Directions in Engineering Research: An Assessment of Opportunities and Needs (1987)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Materials Systems Research in the United States: An Overview Executive Summed Leadership in the materials field is essential to compete suc- cessfully in areas of both conventional and high technology. Thus, the economic well-being and security of the United States require that actions be taken to ensure world leadership in the devel- opment of materials systems that can meet the demands of a technology-intensive future. This report describes the results of a study conducted by the Engineering Research Board's Panel on Materials Systems Re- search.* The study encompassed identification of the important classes of materials, emerging areas for future research, and the national climate and infrastructure governing the conduct of ma- terials systems research. In the United States the science of materials has developed into a recognized discipline with its own group of practitioners, facilities, societies, meetings, and publications. Large amounts of money are spent on basic scientific research in the materials field. *We use the term "materials systems to denote advanced materials re- quiring highly tailored design and specialized processing. "Materials systems research refers to the closely interlinked science and engineering research that underlies the development of these complex systems. 239

240 DIRECTIONS IN ENGINrEERING RESEARCH The result is a greatly increased understanding of the relationship between the structure and properties of materials. However, we have failed to place a comparable emphasis on research directed at the application of this knowledge. Because of the economic importance of those applications, that shortcoming has placed the nation at considerable risk. Other countries have spent com- paratively little on expensive basic scientific research, and have concentrated their effort instead on the relatively inexpensive en- gineering research needed to bridge the gap between pure science and applications. Thus, they have been able to capitalize econom- ically on the scientific knowledge for which we have paid a high price. By comparison, our ability to put fundamental knowledge to work in order to devise optimal designs for the processing of materials remains seriously inadequate. We are suffering the con- sequences in terms of declining international competitiveness not only in the mineral, metal, automobile, and textile industries, but now across an even broader spectrum of manufacturing. If this trend is to be reversed, we must make at least as much use of our basic material science as others `do. That will require a focused and interdisciplinary effort on materials systems research comparable to the investments made earlier in materials science. We must attend not only to specific materials with special characteristics important for our economic well-being and for our national defense, but also to activities important to more than one specific material. Thus, among the conclusions reached in the report the following seven are particularly important: 1. We must increase the emphasis on engineering research in: advanced ceramics; semiconducting materials; magnetic materials; polymers; high-performance composites; performance-driven metallic materials; and biomaterials. 2. There must be a recognition, through special research pro- grams, of the importance of the following subjects that are generic to materials research as a whole: · processing; durability and lifetime prediction; tribology (friction, lubrication, and wear);

MATERIALS SYSTEMS 241 computer modeling; and . materials property data base. 3. Research on materials systems—especially research di- rected at the engineering application of basic knowledge and the processing of materials is seriously underfunded relative to the opportunities and potential benefits that current basic knowledge presents. Engineering research on problems in this area would produce knowledge of immediate economic importance, but its po- tential cannot be adequately tapped without substantially greater federal funding for such research. 4. Expenditures on the order of $150 million are required for materials processing research facilities of various kinds. This amount could purchase roughly a dozen each of facilities for re- search on composites processing, semiconductor processing, and molecular beam epitaxy or organometallic chemical vapor deposi- tion. 5. Individual-investigator, single-discipline research must con- tinue to be nourished even as we provide a better climate and facilities for multidisciplinary approaches to materials systems re- search. 6. The stability of federal agency funding policies is vital if we are to obtain the maximum benefit from the modest support for fundamental engineering research and if we are to ensure the needed continuity of valuable research programs. Introduction and Background INTRODUCTION THE NEED As a result of dramatic technological advancements in many fields over the past two decades, the discovery and development of new materials has become a national imperative not only for the United States, but also for our major foreign competitors and adversaries. To give just one example, the United States is now engaged in fierce international competition to maintain our supremacy in electronics, the loss of which to the Japanese or others might be a more serious blow to national productivity

242 DIRECTIONS IN ENGINEERING RESEARCH and security than were the earlier setbacks in the automobile and steel industries. Electronic materials are a key element in this competition. Without the continued development of a leading position in semiconductor materials and fabrication technology, the United States will not even have the opportunity to compete in the information and computer fields (see, for example: General Accounting Once, 1985~. Materials are equally fundamental to many other commercial areas in which our econorn~c fortunes are at stake. They constitute a "gateway" technology; that is, advances in materials open up a cascade of new possibilities in other technologies. A few examples from the past, such as plastics, semiconductors, and the whole range of useful metal alloys, suggest how sweeping the impact of new materials can be. SCOPE OF THE REPORT Materials systems research involves the development of new materials and the ~rnprovement of existing materials for appli- cation in a wide range of advanced engineering systems. The term "materials systems" recognizes the complex interrelation- ships among structure, properties, processing, performance, and reliability when viewed in the context of an engineering appli- cation. It encompasses the activities of materials science and engineering, which are "concerned with the generation and ap- plication of knowledge relating the composition, structure, and processing of materials to their properties and uses" (National Research Council, 1974~. In examining the engineering research opportunities and is- sues associated with materials systems, the pane! addressed topics related to the availability, application, and fundamental under- standing of present and future engineering materials such as met- als, ceramics, polymers, composites, sern~conductors, magnetic and biomaterials. In addition, the pane] addressed questions im- portant to the overall health of the materials field, including policy, research funcling, and human resources issues. BACKGROUND At the outset, it is important to emphasize that the field of materials science and engineering has undergone a fundamental

MATERIALS SYSTEMS 243 and dramatic change over the past two decades. As characterized by William Baker and Morris Cohen in the 1974 COSMAT report of the National Academy of Sciences, the central element of this change has been the emergence of the science- of the solid state as a field of major importance in the latter half of the twentieth century. Also contributing to this change has been the joining of technologies from the ancient fields of metallurgy and ceramics with the more recent fields of synthetic polymers (e.g., rubbers, plastics, and fibers) and modified bioorganic substances. In order to stimulate the exchange of ideas among experts in these fields and to accelerate the development of new materials, the Defense Advanced Research Projects Agency (DARPA) of the Department of Defense (DOD) established several interdisciplinary laboratories in materials sciences and engineering at major American universi- ties over two decades ago. These Materials Research Laboratories, with the continuing support of the National Science Foundation (NSF), have had a significant influence in unifying the subfields of materials science at universities over the past 20 years. Perhaps the most dramatic change has been a stronger sys- tems orientation in the materials research community at large. Instead of the formerly dominant focus on structure-property relationships (i.e., understanding how structure controls specific material properties), a greater integration of design, performance, and processing requirements with materials properties and mi- crostructures has occurred. As a result, a more rational tailoring of final microstructures to specific applications has been achieved. Materials themselves can now be designed (within limits) to fit the desired use. A dramatic example is seen in large-scale integrated circuits, in which, in addition to the circuit elements, materials structures and transport properties are intricately integrated. The ability to tailor applications to the properties of new materials has also given new degrees of freedom to the product designer- "direct replacements is not the only option. In commercializing the new developments in materials, a higher order of functional integration, involving still more dis- ciplinary groupings, has been necessary. New requirements for materials performance and reliability have emerged not only from society's gradual shift from a consumption to a conservation ori- entation, but also from world competitive forces that increasingly place high market premiums on quality, reliability, and low oper- ating and maintenance costs over the product's expected lifetime.

244 DIRECTIONS IN ENGINEERING RESEARCH The additional requirements for greater productivity, man- ufacturability, and profitability have demanded a broader tech- nology base and detailed attention to a number of new technical considerations. For example, productivity gains and cost advan- tages have been achieved by some American manufacturers by tailoring their processes to an optimal equipment design rather than by insisting on customized equipment to accommodate an optimal process design. Advances in quantitative nondestructive evaluation have not only facilitated in-process inspection but have also sometimes led to viable rework and recycling strategies for high-value materials and components. These advances have been further spurred by the advent of computer-aided design and manufacturing, which have presented the design engineer with new flexibilities for patterns of materials use, as well as opportunities for design optimization based on new materials or improvements of existing materials. Artificial intelligence-based systems are emerging to improve the quality of decision modeling and to reduce human error in production, inspection, and testing activities. New and improved materials have a vital effect on the na- tion's welfare. The ability to maintain a competitive edge in most engineering technologies depends on the development of am propriate materials. Ceramics and coatings for use in adiabatic diesel engines, advanced gas turbines, Stirling engines, and fuel cells are pacing the rate of development of those systems; inter- national competition is fierce (see, for example: U.S. Department of Commerce, 1984; National Research Council, 1986a). Com- posite materials are the key to developing lighter, stronger a~r- frames and ground vehicles. Near net-shape processing combined with advanced nondestructive evaluation methods and automated process control are leading to cheaper and more reliable struc- tures. New materials can contribute (along with construction de- sign and codes) to upgrading the nation's infrastructure of roads, bridges, railroads, and buildings through their characteristics of high strength, low weight, and resistance to harsh environments. Emerging requirements in manufacturing, computation, en- ergy development, and defense are all based on the greater com- plexity of technologies; these higher requirements will not be re- aTized without new materials. Therefore, it ~ essential to the nation's general welfare that a strong national materials technol- ogy base be maintainer!. In addition, the bright young talent and .

