Intro. To Aerospace Eng. Design: - Design Modeling Part II
Intro. To Aerospace Eng. Design: - Design Modeling Part II
To Aerospace
Eng. Design
March 8, 2019 | 2
Design optimization
March 8, 2019 | 3
Design optimization
March 8, 2019 | 4
Design optimization
March 8, 2019 | 5
Design optimization
March 8, 2019 | 6
Design optimization
March 8, 2019 | 7
Design examples
m m
8
Comparing FRPs with metals
9
Comparing FRPs with metals
10
Comparing FRPs with metals
11
Design optimization
March 8, 2019 | 12
Local minima and maxima
March 8, 2019 | 13
Example
March 8, 2019 | 14
Example (Cont.)
March 8, 2019 | 15
Example (Cont.)
March 8, 2019 | 16
Example (Cont.)
March 8, 2019 | 17
Question
March 8, 2019 | 18
Question
March 8, 2019 | 19
General optimization problem
March 8, 2019 | 20
General optimization problem
• Note that we must check the extremes x=a and x=b for local
minima. Equality h(x) and inequality g(x) constraints may be
systems of equations.
March 8, 2019 | 21
Example
March 8, 2019 | 22
Example
• We note that:
• Setting f’(x)=0 we find a local maximum at x = π/2.
• There are no local minima within the bounds of x.
• At the extremes we find f(π/4)=0.707 and f(5π/8)=0.924.
• The global minimum for this problem occurs at x=π/4.
March 8, 2019 | 23
A simple optimization problem.
March 8, 2019 | 24
Global optimization techniques
• Gradient-based methods
• Optimality criterion
• Topology optimization
• Direct search method
• Multi-level optimization
• Hybrid methods
March 8, 2019 | 25
Optimization using a spreadsheet
March 8, 2019 | 26
Optimization using a spreadsheet-procedure
March 8, 2019 | 27
Spreadsheet for optimization of can area for a
given volume
March 8, 2019 | 28
Revised spreadsheet for optimization of a can
March 8, 2019 | 29
March 8, 2019 | 30
Minimization with MATLAB
March 8, 2019 | 31
Minimization with MATLAB
March 8, 2019 | 32
• We can perform a similar optimization by minimizing the cost
of a can, where the ends of the can cost 1.92 times the cost of
the sides.
March 8, 2019 | 33
• We can perform a similar optimization by minimizing the cost
of a can, where the ends of the can cost 1.92 times the cost of
the sides.
March 8, 2019 | 34
• Another more powerful MATLAB function is fmincon. This
function can be used to find the minimum of a constrained
non-linear multivariable function.
March 8, 2019 | 35
• Another more powerful MATLAB function is fmincon. This
function can be used to find the minimum of a constrained
non-linear multivariable function.
March 8, 2019 | 37
• The fmincon function can be called in a number of ways;
however, passing all of the previous parameters is done using
March 8, 2019 | 38
Example
March 8, 2019 | 39
Example
March 8, 2019 | 40
Example (Cont.)
March 8, 2019 | 41
Example (Cont.)
March 8, 2019 | 42
Example (Cont.)
March 8, 2019 | 43
Example (Cont.)
March 8, 2019 | 44
Example (Cont.)
March 8, 2019 | 45
Example (Cont.)
March 8, 2019 | 46
Example (Cont.)
March 8, 2019 | 47
Example (Cont.)
March 8, 2019 | 48
• Local minimum found that satisfies the constraints.
• Optimization completed because the objection function is non-
decr feasible directions, to within the default value of the
function t and constraints are satisfied to within the default
value if the c
<stopping criteria details>
Active inequalities (to within options. TolCon = 1e-6):
Lower upper ineqlin ineqnonlin
1 1
March 8, 2019 | 49