Structural Optimization Assignment Final Print
Structural Optimization Assignment Final Print
Structural Optimization Assignment Final Print
(BDU)
Structural Optimization
Assignment
Submitted By:
Mulualem Atalel
ID: BDU1022147
Submitted To:
Instructor Seyfe N.
Part -1
The term structural optimization is commonly used for the optimization of engineering structures,
such as building, automobile, or airplane structures for improved strength or stiffness properties
and reduced weight or cost. It is the subject of making an assemblage of materials sustain loads in
the best way. In Structural optimization, the purpose is to find the optimal material distribution of
structures or structural components to achieve certain objectives according to some given demands
of a structure.
3) What are the application areas of linear programming model in your field of specialization?
LP problems arise in some fields of engineering such as water resources, systems engineering,
traffic flow control, resources management, and transportation engineering. In the areas of
aerospace, automotive, structural, or mechanical system design, most problems are not linear.
Structural engineers use linear programming
For reducing the environmental impacts of civil engineering structures, such as greenhouse
gas emission and energy consumption
To minimizing the total cost of structures by reducing structure weight or volume.
To find the minimum weight truss designs satisfying equilibrium
To make decisions about the most efficient use of limited resources – like money, time,
materials, and machinery (cost estimation).
4) List the basic requirements for linear programming problem.
Basic requirements for Linear Programming Problem
a. Decision Variable and their Relationship
The decision variables refer to candidates that are completing with one another for sharing
the given limited resources. These variables are usually interrelated in terms of utilization of
Page | 2
STRUCTURAL OPTIMIZATION ASSIGNMENT
resources and need simultaneous solutions. The relationship among these variables should
be linear.
b. Well-Defined Objective Function:
A linear programming problem must have a clearly- defined objective function to optimize
which may be either to maximize contribution by utilizing available resources, or it may be
to produce at the lowest possible cost by using a limited amount of productive factors. It
should be expressed as a linear function of decision variables.
c. Presence of Constraints or Restrictions
There must be limitations or constraints on the use of limited resources which are to be
allocated among various competitive activities. These constraints must be capable of being
expressed as linear equalities or inequalities in terms of decision variables. The solution of
on LP model satisfy these constraints.
d. Alternative Courses of Action
There must be alternative courses of action. For example, it must be possible to make a
selection between various combinations of the productive factors such as men, machines,
materials, market, etc.
e. Non-Negative Restriction:
All decision variables must assume non-negative values as negative values of physical
quantities is an impossible condition
5) Suppose we want to maximize the area of an object, but we have a choice between a square and a
circle, where the length of the square is equal to the radius of the circle, and the radius can be at
most 4 cm. Formulate this as an optimization problem.
Solution
Let As, Ac corresponds to the area of the square and circle. Let x be the length of the radius. Let
y1, y2 be a binary variable that is y1=1 if the object chosen is square or 0 otherwise.
Minimize: 𝑓(𝐴, 𝑥, 𝑦) = −𝐴1 − 𝐴2
Subjected to: 𝐴1 ≤ 𝑥 2
𝐴2 ≤ 𝛱𝑥 2
Page | 3
STRUCTURAL OPTIMIZATION ASSIGNMENT
0<𝑥<4
𝑦1 + 𝑦2 = 1
6) A 100 × 100-m lot is available to construct a multistory office building. At least 20,000 m2 of total
floor space is needed. According to a zoning ordinance, the maximum height of the building can
be only 21 m, and the parking area outside the building must be at least 25% of the total floor area.
It has been decided to fix the height of each story at 3.5 m. The cost of the building in millions of
dollars is estimated at 0.6h + 0.001A, where A is the cross-sectional area of the building per floor
and h is the height of the building. Formulate the minimum-cost design problem.
Solution
Given: The lot size, building floor space and parking area requirements, and the data given in
the problem statement.
Required: It is desired to find the building cross-sectional area and its height to meet all the
requirements and minimize cost of the building.
Procedure: We follow the five step process to formulate the problem as an optimization
problem.
Step 1: Project/Problem Statement
Shown above
Step 2: Data and Information Collection
Area of the lot =100×100 = 10,000 m2
Area available for parking = (10,000 – A), m2
ℎ
Total floor area = (number of floors) ×A = 3.5 𝐴, 𝑚2
Page | 4
STRUCTURAL OPTIMIZATION ASSIGNMENT
ℎ
𝐴 ≥ 20,000, 𝑚2 ------------------------------------------------------------------ (b)
3.5
Parking Constraint:
ℎ
(10,000 − 𝐴) ≥ 0.25 𝐴, 𝑚2 --------------------------------------------------- (c)
3.5
𝐴 ≥ 0, 𝑚2 -------------------------------------------------------------------- (f)
Final Formulation:
Find h and A to minimize the cost function of Eq. (a) subject to the constraints in Eqs. (b) to (g).