MATERIALS SYSTEMS 245 scientific discoveries needed to advance this technology base must be nurtured through support of university education and research in materials science and engineering. Especially linport ant or Emerging Areas of Materials Systems Research RATIONALE FOR SELECTION Out of the enormous range of possibilities that materials sys- tems research represents, it seems almost presumptuous to identify a few "especially important or emerging research areas. It is the very pervasiveness of materials throughout a large fraction of en- gineering that has caused the field of materials research to be fragmented, duplicative, and difficult to codify. Despite these im- pediments, we believe that the exercise of identifying important areas for future research can actually have a unifying result, as we shall point out next. Out of the large number of valid candidates for research em- phasis, we have selected those that appear particularly promising in terms of three criteria: (1) a present or perceived future market for the ultimate products of that research; (2) a current state of the underlying science such that an advance of appropriate magni- tude appears realistic; and (3) a sense of national priority for the product. Given those criteria, we can list the high-priority areas that, in our opinion, are especially well positioned for immediate and intensive engineering research. EMERGING RESEARCH AREAS Seven classes of materials that lead to specific end uses need the focused attention of the engineering research community. A failure to focus that attention will prevent the nation from realizing the potential for commercialization of products emanating from a science base in which it has long held a position at or near the forefront. With no ranking implied, the seven product-specific materials research areas are

246 DIRECTIONS IN ENGINEERING RESEARCH 1. advanced ceramics; 2. semiconducting materials; 3. magnetic n~aterials; 4. polymers; 5. high-performance composites; 6. performance-driven metallic materials; and 7. biomaterials. The pane! strongly believes that there should be a greater uni- fication of effort in materials systems research. This integration can be accomplished by recognizing that it is wasteful and dan- gerous to neglect areas of research that transcend specific product or application lines. These five "generic" research areas, ranked in order of importance for progress in materials, are 1. processing (including sensor research); 2. durability and lifetime prediction; 3. tribology; computer modeling; and materials property data base. The remainder of this section is devoted to a discussion of these research areas. MATERIALS FOR SPECIFIC END USES Advanced ceramics Advanced ceramics are a rapidly evolving class of materials whose usefulness is based on their ability to fill both functional and structural roles. They can be tailored to have combinations of electrical, mechanical, and chemical properties that make them essential and irreplaceable in many new engineering applications. Functional roles include (1) semiconductor device components such as coupling capacitors and substrates, optical communica- tion waveguides, optoelectronic modulators and demodulators, magnetic components, and sensors and transducers; and (2) uses in chemical processes for separations and catalysis. Structural roles include service as load-carrying and wear-resistant compo- nents, especially in circumstances requiring corrosion resistance and high-temperature service. Ceramic materials (like other advanced materials) are a "gate- way" technology; rather than being an end in themselves, they

MATERIALS SYSTEMS 247 open up new potentials in other technologies, thereby achieving high degrees of leverage. The competitive performance of many devices and large systems depends on ceramic components that may make up only a small but vital part of the total. The fact that many ceramics involve nonstrategic or nonimported materials is another factor in their attractiveness. Electronic ceramics are a large family of crystalline and glassy materials serving as dielectric, semiconducting, magnetic, and op- tical materials and components. Current limitations in both the performance of devices and the growth of the field as a whole are in many cases due to insufficient research on the actual ceramic materials or components. The impetus for the development of structural ceramics and ceramic composites has been the promise of higher efficiencies through the use of ceramics in heat engines. There are, however, many other applications, both military and civilian, for struc- tural ceramics that take advantage of their good specific strength (strength-to-density ratio), specific stiffness, hardness, and wear and corrosion resistance. Ceramics present special requirements for successful applications under high stress because they charac- teristically undergo brittle failure. Extreme control of processing to minimize flaw size is required, which presents a special challenge for large parts and complex shapes. Areas of advanced ceramics more closely allied to process technology than to ceramics per se may play additional func- tional roles. For example, improved control over pore structure has opened up new routes for selective catalysis, as well as novel separations based on selective adsorption. New classes of zeolites offer promise as shape-selective catalysts for reactions heretofore limited to a small scale with homogeneous catalysts or biological systems. One example is a new catalyst used for selective iso- merization of xylenes. [n addition, the combination of ceramic materials with organometallic complexes is leading to a new class of catalysts with the selectivity of homogeneous catalysts and the robustness of heterogeneous catalysts. Despite the great diversity of applications for ceramics, a few common technical themes stand out. In almost all cases one is dealing with the behavior of materials near interfaces. Thus one may be concerned with tnin-fiIm coatings to provide proper wear resistance for ceramics, or with the behavior of coatings and layer structures that in turn determine the behavior of ceramics

248 DIRECTIONS IN ENGINEERING RESEARCH for functional applications (e.g., electronics or catalysis) and other aspects of surface modification. Whenever one is faced with joining ceramics, the preparation and properties of surface layers are, of course, critical. These questions of interface behavior may apply either at a near-atomic level (e.g., in assemblies of components for microcircuitry) or on the bulk scale (e.g., in the manufacture of large monoliths). There is an urgent need to learn how to optimize collections of properties simultaneously rather than a single property at a time. Because of the dependence of all of these properties on the fundamental constitution of a given material, there is also the need to develop predictive methods that are rooted in a basic understanding of material behavior. Ceramic processing is in a state of rapid change. This change is clue to improvements in the predominant technology based on powder processing and to new chemical routes that combine or by- pass some of the traditional steps. Ceramic processing increasingly needs to be studied as a system extending from the raw material to the finished product, with product design requirements taken into account by the processor and processing limitations taken into account by the designer. Semiconducting Material Over the past three decades, no single field of science or engi- neering has had a greater impact on the national productivity and quality of life in the United States than has semiconductor mi- croelectronics. Semiconductors have revolutionized the communi- cation, entertainment, and transportation industries, and created the computer industry. Further advances that will have profound consequences are possible through the development of optoelec- tronics and lightwave technology. There must be an increased emphasis on materials for use in aclvanced generations of computers, optical storage elements, and high data rate transmission and processing, because these new materials may be critical for specialized applications involving high-speed processing, optoelectronics, and resistance to radiation. The following discussion summarizes briefly some of the most important areas of research on semiconducting materials.

MATERIALS SYSTEMS 249 Material for Optoelectronic and Microwave Devices Materi- als for these uses, such as the binary compounds gallium arsenide (GaAs) and indium phosphide (InP) must be improved in conduct- ing and semi-insulating forms, and grown in large (at least 3-in. diameter), single-crystal ingots. Specific research areas needing emphasis are . . . Improved growth methods need to be developed for semi- ~nsulating GaAs and InP ingots. This effort will require better modeling of the growth techniques. . A determination of the necessary technology for produc- ing ultra-flat polished GaAs wafers should be pursued. All as- pects of wafer preparation should be included. Modern struc- tural/analytical techniques should be extensively used to assess wafer quality. Compound Semiconductor Helerostructures This is one of the most promising new materials technologies that has emerged in recent years. It offers a high probability for truly revolution- ary discoveries. In the area of optoelectronics, for example, ITI-V quantum-well heterostructures and superiattices promise to have a profound effect. With the advent of advanced crystal growth tech- niques such as molecular beam epitaxy and organometallic chem- ical vapor deposition, radically new structures have become avail- able, with profound device applications. For the first time, crystals can be synthesized layer by layer to yield entirely new compositions of matter. Such structures, typified by superIattices, metastable films, and complex multimaterial/multilayer sandwiches, are fre- quently called artificially structured materials. In many cases these structures possess unusual properties that offer promise for future electronic and optoelectronic devices. Considerable research in crystal growth, characterization of layers and interfaces, carrier transport, and processing methods will be required. Silicon Silicon is still the basic, largest-volume semiconduc- tor material. Despite our considerable experience with it, many problem areas still need research attention. In addition to im- proving the art of crystalline growth generally, there are also more specialized needs. For example, high-purity, floating-zone sili- con serves a segment of the device market in which compositional

