines manufactured each day
Objective function
The objective is to maximize daily profit, which can be expressed in terms of design variables
as f(x) = 400x1+ 600x2
Identification of constraints:
Design constraints are placed on manufacturing capacity, on sales personnel, and on the
shipping and handling facility.
The constraint on the shipping and handling facility is quite straightforward:
x1+ x2 ≤ 16 (shipping and handling constraint) – Behaviour constraint.
x1 x2
+ ≤ 1 (Manufacturing constraint) – Behaviour constraint.
28 14
x1 x2
+ ≤ 1(Sales Dept. limitations) – Behaviour constraint.
14 24
Finally, the design variables must be non-negative asx1, x2 ≥ 0 – Side constraint.
[Here the formulation remains valid even when a design variable has zero value. The problem
has two design variables and five inequality constraints. All functions of the problem are
linear in variables xl and x2. Therefore, it is a linear programming problem.]
Graphical Solution:
x1+ x2 ≤ 16 → g1
x1 0 16
x2 16 0
Fig. 6.10 Constraint boundary and Feasible/infeasible sidefor the inequality x1+ x2 ≤ 16 in the
profit maximization problem.
x1 x2
+ ≤ 1 → g2
28 14
x1 0 28 2 14
x2 14 0 13 7
Fig. 6.11Feasible region and Graphical solution tothe profit maximization problem
x1 x2
+ ≤1 → g3
14 24
x1 0 14
x2 24 0
x1 ≥ 0 → g4
x2 ≥ 0 → g5
Get the optimum function solution (To locate the optimum point)
f(x) = 400x1+ 600x2 Let at (6,4)
= 400 x 6+ 600 x 4
= 4800
f(x) = 400x1+ 600x2= 2400
f(x) = 400x1+ 600x2= 7200
As the contours move up toward point D, feasible designs can be found with larger values for
f(x). It is clear from observation that point D has the largest value for f(x) in the feasible
region. Now simply read the coordinates of point D (4, 12) to obtain the optimum design,
having a maximum value for the profit function as
f(x) = 400x1+ 600x2= 8800.
Example 6.4: A manufacturing firm produces two machine parts using lathes, milling
machines and grinding machines. The different machining times required for each part, the
machining time available on different machines and the profit on each machine part are given
in the following table.
Machining Time required ( in min) Maximum time
Type of Machine available per week
I (x) II (y)
(minutes)
Lathe 10 5 2500
Milling machine 4 10 2000
Grinding machine 1 1.5 450
Profit per unit Rs. 50 Rs. 100
Determine the number of parts I and II to be manufactured per week to maximize the profit.
Coordinates of point M :
Point M can be obtained as intersection of lines,
A B A B
+ =1 + =1
250 500 500 200
500 A + 250 B = 125000 and 200 A + 500 B = 100000
5 A + 2.5 B 1250 and 2 A + 5 B = 1000
10 A + 5 B = 2500
+ 2 A + 5B = + 1000
8 A = 1500
A =187.5
B = 125
Objective function curve through M :
For objective function curve through M (187.5, 125),
f(X) = P = 50A+ 1008
= 50 x 187.5 + 100 x 125
= 21875
Equation of objective function curve (line) through M is,
50 A + 100 B = 21875
50 A + 100 B
=1
21875
A B
+ =1
437.5 218.75
Optimum solution:
As P = f(X) is maximum at point M, the optimum solution is given by point M with,
A = 187.5 ≈ 187 (As A has to be an integer)
and B = I25
P = f(X) = 50 A + 100 B
= 50 x 187 + 100 x 125
= 21850
Number of parts I, A = 187
Number of parts II, B = 125
Profit, P = Rs.21850
Example 6.5: A heat exchanger is to be designed for minimum cost. The design
requirement indicated that 300 m of tube length in all be installed for achieving necessary
heat transfer surface.
The total cost of installation has been approximated as:
1) Total cost of tubes = Rs.24500
2) Cost of shell = Rs.875D2.5L
where D = shell diameter and
L = length of each tube ≈ Length of heat exchanger
3) Cost of floor space occupied by the heat exchanger = 700 D L.
The spacing of the tubes is such that 20 tubes can be accommodated in cross sectional area of
l m2insidetheshell.Determine the diameter D and length L of the heat exchanger to minimize
costs.
Example 6.6: Find the solution using the Lagrange multiplier method:
Minimize f (x, y) = kx−1y−2
(x, y) = x2 + y2 − a2 = 0
Subject to g(x,
Example 6.7: Find the dimensions of a cylindrical tin (with top and bottom) made up of
sheet metal to maximize its volume such that the total surface area is equal toA0 = 24π.
If x1 and x2 denote the radius of the base and length of the tin, respectively, the problem can
be stated as
Maximize f (x1, x2) = πx12x2
equation for an airplane shaft may be the one which expresses its weight, which is an
undesirable effect and needs to be minimized.
(b) Subsidiary Design Equations (SDE)
In a mechanical element, design equations other than PDE are called Subsidiary Design
Equations (SDE). SDE express either functional requirements or significant undesirable
effects, whether they are directly specified or indirectly implied. The most important SDE are
the stress equations, which are generally implied.