Note that for a meaningful design, h must be a multiple of 3.5.
Therefore the final Formulation is:
Cost = (0.6h + 0.001A),
ℎ
Subjected to: 𝐴 ≥ 20,000
3.5
ℎ
(10,000 − 𝐴) ≥ 0.25 𝐴
3.5
ℎ ≥ 3.5
ℎ ≤ 21
𝐴 ≥ 0
𝐴 ≤ 10,000
Page | 5
STRUCTURAL OPTIMIZATION ASSIGNMENT
1. An overhead crane (as shown on Figure -1) is used to transport concrete from a mix plant
to a casting site. Lifting yokes facilitate placing concrete and moving cured member. What
is the maximum concrete (W1 + W2 + W3) that can be carried in the three buckets at the
positions shown? Cable 1 and 2 each have 8KN capacity, cable 3 and 4 each have a 16KN
capacity, and each crane rail (5 and 6) can support 20KN.the crane and yokes are of
negligible weight. All dimensions of the beam length is given in centimeter.
Solution
Let 𝐹1, 𝐹2, … , 𝐹6 represents the force in cables 1,2,3,and 4 supports 5 and 6, respectively. For
equilibrium, we set the sums of the moments about each of the six support position equal to zero.
This yields
𝐹1 = 1⁄3 𝑊2 𝐹4 = 2⁄15 𝑊2 + 3⁄5 𝑊3
Page | 6
STRUCTURAL OPTIMIZATION ASSIGNMENT
8⁄ 𝑊 + 2⁄ 𝑊 ≤ 16 3⁄ 𝑊 + 7⁄ 𝑊 + 1⁄ 𝑊 ≤ 20
15 2 5 3 4 1 12 2 4 3
2⁄ 𝑊 + 3⁄ 𝑊 ≤ 16 1⁄ 𝑊 + 5⁄ 𝑊 + 3⁄ 𝑊 ≤ 20
15 2 5 3 4 1 12 2 4 3
𝑊1 , 𝑊2 , 𝑊3 ≥ 0
The second constraint clearly implies the first. Therefore the second can be omitted and the number
of constraints reduced to five.
Maximization of functions can be treated routinely. For example, if the objective is to maximize a
function, we simply minimize its negative. Therefore
To solve this linear programming model using the Simplex method the following steps are necessary:
1. Express the linear programming problem in its standard form (Introducing slack variables).
Basic variables: S1 = 8 S2 = 16 S3 = 16 S4 = 20 S5 = 20
Non basic variables: W1 = W2 = W3 = 0
Function: f = 0 from the last row of the tableau after substituting for𝑊1 , 𝑊2 𝑎𝑛𝑑 𝑊3 .
Basic
variable W1 W2 W3 s1 s2 s3 s4 s5 RHS
s1 0 1/3 0 1 0 0 0 0 8
s2 0 8/15 2/5 0 1 0 0 0 16
s3 0 2/15 3/5 0 0 1 0 0 16
s4 3/4 7/12 1/4 0 0 0 1 0 20
s5 1/4 5/12 3/4 0 0 0 0 1 20
f -1 -1 -1 0 0 0 0 0 f
Page | 7
STRUCTURAL OPTIMIZATION ASSIGNMENT
Entering
2nd Tabluea variabel
Basic
variable W1 W2 W3 s1 s2 s3 s4 s5 RHS ratio
W2 0.00 1.00 0.00 3.00 0.00 0.00 0.00 0.00 24.00
min(Leaving
s2 0.00 0.00 0.40 -1.60 1.00 0.00 0.00 0.00 3.20 8 variabel)
s3 0.00 0.00 0.60 -0.40 0.00 1.00 0.00 0.00 12.80 21.33
s4 0.75 0.00 0.25 -1.75 0.00 0.00 1.00 0.00 6.00 8.00 24
s5 0.25 0.00 0.75 -1.25 0.00 0.00 0.00 1.00 10.00 40.00 13.33
f -1.00 0.00 -1.00 3.00 0.00 0.00 0.00 0.00 f +24
3rd Entering
Tabluea variabel
Basic
variable W1 W2 W3 s1 s2 s3 s4 s5 RHS ratio
W2 0.00 1.00 0.00 3.00 0.00 0.00 0.00 0.00 24.00
W3 0.00 0.00 1.00 -4.00 2.50 0.00 0.00 0.00 8.00
s3 0.00 0.00 0.00 2.00 -1.50 1.00 0.00 0.00 8.00
min(Leaving
s4 0.75 0.00 0.00 -0.75 -0.63 0.00 1.00 0.00 4.00 5.33 variable)
s5 0.25 0.00 0.00 1.75 -1.88 0.00 0.00 1.00 4.00 16
f -1.00 0.00 0.00 -1.00 2.50 0.00 0.00 0.00 f +32.0
Page | 8
STRUCTURAL OPTIMIZATION ASSIGNMENT
Reduced cost coefficients in the non-basic columns are nonnegative; the tableau
4th Tabluea gives optimum solution.