250 DIRECTIONS IN ENGINEERING RESEARCH purity and crystalline perfection are prime requirements. The important research areas include . growth of large' uniform crystalline ingots to provide the silicon wafers needed in integrated circuit manufacturing; . bettering techniques for isolating and immobilizing impu- rities. Most Bettering techniques now rely on high-temperature processing for activation of the Bettering mechanism. Yet as cir- cuit densities continue to increase, lower temperature processing will be inevitable; deposition and characterization techniques for fabrication of thin films and interfaces; and . doping techniques including diffusion and ion implanta- tion (e.g., through the use of focused ion beams). Small Geometry Effects These effects are unportant in semi- conductors, conductors, and insulators. As linewidths shrink to submicron geometries, a host of materials problems emerge. Much of the carrier transport occurs at high fields, and traditional device structures must be modified or abandoned as field gradients in- crease. Research in fine-line lithography and dry etching, surface and interface effects, hot carrier transport, new conductor and contact materials, and multilayer growth of silicon and GaAs with interspersed insulators will be required. Magnetic Materials Magnetic materials are an integral part of our modern in- dustria] society, often rivaling semiconductors for breakthroughs in high-technology capabilities. They play a key role in power distribution; they permit the interconversion of electrical and me- chanical energy; they underlie microwave communication; and they provide both the transducers and the active storage mate- rial for data storage in computer-based information systems. The properties of magnetic materials are continually being improved, so that many new applications for these materials are now possi- ble. Indeed, magnetic materials seem to offer an endless variety of applications. When one is displaced, another arises. For ex- ample, the first generation of computers used magnetic drums for memory. These devices were displaced by matrices of ferrite cores.

MATERIALS SYSTEMS 251 As these cores were displaced by semiconductor technology, mag- netic materials reappeared in magnetic bubble devices and in disk drives. Despite the many practical uses for these materials, and de- spite their critical importance to the nation's industry and defense, it has become increasingly clear in recent years that the role of the United States in the science of magnetic phenomena, in mag- netic materials, and in magnetic technology has been declining. Since the mid-1970s, American manufacturers have looked increas- ingly to foreign sources for newer, better, and cheaper magnetic materials and devices. Nations such as Japan have invested far more than the United States has in the R&D needed to advance the performance of magnetic materials. As a result, U.S. system manufacturers are dependent on systems produced by the foreign component suppliers. To begin to rectify this situation, research is needed on both the science and technology of magnetism and magnetic materials. The recent report of a Committee on Magnetic Materials (National Research Council, 1985) identifies several new long-term research areas with promise for future payoffs in magnetic technology, such as magnetic superiattices, magnetic phenomena at surfaces and interfaces, and semimagnetic semiconductors. An NSF report (National Science Foundation, 1984) points to the need for fur- ther magnetic and related studies of: (1) valence instabilities in magnetic materials; (2) the electronic structures of magnetic sys- tems of restricted dimensionaTity, including surfaces and bimetal multilayers; (3) the onset of structural instability as heralded by the appearance of strong magnetoelastic effects; (4) the coexis- tence of stable or unstable magnetic states and superconductivity; (5) magnetic ferroelectrics; and (6) magnetic phenomena in the organic and polymeric solid state. Research in those magnetic technologies with strong growth potential or strategic value should be increased. At present the greatest emphasis is on compact information storage. Other areas with a critical need for research are new magnetic materials for special purpose magnets for low-power or high-field applications. There are classes of magnetic materials that have more strategic military value than economic potential at present, and in which modest research support would be appropriate. Two decades ago the DOD supported development of microwave ferrites; today, 85 percent of that market is still military. Magnetostrictive sensors

252 DIRECTIONS IN ENGINEERING RESEARCH for sonar devices are a more recent example. Certain rare earth elements and alloys have extraordinarily large values of magne- tostriction. If the resistivity of these materials could be increased significantly without affecting the magnetostriction, the resulting suppression of eddy currents would greatly extend the frequency range of such materials. Magnetic Bubble memories" are an- other example, in that they are able to compete economically with semiconductor memory in only a few applications; yet their combination of nonvolatility and ruggedness makes them attrac- tive for certain military applications. Bubble devices potentially have much higher storage densities than semiconductor memories. The use of modern lithographic techniques and other procedures for forming magnetic microstructures provides new approaches to recording and storage. The processing of magnetic materials in general has become an important, high-tech field toward which a change in national priorities is in order. Polymers Over the past 50 years the introduction of polymers for struc- tural, engineering, packaging, elastomeric, and fiber applications has been one of the spectacular changes in the history of materials technology. Most recently there has been a change in the direction of polymer research and development, away from the synthesis of new commodity materials and toward the formulation of polymer systems engineered for specific end-use requirements. Examples, in various stages of commercialization, include . fibers with very high polymer chain orientation, with the strength and stiffness of steel, at a fraction of steel's weight; . blends and alloys of polymers designed for optimum com- binations of stiffness and toughness; . polymers compounded with inorganic fillers to improve properties such as stiffness, dimensional stability, electrical con- ductivity, etc.; polymer structures (e.g., laminates or films with controlled porosity), to control the diffusion of small molecules, for barrier purposes in packaging, or for separation and purification of mate- rials; . polymers tailored for application in hostile environments, such as high temperatures or in the presence of solvents or degrad- ants; and

MATERIALS SYSTEMS 253 . liquid crystal polymers (e.g., aromatic polyesters) capa- ble of high orientation, leading to strength and stiffness during conventional plastics forming processes, and/or to unique optical properties. Most of these developments are characterized by a strong in- teraction between processing and the material itself. Second, many of these developments require the combination of several different materials in order to perform their functions. The emergence of these materials therefore challenges materials science and engi- neering to understand, control, and optimize on the one hand the interactions of the components and their dependence on size scale, and on the other hand the manufacturing and forming processes. Nigh-Performance Composites The development of new composite material systems consist- ing of high-performance fibers unified by advanced organic and inorganic binders could revolutionize the technologies associated with the aerospace industry, ground transportation, industrial ma- chinery en cl robotics, and general consumer products. Although raw material costs for high-performance composites exceed those of contemporary materials, their potential for eventual economy lies in the use of processing technologies that allow for near net- shape forming and part consolidation. The term "composites embraces any material made up of two or more discrete components that combine to yield enhanced properties. In earlier applications, Composite material" usually referred to arrays of continuous fibers bound together by a poly- meric matrix—fiberglass is a familiar example. The broadened application of composite materials encompasses a wider range of reinforcing geometries and binders, commonly including ceramics and metals. These combinations may yield performance characteristics that cannot be achieved by the various components acting in- dividually. In addition, such material systems offer the advantage of being designed to match the requirements of each application. The newer composites, such as graphite-epoxy composites, have already made great inroads in defense aerospace vehicles (fighter aircraft, missiles, and spacecraft) and are beginning to penetrate

254 DIRECTIONS IN ENGINEERING RESEARCH the commercial automotive and industrial machinery manufac- turing industries (e.g., automobile grill openings and panels, and robot arms and grippers). Because the microstructure of these materiab is determined by the manufacturing process to a greater degree than is the case with contemporary materials, the design process must integrate both the macroscale geometry of the finished product and the material microstructure in order for the full potential of these materials to be exploited. Thus, new design tools that are likely to be based on computer-aided methodologies must be developed so that both the manufacture and behavior of the product can be simulated. Although composite materials are established in several indus- tries and their use is growing, the realization of their full benefits will require fundamental research on the nature of these materials systems. Predicting the performance of the manufactured product requires an understanding of the failure processes, which initiate in defects and structural inhomogeneities ranging in size from a few nanometers to centimeters. The development and substan- tiation, through experimental observation, of failure prediction models pose a significant challenge for future research. Contemporary manufacturing methods are often labor-inten- sive when high-performance materials are achieved, yet contem- porary automates] methods do not produce high-performance ma- terials. Thus, another key challenge for research in this important technology is to develop the understanding and tools necessary to produce high-performance properties via automated manufactur- ing methods. Performance-Driven Metallic Material Performance-driven materials are materials that are selected on the basis of one or more special properties (as opposed to cost). Although many metallic materials may someday be replaced by ceramics, composites, and polymers, modern metals will be an important part of most high-performance structures for the foreseeable future. Moreover, major performance gains can still be made through research on metallic materials. In particular, considerable promise is seen in research in the following areas: high-strength/high-density materials; lightweight materials;

