|
SUMMARIES OF FY 1996
RESEARCH IN THE CHEMICAL
SCIENCES |
Chemical Engineering Sciences:
National Laboratory Projects
Lawrence Berkeley National Laboratory
University of California
Berkeley, CA 94720
Chemical Sciences Division
Molecular Thermodynamics for Phase Equilibria in Mixtures
Phase equilibria are required for design of efficient separation processes (e.g., polymer
devolatilization or selective precipitation of a target protein from an aqueous protein mixture) in
the chemical and related industries and for development of new chemical products and materials.
In this context, "efficient" refers to optimum use of raw materials and to conservation
of energy. Since the variety of technologically important fluid mixtures is extremely large, it is not
possible to obtain all equilibria from experiment. Therefore, the objective of this research is the
development of molecular thermodynamics for interpretation and correlation of reliable
experimental data toward confident prediction of phase equilibria for chemical process and
product design. The correlations are expressed through semi-theoretical, physicochemical models
based on statistical and classical thermodynamics and on our own as well as published
experimental data. Particular attention is given to those materials that may be useful for innovative
low-energy-consuming separation processes such as polymers, gels, and polyelectrolyte systems
with applications in biotechnology. However, attention is also devoted to conventional materials
for applications in fuel technology, for recovery of salts from large-scale gas-scrubbing processes,
and for proteins used in nutrition or pharmaceutics. Development of molecular thermodynamics
calls for a combination of theoretical, computational, and experimental work. Further, it requires
simultaneous awareness of progress in molecular science and of realistic requirements for
engineering application.
Lawrence Berkeley National Laboratory
University of California
Berkeley, CA 94720
Energy and Environment Division
Turbulent Combustion
The objective of this program is to investigate, primarily experimentally, the interaction and
coupling between turbulence and combustion. These turbulent combustion processes are
characterized by scalar and velocity fluctuations with time and length scales spanning several
orders of magnitude. Our approach is to gain a fundamental understanding through detailed
investigation of idealized laboratory flames with flow and flame geometries amenable to
theoretical analysis. These laboratory flames are accessible to detailed interrogation by laser
diagnostics. The programmatic emphasis is on gaining a physical understanding of the coupled
effects of turbulence and flame geometry on heat release, turbulent burning rate, stability limits
and flame extinction. Such knowledge is essential for the development of turbulent combustion
theories. A primary and continuing effort is the investigation of flames with moderate turbulence
under which the chemical reaction rates are high compared to those of turbulence. The burning
rate can be inferred from the flame wrinkles scales for direct comparison to turbulence scales. Our
current focus is on investigating the relative effects of shear and non-shear turbulence. The
research on moderate turbulent flames forms the foundation for the investigation of flames with
intense turbulence. Intense turbulent flames are closer simulations of the combustion processes in
practical systems where turbulence may alter the reaction rates, cause flame quenching and affect
the formation of pollutants. The use of planar laser induced fluorescence techniques to study
intense turbulent flames is our current emphasis.
Los Alamos National Laboratory
University of California
Los Alamos, NM 87545
Theoretical Division
Thermophysical Properties of Mixtures
| Investigator(s) |
Erpenbeck, J.J.
|
$50,000 |
| Phone | 505-667-7195 |
| E-mail |
jje@lanl.gov |
The thermophysical properties of mixtures of particles interacting through simple interaction
potentials are evaluated, using both equilibrium Monte Carlo and molecular dynamics, as well as
nonequilibrium molecular dynamics. The properties under investigation include the equation of
state and transport properties, including mutual and self diffusion, shear and bulk viscosity,
thermal conductivity, and thermal diffusion. Fundamental questions arising in the theory of fluids
and fluid mixtures are addressed where numerical "experiments" are appropriate.
