Optimization design

The material selection procedure for design described above is a method of rating materials for a particular design application that is based on a single design variable such as varying the depth h or diameter d in Table 6.5. This allows use of the design variable in a linear form. 

In this section, design optimization theory is presented, where multiple design variables are allowed and where nonlinear constraints can be imposed. The following is the standard format for posing a design optimization problem. 

Assume that there are some inequality constraints given by 

Assume that there are side constraints given by 

---

Computer simulation programs for process design optimization have been developed for the PUREX process utilizing these relationships (22). A subroutine has also been developed which describes the behavior of fission products (23). 

Dong, Z. 1997 Tolerance Synthesis by Manufacturing Cost Modeling and Design Optimization. In Zhang, H. (ed.). Advanced Tolerancing Techniques. NY Wiley Interscience. 

In addition to evolutionary search strategies such as GAs, there are a number of other search techniques that are employed for design optimization. 

The minimum number of the tube rows recommended to establish a proper air flow pattern is 4, although 3 rows can be used. The typical unit has 4-6 rows of tubes, but more can be used. Although more heat can be transferred by increasing the number of tubes, the required fan horsepower will be increased however, this balance must be optimized for an effective economical design. Tubes are laid out on transverse or longitudinal patterns however, the transverse is usually used due to the improved performance related to pressure drop and heat transfer. The tube pitch is quite important for best air-side performance. A typical representative tube arrangement for design optimization is for hare-tube O.D., tinned-tube O.D., and tube pitch ... 

Kalamkarov, A. L., et al., Analysis, Design, Optimization of Composite Structures, Wiley, 1997. 

A related problem in eomposites is the need to design optimal fiber orientations for a eomposite part given the set of stress vectors and levels to whieh the part will be subjected. These design eonsiderations would be useful in designing airframe eomponents sueh as parts for the tail, wing, or fuselage. A similar problem is assessment of the peiformanee penalties that might result from imperfections in manufacture. 

---