MATERIALS SYSTEMS . high-temperature materials; 255 metallic glasses; improved efficiency of manufacturing by means of closer process control; improved productivity by means of continuous processing; chemical assembly and plasma spray build-up of complex and graded alloy compositions; materials synthesis using process-structure-property mod- eling and experimentation; improved understanding of nonequilibrium; an understanding of the effects of adverse environments on metallic materials performance; and improved modeling of structure/fracture properties. Biomaterials Biomaterials are used in products that repair, restore, or re- place functions (or tissues) of the human body. They are, at present, metallic alloys (e.g., stainless steel, cobalt-chrome, and tantalum and titanium alloys); plastics, such as polyurethane and teflon; some ceramics; and some forms of pyrolytic carbon. As we learn more about controlling acute disease, and thus can extend human life, the problems of degenerative failure of the human body are bound to increase—as indeed they are already doing. The idea of retrofitting the human body may become as commonplace as the preventive and corrective maintenance applied by plant man- agers to the capital investments for which they are responsible. However, the technical (and human) problems are infinitely more subtle and complex In the case of the human Unphysical plants We need to better understand the interactive processes be- tween living and inanimate materials, so that the reliability, safety, and longevity of materials can be increased. Specific areas needing improvement are corrosion resistance (in metals); biodegradation of polymers; the development of nonthrombogenic surfaces for prosthetic materials to prevent blood clotting; design of device structures to reduce hemolysis (destruction of red blood celIs); better understanding of cancer-producing factors for materials; and closer simulation of tissue weight, strength, and modulus characteristics. In addition, specific mechanical properties appro- priate for various retrofit functions can be specified, and in many cases, clesigned.

256 DIRECTIONS IN ENGINEERING RESEARCH AREAS OF GENERIC RESEARCH* Processing (Including Sensor Research} Because of the historical emphasis placed by the materials community on materials science and on structure-property rela- tions, there has been far too little coordinated effort on the rela- tion between product performance and the many steps by which that product evolved from some native state or from chemically different raw materials. Perhaps because of the complex chem- istry involved (and perhaps also because of its relative newness), the plastics industry provides examples from which some impor- tant lessons can be learned. When attempting to manufacture a polymer product to match certain performance specifications, one often seeks leverage from each step in a flowsheet that begins with unreacted monomer. Reaction conditions, types and amounts of additives, chemical distribution of products, and orientation and modification during fabrication all come under scrutiny. From intensive study of the full process, one arrives at various sets of design alternatives and proceeds to find an optimum. The same thought processes need to be applied to materials production in general (see, for example, National Research Council, 1979~. If the approach outlined above is to be successful, new diag- nostic tools will need to be more closely integrated into processing steps. We shall have to develop in-line sensors that will inform the manufacturer of changes in processing results long before these changes become apparent as o~-specification products. In the ideal case, one would wish to design in-place sensors that can re- flect undesirable trends either in processing or, subsequently, in the product during service. Finally, it should be noted that processing refers not only to steps in the preparation of classic structural materials, but also to the broader range of systems already noted in this report. Hence, we include electronic materials and the manufacture of materials associated with biotechnology, including such processing steps as bioseparations. (See the relevant reports of other panels of the Engineering Research Board, as well as the board's own report.) We are now in a position to think realistically about pro- cessing capabilities undreamed of just a few years ago. An active National Aeronautics and Space Administration (NASA) program *These areas are not restricted to specific materials.

MATERIALS SYSTEMS 257 of materials processing in space has already shown some of the po- tential inherent in gravity-free processing of structural, electronic, and biological materials. In space we could study and improve on methods for creating new forms of metals and ceramics. The weightlessness of space also provides a means of separating bio- logically important materials with methods that, because of the adverse effects of gravity, are impossible on Earth. There are al- ready examples of casting methods in conventional foundries that were improved with knowledge gained from crystal-growth exper- · — laments In space. Durability and Lifetime Prediction The applications of modern materials are often paced by per- formance characteristics over their entire lifetime. This is because life-cycle cost is frequently the only factor offsetting the higher initial cost of incorporating modern materials into structures. In order to calculate life-cycle cost accurately, documented models of the lifetime and ultimate failure mechanisms of modern ma- terials are needed. This requirement points to several aspects of time-dependent behavior as promising areas for research: models for failure mechanisms; corrosion of monolithic metallic materials; environmental degradation of composites, polymers, and ceramics; . models for time-dependent fracture (e.g., creep, fatigue, and stress-corrosion cracking); and . methods for assessing residual life of materials in use. Tribology (Friction, Lubrication, and Wear) Tribology is the science and technology of interacting surfaces in relative motion. It encompasses the fields of friction, lubrication, and wear. ~ibological processes involve the interaction of the surfaces in relative motion, the lubricant, arid the environment in which they must operate. Tribological processes are highly coupled and require a knowledge of chemistry, mechanics, and transport phenomena occurring under transient conditions and at lengths measured in microns or less. Models are still very tenuous, partly because critical data are hard to obtain. Tribology is an interdisciplinary field. Many of the research problems in tribology are associated with the need for better sensor

258 DIRECTIONS IN ENGINEERING RESEARCH development. In particular, we need noninvasive means to monitor chemical and thermal changes over short times and small volumes or surfaces. It is also important to be able to associate research results with predictions of service performance under a variety of conditions. Materials can fad! in response to environmental as well as me- chanical attack. There Is a need to systematize our approach to understanding pervasive causes of failure, such as erosion and cor- rosion. These causes are also highly coupled phenomena requiring multidisciplinary approaches. Computer Modeling Computer modeling refers to the link between large-scale com- putation as a modern research discipline (as demonstrated, for example, in the NSF's recent awards to four universities for su- percomputer centers) and the need to predict how a material will function under given initial and boundary conditions. For exam- ple, for semiconductors, the progress in thin-fiIm research that has improved our ability to fabricate artificially structured materials has in some respects outstripped our theoretical models. So many potential materials configurations can, in principle, be prepared that very sophisticated predictive models are needed to select the more prorn~sing structures. Not only is this the essence of materials processing, it is also the final object of engineering research—namely, engineering de- sign. Through sophisticated computational algorithms designed by persons knowledgeable about the physics, chemistry, and the- oretical structural mechanics of materials, one can predict how a material, a surface, or a composite will behave under complex state conditions. Such modeling need not utilize the enormous power of supercomputers in order to be valuable. Computer modeling in general has not been adequately developed as a materiab systems research tool, and continues to offer great promise for materials process development. Materials Property Data Base Existing data bases are inadequate for characterizing both physical and mechanical properties of modern materials. Yet the sophisticated computer modeling we are on the verge of developing can never be more reliable than the reliability and integrity of

MATERIALS SYSTEMS 259 fundamental property data supplied to the computer's data bank. A generation ago, many laboratories were engaged in measuring physical properties; that activity has long been out of favor. Now, with new materials and advances in computational technology, the need for such data has grown rapidly. We are on the verge of being unable to use computer models of modern material behavior for lack of a sufficiently thorough data base on properties. In some instances empirical relationships, equations of state, and property "maps" from which reliable estimates of needed data can be derived by computer will be more useful and cost effective than tabulations of discrete data. Such data bases should (ideally) be interactive to permit cross-checking, comment, and a degree of correction and update by its users. Communication among researchers in a field would be enhanced by such interaction. Issues Determining the Health of Materials Systems Research NEED FOR PRIORITIES Materials science and engineering issues are pervasive in engi- neering applications, yet the materials community is fragmented by its very range. Various mission-oriented government agencies support overlapping programs of materials research. As a result, findings resulting from the nation's investment in research are of- ten not efficiently assembled, analyzed, or reduced to practice, and some important areas are inadequately supported. Because materials science and engineering is such a broad area, mechanisms for setting priorities are not well established. For example, the processes by which materials systems research programs are selected by the federal government are not uniform. They range from peer review through internal agency review to, in the extreme, "pork barrel" techniques. Differences in mission account for some of the differences among agencies in selection criteria and procedures, to be sure; but this fact does not explain or justify the degree of variability that is seen. Although there are some interagency coordinating mechanisms, no true clearing- house yet exists for setting overall national priorities or directions.