Current research is concentrated on the phase diagram of hard-sphere mixtures having diameter
ratios no greater than 0.2 and compositions in which the molecular volume of the two species are
similar. Theoretical predictions of fluid phase demixing have been made for such mixtures in
recent years, contradicting the conventional understanding arising from the Lebowitz solution to
the Percus-Yevick integral equation. The Gibbs ensemble Monte Carlo method permits the study
of the coexistence of disordered phases, but current implementations are inadequate to overcome
the ergodic difficulties attendant upon the disparate hard-sphere interactions. Advanced methods
have been development to permit the study of this mixture as well as others having large
disparities in the interactions. For the binary mixture of hard spheres having a diameter ratio of
0.2 and a mole fraction of 0.008 for the large spheres, calculations in the constant pressure
ensemble at a reduced pressure, 
=pv0/kT, (where p is the pressure, T is the temperature, k is
Boltzmann's constant, v0=(x1
12+x222),
xi and i are the mole fraction and diameter of species i) of 6 exhibit
demixing for systems of 5832 and 8000 particles but preliminary results for 10648 particles shows
no evidence thereof. Additional calculations have been initiated for =8. While all these calculations approach equilibrium very slowly, the
number of large particles in each subsystem rapidly becomes essentially fixed. Nonetheless, the new
algorithm still enables small fluctuations in these numbers, leading one to believe the Monte Carlo
procedure to be ergodic, i.e. to explore all important regions of configuration and composition
space.
National Institute for Petroleum and Energy Research
Bartlesville, OK 74005
Department of Fuels Research
Thermodynamic Properties for Polycyclic Systems by Noncalorimetric Methods
| Investigator(s) |
Steele, W.V.; Klots, T.D.
|
$194,000 |
| Phone | 918-337-4210 |
| E-mail |
bsteele@bpo.gov |
This project develops thermodynamic properties for petroleum processing polycyclic molecules
via assigned spectra and statistical mechanics. Polycyclics studied fall into two categories, a rigid
class built of ring systems that contain some degree of aromaticity, such as furan and indole, and a
second class that differs by hydrogenation of a single ring bond, e.g., 2,5-dihydrofuran and
indoline, which increases ring flexibility. The experimental work entails measurement of
vibrational frequencies by the recording of infrared and Raman spectra, while theoretical models
are sought to assist and verify the assignment of fundamentals and evaluation of thermodynamic
data. The primary thrust during the next two years is to simplify the vibrational frequency
calculations by incorporating the empirical results from the past three years and also to examine
improved modeling methods. This will be accomplished firstly by vertically integrating the many
steps in the presently used semi-empirical scaled quantum mechanical methodology and secondly,
by launching full ab initio techniques to generate the initial vibrational force constant matrix, given
the anticipated acessibility of higher speed computers. Highlights of the past year's research
include: (1) A series of papers published on the vibrational spectra and thermodynamics of indene
and derivatives: In part I, Spectrochimica Acta 1995, 51A, 1255, compelling justification of
the vibrational assignment for indole, benzofuran, benzothiophene, benzoxazole, and
benzothiazole was provided using the scaled AM1 force field technique; in part II, Spectrochimica
Acta 1995, 51A, 1273, the spectra, assignment, and vapor-liquid wavenumber shifts for
benzothiophene and benzothiazole were discussed; in part III, Spectrochimica Acta 1995, 51A, 1291,
the spectra and assignment for indole, benzofuran and benzoxazole were described, and
dramatically improved ideal-gas entropies were demonstrated for benzoxazole. (2) The paper
"Vibrational Assignment and Analysis for 2,3-Dihydrofuran and 2,5-Dihydrofuran"
appeared in print, Spectrochimica Acta 1994, 50A, 1725.
Sandia National Laboratories, California
Livermore, CA 94551
Combustion Research Facility
Analysis of Turbulent Reacting Flows
| Investigator(s) |
Ashurst, W.T.; Kerstein, A.R.
|
$414,000 |
| Phone | 510-294-2274 |
| E-mail |
ashurs@ca.sandia.gov |
The goal of this project is to develop numerical simulation techniques for the understanding of
reacting turbulent flows. The objective is to show the mechanisms of turbulent mixing and
reaction. A model of premixed flame propagation along the core of a vortex indicates that the
generation of new vorticity may cause significant enhancement of the flame speed. This generation
occurs when the pressure gradient created by the vortex swirling motion is normal to the density
gradient across the flame, known as the baroclinic effect. Previous experiments with a burning
vortex ring showed a factor of twenty enhancement in the flame speed. This behavior may also
occur within the intense vortex structures observed in turbulent flows. The linear-eddy mixing
model, unique in its representation of the distinct influences of convective stirring and molecular
diffusion, is being used to investigate the spatial correlations of chemical species in turbulent
reacting flows.
Last updated by Harry J. Dewey, (hd@lanl.gov) on December 16,
1996.