With respect to this document, manufacturing science is defined as the understanding of the process by which material, labor, energy, and equipment are brought together to produce a product having a greater value than the sum of the individual inputs. The emphasis here is on the approach needed to understand and control inputs required for a specific output. In itself, it recognizes the coupling of the three decision areas of materials, configuration, and processes, in that a choice of a specific aspect or attribute within any of these classes necessitates simultaneous choices in the others, due to their close interdependence. The concurrency of decisions necessitates the development of an integrated science base for the manufacturing of composites so as to enable optimization and development of a system before actual manufacture. This is in sharp contrast to the build-test-fix methodology attributed to metals manufacturing. In short, it requires an understanding of the links between various aspects, as shown in Figure 6.1, as well as their integrated effect on manufacturing in terms of metrics, such as life-cycle concerns, fundamental laws of physics and chemistry, reliability, quality, and cost.

Figure 6.1. Interaction of Manufacturing With Other Key Areas (Univ. of Delaware)

At the outset, it is interesting to compare the two models that could hypothetically be proposed for "intelligent manufacturing of polymer-matrix composites" using a manufacturing science base. On the one hand, we have an approach common in the U.S. (Figure 6.2 -- as developed at the University of Delaware, based on work being conducted at the Center for Composite Materials). On the other hand, we could consider one adapted from the Japanese industry at the University of Tokyo (Figure 6.3). The first emphasizes the use of process models and simulation as a short cut and an efficient mechanism to process understanding driven by the latest advances in computers, AI, modeling, and sensor technology. The second emphasizes the integration of design, materials, fabrication, and inspection through knowledge gained by experimentation and routine, detailed inspection.

Figure 6.2 shows the conceptual ideal manufacturing system (the U.S. model), wherein process understanding drives the idea of integrated manufacturing brought about by appropriate use of process models, sensors, control systems, and statistical and quality control measures.

Figure 6.2. Schematic of an Ideal Manufacturing System

Figure 6.3. An Idealized Manufacturing Science System (University of Tokyo version)

Fully developed process models, coupled with empirical results and analysis (Figure 6.4), lead to the quick development of a system wherein simulation serves as an efficient aid to focusing the design space for directed experimentation (Figure 6.5).

Figure 6.4. Use of Various Aspects to Aid in the Development of a Science Base

Figure 6.5. Simulation of Processes Can Aid in Directing Experimentation

It also ensures that experiments conducted for process and materials development and validation are as close as possible to the overall optimum solution rather than in isolated pockets of the design spaces. The U.S. model is based on the premise that the development of advanced models is key to the development of manufacturing systems and must be done hand-in-hand with advances in sensor and control technology. It may even be described as a system that reaches for perfect understanding before product experiments are allowed to occur -- a factor that may be hindering significantly the U.S. capability to "go to product early," since expensive experimentation is often not deemed suitable (the underlying reasons will be explained in the next chapter), even if there is a high-payoff advantage to taking a high-risk gamble.

The Japanese analog (Figure 6.3) does make extensive use of computers, in terms of both computer-aided design (CAD) and computer- aided inspection (CAI), but places direct and primary emphasis on the development of an experimental database that would allow one to go directly to the first- generation product without using detailed process models or control strategies.

As a specific example, scientists at Nippon Steel preferred to directly wind a large number of thick cylindrical parts to gain an understanding of the process and end product quality rather than use existing codes because of time and economical constraints! It was felt that the time taken to adequately train an engineer in the use of the codes could be better spent on actual experimentation on large-scale specimens, and that the results of direct manufacturing would be more believable to the end user (or customer) than the results of a paper simulation. This is not a rule and should not be taken to mean that this specific company (or any other Japanese company) does not use simulations -- only that direct experimentation on large specimens is often deemed more desirable at the early stages of product development. It should be emphasized here that whereas in the U.S. manufacturing science is taken to be an academic or advanced industrial research task, in Japan it is seamlessly integrated into product development.

For the overall development of an integrated manufacturing/processing schema as shown in Figure 6.4, it is essential that the features of models, sensors, and control be integrated into the previously described coupled decision factors of materials, configuration, and processing (Figure 6.6). The model provides an understanding of the physics and chemistry involved and explains the relationship between dependent and independent variables. The sensors provide information related to the state of the materials and process variables so as to facilitate process information feedback. The control function maintains quality through the use of data provided by sensors as input to the models and manages the product realization process to achieve the design profile. The development of a manufacturing science base necessitates the development of all the features discussed above.

Figure 6.6. The Coupling of Design Factors With Processing Science

In contrast, the Japanese analog (Figure 6.7) emphasizes attention to people factors and experience through rules gained by extensive experimentation and detailed study of materials and processes. The study is done with attention to the most minute detail, and the entire system is often designed keeping in mind the need for human interaction. The attention to detail, as alluded to in the earlier chapters, is a focus within such a model.

Figure 6.7. The Coupling of Design Interactions for Intelligent Manufacturing

In applying a concurrent engineering methodology to a process like resin transfer molding, the U.S. approach to the problem is through simulations for tool design, preform design, infusion, and cure. The part is ideally analyzed on a computer before actual processing. In Japan, on the other hand, there is an intrinsic application of concurrent engineering to RTM, simply by virtue of the way the Japanese approach the problem through the use of innovative machines, structures for which they have extensive data, and rules for tool design based on experience and attention to detail. The Japanese experience with experimentation leads them to apply this knowledge in a concurrent engineering environment without complete dependence on simulations. The end goal is always zero defect manufacturing, rather than mere metrics such as quality control, reliability, economics, and maintainable production schemes, which are already intrinsically built in. In short, the Japanese approach can be characterized as the use of a building-block approach wherein extensive experimentation and detailed trade-off analyses are conducted at each level to achieve an in-depth understanding of materials, processing, and design through attention to people factors and detail at all stages of the development of a manufacturing science base.

Before providing specific examples of the developments in Japan under the topic of this chapter, it is interesting to provide a focus by emphasizing that developments in composites will be successful only if an understanding of available technology is facilitated to ensure rapid commercialization and use of the materials and processes in a reliable and cost-efficient manner (which at first glance would seem to be the route being followed by the Japanese). Table 6.1 pinpoints expected developments in a number of important composite processes as driven through the advancement of materials and process understanding.

Table 6.1
Expected Developments in Composite Processes

Table 6.2 depicts the level of the U.S. capability in the broad area of composites manufacturing determined by the DoD Key Technology Plan of July 1992. It will prove useful to keep this in mind throughout the study, as it actually represents a fairly accurate snapshot of the state of the U.S. composites industry vis-a-vis its global counterparts.

Table 6.2
Comparison of Capabilities

Published: April 1994; WTEC Hyper-Librarian