260 DIRECTIONS IN ENGINEERING RESEARCH Whether or not national priorities can realistically be set leads to the question of "industrial policy"; this concept, involving central- ized planning, is anathema to many in the United States. In assessing the magnitude of the problem, the Committee on Materials (COMAT) of the Federal Coordinating Council for Sci- ence, Engineering, and Technology made an inventory of materials research and technology (RAT) development supported by federal funds for the years 1976, 1980, 1982, and an estimate for 1983. In round figures, about $1 billion was spent in each of these years, with the Department of Energy (DOE) dominating, and with sub- stantial expenditures by the DOD, the Department of the Interior, the NSF, NASA, and, erratically, the Environmental Protection Agency (EPA) (see Table 1~. The level of funding since 1983 has been relatively constant, although there was a 5.5 percent decrease in FY86. However, it is estimated that the proposed overall al- location for FY87 will again be about $1 billion (Federation of Materials Suppliers, 1986; Margolin, 1985~. Private investments in materials research and development are hard to quantify, but estimates of about $4 billion are heard. These figures seem to suggest a field that is, if not expanding, at least holding its own. However, they are not otherwise very illuminating, en c! may actually be misleading because of the lack of consistent principles for aggregating the data. To begin with, the portion spent for research is unknown, as is the relative allocation between materials science and engineering. It is generally believed that the majority of materials engineering work is performed in other engineering fields (e.g., electronics), and particularly at the interfaces of various engineering disciplines. Funding for this work does not appear in the table. The DOD may sponsor a good deal more materials research than is reflected in the table (some say as much as $500 million more). The interdisciplinary nature of materials systems research makes it virtually impossible to gain a clear understanding of how much is spent, where it is spent, by whom, and for what purposes. The same statistical difficulties are found in attempting to assess student enrollments, degrees, and faculty numbers and specializations within the materials field. Better data on a national basis are certainly needed to support policy studies in this crucial field. The federal government should sponsor an effort to determine which specific areas of the field need better definition and data, and how this can best be accomplished.

MATERIALS SYSTEMS TABLE 1 Federal Materials R&T Funding (in Sthousands) 261 Funding Source 1983a 1982 1980 1976 D OE 276,207 285,262 514,100 332,897 DOD 162,200 147,000 160,200 131,881 EPA 179,426 146,693 2,400 88,398 Interior 104,609 113,667 119,686 165,350 NSF 102,414 99,721 88,920 68,700 NASA 102,305 101,415 78,582 51,533 Health 34,260 31,920 6,070 16,625 Agriculture 28,551 29,759 64,598 38,254 NRC b 18,740 17,710 13,674 7,028 Commerce 13,856 14,201 35,795 21,080 Transportation 6,854 5,682 3,442 6,153 TVA- 2,607 2,792 9,650 9,226 Treasury 1,258 1,080 2,516 790 Smithsonian 825 750 1,000 1,000 FEMA d 0 107 50 0 HUDe 0 0 0 6,669 State 0 0 0 540 Labor 0 0 3,000 4,063 GSA f 0 0 0 132 Total 1,034,112 998,329 1,103,683 961,320 -Estimated amounts. bNuclear Regulatory Commission. dTennessee Valley Authority. eF`ederal Emergency Management Agency. fDepartment of Housing and Urban Development. -Government Service Administration. SOURCE: COMAT, 1984. Given the uncertainties, the main point to be made with re- gard to funding in the materials field is not whether there is enough money spent, but whether it ~ spent in the right places, whether the balance between materials science and engineering research funding (to the extent that they can be clearly differentiated) is appropriate, and whether there is continuity of effort along the entire R&D chain. It should also be noted that the apparent pro- portion of government-to-industry funding in this field (20 percent government, 80 percent industry) is considerably smaller than the proportion in most other technical fields (an average of about 40 percent government, compared with 60 percent industry). This fact suggests that not enough of the kind of long-range, funda- mental engineering research normally supported by government is

262 DIRECTIONS IN ENGINEERINrG RESEARCH being pursued. It is likely that the largely developmental work pursued by industry could be leveraged and made more effective through a greater effort directed at understanding how materials processes work. Although definitional inconsistencies make it difficult to sort out the relative expenditures on different aspects of materials sci- ence and engineering, two points stand out: (1) government fund- ing in this field is unusually low in comparison to industrial R&D spending, and (2) there is a strong perception that basic scientific research and expensive major facilities account for the majority of federal R&D funds, with only a fraction of the total going into ma- terials systems and processing research. Recent studies (e.g., U.S. Department of Commerce, 1984; Government Accounting Office, 1985) point out that other nations especially Japan are placing much greater relative emphasis on the processing and aspects of materials and are poised to reap great benefits as a result. The question of proper balance should be examined and, if necessary, priorities should be altered accordingly. To address these and other questions, Title II of the National Critical Materials Act of 1984 establishes a National Critical Ma- terials Council under and reporting to the Executive Office of the President. This Council is composed of three members, with one member having ha background in and understanding of environ- mentally related issues. The bill spells out the purposes, responsibilities, and author- ities of the council. They are, in a word, sweeping, including the establishment of a national Federal program plan for advanced materials R&D. There is concern within the materials science and engineering community that this small group will autonomously (and perhaps un(luly) influence the allocation of funding in the materials field. The pane! doubts that this mechanism-can ac- curately represent the needs and priorities of a constituency as broad as the materials community. We therefore recommend that another, less centralized mechanism be found for achieving these desirable goals. RESEARCH FACILITIES Another set of priorities of importance to materials scien- tists and engineers involves the division of expenditures between

MATERIALS SYSTEMS 263 salaries and facilities for carrying out research. Facilities fall into three general categories: 1. Major: large national or regional facilities, available on an open or proprietary basis, with or without payment of operating fees (examples are facilities for synchrotron radiation, neutron scattering, and high-resolution electron microscopy); 2. Intermediate: local facilities shared among users at a single lo- cation (e.g., electron microscopes, lasers, molecular-beam epitaxy or chemical vapor deposition apparatus, clean rooms, specialized magnets, mid-size computers, etc.~; and 3. Small-scale: individual group equipment to be used ordinarily by one principal worker and that person's colleagues and students. The allocation of resources among these three classes and within each class is obviously a very subjective and controversial matter. As in other areas, the diversity of backgrounds of people who identify with materials research makes it difficult to achieve a consensus that represents all of the needs of the community. Some- times the needs of the "best-connected" people and institutions therefore prevail. Such needs, if narrowly defined or self-serving, will almost invariably fail to represent the mainstream of the ma- terials community. This is especially clear if the mainstream" is defined as that part of the community that is central to the com- petitiveness of U.S. producers and users of performance-driven materials. An attempt has recently been made (by the Commission on Physical Sciences, Mathematics, and Resources of the National Re- search Council) to assess priorities within the major-facility cate- gory (National Research Council, 1984b). Consensus was reached among members of a pane! chosen for diversity of interests, re- search styles, and institutions. The committee's recommendations for new facilities, and their estimated costs in FY85 dollars, are (in order of priority): 1. a Gem synchrotron radiation facility ($160 million); 2. an advanced steady-state neutron source facility ($260 mil- lion); 3. a 1- to 2-GeV synchrotron radiation facility ($70 million); and 4. a high-intensity, pulsed-neutron source facility ($330 mil- lion).

264 DIRECTIONS IN ENGINEERING RESEARCH For comparison, the pane] hap estimated the cost of setting up facilities for materials processing research, which is currently a neglected area. These facilities are classed as "intermediate facilities." Estimates of their costs are (with no priority implied by ranking): . a composite processing laboratory for metals, ceramics, or polymers ($5 million each); ~ a semiconductor processing research facility ($7-$8 million each); and . a molecular-beam epitaxy or metalorganic chemical vapor deposition research facility ($2 million each). There is an urgent need for facilities such as these in uni- versities across the country. We believe that policymakers have commonly failed to recognize research needs that relate directly to the competitiveness of the materials sector of U.S. industry. To be sure, the new concept of funding major cross-disciplinary research centers is intended to be a step ~ this direction. The Center for Advanced Materials at the University of California at Berke- ley and the NSF Engineering Research Centers at the University of Delaware and the University of California at Santa Barbara have the potential to make strong contributions in the materials processing area. But these as-yet unproven centers are only a beginning. More processing-oriented research facilities are needed for many types of advanced material. The cost of these greatly needed facilities, which will produce fundamental knowledge that will rapidly translate into economic gains, is quite low in compar- ison to the costs of major facilities used to conduct basic research in materials science. Indeed, much of the approximately $1 billion in federal funding depicted earlier in Table 1 supports the physical science aspects of materials research, which often requires costly equipment and facilities. In 1986, for example, 24 percent of the budget of DOE's Materials Sciences basic research subprogram (by far the largest unit for materials R&D in DOE) ~ devoted to facilities and equipment; in 1985 the figure was 31 percent (Mar- golin, 1985~. The facilities needed to fill a significant gap in our current materials science and engineering research capabilities and produce knowledge of key economic importance would have a total cost in the range of $150 million.

