Models/modeling with supercomputers

The development of supercomputers has made it possible to apply the idea of potential flow outside the boundary layer and viscous flow inside the boundary layer to yery complex structures and situations. Currently commercial aircraft designs are. tested with supercomputers using the combination of potential flow and jthe boundary layer, much more quickly and cheaply than they could be tested with models in wind tunnels or in flight tests [8]. 

The Linked Atom Least Squares (LALS) program (18,19) by Arnott and coauthors is used at several institutions, including Purdue, where it was developed, and Case Western Reserve University, where it was used by Gardner and Blackwell. It is best Implemented on a large computer the Control Data Corporation supercomputer at Purdue handles it very rapidly. Setting up a new model with LALS is more complicated than with the other programs, but LALS appears to provide realistic molecular flexibility. 

Last but not least, the reactor-level modeling is expected to further interact with 3D flow modeling in the exhaust system, especially with respect to flow homogeneity. State-of-the-art 3D CFD exhaust system models are able to provide useftil conclusions on the flow and NH3 uniformity at steady-state operating modes. However, the computational effort required for transient cycle calculations remains formidable even with supercomputers. Couphng of 3D flow calculations with transient reactor models remains one of the main challenges for exhaust system modelers in the coming years. 

Correlated calculations, such as configuration interaction, DFT, MPn, and coupled cluster calculations, can be used to model small organic molecules with high-end workstations or supercomputers. These are some of the most accurate calculations done routinely. Correlation is not usually required for qualitative or even quantitative results for organic molecules. It is needed to obtain high-accuracy quantitative results. 

This kind of simulation requires massive computer power, and much of it is done on so-called supercomputers . This is a reason why much recent research of this kind has been done at Los Alamos. In a survey of research in the American national laboratories, the then director of the Los Alamos laboratory, Siegfried Hecker (1990) explains that the laboratory has worked closely with all supercomputer vendors over the years, typically receiving the serial No. I machine for each successive model . He goes on to exemplify the kinds of problems in materials science that these extremely powerful machines can handle. 

Successful systems design and fabrication depend on understanding the connections between microscale phenomena and macroscale behavior of materials. For example, with sufficient insight into intermolecular interactions, appropriate models, and the computational power of supercomputers, it may be possible to predict changes in macromolecular configurations when loads are imposed on polymers or changes in the properties of a material as a result of... 

Before the advent of modem computer-aided mathematics, most mathematical models of real chemical processes were so idealized that they had severely limited utility— being reduced to one dimerrsion and a few variables, or Unearized, or limited to simplified variability of parameters. The increased availability of supercomputers along with progress in computational mathematics and numerical functional analysis is revolutionizing the way in which chemical engineers approach the theory and engineering of chemical processes. The means are at hand to model process physics and chenustry from the... 

To address these challenges, chemical engineers will need state-of-the-art analytical instruments, particularly those that can provide information about microstmctures for sizes down to atomic dimensions, surface properties in the presence of bulk fluids, and dynamic processes with time constants of less than a nanosecond. It will also be essential that chemical engineers become familiar with modem theoretical concepts of surface physics and chemistry, colloid physical chemistry, and rheology, particrrlarly as it apphes to free surface flow and flow near solid bormdaries. The application of theoretical concepts to rmderstanding the factors controlling surface properties and the evaluation of complex process models will require access to supercomputers. 

The program can solve both steady-state problems as well as time-dependent problems, and has provisions for both linear and nonlinear problems. The boundary conditions and material properties can vary with time, temperature, and position. The property variation with position can be a straight line function or or a series of connected straight line functions. User-written Fortran subroutines can be used to implement more exotic changes of boundary conditions, material properties, or to model control systems. The program has been implemented on MS DOS microcomputers, VAX computers, and CRAY supercomputers. The present work used the MS DOS microcomputer implementation. 

Supercomputers become more and more useful, and the Insights they can generate become more and more unique, as the complexity of the system modelled Is Increased. Thus Interfaclal phenomena are a very natural field for supercomputation. In addition to the examples In this volume It may be useful to mention the work of Llnse on liquid-liquid benzene-water interfaces, which he studied with 504 H2O molecules, 144 CgHg molecules, and 3700 Interaction sites. He generated over 50 million configurations In 56 hours on a Cray-lA, and he was able to quantitatively assess the sharpness of the Interfaclal density gradient, which Is very hard to probe experimentally. Similarly Spohr and Helnzlnger have studied orientational polarization of H2O molecules at a metallic Interface, which is also hard to probe experimentally. 

Mathematical models require computation to secure concrete predictions. Successes in relatively simple cases spurs interest in more complex situations. Somewhat specialized computer hardware and software have emerged in response to these demands. Examples are the high-end processors with vector architecture, such as the Cray series, the CDC Cyber 205, and the recently announced IBM 3090 with vector attachment. When a computation can effectively utilize vector architecture, such machines will out-perform even the most powerful conventional scalar machine by a substantial margin. Such performance has given rise to the term supercomputer. ... 

Supercomputers together with software to use them effectively, provide capabilities to solve important problems that combine both scientific complexity and engineering practicality. The supercomputing environment allows a level of sophistication in physical models that is not possible analytically and is prohibitively expensive without supercomputing speed and storage capabilities. 

The climate is an important aspect of the environment, an aspect that interacts strongly with the composition of the ocean and atmosphere. This interaction works in two ways Climate is influenced by composition through the greenhouse effect, and climate also influences composition through its effect on reaction rates, particularly on weathering and the flux of dissolved constituents into the sea. Full-scale climate models are exceedingly complicated and can run only on supercomputers. But here I shall demonstrate how one aspect of the climate system—average tern-... 

Water Potentials. The ST2 (23), MCY (24), and CF (2J5) potentials are computationally tractable and accurate models for two-body water-water interaction potentials. The ST2, MCY and CF models have five, four, and three interaction sites and have four, three and three charge centers, respectively. Neither the ST2 nor the MCY potentials allow OH or HH distances to vary, whereas bond lengths are flexible with the CF model. While both the ST2 and CF potentials are empirical models, the MCY potential is derived from ab initio configuration interaction molecular orbital methods (24) using many geometrical arrangements of water dimers. The MCY+CC+DC water-water potential (28) is a recent modification of the MCY potential which allows four body interactions to be evaluated. In comparison to the two-body potentials described above, the MCY+CC+DC potential requires a supercomputer or array processor in order to be computationally feasible. Therefore, the ST2, MCY and CF potentials are generally more economical to use than the MCY+CC+DC potential. 

---