Design Optimization

Contents (13.09, 2024)

 Introduction: What is design optimization?
 Trends of design approach: Conventional vs. optimized
 Optimization-based design process – simple examples
 Standard Optimization statement
 General optimization procedure
 Classification Design optimization problems
 Classification of solution techniques
 Optimization solvers
 Graphical methods
 Summary
References:
 J.S. Arora, Introduction to Optimum Design, Academic Press, Elsevier, 3rd Ed. 2012.
 K.-H. Chang, Design Theory and Methods using CAD/CAE, Chap. 3 – 5;

Academic Press, Elsevier, 2015.

CAD/CAM and Additive Manufacturing 1
 What is design optimization?

Defn:
An optimization problem in design is a process of selecting the “best”
alternative design within available resources.

The design of a product/system can be formulated as an optimization problem in

which a measure of performance is to be optimized while satisfying all constraints!

Three fundamentsl elements of optimization

Design variables How are different designs described?

Objective function Which criteria is used to select the «best» design?

Constraints What means or resources are available for the design?

(Design requirements)

CAD/CAM and Additive Manufacturing 2

 Current Trends of design Approach
Simulation-based design Optimization

A. Safari, PhD (2010 – 2014)

CAE in design optimization

Courtesy: B. Louhichi, et al 2014: DOI: 10.1007/s00170-014-6248-y

CAD/CAM and Additive Manufacturing 3

 Conventional vs. Optimum Design ....

Optimum design The key features of the two processes:

1. The optimum design method has block 0,
Conventioanl design where the problem is formulated
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.
CAD/CAM and Additive Manufacturing 4
 The Optimization-based Design Process
The Design process
(How do I begin to design a system/ product?)

 Design is a complex task:

Many assumptions, limitations, possibilities, and economical factors should be considered in an
interdisciplinary environment.

 Design begins with analysis of options:

subsystems are identified, designed and tested in an iterative process.

 Effective design involves intitution, use of experience data, making

mathematical analysis, etc.

 Final goal is to achieve cost-effective, efficient, reliable, and durable design.

 The design process should be well organized.

CAD/CAM and Additive Manufacturing 5

 The Optimization-based design process …

 Design is a process that gives the “best” physical size (optimal size) or the
“best” product performance, obtained through iterative process
 optimization is a part of a design process, not the design process
Example: Formulation of design problem for optimization
Optimization
variables

Objective function

NB: Selection of the optimization variables and the objective functions is dictated by the design intent
CAD/CAM and Additive Manufacturing 6
 The Optimization-based design process …

 Engineering design problem: mathematical formulation of the

D
performance parameter as a function of the design variables
to minimize (maximize) under given constraint(s)

h
 Example: Soda can design to maximize the volume, given that
the surface area must not exceed a given limit (A0)
 Design variables: r = D/2 and h
 Objective: Maximize V(r, h),
where: V(r,h) =
 Constraint: S(r, h) = A0 - (Equality constraint);
where S(r,h) =
 r >0; h > 0 - (Side constraint)

 Formulate mathematical expression for V(r,h) and S(r, h).

 Assume A0 = π and formulate an unconstrained function to

solve max. volume
CAD/CAM and Additive Manufacturing 7
 Design Optimization …
 Every optimization problem consists of the following components
Given Note: The general design model
 independent variables: x1, x2 …… xn ∈ R treats only minimization problems.
If f(x) is minimization,
 An objective function f(x) -f(x) becomes maximization
 Constraint functions g(x)
f(x) and g(x) are classified as problem functions.
 Simple formulation of the optimization problem,
f. ex. a minimization problem
min f ( x)
x
subject to (s.t.): x ∈ G
a ≤ g(x) ≤ b
where a and b are the lower and upper limits of the
constraints and G ∈ R is the feasible set.
The above formulation gives a constrained optimization problem.

If the constraint g(x) does not exist, then

we have an unconstrained optimization problem.

CAD/CAM and Additive Manufacturing 8

 Standard Optimization statement

 Objective function

 Inequality constraint

 Equality constraint

 Design variable

NOTE:
Both objective functions and constraint functions must be (explicitly or
implicitly) expressed as a function of design variables, which are controlled
by the designer
CAD/CAM and Additive Manufacturing 9
 Standard Optimization statement …

 Objective function  Design variable

Where, for example f(x = mass)

OR
 Design variables

f(x = section area) = f(b, h)

 X can be manufacturing cost, mechanical strength, displacement, natural

frequency, etc.