MATERIALS SYSTEMS UNIVERSITY-GOVERNMENT RELATIONS 265 Although support from industry, state governments, and vari- ous other sources has become increasingly important to the mate- rials science and engineering research enterprise, federal agencies continue to be major factors in supporting this research. The government agencies that are concerned primarily with basic re- search (particularly NSF, DOE, NASA, DARPA, the Air Force Office of Scientific Research, the Army Research Office, and the Office of Naval Research) have generally understood the nature of university research and its importance to national goals. University research should address the fundamental aspects of materials, devices, and processes. It should be recognized that the accompanying education of Ph.D.s in these areas is of substantial importance to the country's economic health and defense. To the extent that university-government interactions stimu- late, support, and facilitate that type of fundamental investigation and excellence of education, such interactions can be extremely healthy. On the other hand, restrictions and impediments to such research and education can result in an eventual degradation of the economic and defense enterprises which these agencies are charged with supporting. MISSION ORIENTATION AND OVERMANAGEMENT Federal funding of university research in the postwar period has transformed engineering colleges into major research institu- tions. However, this support from federal agencies has carried with it a potential for conflict between the free, creative exploration of new ideas on the one hand and these agencies' needs for program development on the other. New materials and new processing tech- niques important to the country's needs have frequently emerged through these funding programs. Yet if such programs are too narrowly focused, the hoped-for objectives of creative innovation and discovery may be lost. In materials research, a case in point is the current techno- logical need for a metaI-insulator-semiconductor structure in com- pounds such as GaAs or InP to provide a metal-oxide-sem~conduc- tor (MOS) technology similar to that which has been so produc- tive in silicon. Recognizing the importance of this technology to future electronic devices, government agencies may fund work at

266 DIRECTIONS IN ENGINEERING RESEARCH universities to study the properties of interfaces between com- pound semiconductors arid various insulators. On the other hand, a narrowly defined contract on MOS structures between specific compounds and specific insulators, with deliverables at the end of the contract period, would probably fad! because the basic work on interface properties has not been done in detail. Researchers should instead be encouraged to look carefully at the basic mate- rials properties before trying to develop devices. The problem of overmanagement from Washington need not be associated only with applied research having short-term payoffs. Even in pursuit of basic research, in which a far-sighted approach is being taken by the sponsoring agency, research management that is too detailed can stifle the investigator's ability to capitalize on emerging results. Overmanagement of a project by the sponsor' whether in basic or applied research, reduces the chance of success and should be avoided. The investigators themselves should be permitted to de- fine and follow their own research strategy. If reviews of a specific research effort after a reasonable period of time (often 3 or more years) indicate a lack of progress or loss of focus, such research should be terminated with the provision of adequate funding to allow any graduate students to finish their thesis work. DILuTIoN OF BASIC RESEARCH Because the program needs of DOD and other agencies of the government often bias these programs toward specific problems, there is a tendency to dilute the pool of basic research funding with mission-oriented projects. Although many faculty members are able to extract good fundamental research from a program with a rather applied mission focus, the risk remains that such well- defined programs may stifle the creative enterprise so important to university work. The increase ~ theme-oriented basic research is an unwelcome trend. If "suggested areas of research" predominate, with no increase in total resources, then truly unsolicited research becomes harder to find. Agencies must carefully protect the rather modest amount of money going to fundamental research and avoid the temptation of putting highly mission-oriented projects on the list of basic research programs.

MATERIALS SYSTEMS 267 NEW THRUSTS One of the major responsibilities of support agencies is to re- main alert to new developments in the fields they are supporting; the best way is by involvement with productive researchers. These agencies can and should alter their programs to accommodate new areas that emerge from ongoing research. On the other hand, it is important that research funding remain relatively stable at univer- sities, at least on the timescale required for Ph.D. thesis research. In many cases, considerable time is required to bring projects to completion. Particularly in the areas of fundamental materials re- search and development of new processes, stable research funding over Tong periods of time (e.g., 5 years) is an almost universal requirement. As a result, funding agencies should be careful not to abandon productive areas of work to chase every new idea. Given the inevitable restrictions on a particular agency's fund- ing, balancing stable funding for ongoing productive research against the need to support new thrusts and ideas is a delicate matter. Each agency should have some portion of its budget avail- able for pursuing unexpected developments without detracting from support for productive ongoing projects. As is often done, reason- able deviations from the originally proposed approaches should be encouraged. At the same time and perhaps partly as an offset- less productive programs of research should be phased out and terminated. In every case the importance of creativity, innovation, and productivity should be emphasized over a strict adherence to preconceived notions about the direction and methodologies of the research. USE OF GOVERNMENT RESEARCH FACILITIES Universities can greatly benefit from the use of specialized ma- terials research facilities at government and government-supportect laboratories, such as Oak Ridge and Sandia National Laborato- ries. This is particularly true in the case of large or expensive equipment and facilities. In addition, the expertise of a labora- tory's staff can complement the experience of university faculty in materials research. These facilities can also be very effective as sites for the conduct of graduate student thesis research. Such arrangements must, however, be organized to maintain academic control over thesis topics.

268 DIRECTIONS IN ENGINEERING RESEARCH There has been good experience at many laboratories in spon- soring such thesis research visits from universities. In general, universities who are able to use these facilities should take full advantage of the equipment and expertise resident within them. RESTRICTED DATA On rare occasions, a new material or process emerging from a university research project warrants protection in the national interest. It is extremely important, however, that funding agencies restrict data reluctantly. Generally, a free exchange of research data and the publication of results enhances the entire research and development enterprise. This is particularly true in the case of universities, which depend on publications and technical meetings as their research output, rather than providing products for the marketplace. Given the fundamental nature of most materiab research at universities, it is rare that results and data are of such immediate use to defense objectives as to require restriction from publication. There are developments that permit optimism in this re- gard. In October 1984, then Under Secretary of Defense Richard DeLauer issued a memorandum specifying that no restrictions should be placed on the publication of unclassified fundamental research sponsored by DOD. The new rule thus applies to all DOD- supported research on university campuses; definitions generally cover all 6.1 research and any unclassified research performed on- campus from 6.2 funds. Consideration is being given at the White House to establishing these policies in all other federal agencies. Thus, for the time being, the classification prior to the granting of a DOD research contract is the deciding factor, and most univer- sities will not undertake classified work. The assumption should be that research results are freely publishable and that restriction of these results is an exception requiring careful consideration by the agency and discussion with the principal investigator.

MATERIALS SYSTEMS Assessment of Personnel Resources INTRODUCTION 269 The success of research in materials systems clearly depends on having the talent available to conceive and carry out that re- search. This section deals with the availability of new research talent, primarily from graduate programs in materials science and engineering in the United States. These disciplines and research areas include metallurgy, ceramics, polymers, electronic materials, and portions of solid state physics and the chemistry of materials. The industries that will benefit most from the availability of ca- pable graduates in those fields are those producers and users with R&D programs looking ahead to new materials and processing methods for the next generation of materiab systems. Creating the environment, identifying the appropriate prom lems, and providing the tools for inventive and imaginative re- search by Ph.D. students is one of the strongest challenges facing the universities with programs in these fields. It is also one of the most crucial issues determining the future of materials systems development in this country. ASSESSMENT OF STUDENTS AND GRADUATES It is extremely difficult to measure the quality of students by any objective criteria, particularly in research areas requiring not only academic skills but also the ability to understand and cre- atively use complex fabrication and measurement facilities. Data on total enrollments are also difficult to pin down, because the materials field is highly interdisciplinary. Therefore, this section largely summarizes the panel's general impressions on the avail- ability and characteristics of students in materials research fields' as compared with engineering and science students generally. ABILITY OF GRADUATE STUDENTS Most candidates for graduate degrees in materials programs come from undergraduate engineering curricula, with a smaller