Basic
variable W1 W2 W3 s1 s2 s3 s4 s5 RHS
W2 0.00 1.00 0.00 3.00 0.00 0.00 0.00 0.00 24.00
W3 0.00 0.00 1.00 -4.00 2.50 0.00 0.00 0.00 8.00
s3 0.00 0.00 0.00 2.00 -1.50 1.00 0.00 0.00 8.00
W1 1.00 0.00 0.00 -1.00 -0.83 0.00 1.33 0.00 5.33
s5 0.00 0.00 0.00 2.00 -1.67 0.00 -0.33 1.00 2.67
f 0.00 0.00 0.00 -2.00 1.67 0.00 1.33 0.00 f +37.33
The solution obtained after four iteration
2. A three member rigid frame is shown in the Figure -2. Determine plastic moment capacity of the
beam and column (Mpa and Mpc) for minimum total frame weight. Use limit analysis and assume a linear
relationship between the plastic moment and weight per unit length of the member.
Solution
In the plastic analysis of this frame, the relationship between external force and internal moment can
be obtained by considering the balance of external work W and internal energy U in each of the
Page | 9
STRUCTURAL OPTIMIZATION ASSIGNMENT
deformed shape corresponding to the different collapse modes. The three possible collapse mechanism
are illustrated in the following figure where the plastic hinge are indicated at location of maximum
positive/ negative moments and the corresponding limiting value of W and V are given.
The figure 2.1 shows three collapse mechanism and the six corresponding linear inequalities
expressed in terms of the plastic moments in the column and beam. The mechanism constraints are
obtained using principle of virtual work. The virtual work W and the maximum internal energy U are
given in fig. 2.1.
Page | 10
STRUCTURAL OPTIMIZATION ASSIGNMENT
Fig 2.1 Collapse modes for the three member rigid frame; (o) denotes a plastic-hinge location.
4𝑀𝑝𝑏 ≥ 30
Subjected to: } Beam mechanism
2𝑀𝑝𝑐 + 2𝑀𝑝𝑏 ≥ 30
2𝑀𝑝𝑐 + 2𝑀𝑝𝑏 ≥ 20
} Sway mechanism------------------- (b)
4𝑀𝑝𝑐 ≥ 20
2𝑀𝑝𝑐 + 4𝑀𝑝𝑏 ≥ 50
} Combined mechanism
4𝑀𝑝𝑐 + 2𝑀𝑝𝑏 ≥ 50
Page | 11
STRUCTURAL OPTIMIZATION ASSIGNMENT
To illustrate the duality theorem, we solve the dual of the above problem using simplex technique.
Since there are six primal constraints, there are six dual variables for the problem. Let y1, y2, y3, y4,
y5 and y6 be the dual variables associated with the six constraints.
The dual for the standard primal is defined as follows: Find the dual variables y1,y2,……ym to
minimize a dual objective functionn:
𝑚
𝑓𝑑 = 𝑒1 𝑦1 +. . . +𝑒𝑚 𝑦𝑚= ∑ 𝑒𝑖 𝑦𝑚 = 𝒆𝑻 𝒚
𝑖=1
… … … .. (𝑨𝑻 𝒚) ≥ 𝒅
𝑎1𝑛 𝑦1 +. . . +𝑎𝑚𝑛 𝑦𝑚 ≥ 𝑑𝑛
𝑦𝑖 ≥ 0: 𝑖 = 1 𝑡𝑜 𝑚
1. Express the linear programming problem in its standard form (Introducing slack variables).
Consider expression (a) and (c) as the primal problem, the dual problem can be formulated as
Max: 𝑧 = 30𝑦1 + 30𝑦2 + 20𝑦3 + 20𝑦4 + 50𝑦5 + 50𝑦6 + 0𝑠1 + 0𝑠2 ------------------------------- (a)
𝑦𝑖 ≥ 0 𝑖 = 1,2, … … ,6
Slack variables 𝑠1 and 𝑠2 are added to inequation to obtain a canonical form. The resulting simplex
tableaus are shown in tabel below.
