SCI Publications

The thermal decomposition of hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) has been studied by thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). Activation energies as a function of the extent of conversion, α, have been determined by model-free isoconversional analysis of these data. In open pans, evaporation is a prevalent process with an activation energy of ∼100 kJ mol^-1. Confining the system in either a pierced pan or a closed pan promotes liquid state decomposition of RDX that occurs with an activation energy of ∼200 kJ mol^-1, which suggests scission of an N-N bond as the primary decomposition step. In such a confined environment, gas phase decomposition is a competing channel with an activation energy estimated to be ∼140 kJ mol^-1. In a closed pan, RDX generates a heat release of ∼500 kJ mol^-1 that is independent of both the heating rate, β, and the mass.

By partitioning the total stresses in a damaged composite into either mechanical and residual stresses or into initial and pertubation stresses, it was possible to derive several exact results for the energy release rate due to crack growth. These general results automatically include the effects of residual stresses, traction-loaded cracks, and imperfect interfaces. By considering approximate solutions based on admissible stress states and admissible strain states, it was possible to derive rigorous upper and lower bounds to the energy release rate for crack growth. Two examples of using these equations are mode I fracture in adhesive double cantilever beam specimens and analysis of microcracking in composite laminates.

Multigrid methods for solving the steady-state incompressible Navier-Stokes equations require an appropriate smoother and coarse grid solution strategy to be effective. Classical pressure-correction methods, such as SIMPLE and SIMPLER, are widely used as solvers in engineering analysis codes, but can also be used as effective multigrid smoothers. An inexact Newton method preconditioned by a linear multigrid method with a pressurecorrection smoother can serve as a coarse grid solver. A hybrid nonlinear multigrid scheme based on combinations of these components is described. A standard benchmark problem is used to demonstrate the effectiveness of SIMPLER smoothing and the impact an inexact Newton coarse grid solver has on the resulting nonlinear multigrid scheme.

Neutron scattering and computer simulations are powerful tools for studying structural and dynamical properties of condensed matter systems in general and of polymer melts in particular. When neutron scattering studies and quantitative atomistic molecular dynamics simulations of the same material are combined, synergy between the methods can result in exciting new insights into polymer melts not obtainable from either method separately. We present here an overview of our recent efforts to combine neutron scattering and atomistic simulations in the study of melt dynamics of polyethylene and polybutadiene. Looking at polymer segmental motion on a picosecond time scale, we show how atomistic simulations can be used to identify molecular motions giving rise to relaxation processes observed in experimental dynamic susceptibility spectra. Examining larger length and longer time scale polymer dynamics involving chain self-diffusion and overall conformational relaxation, we show how simulation results can motivate experiment and how combined results of scattering and simulation can be used to critically test theories that attempt to describe melt dynamics of short polymer chains.

A concept and algorithm are presented for non-iterative robust estimation of piecewise smooth curves of maximal edge strength in small image windows-typically 8/spl times/8 to 32/spl times/32. This boundary-estimation algorithm has the nice properties that it uses all the data in the window and thus can find locally weak boundaries embedded in noise or texture and boundaries when there are more than two regions to be segmented in a window; it does not require step edges-but handles ramp edges well. The curve-estimates found are among the level sets of a dth degree polynomial fit to "suitable" weightings of the image gradient vector at each pixel in the image window. Since the polynomial fitting is linear least squares, the computation to this point is very fast. Level sets then chosen to be appropriate boundary curves are those having the highest differences in average gray level in regions to either side. This computation is also fast. The boundary curves and segmented regions found are suitable for all purposes but especially for indexing using algebraic curve invariants in this form.

We present a new method called Reaction Class Transition State Theory (RC-TST) for estimating thermal rate constants of a large number of reactions in a class. This method is based on the transition state theory framework within the reaction class approach. Thermal rate constants of a given reaction in a class relative to those of its principal reaction can be efficiently predicted from only its differential barrier height and reaction energy. Such requirements are much less than what is needed by the conventional TST method. Furthermore, we have shown that the differential energetic information can be calculated at a relatively low level of theory. No frequency calculation beyond those of the principal reaction is required for this theory. The new theory was applied to a number of hydrogen abstraction reactions. Excellent agreement with experimental data shows that the RC-TST method can be very useful in design of fundamental kinetic models of complex reactions.

We present a new practical computational methodology for predicting thermal rate constants of reactions involving large molecules or a large number of elementary reactions in the same class. This methodology combines the integrated molecular orbital+molecular orbital (IMOMO) approach with our recently proposed reaction class models for tunneling. With the new methodology, we show that it is possible to significantly reduce the computational cost by several orders of magnitude while compromising the accuracy in the predicted rate constants by less than 40% over a wide range of temperatures. Another important result is that the computational cost increases only slightly as the system size increases.