 Mathematical expressions consisting of at least one design variable must

be formulated

CAD/CAM and Additive Manufacturing 10

 Design Optimization … Optimum design formulation
 The design problem must be
parametrized

Example: Minimization of beam cross section

 Optimization (independent) variables: X = (b, h)
 Performance index (objective function):

A = f(X) = b*h

Optimization variable:
Parameters to be optimized (i.e. minimized or maximized)
Objective function: performance index derived from optimization variables
Composite objective function:
when it is impossible to express the problem using a single optimization function.
Composite objective functions are often used because many design problems
cannot be expressed by a single function.
CAD/CAM and Additive Manufacturing 11
 Design Optimization … Optimum design formulation …

 Constraints:

functions that give geometrical and functional restrictions on the optimization

problem. Constraints can be considered as measures of performance.

Example, minimizing the x-section (or the weight) of a beam and limiting the size of
the bending stress in the beam.
 Stress in the beam should be constrained to a value less than the allowable stress
of the material

Inequality constraint

 Side constraint:
a constraint that restricts the boundary limits on the optimization variables.
This will be the case when the independent variables have a range of validity
Example: b, h > 0
CAD/CAM and Additive Manufacturing 12
 General Optimization Procedure

CAD/CAM and Additive Manufacturing 13

 Classification of Design Optimization Problems

Knowing problem class (type) is important from computational point of view

- Efficient and appropriate algorithms can be selected for the particular problem class
- The problem class dictates the adopted solution technique

 Diverse forms of classifications exist based on:

1. Existence of constraints:
unconstrained & constrained problems.
Constraints can be expressed as equality or inequality functions.

CAD/CAM and Additive Manufacturing 14

 Classification of Design Optimization problems
Diverse forms of classifications …
2. Optimization objectives:
Single objective vs. multiobjective optimization problems (SOP vs. MOP).

3. Nature of expression:
Linear, nonlinear and
quadratic programming problems

4. Discipline based:
Single disciplinary and multidisciplinary optimization problems (SDO vs. MDO)
Example: structural optimization (SDO) and structural and thermal (MDO)

5. Value of design variables:

Discrete optimization (integer programming problem) vs.
continuous optimization (real number variables)
For discrete optimization, its value must be selected from a given finite set of values
6. Deterministic nature of variables:
Deterministic vs. stochastic programming problems
CAD/CAM and Additive Manufacturing 15
 Classification of solution techniques

1. Optimality criteria (classical) method:

- based on differential calculus (contributed by Newton, Lagrange and Bernoulli Euler)
- assumes that the function is differentiable twice with respect to the design
variables and the derivatives are continuous.

CAD/CAM and Additive Manufacturing 16

 Optimum design problem formulation
Optimization Methods: Classification of Solution Techniques

2. Graphical method:
- provides clear picture of the feasible region and
iso-lines of the objective function (f(x) & g(x)
are plotted in the design space)
- limited application – difficult to handle
more than two design variables.
Solution without numerical calculation. Graphical solution of the beam section
CAD/CAM and Additive Manufacturing 17
 Optimum design problem formulation
Optimization Methods: Classification of Solution Techniques

3. Search methods:
- Involve numerical calculation through iterative process
- Derivatives of objective function and constraint function guide the search
( gradient-based).
Example; steepest descent method for unconstrained optimization (invented by Cauchy)
- Non-gradient approaches use certain rules not based on derivatives, e.g. SA, GA, ..
CAD/CAM and Additive Manufacturing 18
 Optimum design problem formulation
Optimization Methods: Classification of Solution Techniques

Can be constrained or
unconstrained

CAD/CAM and Additive Manufacturing 19

 Optimality Criteria
 Optimality criteria are the conditions a function must satisfy at its minimum point.

Basic calculus concepts regarding maximum and minimum

 Stationary point (x*): a point at which a continuous and differentiable function
f(x) vanishes  f’(x) = df(x)/dx = 0 at x = x*
 f’’(x = x*) > 0  the stationary point is a minimum
 f’’(x = x*) < 0  the stationary point is a maximum
 f’’(x = x*) = 0  the stationary point is an inflection point

CAD/CAM and Additive Manufacturing 20

 Optimality Criteria
Basic calculus concepts …
 Global and local m inim um :
 a function f(x) has a global (absolute) minimum at x = x* if f(x*) ≤ f(x) for allx in the
domain over which f(x)is defined
 a function f(x) has a local (relative) minimum at x = x* if f(x*) ≤ f(x* + δ) for all
sufficiently small positive or negative values of
δ in the neighborhood of the stationary point x*