Basic
variable y1 y2 y3 y4 y5 y6 s1 s2 z RHS
s1 0 2 2 4 2 4 1 0 0 40
s2 4 2 2 0 4 2 0 1 0 30
z -30 -30 -20 -20 -50 -50 0 0 1 0
Page | 12
STRUCTURAL OPTIMIZATION ASSIGNMENT
Entering
1st Tabluea variabel
Basic
variable y1 y2 y3 y4 y5 y6 s1 s2 z RHS ratio
s1 0 2 2 4 2 4 1 0 0 40 20 10
s2 4 2 2 0 4 2 0 1 0 30 7.5 15 min(Leaving variable)
z -30 -30 -20 -20 -50 -50 0 0 1 0
2nd Entering
Tabluea variabel
Basic
variable y1 y2 y3 y4 y5 y6 s1 s2 z RHS ratio
min(Leaving
s1 -2 1 1 4 0 3 1 -0.5 0 25 8.33 variable)
y5 1 0.5 0.5 0 1 0.5 0 0.25 0 7.5 15
z 20 -5 5 -20 0 -25 0 12.5 1 375
3rd Tabluea
Basic
variable y1 y2 y3 y4 y5 y6 s1 s2 z RHS
y6 -0.667 0.333 0.333 1.333 0 1 0.333 -0.167 0 8.333
y5 1.333 0.333 0.333 -0.667 1 0 -0.167 0.333 0 3.333
z 3.333 3.333 13.33 13.333 0 0 8.333 8.333 1 583.33
𝑦1 = 𝑦2 = 𝑦3 = 𝑦4 = 0
The solution to the primal problem can be obtained directly from the cost coefficients in the last
tableau, i.e., corresponding to the slack variables s7 and s8, Mpc=8.333 and Mpb=8.333. Alternatively,
corresponding to the basic dual variables 𝑦5 𝑎𝑛𝑑 𝑦6 , we must have equalities in the primal constraints,
that is, 2𝑀𝑝𝑐 + 4𝑀𝑝𝑏 = 50 and 4𝑀𝑝𝑐 + 2𝑀𝑝𝑏 = 50. Solving this set of equation simultaneously
yields the optimal solution.
Page | 13
STRUCTURAL OPTIMIZATION ASSIGNMENT
1. Find the absolute maximum and minimum values of the function f (x, y) y2 x2 over
the region given by x2 y2 4 (Hint: Use Lagrange multipliers).
Solution
f𝑥 = 𝜆g 𝑥 f𝑦 = 𝜆g 𝑦 g(x, y)= 4
Which becomes
2𝑦 =2𝑦𝜆 … … … … … … … … … … … . . … … … … … (𝑏)
x 2 y 2 4… … … … … … … … … … … … . … … … … (𝑐)
𝑓(0,2) = 4 𝑓(2,0) = −4
Therefore the maximum value of f on the circle x 2 y 2 4 is 𝑓(0, ±2) = 4 and the minimum
value is 𝑓(±2,0) = −4
2. Find the maximum area of a rectangle with sides measuring x and y if the perimeter is 14
units. Is there a minimum value of the area?
Solution
Page | 14
STRUCTURAL OPTIMIZATION ASSIGNMENT
Constraints g(x, y) = 2𝑥 + 2𝑦 14
f𝑥 = 𝜆g 𝑥 f𝑦 = 𝜆g 𝑦 g(x, y)= 14
Which becomes
𝑦 =2𝜆 … … … … … … … … … … … … … … … … (𝑎)
𝑥 =2𝜆 … … … … … … … … … … … . . … … … … … (𝑏)
2𝑥 + 2𝑦 14… … … … … … … … … … … … . … … … … (𝑐)
14
From (𝑎) we have 𝑦 = 2 𝜆 and from(𝑏), 𝑥 =2𝜆 so then (𝑐) gives 4𝜆+4𝜆 14. , 𝜆 = ,𝑥 =
8
7 7 7
𝑦 = 2 Therefore f has possible extreme values at the point(2 , 2)
7 7 49
𝐴 = 𝑓( , ) =
2 2 4
If one of the sides is zero, the area is zero, and areas generally can’t get smaller than that. If we
allow negative sides and signed area there’s is no minimum.
3. A cylindrical can is to be designed with a capacity of holding 475cm3 of fluid. What are the
dimensions of the can that would require the least amount of material?
Solution
Let height = h and radius of the base = r. Then, the volume of the tank (𝜋𝑟 2 ℎ) is required to be V.
