Читать книгу Computer Aided Design and Manufacturing - Zhuming Bi - Страница 20

In this section, the role of CATs in the design aspect is analysed, the nature of engineering designs is discussed, and the potential for using CAT technologies for design problems with different creativity levels is explored. Based on the level of creativity, Designwork (2016) classifies the engineering design problems into types: routine design, redesign, selection design, parametric design, integrated design, and new design (Table 1.2).

Table 1.2 Types of design activities.

Type	Description	Example
Routine design	To perform a design by following existing standards and codes that outline the steps and computations for certain products or systems.	Follow the American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Codes to design a pressure vessel.
Redesign	To revise an existing design when FRs have been changed in a dynamic environment.	Reprogram a robot when the tag points on the motion trajectory are changed.
Selection design	To find a solution by selecting appropriate components from an existing design inventory.	Select standardized fasteners to join two metal plates.
Parametric design	To determine design variables in a given conceptual structure for an optimized performance.	Minimize the materials usage of a cylindrical container subject to the given volume.
Integrated design	To design and assemble components as an integrated product or system to meet strongly coupled FRs.	Design a robotic configuration for a given task in a modular robot system.
New design	To design a product or system from scratch to meet emerging FRs.	Design a patentable product or system.

Corresponding to the needs to involve computers, human–machine interaction, and human designers, a rougher classification based on the level of creativity and innovation may facilitate further discussion. To this end, the designs can be regrouped as:

1 New design for a design from scratch to conceptual design, detailed design, and assembly design to the final product to meet specified design requirements.

2 Incremental design for a design subjected to given product structures, change individual parts, and functions to meet additional requirements of products.

3 Routine design for a design subjected to given functionalities, topological relations, layouts, and parametrize dimensions for the design of product families.

The level of creativity and innovation can be characterized based on the types of solution space and design variables. Goel (1997) corresponded routine design, innovative design, and new creative design to different combinations of solution space and design variables (Table 1.3).

Table 1.3 Characteristics of solution spaces and design variables.

Level of creativity	Routine design	Incremental design	New design
Solution space	Structure	Known	Known	Unknown
Search procedure	Known	Unknown	Unknown
Design variables	Types	Fixed	Fixed	Changed
Ranges	Fixed	Changed	Changed

The characteristics of the solution space and design variables determines whether or not an engineering design belongs to a ‘routine’, ‘incremental’, or ‘new’ design. Both humans and computers can compete to accomplish some design activities; however, computers and human designers are good at different things, and it is desirable to synergize the strengths of designers and computers to achieve the effectiveness of engineering designs. Many researchers have discussed the differences between humans and computers. Table 1.4 summarizes the role differences of human beings and computers in engineering design processes.

Table 1.4 Comparison of human designers and computers.

	Human designers	Computers
Strengths	Identifying design needsBrainstorming to think solutions ‘out of the box’Engineering intuition and a big knowledge baseSelecting design variationsThe flexibility to deal with changesQualitative reasoningPsychologically, human decision is more trusted than artificial intelligencePredict trends, patterns, or anomalies, andLearn from experience	Fast speed, reliable, endurance, and consistentCapable of exploring a large number of optionsCarry out long, complex, and laborious calculationsStore and efficiently search large databases andProvide information on design methodologies, heuristic data, and stored expertise
Weaknesses	Easily tired and boredCannot do micro manageBiased and inconsistentProne to make errorsNot good at quantified reasoningIncapable of utilizing the data presented in an awkward manner	Difficult to synthesize new rulesLimited knowledge baseNo common sense