 Necessary condition: For f(x*), x ∈ (a, b) to be

minimum for f(x), the necessary condition is
f’(x = x*) = 0
 Sufficient condition: For f(x*), x ∈ (a, b), if f’(x = x*) = 0
 the sufficient condition to be minimum value of f(x): f’’(x = x*) > 0
 the sufficient condition to be maximum value of f(x): f’’(x = x*) < 0

CAD/CAM and Additive Manufacturing 21

 Optimum design problem formulation

Proper definition and formulation of the problem takes roughly 50% of the total
effort needed to solve it!

In other words:
Process of transforming the current design of a system/ product into an optimum
design problem is crucial.

 Design problem formulation:

transforming verbal description (often qualitative) into a quantitative statement

in a mathematical form that defines the optimization problem

 Converting qualitative design requirements into

quantitative performance measures

CAD/CAM and Additive Manufacturing 22

 Optimum design problem formulation …

 Five-step formulation procedure:

1. Project/problem statement.

2. Data/info collection.

3. Identification/definition of design variables.

 Should be independent as far as possible
 A minimum number of design variables required
 The numerical values could be given to these variables

4. Description of a criterion that is to be optimized

 Driving the objective function of the design variables.
Examples: cost, profit, weight, energy consumption, aerodynamic efficiency
of a blade, etc.
 Both single- and multiobjective fitness functions.
 Complex systems may need establishing several subsystems.

CAD/CAM and Additive Manufacturing 23

 Optimum design problem formulation …
 Five-step formulation procedure …

5. Identification of constraints  restrictions on design variables

 Identifying constraints and developing expressions A meaningful constraint

must be a function of at
 Consider resources and performance requirements least one design variable

 Feasible design (point): meets all the design requirements.

 Feasible region: The domain composed of feasible design points.
All the constraints are satisfied in the feasible region.
 An infeasible design
does not meet one or
more of the requirements.

CAD/CAM and Additive Manufacturing 24

 Optimum design problem formulation …
 Five-step formulation procedure …

5. Identification of constraints

 Equality and inequality constraints

CAD/CAM and Additive Manufacturing 25

 Optimum design problem formulation …

Example 1: Design of a Soda Can

1) Problem statement: To design a soda can that holds at least400 ml soda. The cost of
sheet metal is to be minimized. Diameter no more than 8 cm and no less than 3.5 cm,
height no more than 18 cm and no less than 8 cm.

D
2) Data collection: (stated in problem statement)
3) Identification/definition of design variables:

h
Two design variables: x1 = D and
x2 = h.
4) Criterion to Be Optimized:
minimizing the total surface area S of the sheet metal
for the three parts of the cylindrical can

CAD/CAM and Additive Manufacturing 26

 Optimum design problem formulation …

Example 1: Design of a Soda Can …

D
4) Criterion to Be Optimized:

h
minimizing the total surface area S
πD 2 1
h ≤ 400 ml (= cm ) ⇒ g ( x) = π x12 x2 ≤ 400
3

4
Objective function : 4

πD 2 π x12
F = πDh + ⇒ f ( x) = π x1 x2 +
2 2
5) Constraints: the can must hold at least 400 ml
𝜋𝜋𝐷𝐷2 1 2
h ≥ 400 𝑚𝑚𝑚𝑚 (= 𝑐𝑐𝑐𝑐3 ) 𝑥𝑥1 𝑥𝑥2 ≥ 400)
4 4

Side constraints (Sizing constraints) on

design variables:

3,5 ≤ x1 ≤ 8 cm and 8 ≤ x2 ≤ 18 cm

CAD/CAM and Additive Manufacturing 27

 Optimum design problem …

 Exercise (homework)

 Formulate the soda can design optimization problem to maximize the

volume subject to a constrained (specified) sheet metal area A0

CAD/CAM and Additive Manufacturing 28

 Optimum design problem …
Example 2: design of a two-bar truss structure

Step 1: Problem statement

• Problem: Designing a two-bar bracket (truss
structure) to support a force W without failure
• Angle θ is between 00 and 900
• Total cost is directly related to the size
of the 2 bars

• Objective:
minimize the total mass of the structure while
satisfying desired performance

CAD/CAM and Additive Manufacturing 29

 Optimum design problem …
Step 2 : Data and information collection
o Predetermined force W and its angle of application ϴ.
o Predefined space limitations (s, h, l, …)
o Material type for the bars is also assumed specified
o Structural performance depends on the bar forces.
From static equilibrium:

CAD/CAM and Additive Manufacturing 30

 Optimum design problem …
Step 2 : Data and information collection …
Expression for
• Stress level in bars (to avoid failure)

• What form of the area?