Therefore, volume 𝑉 = 𝜋𝑟2 ℎ = 475 and surface area of the can A = 2𝜋𝑟2 + 2𝜋𝑟ℎ
Therefore, the cost function for the problem is given as
𝑓 = 𝑐(2𝜋𝑟 2 + 2𝜋𝑟ℎ)
Constraints g(h, r)= 𝜋𝑟2 ℎ = 475 ∇𝑓 = 𝜆∇g and g(h, r)= 475
2𝜋𝑟 =𝜆𝜋𝑟^2 … … … … … … … … … … … … … … (𝑎)
4𝜋𝑟 + 2𝜋ℎ =𝜆2𝜋𝑟ℎ… … … … … … … … … . … … . (𝑏)
𝜋𝑟^2 ℎ = 475 … … … … … … … … … … … … … … . (𝑐)
16𝜋
From (𝑎) 𝑟 = 2/𝜆 and substituting this in to (𝑏) ℎ = 4/𝜆. And from (𝑐)𝜆3 = 475 , 𝜆 = 0.4729
2 4
𝑟 = = 4.229𝑐𝑚 ℎ = = 8.458𝑐𝑚
𝜆 𝜆
Page | 15
STRUCTURAL OPTIMIZATION ASSIGNMENT
The can with volume 475 cm3 that has the smallest surface area is 4.23 cm in radius and 8.46 cm
in height. The surface area of the can is 337 cm2. (In general, a cylinder where diameter = height
has the smallest surface area given constant volume.)
4. Maximize f(x) 12x 3x4 2x6 using Newton-Raphson technique with error margin of
0.0001.
Solution
A numerical method for solving of optimization of a nonlinear function and using Newton-Raphson
technique to maximize the function.
𝑓 ′ ( 𝑥𝑖 )
𝑥𝑖+1 = 𝑥𝑖 −
𝑓′′ (𝑥𝑖 )
Page | 16
STRUCTURAL OPTIMIZATION ASSIGNMENT
5. Find the location of the maximum of f(x) x5 x on [0, 20] to within error margin of
1% using Golden section search technique.
Solution
Iteration 1:
Given the values for the boundaries of 𝑥𝑙 = 0 and 𝑥𝑢 = 20 and 1% error.
Step 1: Determine two intermediate points x1 and x2
𝑥1 = 𝑥𝑙 + 0.618(𝑥𝑢 − 𝑥𝑙 ) = 0 + 0.618 ∗ (20 − 0) = 12.360
𝑥2 = 𝑥𝑢 − 0.618(𝑥𝑢 − 𝑥𝑙 ) = 20 − 0.618 ∗ (20 − 0) = 7.640
Step 2: Evaluate f(x1) and f(x2).
Page | 17
STRUCTURAL OPTIMIZATION ASSIGNMENT
6. A minimum weight tubular column design problem is formulated in chapter one using
mean radius R and thickness t as design variables. Solve the KKT conditions for the
problem imposing an additional constraint R/t ≤ 50 for this data: P = 50kN, l = 5.0m,
E = 210GPa, σa = 250MPa and ρ = 7850kg/m3. Interpret the necessary conditions at
the solution point graphically.
Solution
In chapter one using mean radius R and thickness t as design variables a minimum weight tubular
column design problem is formulated. There is an additional constraint R/t ≤ 50 is given. Therefore
from chapter one we have the following formulations with constraints.
𝑅/𝑡 ≤ 50
tmin ≤ t ≤ tmax
𝑀𝑎𝑠𝑠 = 0.2466𝑅𝑡
Subjected to: 7.9577/𝑅𝑡 ≤ 0.25
767893.7 ≤ 𝑅 3 𝑡
𝑅/𝑡 ≤ 50
Rmin ≤ R ≤ Rmax
tmin ≤ t ≤ tmax
𝑔1 = 1/𝑅𝑡 − 0.0314 ≤ 0
𝑔2 = 767893.7 – 𝑅 3 𝑡 ≤ 0
𝑔3 = 𝑅/𝑡 − 50 ≤ 0
1 𝑅
𝐿 = 0.2466𝑅𝑡 +𝜆1 (𝑅𝑡 – 0.0314 + 𝑠12 ) +𝜆2(767893.7– 𝑅 3 𝑡 + 𝑠22 ) +𝜆3 ( 𝑡 – 50 + 𝑠32 ) . . (a)
Page | 18
STRUCTURAL OPTIMIZATION ASSIGNMENT
Case 1:-
𝜆1 = 0, 𝑠2 = 0, 𝑠3 = 0, g1 is in active; g2 and g3 are active.
