ALBERT

Description: The issues of industrial productivity and economic competitiveness are of major significance in the U.S. at present. By advancing the science of design, and by creating a broad computer-based methodology for automating the design of artifacts and of industrial processes, we can attain dramatic improvements in productivity. It is our thesis that developments in computer science, especially in Artificial Intelligence (AI) and in related areas of advanced computing, provide us with a unique opportunity to push beyond the present level of computer aided automation technology and to attain substantial advances in the understanding and mechanization of design processes. To attain these goals, we need to build on top of the present state of AI, and to accelerate research and development in areas that are especially relevant to design problems of realistic complexity. We propose an approach to the special challenges in this area, which combines 'core work' in AI with the development of systems for handling significant design tasks. We discuss the general nature of design problems, the scientific issues involved in studying them with the help of AI approaches, and the methodological/technical issues that one must face in developing AI systems for handling advanced design tasks. Looking at basic work in AI from the perspective of design automation, we identify a number of research problems that need special attention. These include finding solution methods for handling multiple interacting goals, formation problems, problem decompositions, and redesign problems; choosing representations for design problems with emphasis on the concept of a design record; and developing approaches for the acquisition and structuring of domain knowledge with emphasis on finding useful approximations to domain theories. Progress in handling these research problems will have major impact both on our understanding of design processes and their automation, and also on several fundamental questions that are of intrinsic concern to AI. We present examples of current AI work on specific design tasks, and discuss new directions of research, both as extensions of current work and in the context of new design tasks where domain knowledge is either intractable or incomplete. The domains discussed include Digital Circuit Design, Mechanical Design of Rotational Transmissions, Design of Computer Architectures, Marine Design, Aircraft Design, and Design of Chemical Processes and Materials. Work in these domains is significant on technical grounds, and it is also important for economic and policy reasons.

Description: Millisecond pulsars are galactic objects that exhibit a very stable spinning period. Several tens of these celestial clocks have now been discovered, which opens the possibility that an average time scale may be deduced through a long-term stability algorithm. Such an ensemble average makes it possible to reduce the level of the instabilities originating from the pulsars or from other sources of noise, which are unknown but independent. The basis for such an algorithm is presented and applied to real pulsar data. It is shown that pulsar time could shortly become more stable than the present atomic time, for averaging times of a few years. Pulsar time can also be used as a flywheel to maintain the accuracy of atomic time in case of temporary failure of the primary standards, or to transfer the improved accuracy of future standards back to the present.

Description: Conventional sandwich structure fabrication methods are labor intensive and high in cost. A low cost method is needed to produce lightweight sandwich structures. Sundstrand has developed a series of in situ composite fabrication methods in which the raw materials (skin and core materials) are placed in a closed mold, and the component is produced in one heating cycle. Internal pressure is generated by chemical agents during the thermal cycles, which consolidates the skins and produces the foam core. The finished part is a net-shape composite sandwich structure with skins and a foamed core. The in situ process reduces cost by eliminating several secondary operations that are used in conventional fabrication methods. Further, a strong molecular bond is produced between the core and skin, which eliminates adhesive bonding and prevents a weak bond section in the sandwich structure. In this investigation, we evaluated the feasibility of the in situ process using thermoset materials currently under consideration for commercial airplane fuselage applications, such as keel sections. The materials used were Hercules 855340 toughened epoxy resin in both liquid and powder forms, and 3M Scotchply PR500 resin, manufactured by 3M Corporation, in powder form. We successfully foamed these resins and produced experimental panels with AS-4/855340 Hercules prepreg skins. Chopped fibers were added to the core to increase performance of the foam. Mechanical property testing on these panels showed properties competitive with other foams. Additional experiments are required to optimize the in situ foam core sandwiches for specific properties and applications.

Description: Research and development in light-weight, high-temperature composite materials for ultrahigh-bypass engines to be used in high-speed civil transport/rotocraft is presented. It is noted that the expected benefits to be attained by this R&D include weight reduction, lowered fuel consumption, and lower direct operating costs. A major effort underway in this area is the Advanced High Temperature Engine Materials Technology Program (HITEMP) of NASA, which focuses on providing revolutionary high-temperature composite materials: to 425 C (800 F) for polymer-matrix composites (PMCs), to 1250 C (2280 F) for metal-matrix/intermetallic-matrix composites (MMCs/IMCs), and to as high as 1650 C (3000 F) for ceramic-matrix composites (CMCs). Analytical modeling is being used to investigate the structural behavior of these advanced materials in six distinct areas: micromechanics, deformation and damage, fatigue, fracture, trade-off studies, and load definition. It is concluded that the development of advanced materials such as high-temperature composites is highly dependent on the availability of high-temperature fibers. The wide range of fiber characteristics needed will require the development of more than one fiber. In general, a candidate fiber should have low density, high strength, high stiffness, a CTE matching the matrix, chemical compatibility with the matrix, environmental stability and appropriate fiber diameter.