• Design stress – allowable stress accounting for design safety factor.

Yield stress Safety factor

CAD/CAM and Additive Manufacturing 31
 Optimum design problem…
Step 3: Definition of design variables
o Several sets of design variables may be identified:
height h and span s,
o cross-sectional shape of bars 1 & 2
o A possible set of parameters (for tubular x-section):
x1= height h of the bracket
x2= span s of the bracket
x3= outer dia. of bar 1
x4= inner dia. of bar 1
x5= outer dia. of bar 2
x6= inner dia. of bar 2

CAD/CAM and Additive Manufacturing 32

 Optimum design problem…
Step 3: Definition of design variables …

• Identifying cross-sectional area for the

selected tubular section

Assuming the x-sections of bar 1 and 2 are not equal

CAD/CAM and Additive Manufacturing 33

 Optimum design problem…

Step 4: Identify criteria to be optimized

o Objective function:
The mass of the structure is to be minimized
 the mass is formulated as the cost function of the problem.
o Expression for the mass:

x1
0.5x2

CAD/CAM and Additive Manufacturing 34

 Optimum design problem…

Step 5: Identify constraints

o The acceptable stress in the bars without over-stressing

o Side constraints

Xil = lower limit & Xiu = upper limit values for design variables

CAD/CAM and Additive Manufacturing 35

 Optimum design problem …
 Summary of the formulation

 Specified data: W, ϴ, σa > 0;

xil and xiu, i = 1 to 6
 Design variables: x1, x2, x3, x4, x5, x6
 Cost function: Minimize total mass of the bars

 Constraints: stress in bars

σa = allowable stress

CAD/CAM and Additive Manufacturing 36

 Optimum design problem ….
Example 3: Formulation of design optimization problem
 Four identical helical coil springs are used to support a
a vehicle weighing 2000 kg. Formulate the problem of
finding the wire diameter (d), coil diameter (D), and
the total number of turns (N) of each spring for
minimum weight by limiting the spring deflection to 2,5
mm and the shear stress to 70 MPa.

 In addition, the natural frequency of vibration of the

spring is to be greater than 100 Hz.

CAD/CAM and Additive Manufacturing 37

 Optimum design problem ….
Example 3: Formulation of design optimization problem …

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
Gd 4 where
k=
8ND 3 G is the shear modulus,
8 FD F is the compressive load on the spring,
τ = ks m is the mass of the spring,
πd 3

ρ is the weight density of the spring, and

1 k Ks (= 1,05) is the shear stress correction factor.
fn =
2π m
Assume that the material of the spring is steel.

CAD/CAM and Additive Manufacturing 38

 Optimum design problem ….
Solution: Five step Formulation of design optimization problem
Step 1: Problem statement
The objective is to design support springs with minimum weight
(mass) that can carry a given axial load without material failure
and satisfy the given constraints.

Step 2: Data collection (as given in problem description)

Step 3: Definition of design variables

The following independent variables are identified: wire
diameter (d), mean coil diameter (D) and number of coils (N)

d   x1 
  
x =  D  =  x2  
 N   x3 

CAD/CAM and Additive Manufacturing 39

 Optimum design problem ….
Solution: Five step Formulation of design optimization problem
Step 4: Definition of the fitness function (optimization
criteria), f(x)

f ( x) = A ⋅ L ⋅ ρ ρ = mass density
πd 2
π 2
2
= ⋅ πDN ⋅ ρ = d DNρ
4 4
π 2ρ
⇒ f ( x) = x12 x2 x3
4

CAD/CAM and Additive Manufacturing 40

 Optimum design problem ….
Solution: Five step Formulation …
Step 5: Constraints, g(x)
i) Deflection constraint:
4
F Gd
δ = ≤ 2,5 mm where k = 3
k 8 ND
8FND 3
= 4
≤ 2,5 mm
Gd
where F = 2000 * 9,81 / 4 ≈ 4905 N
and G = 80,7 GPa
4 −3 −1
⇒ g1 ( x) = 5,1x x x 1 2 3 >1

CAD/CAM and Additive Manufacturing 41

 Optimum design problem ….
Solution: Five step Formulation … Torsional moment

Step 5: Constraints, g(x) …

ii) Shear stress constraint:

Direct shear force

4 Fs 16 F
τs = =
3 A 3π d 2

8FD
τ = ks ≤ 70 MPa
πd 3

Max. shear stress in direct shear

For circular and rectangular x-sections ⇒ g 2 ( x) = 5,24 *10 −3 x13 x2−1 > 1
CAD/CAM and Additive Manufacturing 42
 Optimum design problem ….
Solution: Five step Formulation …
Step 5: Constraints, g(x)
iii) Natural frequency constraint:
Note: Natural frequency is the rate at which a body vibrates
when disturbed without being subject to a driving or damping
force

1 k
fn = ≥ 100
2π m
Damping factor
Gd 4
πd 2
where k = 3
and m = DNρ
8 ND 4
Damping ratio
G 2 −1
⇒ g 3 ( x) = x x 2 3 >1
200π 2 ρ

iv) Diameter constraint (side constraint):

D + d ≤ D0
CAD/CAM and Additive Manufacturing 43
 Optimization methods and Solvers

Example
illustrations

CAD/CAM and Additive Manufacturing 44

 Optimization Solvers: EXCEL Solver

Soda can
parameters

CAD/CAM and Additive Manufacturing 45

 Optimization Solvers: MATLAB Toolbox functions

CAD/CAM and Additive Manufacturing 46

 Classification of Solution Techniques

2. Graphical method:
- provides clear picture of the feasible region, iso-lines of the objective function
f(x) & the constraint functions gi(x), which are plotted in the design space)

- has limited application

–> difficult to handle more than two design variables.

47
 Graphical method: Introduction
 Graphical method is best convenient
to visualize in solving optimization
problems of few variables
 (cumbersome for problems with more
than two variables)
 The solution is obtained quickly without
involving numerical algorithms
 The approach involves:
 Plotting all constraint functions
 Identifying a set of feasible designs (the
feasible set) for the problem
 Drawing the objective function contours, and
 Determining the optimum design by
inspection
 We use the five-step procedure to
illustrate the solution approach using
examples

48
 Graphical method …
Example 1: Optimization of beam cross-section
Step 1: Project/problem description
 A beam of rectangular cross-section is subjected to
a bending moment M (Nmm) and a maximum shear
force V (N). The allowable stresses in bending and
shear are 10 MPa and 2 MPa, respectively. It is also
desirable that the depth of the beam not exceeds twice
its width and that the cross-sectional area of the beam be minimized.
 Formulate and solve the problem using the graphical method.
 The bending stress in the beam is calculated as
6𝑀𝑀
𝜎𝜎 = 2 (MPa), and
𝑏𝑏ℎ
3 𝑉𝑉
 The max. shear stress is calculated as τ= (MPa),
2 𝑏𝑏ℎ
where b is the width and h is the depth of the beam
49
 Example 1…
Step 2: Data and information collection
 Let bending moment M = 40 * 103 kN.mm and

the shear force V=150 kN.

All other data and necessary equations are given in the project statement.

 Step 3: Definition of design variables:

The two design variables are
 h = x1= depth of beam, mm
 b = x2 = width of beam, mm

 Step 4: Definition of optimization criteria: The cost function for

the problem is the cross-sectional area, which is expressed as

A = f(b,h) = bh or f(x1, x2) = x1x2

50
 Example 1…

 Step 5: Formulation of constraints: The constraints for the problem

consist of bending stress, shear stress, and depth-to-width ratio.
6𝑀𝑀 6× 40 × 103 ×103
 Bending stress: 𝜎𝜎 = =
𝑏𝑏ℎ2 𝑏𝑏ℎ2

240 × 106
= (MPa) ≤ 𝜎𝜎𝑎𝑎 = 10 MPa
𝑏𝑏ℎ2

240 × 106
 𝑔𝑔1 = − 10 ≤ 0
𝑏𝑏ℎ2
3 𝑉𝑉 3 150× 103
 Shear stress: τ = =
2 𝑏𝑏𝑏 2 𝑏𝑏𝑏

225 × 103
= (MPa) ≤ τ𝑎𝑎 = 2 MPa
𝑏𝑏𝑏
225 × 103
 𝑔𝑔2 = −2 ≤0
𝑏𝑏ℎ

51
 Example 1…

Step 5: Formulation of constraints: …

 Constraint regarding the limitation that the depth be no more

than twice the width can be expressed as

 𝑔𝑔3 = h −2𝑏𝑏 ≤ 0

 Both design variables

should be non-negative:

 𝑔𝑔4 = −𝑏𝑏 ≤ 0 and

𝑔𝑔5 = -h ≤ 0

52
 Example 1…

Step 5: Formulation of constraints: …

 Note that cost function

is parallel to the constraint

g2 (both functions have
the same form:
bh = constant)
 any point along the curve

A – B represents an
optimum solution, so
there are infinite
optimum designs.

 a wide choice of optimum

solutions is desirable for
the designer

53
 Example 1…

Optimum solution (graphical)

The optimum cross-sectional area:
Area = 112 500 mm2.