270 DIRECTIONS IN ENGINEERING RESEARON number coming from science disciplines such as physics and chem- istry. As in other fields of engineering, attractive salaries in indus- try draw many of the most talented students into the workforce at the B.S. level, so that relatively few continue on toward a Ph.D. and research. Materials researchers must have strong backgrounds in both science and engineering. Thus, one disincentive for many top students is the substantial time and effort involved in pursuing a doctoral program in this field. Because of the wide range of knowledge ant] skills required for success in materials research, materials science and engineering students tend to be very broad in their background and inter- ests. Indeed, some enter the materials field through double-major programs in conjunction with mechanical, aerospace, or other en- gineering majors. Students who pursue advanced study tend to proceed entirely through the Ph.D. program and to do produc- tive work. However, it is generally apparent to faculty members that the best students are not choosing to enter materials pro- grarns, with the exception of certain well-publicized fields such as electronic materials. It is difficult to compete with fields such as computer science and bioengineering, which receive more fa- vorable press than does materials science and engineering. There is an image problem to be overcome in attracting the brightest students at both the undergraduate and graduate levels. Certainly one positive note is the influx of women into the materials field in recent years. Indeed, the rise in the percent- age of female students is the largest in any engineering discipline. Women now make up about 14 percent of all graduate students in materials and metallurgy (the closest relevant category) (En- gineering Manpower Commission, 1985a). At some schools the percentage of women is even higher 17 percent at MIT, for ex- ample (Engineering Manpower Commission, 1985b, 1986~. These women seem to perform well in materials graduate programs and as researchers. A continuing dilemma in materials research is the high per- centage of non-U.S. citizens who are enrolled. Often these students are superbly qualified from a technical standpoint, but are not able to provide the level of interaction with faculty, graduate students, and undergraduate students that leads to a healthy discipline. After receiving an advanced degree, these foreign students often, for economic or political reasons, return to their homeland. As a result, we frequently train personnel who in turn are at the leading

MATERIALS SYSTEMS 271 edge of technology in countries with which we are, or will be, in direct economic competition. ISSUES OF SUPPLY AND DEMAND General Concerns The output of materials science and engineering graduates is adequate on average but inadequate at times of peak industrial expansion. A central problem is the cyclical nature of the demand and the long lead time in the university's response to changes in the demand for engineers from a particular discipline or specialty area. In particular, the longevity of tenured faculty appointments leads to inertia in the ability to change emphasis within a disci- pline. (This is also true in other rapidly developing fields besides materials.) As a result, university administrators are cautious when responding to the call from industry to increase the degree- granting capacity of or shift the emphasis of a program relative to a current "hot" discipline or area such as electronic materials. This cautious attitude has been legitimized in recent years by a growing tendency of employers to treat engineers (in research as well as applications) more like a commodity than a resource. If there is to be a better match of supply and `demand, positive action must be taken. First, an improved method for forecasting the need for engineering graduates at all degree levels must be developed. Second, longer term commitments by employers are needed. Third, some methods for protecting universities against the Downside are required. The resources to implement this protection should come from industry, but probably will have to be provided by government before any real prospect of achieving a better balance between supply and demand is to be achieved. Graduate Degrees The supply of M.S. and Ph.D. graduates in the materials field is, as is suggested earlier, quite low. Even in the burgeoning field of ceramics, for example, annual degree production nationally in 1984 was 87 at the M.S. level and 18 at the Ph.D. level (Engineering Manpower Commission, 1985~. Ceramics is the only specialty of materials for which data on degree production are readily available. In the broader field of materials and metallurgical engineering, in the same year there were 578 M.S. and 241 Ph.D. graduates

272 DIRECTIONS IN ENGINEERING RESEARCH nationwide. It is not possible to determine how many of these students were specialized in the advanced materials systems areas addressed in this report. The balance between supply and demand for those with ad- vanced degrees in materials research appears to vary considerably across specialties. Whereas the demand for Ph.D.s in electronic materials and related fields is very strong, the hiring needs in some other fields are static or declining. For example, virtually all M.S. and Ph.D. graduates in ceramics are hired to fill positions related to electronic and structural ceramics, rather than those related to conventional ceramics. The pane! believes that the growing dependence of the na- tion's economy on new materials ensures a long-term increase in demand for doctoral-level researchers in every area of the materi- als field, for both industry and academe. Therefore, it is necessary to take steps that will encourage more of the brightest undergrad- uate students to enter graduate programs in material science. Students should be counseled on the advantages of graduate edu- cation in materials, particularly regarding the Ph.D. as an entree to a research career. Undergraduate students in relevant disci- plines should be exposed to materials engineering research. At universities with strong graduate programs this could be done by offering undergraduates (especially entering freshmen) opportuni- ties for part-time and/or summer employment in such programs, and by organizing undergraduate thesis programs. A large pool of potential talent also exists at smaller schools unable to offer such opportunities. Undergraduate seminars should be given at these schools by academic and industrial researchers to acquaint students with materials systems research fields. In addition, summer jobs in industrial and academic research centers should be made available to interested students. Federal and in- dustrial funding ought to be provided to support these recruitment activities. Because a lack of awareness of the materials field is a large factor in restricting the supply of graduates, the pane! believes that efforts to increase students' interest should be made as early as possible in the educational experience. Therefore, the existence, importance, and challenge of materials systems research as a disci- pline should also be communicated to high school students. Steps should be taken to ensure that the materials field catches the in- terest of young women as well as young men. Possible mechanisms

MATERIALS SYSTEMS 273 to do this include industry sponsorship of engineering fairs (analo- gous to science fairs); production of suitable video or film programs by professional societies; and tours of academic, government, and industrial laboratories engaged in materials research. ADEQUACY OF GRADUATE PROGRAMS This section deals with two factors perceived to be at the heart of quality graduate programs in materials, namely, new programs and the adequacy of the faculty. DEVELOPMENT OF NEW PROGRAMS There are relatively few established formal graduate programs in materials. In ceramics, for example, only 10 schools produce any advanced degrees at all, and only one program is accredited at the master's level (Engineering Manpower Commission, 1985a). In composites there are five formal graduate programs; in polymers there are even fewer programs. New programs are needed in the materials field to meet the emerging demand. For new graduate programs to be most elective, they should generally be specialized. Such programs should expect to build on a student's strong grounding in fundamentals, and should fo- cus their resources on achieving quality education in a single area such as composites or ceramics. In any field of engineering the development of a high-quality, broad-based program takes con- siderable time and requires extensive resources. It seems unlikely that many new broad-based programs will be able to compete with established programs for full-time graduate students, espe- cially in coming years, when there will likely be fewer students. Without an adequate supply of full-time graduate students, strong research programs of the type that characterize the great research universities in the United States cannot develop. There is probably adequate overall capacity in universities to conduct materials research, at least in a time-averaged sense, but sheer research capacity is not the sole motivation for developing new programs. In fact, the development of new programs in engi- neering is motivated as much by the need for graduate education- especially part-time programs—as it is by research needs. That driver seems to be closely tied to the shifting geographic locations of the "market." This is especially true in the case of part-time

274 DIRECTIONS IN ENGINEERING RESEARCH graduate study, in which case the university program must be near the workplace to enable part-time students to commute from school to work. In many cases, the new locations of industry are not near well-established universities; thus, there is a tendency for new programs to develop at regional universities and colleges. The deceptive aspect of this is the time required for these new pro- grams to achieve adequate academic stature. If the employers in the new location recognize the need for new technology, then this "quality issue becomes a major barrier to acceptance of the de- veloping programs. That is, employers who recruit bachelor-level engineers from the best schools in the country find that these peo- ple have little interest in continuing their education on a part-time basis at institutions of distinctly lower stature than those at which they studied as undergraduates. Thus, whereas the development of some new programs is inevitable, it may be more productive for the existing established institutions to work harder at delivering part-time graduate programs off-site using such means as visiting faculty, teleconferencing, or videotaped instruction. There are a number of prototype off-site programs that could be mentioned in this context. The degree of success that these programs have achieved is as variable as their formats. Therefore, the need for programs offering part-time graduate study in materials in areas where high-tech industries are locating can best be served by established, first-rank institutions offering such programs off-site. . ADEQUACY OF THE FACULTY In the major existing programs the supply of faculty is gener- ally adequate, with the possible exception of new directions within these programs. Two examples of such directions are advanced ce- ramics and electronic materials. Even in the few institutions in which these relatively new activities began and are now expand- ing, there are not enough faculty. Moreover, those new faculty that are available come from only a few schools, so that a degree of inbreeding inevitably occurs. (Here, programs of postdoctoral research abroad offer some opportunity for ~cross-fertilization~; France and West Germany in particular have attractive university research programs.) Similar problems also occur in other emerg- ing areas of materials research. There are probably fewer than