Description: Creep-rupture and tensile tests have been used to evaluate thoriated W-wire reinforced Nb-1 percent Zr alloy matrix composites fabricated via arc-spray monotape technique. A significant creep strength enhancement was observed over the unreinforced matrix alloy while matrix integrity was maintained; the fiber/matrix interface phase is noted to be a strong and ductile W/Nb alloy, which is formed due to the mutual solubility of the constituent metals. High strength, toughness, and thermal stability are demonstrated by this material system, which is also resistant to liquid alkali metal corrosion.

Description: The mechanical behavior of continuous fiber reinforced SiC/RBSN composites with strong and weak interface characteristics is evaluated. Both catastrophic and noncatastrophic failures are observed in tensile specimens. Effects of fiber/matrix interface debonding (splitting) parallel to the fibers are discussed. Micromechanical models incorporating residual stresses to calculate the critical matrix cracking strength, ultimate strength and work of pull-out are reviewed and used to predict composite response. Experimental results are compared to analytical predictions.

Description: Previous studies of a single lot of NiAl powder which had been ground under high intensity conditions in liquid nitrogen (cryomilling) indicated that this processing leads to a high strength, elevated temperature NiAl-AlN composite. Because this was the first known example of the use of the reaction milling process to produce a high temperature composite, the reproducibility of this technique was unknown. Two additional lots of NIAl powder and a lot of a Zr-doped NiAl powder have been cryomilled, and analyses indicate that AlN was formed within a NiAl matrix in all three cases. Compression testing between 1200 K and 1400 K has shown that the deformation resistance of these heats is similar to that of the first lot of NiAl-AlN; thus cryomilling can improve the creep resistance of NiAl by a factor of six. Based on this work, it is concluded that cryomilling of NiAl powder to form high temperature, high strength NiAl-AlN composites is a reproducible process.

Description: Reaction of Ni-Al alloys within the beta-NiAl phase with CrB2 was studied at 1473 K as a function of Al concentration in the alloy. Reaction of 49-50 at. pct Al alloys with CrB2 occurred by interdiffusion of Ni into CrB2 and Cr into the alloy without forming a new product phase. On the other hand, a new product phase, rich in Ni and B, formed by the reaction of alloys having Al concentrations 48 at. pct or lower with CrB2. The reaction product was observed both at the CrB2/alloy interface and along the alloy grain boundaries.

Description: We study the transfer of momentum from photons to dust grains to (molecular) gas in the outflow around cool giants (carbon-stars, Mira variables and OH/IR stars) beyond the radius where the dust grains condense. The problem is circular: radiation pressure determines the outflow velocity of the dust and thus also the dust density; on the other hand the dust density determines, via radiative transfer effects, the spectrum of the photons and thus the effective radiation pressure. This circular problem is solved by a rapidly converging iterative procedure. We compare our predictions with observed properties of a large sample of OH/IR stars and of Miras and find a good qualitative and quantitative agreement. We confirm a conclusion by Wood et al. (1993) that very luminous OH/IR stars in the Large Magellanic Cloud (LMC) owe their low outflow velocity to the low dust-to-gas ratio, a consequence of the low metallicity of the LMC. Similarly we consider a sample of about 100 OH/IR stars within 200 pc from the galactic center that has an average asymptotic giant branch (AGB) luminosity and an uncommonly high value of v(sub out); we conclude that these stars are probably very metal rich, perhaps even more than the stars in the Baade window studied by Rich (1990).

Description: Magnetic field data measured by the MAGMA instrument in the Martian magnetotail lobes are compared with the ram pressure of the upstream solar wind observed by the TAUS instrument in the circular orbits of the Phobos 2 spacecraft. High correlation was found between the magnetic field intensity in the Martian magnetotail lobes and the solar wind ram pressure. From this relationship the average flaring angle of the Martian magnetotail was determined as approximately 13 deg, and the average magnetosonic Mach number was estimated as approximately 5. The observed relationship between the Martian magnetotail magnetic field intensity and the solar wind magnetic field reflects the correlation of the solar wind magnetic field to the ram pressure providing a value of approximately 7 for the average Alfvenic Mach number. The flaring angle obtained for the Martian magnetotail was found to be an intermediate value between the flaring angle of the magnetotail of the Earth and that of Venus at comparable distances.