From the above equations 𝑅 = 50𝑡 and
767893.7 1/4
767893.7 = 503 𝑡 4 ; 𝑡 = ( ) = 𝟏. 𝟓𝟕𝒎𝒎
503
Page | 19
STRUCTURAL OPTIMIZATION ASSIGNMENT
𝑑 2 𝐿 2𝜆1
= + 2𝜆3𝑅/𝑡 3 = 2𝜆3𝑅/𝑡 3 = 12.356
𝑑𝑡 2 𝑅𝑡 3
𝑑2𝐿
= 0.2466 − 2𝜆1𝑅 2 − 3𝜆2𝑅 2 − 𝜆3/𝑡 2 = −5.555
𝑑𝑅𝑑𝑡
The sufficient condition for the minimum is not satisfied
Case 2
𝑆1 = 0; 𝜆2 = 0; 𝑠3 = 0
From 𝑅 = 50𝑡; 1/50𝑡 2 = 0.0314
𝑡 = 0.8𝑚𝑚
𝑅 = 39.9𝑚𝑚
From eqn (e); 767893.7 – 39.93 ∗ 0.8 + 𝑠22 = 0; 𝑠22 = −717255 < 0
The KKT condition is not satisfied. So this can’t be the candidate for minimum point.
Case 3; 𝑠1 = 0; 𝑠2 = 0; 𝜆3 = 0;
From eqn (d); 1/𝑅𝑡– 0.0314 and from eqn (e); 767893.7 – 𝑅 3 𝑡
𝑅 = 155.32𝑚𝑚 𝑎𝑛𝑑 𝑡 = 0.2𝑚𝑚.
𝑅/𝑡 – 50 + 𝑠32 = 0; 𝑠32 = 50 − 155.32/0.2 = −707.88 < 0
The KKT condition is not satisfied. So this can’t be the candidate for minimum point.
Page | 20
STRUCTURAL OPTIMIZATION ASSIGNMENT
7. Using secant and Newton-raphson method find the root of the following equations. Start from a
reasonable initial point.
a) 𝑓(𝑥) = 𝑎𝑥 6 − 𝑏𝑥 5 − 𝑐𝑥 3 + 𝑑𝑥 + 𝑒
b) 𝑓(𝑥) = 𝑔 sin(𝑥) + ℎ𝑥 4
a b c d e g h
0.3 3.9 2.2 2.1 40 0.9 0.16
Therefore,
Solution
A) Secant Method
a) 𝑓(𝑥) = 0.3𝑥 6 − 3.9𝑥 5 − 2.2𝑥 3 + 2.1𝑥 + 40
Calculate the next estimate of the root from two initial guesses
𝑓(𝑥𝑖 )
𝑥𝑖+1 = 𝑥𝑖 − (𝑥 − 𝑥𝑖−1 )
𝑓(𝑥𝑖 ) − 𝑓(𝑥𝑖−1 ) 𝑖
𝑓(𝑥0 )
𝑥1 = 𝑥0 − (𝑥 − 𝑥−1 )
𝑓(𝑥0 ) − 𝑓(𝑥−1 ) 0
𝑓(𝑥𝑖 ) = 𝑓(3) = 0.3 ∗ 36 − 3.9 ∗ 35 − 2.2 ∗ 33 + 2.1 ∗ 3 + 40 = −702.5
𝑓(𝑥𝑖−1 ) = 𝑓(2) = 0.3 ∗ 36 − 3.9 ∗ 35 − 2.2 ∗ 33 + 2.1 ∗ 3 + 40 = −70.20
−702.5
𝑥1 = 3 − (3 − 2 ) = 1.8889
−702.5 + 70.20
|𝑒 | = |1.8889 – 3| = 1.111 > 0.0001 Not ok
Page | 21
STRUCTURAL OPTIMIZATION ASSIGNMENT
Let us assume the initial guesses of the root of f (x) = 0 as 𝑥−1 = 8 𝑎𝑛𝑑 𝑥0 = 3
𝑓(𝑥0 )
𝑥1 = 𝑥0 − (𝑥 − 𝑥−1 )
𝑓(𝑥0 ) − 𝑓(𝑥−1 ) 0
𝑓(𝑥𝑖 ) = 0.9 sin(𝑥) + 0.16𝑥 4
𝑓(𝑥𝑖 ) = 𝑓(3) = 0.9 sin(3) + 0.64(3)4 = 13.08701
𝑓(𝑥−1 ) = 𝑓(8) = 0.9 sin(8) + 0.64(8)4 = 656.2504
13.08701
𝑥1 = 3 − (3 − 8 ) = 2.8982
13.08701 − 656.2504
|𝑒 | = |2.8982 – 3| = 0.10174 > 0.0001 Not ok
Page | 22
STRUCTURAL OPTIMIZATION ASSIGNMENT
B) Newton-Raphson Method
a) 𝑓(𝑥) = 0.3𝑥 6 − 3.9𝑥 5 − 2.2𝑥 3 + 2.1𝑥 + 40