Extreme optimum solutions:

Point A: b = 527.3 mm and h = 213.3 mm.
Point B: b = 237 mm and h = 474 mm.

All other solutions lie between these two

points on the curve A - B.

54
 Example 1…
Summary of the optimization problem
Minimize: f(b,h) = bh
240 × 106
Such that: 𝑔𝑔1 = − 10 ≤ 0
𝑏𝑏ℎ2
225 × 103
𝑔𝑔2 = −2 ≤0
𝑏𝑏𝑏
𝑔𝑔3 = h −2𝑏𝑏 ≤ 0
𝑔𝑔4 = −𝑏𝑏 ≤ 0 and
𝑔𝑔5 = -h ≤ 0

Homework: Solve the optimization problem and find the optimized value using
a) EXCEL Solver (Try)
b) MATLAB code

55
 Graphical method …
Ex am ple 2: P rofit m ax im ization problem (graphical method)
1. Problem description:
A company manufactures two machines, A and B, with a capacity of either 28 parts
of A or 14 parts of B per day.
The sales department can sell up to 14 machines of A type or 24 machines of B
type, while 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?

2. Data and information are defined in the project statement

3. Definition of design variables

Let x1 = nr. of type A machines manufactured per day
x2 = nr. of type B machines manufactured per day

4. Optimization criterion: Cost function: P = f(x) = 400x1 + 600x2

56
 Example 2 …
5. Formulation of constraints: Objective function: P = f(x) = 400x1 + 600x2
Shipping and handling constraints : A contour is a curve on the graph connecting all
x1 + x2 ≤ 16 ⇒ g1 ( x) = x1 + x2 − 16 ≤ 0 points having the same objective function
Manufacturing constraint : x2
value.
x1 x2 x x
+ ≤ 1 ⇒ g 2 ( x) = 1 + 2 −1≤ 0 Contours of minimization are called Iso-cost
28 14 28 14 curves
Limitations of the sales dept.
x1 x2 x x
+ ≤ 1 ⇒ g 3 ( x) = 1 + 2 −1≤ 0
14 24 14 24
Side constraints (material flow bound) : Solution point (4,12)
x1 ≥ 0 and x2 ≥ 0
with P = $8800

5. Graphical solution – steps

1: Set up the coordinate system Infeasible
reagion
2: Plot the inequality constraints g1
3: Identify the feasible region for each
g2
inequality constraint
4. Identify the feasible region for
the problem
5: Plot the objective function contours ,
starting from a point in the feasible Feasible Infeasible
region reagion reagion
6: Identify the optimum solution
x1
57
 Production process optimization
Ex ercise: Production m ix optim ization
Production mix problem is a common issue when several products are machined at a
single machining center.
A machining company produces two products (designated as product A and B) using 3
machine centers (designated as M1, M2 and M3). The company needs to optimize
(maximize) the number of units x1 and x2 of products A and B respectively that pass
through the machining centers. The engineers have made analysis of the existing
situation and formulated the problem as follows:
2 2
Cost function : f ( x) = x + 4 x
1 2

Constraints : g1 ( x) = 2 x1 + x2 − 4 ≤ 0
g 2 ( x) = 3 x1 + 2 x2 −12 ≤ 0 Solve this optimization
g 3 ( x) = 5 x1 + 4 x2 − 20 ≤ 0 problem graphically!
Side constraints (material flow bound) :
0 ≤ x1 ≤ 3 and 0 ≤ x2 ≤ 6
58
 Summary and Questions

In this lecture, among others, the following issues are discussed

 Design optimization concept

 Current trends of design optimization
 Standard problem statement and procedures
 Classifications of optimization problems and the solution techniques
 Local and global optimization functions
 Optimization methods and solvers
 Graphical methods to solve an optimization with two variables
 …

?
Next: Chap. 5.3: Structural optimization and introduction to search tools
59