SCI Publications

Improvements in hardware have recently made interactive ray tracing practical for some applications. However, when the scene complexity or rendering algorithm cost is high, the frame rate is too low in practice. Researchers have attempted to solve this problem by caching results from ray tracing and using these results in multiple frames via reprojection. However, the reprojection can become too slow when the number of samples that are reused is high, so previous systems have been limited to small images or a sparse set of computed pixels. To overcome this problem we introduce techniques to perform this reprojection in a scalable fashion on multiple processors.

We have performed molecular dynamics (MD) simulations of a melt of 1,4-polybutadiene (PBD, 1622 Da) over the temperature range 400-273 K. ¹³C NMR spin-lattice relaxation times (T₁) and nuclear Overhauser enhancement (NOE) values have been measured from 357 to 272 K for 12 different resonances. The T₁ and NOE values obtained from simulation C-H vector P₂(t) orientational autocorrelation functions were in good agreement with experiment over the entire temperature range. Analysis of conformational dynamics from MD simulations revealed that T₁ depends much less strongly on the local chain microstructure than does the mean conformational transition time. Spin−lattice relaxation for a given nucleus could not be associated with the dynamics of any particular dihedral; instead, spin−lattice relaxation occurs as the result of multiple conformational events. However, a much closer correspondence was found between torsional autocorrelation times and the C-H vector P₂(t) autocorrelation times upon which T₁ depends. Both processes exhibited stronger than exponential slowing with decreasing temperature. The non-Arrhenius temperature dependences of these relaxation times as well as the stretched-exponential character of the autocorrelation functions themselves were found to be consistent with increasing dynamic heterogeneity in conformational transition rates with decreasing temperature.

We have investigated chain dynamics of an unentangled polybutadiene melt via molecular dynamics simulations and neutronspin echo experiments. Good short-time statistics allows for the first experimental confirmation of subdiffusive motion of polymer chains for times less than the Rouse time (T_R) confirming behavior in this regime observed in simulations. Analysis of simulation trajectories obtained over several Rouse times reveals non-Gaussian segmental displacements for all time and length scales. These results, particularly non-Gaussian displacements on large time- and length scales, demonstrate the importance of intermolecular correlations on chain dynamics. Rouse-type analytical models fail to account for this non-Gaussianity leading to large deviations between the experimental dynamic structure factor and model predictions.

Soot samples, including the associated organics, produced from an Illinois No. 6 coal (five samples) and two model compounds, biphenyl (three samples) and pyrene (two samples), have been studied by 13C NMR methods. The coal soot data served as a guide to selection of the temperature range that would be most fruitful for investigation of the evolution of aerosols composed of soot and tars that are generated from model compounds. The evolution of the different materials in the gas phase followed different paths. The coal derived soots exhibited loss of aliphatic and oxygen functional groups prior to significant growth in average aromatic cluster size. Between 1410 and 1530 K, line broadening occurs in the aromatic band, which appears to have a Lorentzian component that is observable at the lower temperature and is quite pronounced at the higher temperature. The data indicate that the average aromatic cluster size (the number of carbon atoms in an aromatic ring system where the rings are connected through aromatic bridgehead carbon atoms) may be as large as 80-90 carbons/cluster. The data obtained for the biphenyl samples exhibit a different path for pyrolysis and soot growth. A significant amount of ring opening reactions occurs, followed by major structural rearrangements, after the initial ring opening and hydrogen transfer phase. The cluster size not only grows significantly, but the crosslinking structure also increases, indicating that soot growth in biphenyl soots consists not only of cluster size growth but also cluster cross-linking. The evolution of pyrene aerosol samples follows still another path. Little evidence is noted for ring opening reactions. Major ring growth has not occurred at 1410 K but cross-linking reactions are noted, indicating the formation of dimer/trimer structures. Although a significant amount of ring growth is noted, the data are inconclusive regarding the mechanism for ring growth in the pyrene aerosols between 1410 and 1460 K.

The Soot Model Development Web Site (SMDWS) is a project to develop collaborative technologies serving combustion researchers in DOE national laboratories, academia and research institutions throughout the world. The result is a shared forum for problem exploration in combustion research. Researchers collaborate over the Internet using SMDWS tools, which include a server for executing combustion models, web-accessible data storage for sharing experimental and modeling data, and graphical visualization of combustion effects. In this paper the authors describe the current status of the SMDWS project, as well as continuing goals for enhanced functionality, modes of collaboration, and community building.

Formation pathways for high-molecular-mass compound growth are presented, showing why reactions between aromatic moieties are needed to explain recent experimental findings. These reactions are then analyzed by using quantum mechanical density functional methods. A sequence of chemical reactions between aromatic compounds (e.g., phenyl) and compounds containing conjugated double bonds (e.g., acenaphthylene) was studied in detail. The sequence begins with the H-abstraction from acenaphthylene to produce the corresponding radical, which then furnishes higher aromatics through either a two-step radical-molecule reaction or a direct radical-radical addition to another aromatic radical. Iteration of this mechanism followed by rearrangement of the carbon framework ultimately leads to high-molecular-mass compounds. This sequence can be repeated for the formation of high-molecular-mass compounds. The distinguishing features of the proposed model lie in the chemical specificity of the routes considered. The aromatic radical attacks the double bond of five-membered-ring polycyclic aromatic hydrocarbons. This involves specific compounds that are exceptional soot precursors as they form resonantly stabilized radical intermediates, relieving part of the large strain in the five-membered rings by formation of linear aggregates.

Thermogravimetry has been used to study the kinetics of the thermal dissociation of solid and liquid ammonium nitrate. Model-fitting and model-free kinetic methods have been applied to the sets of isothermal and nonisothermal measurements to derive kinetic characteristics of the processes. The application of the model-fitting method to the isothermal data has demonstrated that both solid- and liquid-phase kinetics are characterized by a single activation energy of ∼90 kJ mol^-1 and by the model of a contracting cylinder. A model-free isoconversional method has also been applied to isothermal and nonisothermal data and has yielded an activation energy of ∼90 kJ mol^-1, which is essentially independent of the extent of conversion. The obtained kinetic characteristics have been assigned to the process of dissociative sublimation/vaporization.