𝑓 ′ (𝑥) = 1.8𝑥 5 − 19.5𝑥 4 − 6.6𝑥 2 + 2.1
𝑓(1) = 0.3(1)6 − 3.9(1)5 − 2.2(13)3 + 2.1 ∗ 1 + 40 = 36.3(+𝑣𝑒)
𝑓(2) = 0.3(2)6 − 3.9(2)5 − 2.2(2)3 + 2.1 ∗ 2 + 40 = −70.2(−𝑣𝑒)
So the root lies in between (1,2)
Iteration 1:- Assume the pre-specified relative error tolerance (e =0.0001) and the initial guesses of
the root of f (x) = 0 as x0 =1
Page | 23
STRUCTURAL OPTIMIZATION ASSIGNMENT
8. Using the steepest–descent method find the roots of the following function. Start from a
reasonable initial point
a) 𝑓(𝑥1 , 𝑥2 ) = 𝑎𝑥1 − 𝑏𝑥2 + 𝑐𝑥1 2 + 𝑑𝑥1 𝑥2 + 𝑒𝑥2 2
b) 𝑓(𝑥1 , 𝑥2 ) = 𝑔𝑥1 2 + ℎ𝑥2 2 − 𝑘
a b c d e g h k
1 1.1 1.2 1.1 0.9 0.9 0.9 100
Therefore,
Page | 24
STRUCTURAL OPTIMIZATION ASSIGNMENT
Solution
𝜕𝑓
𝜕𝑥1 1 + 2.4𝑥1 + 1.1𝑥2
𝑮𝒓𝒂𝒅𝒊𝒆𝒏𝒕 of 𝑓 is ∇𝑓(𝑥1 , 𝑥2 ) = ∇𝑓 = =[ ]
𝜕𝑓 −1.1 + 1.1𝑥1 + 1.8𝑥2
[𝜕𝑥2 ]
3.4
∇𝑓(1,0) =[ ]
0
𝜕2𝑓 𝜕2𝑓
𝜕𝑥1 2 𝜕𝑥1 𝜕𝑥2 2.4 1.1
𝑯𝒆𝒔𝒔𝒊𝒂𝒏 𝑴𝒂𝒕𝒓𝒊𝒙 𝑖𝑠 𝐻 = =[ ]
𝜕2𝑓 𝜕2𝑓 1.1 1.8
[𝜕𝑥2 𝜕𝑥1 𝜕𝑥2 2 ]
1. Direction of search
3.4 −3.4
𝑑𝑘 = −∇𝑓 = − [ ]=[ ]
0 0
2. Determine the optimal step length α in the direction of 𝒅𝒊
−3.4
𝒅𝟎 𝑻 𝒅𝟎 [−3.4 0] [ ] 11.560
α= = 0 = = 0.4167
𝒅𝟎 𝑻 𝑯𝒅𝟎 2.4 1.1 −3.4
[−3.4 0] [ ][ ] 27.744
1.1 1.8 0
3. Check the optimum
Iteration 2
1. Direction of search
0 0
𝑑2 = −∇𝑓 = − [ ]=[ ]
−1.5584 1.5584
2. Step length α in the direction of 𝒅𝟏
Page | 25
STRUCTURAL OPTIMIZATION ASSIGNMENT
0
𝒅𝟏 𝑻 𝒅𝟏 [0 1.5584] [ ] 2.4286
α= = 1.5584 = = 0.5556
𝒅𝟏 𝑻 𝑯𝒅𝟏 2.4 1.1 0
[0 1.5584] [ ][ ] 4.3715
1.1 1.8 1.5584
Iteration 3
1. Direction of search
0.9524 −0.9524
𝑑2 = −∇𝑓 = − [ ]=[ ]
0 0
2. Step length α in the direction of 𝒅𝟏
−0.9524
𝒅𝟐 𝑻 𝒅𝟐 [−0.9524 0] [ ] 0.8538
α= = 0 = = 0.4167
𝒅𝟐 𝑻 𝑯𝒅𝟐 1.8 0 −0.9524 2.0491
[−0.9524 0] [ ][ ]
0 1.8 0
Iteration 4
1. Direction of search
Page | 26
STRUCTURAL OPTIMIZATION ASSIGNMENT
0.0283 −0.0283
𝑑3 = −∇𝑓 = − [ ]=[ ]
−0.4234 0.4234
2. Step length α in the direction of 𝒅𝟏
−0.0283
𝒅𝟑 𝑻 𝒅𝟑 [−0.0283 0.4234] [ ] 0.1801
α= = 0.4234 = = 0.6038
𝒅𝟑 𝑻 𝑯𝒅𝟑 1.8 0 −0.0283
[−.0283 0.4234] [ ][ ] 0.2984
0 1.8 0.4234
𝜕𝑓
𝜕𝑥1 1.8𝑥1 1.8
𝑮𝒓𝒂𝒅𝒊𝒆𝒏𝒕 of 𝑓 is ∇𝑓(𝑥1 , 𝑥2 ) = =[ ]=[ ]
𝜕𝑓 1.8𝑥2 0
[𝜕𝑥2 ]
𝜕2𝑓 𝜕2𝑓