MATERIALS SYSTEMS TABLE 2 Materials Processing Faculty in the U.S., by Speciality 275 1978 1984 Speciality No. Percent No. Percent Casting/solidification 22 32 16 17 Deformation processing 15 22 18 20 Design/systems 4 6 10 11 Powder metallurgy 13 19 14 15 Welding 14 21 16 17 Ceramic processing O O 10 11 Polymer processing 0 0 8 9 Total 68 92 PA different and probably more reliable census yielded numbers of 28 and 36 in this category for 1978 and 1984, respectively. Source: American Society for Metals, 1985. 40 full-time-equivalent faculty involved in composites research, for example. Table 2 shows a recent estimate of the specialization of all U.S. faculty in materials processing (American Society for Metals, 1985~. The absolute numbers cannot be assumed to be accurate (see, for example, the notation regarding ceramic processing), but their magnitudes and relative figures are probably reasonable. One sees that even in this currently much-emphasized field, the overall numbers are still quite small, and that the fields showing high relative growth rates are in highly publicized speciality areas within materials science. There is some concern about the loss of faculty to industry, especially in shots areas. This has become less troublesome since universities began to realize that they must offer more competitive salaries. However, improvements have been most marked in the first-rate schools; qualified faculty members are still leaving many institutions for better paying industrial jobs. To help alleviate these problems, industry should be encouraged to support the entry of new graduates into acacleme. The "forgiveable loans program for graduate students sponsored by General Electric is an example of such encouragement. In this program, loans of up to $5,000 are made to Ph.D. candidates, and repayment is forgiven if

276 DIRECTIONS IN ENGINEERING RESEARCH on graduating the student pursues an academic career for at least 4 years. It is particularly hard for developing institutions to compete for high-quality faculty, both on the basis of the environment they can provide (facilities, colleagues, etc.) and on the basis of salary. Therefore, in general it is probably true that quality is a bigger problem than quantity in these programs. In the most competitive areas of materials specialization, it may be difficult for emerging programs to compete for any of the limited number of qualified faculty candidates. In these areas, both quality and quantity are a problem. New Ph.D.s are especially hard for universities to recruit. There are many disincentives that a recent Ph.D. graduate must face as a new faculty member. Some candidates are uncertain about widely publicized issues such as "publish or perish," ob- taining research support, and the promotion and tenure decision processes. The NSF's Presidential Young Investigator Awards help somewhat to ease the question of research support, even though the required industrial matching money has been difficult to oW fain in many instances. In addition, it appears that university administrators are beginning to recognize the need to provide a transition period for young faculty to adjust to the spectrum of ac- tivities they must learn to deal with simultaneously. The bottom line is that outstanding young faculty can adjust to the university climate ancI thrive. The ones who have succeeded serve as role models for others and are a source of encouragement. Programs that have no such role models often find it harder to present a convincing case for their environment being "penetrable." These programs may have more difficulty recruiting young faculty. Continuing efforts should be made by universities to improve the attractiveness of academic life for entry-levelfaculty, especially in emerging or high-priority areas such as electronic materials, advanced ceramics, and processing. The Presidential Young Inves- tigator Awards or comparable long-term research grants represent one effective mechanism for doing this. Government and indus- try should both be willing to fund this type of grant in targeted specialized areas.

A~4TERIALS SYSTEMS References 277 American Society for Metals. Mctallurgy/Matcr~als Education Ycar6001c. Metals Park, OH: American Society for Metals, 1985. Committee on Materials. Inventory of Federal Materials Research and Tech- nology: FY82. Federal Coordinating Council for Science, Engineering, and Technology. Washington, DC: Office of Science and Technology Policy, June 1983. Engineering Manpower Commission. Engineering and Technology Degrees, 1984. Part III: By Curriculum. Engineering Manpower Commission of the American Association of Engineering Societies, Inc., 1985a. Engineering Manpower Commission. Engineering and Technology Degrees, 1984. Part II: Minorities. Engineering Manpower Commission of the American Association of Engineering Societies, Inc., 1985b. Engineering Manpower Commission. Engineering and Technology Degrees, 1985. Part I: By School. Engineering Manpower Commission of the American Association of Engineering Societies, Inc., 1986. Federation of Materials Societies. R&D in Materials Science and Engineering. March 1986. Government Accounting Office. Support for Development of Electronics and Materials Technologies by the Materials Technologies by the Gov- ernments of the United States, Japan, W. Germany, France, and the United Kingdom (GAO/RCED-85-63~. Washington, DC: Government Accounting Office, September 1985. Margolin, S. V. ROD in Materials Science and Engineering. Report prepared by Federation of Materials Societies, May 1985b. National Research Council. Mat Trials and Marie Needs: Mat Trials Scicnec and Enginecnug. National Research Council Committee on Strategic Materials. Washington, DC: National Academy Press, 1974. National Research Council. Science Base for Materials Processing: Selected Topics (NMAB-355~. National Materials Advisory Board of the National Research Council. Washington, DC: National Academy Press, 1979. National Research Council. High-technology Ceramics in Japan (NMAB- 418~. National Materials Advisory Board, Committee on the Status of High-technology Ceramics in Japan. Washington, DC: National Academy Press, 1984c. National Research Council. Major Facilities for Materials Research and Related Disciplines. National Research Council Committee on Major Materials Facilities. Washington, DC: National Academy Press, 1984b. National Research Council. Magnetic Materials (NMAB-426~. Report of the Committee on Magnetic Materials, National Materials Advisory Board. Washington, DC: National Academy Press, 1985. National Research Council. Surface Modification of Electronic Materials in the United States and Japan: A State-of-the-Art Review (NMAB-443~. Report of the Panel on Materials Science, National Materials Advisory Board. Prepared for the National Science Foundation, March 1986. National Science Foundation. In lieu of the original. Mosaic 2~33:18-25, 1971.

278 DIRECTIONS IN ENGINEERING RESEARCH National Science Foundation. Trends and Opportunities in Materials Re- search. Report of the Materials Research Advisory Committee, Division of Materials Research. Washington, DC: National Science Foundation, 1984. U.S. Department of Commerce. A Competitive Assessment of the U.S. Advanced Ceramics Industry (NTIS-PB84-162288~. Industry Analysis Division, Office of Industrial Assessment. Washington, DC: National Technical Information Service, March 1984.

MATERIALS SYSTEMS Appendix Responses to the Engineering Research Board's Request for Assistance from Universities Professional Societies, and Federal Agencies and Laboratories 279 Requests for assistance were sent by the Engineering Research Board to a number of universities, recipients of Presidential Young Investigator Awards, professional societies, and federal agencies and laboratories in order to obtain a broader view of engineering research opportunities, research policy needs, and the health of the research community. Some of the responses included comments on engineering research aspects of materials systems research; these were reviewed by this panel to aid in its decision-making process. The pane! found the responses most helpful and wishes that it were possible to individually thank all those who took the time to make their views known. Because that is not practical, we hope nevertheless that this small acknowledgment might convey our gratitude. Responses on aspects of materials systems research were re- ceived from individuals representing 46 different organizations, listed in Table A: 22 universities (including 9 represented by re- cipients of NSF Presidential Young Investigator Awards), 11 pro- fessional organizations, and 13 federal agencies or laboratories. Some comments covered specific aspects of the panel's scope of activities, whereas others provided input on a variety of subjects. Although most of the responses addressed priority research needs, several respondents did reflect on policy issues. Many of the research needs and issues of policy and health addressed by the respondents were similar to those noted by pane} members. The broadened! perspective provided by the responses to the survey was most beneficial in the panel's deliberations.

280 DIRECTIONS IN ENGINEERING RESEARCH TABLE A-1 Organizations Responding to Information Requests Relevant to Materials Systems Research UNIVERSITIES Clarkson University Lehigh University Massachusetts Institute of Technology North Carolina State University Northwestern University Ohio State University Oregon State University Rensselaer Polytechnic Institute State University of New York, Buffalo Texas A&M University University of Connecticut University of California, D avis University of California, Los Angeles University of Florida University of Illinois—Urbana/Champaign University of Kansas University of Michigan University of Oklahoma University of Pennsylvania University of Texas at Austin University of Utah Washington University at St. Louis PROFESSIONAL ORGANIZATIONS American Chemical Society American Institute of Aeronautics and Astronautics American Institute of Chemical Engineers American Society of Civil Engineers American Society of Mechanical Engineers Council for Chemical Research Institute of Electrical and Electronic Engineers Institute of Industrial Engineers Industrial Research Institute Society of Automotive Engineers Society of Engineering Science, Inc. AGENCIES AND LABORATORIES Air Force Institute of Technology Air Force Office of Scientific Research Army Materials and Mechanical Research Center Army Research Office Brookhaven National Laboratory Lawrence Livermore National Laboratory NASA Jet Propulsion Laboratory NASA Langley Research Center NASA Lewis Research Center Naval Research Laboratory Office of Naval Research Oak Ridge National Laboratory Sandia National Laboratory