(c) Limit Equations or Constraints (LE)
As mentioned previously, various parameters have certain ranges which are expressed in
mathematically simple limit equations. For example, the limitations in stress are imposed by
the strength of the material. Limitations on geometry are imposed by certain functional
requirement or space constraints.
Example 6.8: Design a tensile bar of length L = 200 mm to carry a tensile load of 5 kN
forminimum cost, out of the following materials:
cρ
Material MSF =
Sy
16 × 7500
Steel = 923.077
130
32 × 3000
AL-alloy = 1920
50
480 × 4800
Titanium Alloy = 25600
90
32 × 2100
Magnesium Alloy = 3360
20
It is seen that steel has the lowest MSF and hence should be used to cost for achieve
optimization in cost for given conditions
Assuming F.S. = 2, the free variable A (area) can be calculated by using SDE i.e.
Example 6.9: In a light weight equipment, a shaft is transmitting a torque of 900 Nm and is
to have a rigidity of 90 Nm/degree. Assume a factor of safety of 1.5 based on yield stress.
Design the shaft with minimum weight. What will be the change in design for minimum cost?
Assume maximum shear stress theory of failure. Use the following data for the materials:-
(1) PDE:
W = ρ x L x g x d2 x π/4
MT, F.S. and K aree specified functional requirements. Hence, the minimum
imum MSF will mean
minimum weight.
Calculating from the data given
8500 × 80 ×109
Steel = 0.04024 0.6438
(130 ×106 )2
4800 × 40 ×109
Titanium Alloy = 0.02370 11.3760
(90 ×106 )2
Minimum MSF (0.02370) is for Titanium Alloy and hence it should be used to achieve,
minimum weight for given conditions.
From the table, it is seen that the minimum MSF (0.6438) is tor steel alloy which will result,
in minimum cost.
Example 6.10 : A simple tensile bar is subjected to the specified constant tensile force ‘F’.
Design the bar with the objective of minimizing the material cost using factor of safety ‘Nf’.
The following limitation is specified in the optimum design.
Lmin ≤ L ≤ Lmax
Fig. 6.14
Cm = material cost of the tensile bar, Rs.
c = cost per unit mass for the bar material, Rs /kg
ρ = mass density of the bar material, kg / m3
A = cross-sectional area of the bar, m2
……………(f)
• This is final P.D.E. In order to minimize ' Cm ', the material selection factor [c ρ / Syt ]
should be minimum.
7. Selection of material :
• The material selection factor [c ρ / Syt] is calculated for all the available material and
the material with the lowest value of [c ρ / Syt] is selected.
8. Determination of eliminated parameter :
• The eliminated parameter 'A' can be determined by using S.D.E. [Equation (b)].
Therefore, substituting the limiting value of σt Equation (b),
Example 6.11: A tensile bar of circular cross-section is subjected to the cyclic tensile force
which varies from zero to maximum. Design the bar with the objective of maximizing the
energy absorption capacity using the factor of safety 'Nf'. The following limitations are to be
used in the design due to space and assembly restrictions.
d ≤ dmax
Lmin ≤ L ≤ Lmax
Fig. 6.15
……………………(c)
The maximum tensile stress induced in the bar is given by,
…………………(d)
For the repeated stress, as shown in Fig. 6.16, the mean stress and the stress amplitude are
given by,
The Equations (h), (i) and (j) are the limit equations.
4. Classification of parameters :
Specified Limited Unspecified and
Unlimited
Functional Requirement Parameters Nf U, Fmax, δmax
Undesirable Effect Parameters σmax
Geometrical Parameters L,d A
Material Parameters c, ρ, Syt
………..(k)
Again S.D.E. (d) is combined with P.D.E. [Equation (k)] by eliminating the unspecified and
unlimited parameter Fmax Hence,
………..(l)
This is developed P.D.E. From Equation (l) it is seen that, the effects of all limit equations i.e.
Equations (h), (i) and (r) can be included in the developed P.D.E.
Therefore, this is the case of normal specifications.
6. Combining limit equations with P.D.E. :
The next step is to include the effects of all limit equations in the developed P.D.E.
U α σ2max, hence for maximizing 'U', the parameter σmax, should be placed at its upper limit
2 K .Sut .S yt
i.e.
N f K .Sut + S yt
U α d2, hence for maximizing 'U', the parameter d, should be placed at its upper limit i.e. dmax.
Again U α L, hence for maximizing 'U', the L, should be placed at its upper limit i.e. Lmax.
Substituting these limiting values, Equation (l) can be written as,
…………..(m)
This is final P.D.E. In order to maximize ‘U’, the material selection factor
1 K .S .S 2
ut yt
should be maximum.
E K .S + S yt
ut
7. Selection of material:
1 K .S .S
2
The material selection factor is calculated for all the available materials
ut yt
E K .Sut + S yt
1 K .S .S 2
and the material with the largest value of is selected.
ut yt
E K .Sut + S yt
8. Determination of eliminated parameters :
The eliminated parameters Fmax and δmax are determined by using S.D.E.’s i.e. Equations (c)
and (d).
Substituting limiting values of σmax and ‘d’ in Equation (d),
References :
− Engineering Optimization – S. S. Rao
− Introduction to Optimum Design – J. S. Arora
− Machine & CAD – Farzak Haideri