𝜕𝑥1 2 𝜕𝑥1 𝜕𝑥2 1.8 0
𝑯𝒆𝒔𝒔𝒊𝒂𝒏 𝑴𝒂𝒕𝒓𝒊𝒙 𝑖𝑠 𝐻 = =[ ]
𝜕2𝑓 𝜕2𝑓 0 1.8
[𝜕𝑥2 𝜕𝑥1 𝜕𝑥2 2 ]
Direction of search
1.8 −1.8
𝑑𝑘 = −∇𝑓 = − [ ]=[ ]
0 0
−1.8
𝒅𝟎 𝑻 𝒅𝟎 [−1.8 0] [ ] 3.24
α= = 0 = = 0.555
𝒅𝟎 𝑻 𝑯𝒅𝟎 1.8 0 −1.8 5.832
[−1.8 0] [ ][ ]
0 1.8 0
Page | 27
STRUCTURAL OPTIMIZATION ASSIGNMENT
𝑥𝑖+1 = 𝑥𝑖 + 𝛼𝑑𝑘
1 −1.8 0
𝑥1 = 𝑥0 + α𝑑0 = [ ] + 0.5556 [ ]=[ ]
0 0 0
Page | 28
STRUCTURAL OPTIMIZATION ASSIGNMENT
Dynamic Programming
1. Bahir Dar University want to allocate the available water resource to the three selected campuses. If the
total available water resource is 60000m3 (6 unites) which is to be allocated for the three campuses the
return in supplying the respective unit of water is given below in the table-1. How much unit of water
must be allocated to each users to get the maximum benefit?
State
S3
X3 R3(X3) F3(X3) X3*
0 0 0 0 0
0 0
1 7 1
1 7
0 0
2 1 7 12 2
2 12
0 0
1 7
3 15 3
2 12
3 15
0 0
1 7
4 2 12 16 4
3 15
4 16
0 0
5 1 7 16 4
2 12
Page | 29
STRUCTURAL OPTIMIZATION ASSIGNMENT
3 15
4 16
5 15
0 0
1 7
2 12
6 3 15 16 4
4 16
5 15
6 12
Stage-2: Considering the last two stages together, the sub-optimization function is:
f 2* ( S 2 ) Max R2 x2 f 3* ( S 2 x2 )
0 x2 S 2
0 S 2 Q
State
X2 R2(x2) (S2-X2) f3(S2-X2) f2(X2)= R2(x2)+f3(S2-X2) f2*(S2) X2*
S2
0 0 0 0 0 0 0 0
0 0 1 7 7
1 7 0
1 5 0 0 5
0 0 2 12 12
2 1 5 1 7 12 12 0,1
2 6 0 0 6
0 0 3 15 15
1 5 2 12 17
3 17 1
2 6 1 7 13
3 3 0 0 3
0 0 4 16 16
1 5 3 15 20
4 2 6 2 12 18 20 1
3 3 1 7 10
4 -4 0 0 -4
0 0 5 15 15
1 5 4 16 21
2 6 3 15 21
5 21 1,2
3 3 2 12 15
4 -4 1 7 3
5 -15 0 0 -15
0 0 6 12 12
1 5 5 15 20
2 6 4 16 22
6 3 3 3 15 18 22 2
4 -4 2 12 8
5 -15 1 7 -8
6 -30 0 0 -30
Page | 30
STRUCTURAL OPTIMIZATION ASSIGNMENT
Stage-3: Considering all the three stages together, f 1* (Q) Max R1 x1 f 2* ( S1 x1 )
0 x1 Q
S1 Q
State
X1 R1(X1) Q-X1 f2(Q-X1) f1(S1) = R1(X1) + f2 (Q-X1) f1*(S1) X1*
S1– Q
0 0 6 22 22
1 5 5 21 26
2 8 4 20 28
6 3 9 3 17 26 28 2
4 8 2 12 20
5 5 1 7 12
6 0 0 0 0
Thus, S2 = S1 – X1 = 4
Now, S3 = S2 – X2= 3
Check
X1 = 2 (8)
X2 = 1 (5)
X3 = 3 (15)
Therefore, Max ∑3𝑖=1 𝑅𝑖 = 8+5+15 = 28,
Page | 31