Balanced Truncation Model Reduction of Large-Scale Dense Systems on Parallel Computers

Model reduction is an area of fundamental importance in many modeling and control applications. In this paper we analyze the use of parallel computing in model reduction methods based on balanced truncation of large-scale dense systems. The methods require the computation of the Gramians of a linear-time invariant system. Using a sign function-based solver for computing full-rank factors of the Gramians yields some favorable computational aspects in the subsequent computation of the reduced-order model, particularly for non-minimal systems. As sign function-based computations only require efficient implementations of basic linear algebra operations readily available, e.g., in the BLAS, LAPACK, and ScaLAPACK, good performance of the resulting algorithms on parallel computers is to be expected. Our experimental results on a PC cluster show the performance and scalability of the parallel implementation.     

Computer simulation and replica Ornstein—Zernike integral equation studies of a hard-sphere dipolar fluid adsorbed into disordered porous media

By means of extensive grand canonical Monte Carlo simulations and replica Ornstein-Zernike integral equation calculations, we explore the thermodynamics and dielectric behaviour of a dipolar fluid confined in disordered matrices. Different matrix topologies are modelled using, on the one hand, quenched hard-sphere configurations and, on the other, randomly positioned spheres. This illustrates the influence of the pore size and shape on the properties of the adsorbed fluid. For the same purpose, various sizes of the matrix particles have been considered. The integral equation calculations in the hypernetted chain approximation agree quantitatively with the simulation results. As in other studies on quenched disorder, one observes that the effect of confinement when considering hard-sphere matrices is rather limited, exhibiting however a tendency to facilitate the transition to ferroelectric states, as well as favouring the gas-liquid transition. Additionally, one observes that the sensitivity of the fluid properties to changes in the size of the matrix particles is considerably larger when these are randomly positioned.     

Study of the ferromagnetic transition in a positionally frozen Heisenberg spin system

Using extensive Monte Carlo simulations with both particle and cluster orientational moves, in conjunction with finite size scaling and histogram reweighting techniques, we have determined the Curie temperature for two models of positionally frozen Heisenberg spin systems: a system with spatial correlations corresponding to a hard sphere fluid and a spatially random system. We find that the results for the positionally frozen hard sphere Heisenberg system are fairly similar to those previously obtained for the Heisenberg spin fluid and quantitatively agree with the mean field theory estimates. The random system undergoes the ferromagnetic transition at a higher temperature since the lack of core repulsion increases the spin correlations. In this case however the mean field theory overestimates by far the critical temperature.     

CARBOXYLATED ZINC-PHTHALOCYANINE, INFLUENCE OF DIMERIZATION ON THE SPECTROSCOPIC PROPERTIES. AN ABSORPTION, EMISSION, AND THERMAL LENSING STUDY

Monomer and dimer absorption and emission spectra, and dimerization constants are reported for the diamide of the zinc-tetracarboxyphthalocyanine in pure dimethylformamide (DMF) and in H₂O/DMF mixtures at room temperature. The dimerization constant increases steadily with the water content. The monomer absorption Q-band is insensitive to the solvent composition, whereas dimer spectra show great variations with the water content. Stationary emission measurements show that fluorescence originates exclusively from the monomers. The fluorescence spectrum as well as its absolute fluorescence quantum yield, measured by steady-state thermal lensing, are also insensitive to the solvent composition. The thermal lensing method is discussed for the case of two absorbing species in equilibrium.     

Reexamining the phase diagram of the Gay—Berne fluid

This work reexamines and updates earlier investigations on the phase behaviour of the Gay-Berne liquid crystal model, concentrating on the effect of varying temperature. Constant volume and constant pressure Monte Carlo simulations are combined for systems consisting of N = 500 molecules along different isotherms over the reduced temperature range 0.60 ≤ T ≤ 1.25. As in previous simulation studies of the model, the study identifies nematic and smectic B phases on compressing the isotropic fluid, the particular phase sequence depending on temperature. The nematic phase is found to be stable with respect to the isotropic phase for reduced temperatures T ≥ 0.75. In the temperature range 0.75 ≤ T ≤ 1.25, the phase boundaries of the isotropic-nematic transition are obtained by computing the Helmholtz free energy of both phases from thermodynamic integration. From the simulation data, the relative volume change at the isotropic-nematic transition is about 2%, and this value appears to be rather insensitive to changes in temperature. On compressing the nematic phase, the Gay-Berne fluid undergoes a strong first-order transition to the smectic B phase. This transition is studied by using constant pressure simulation, and the coexistence properties are estimated from the limits of mechanical stability of the nematic phase. Larger relative volume changes are found at the transition than those suggested by previous studies, with typical values increasing up to 10.5% as the temperature is decreased. The results are consistent with the existence of strong coupling between nematic and smectic order parameters. For temperatures T ≤ 0.70 the nematic phase is no longer stable, and the phase sequence isotropic-smectic B is observed. Therefore, the Gay-Berne model exhibits an isotropic-nematic-smectic B triple point. Extrapolating the present simulation data, this triple point is located approximately at reduced temperature T_INB ? 0.70 and reduced pressure P_INB ? 1.825.     

Monitoring foam coarsening using a computer optical mouse as a dynamic laser speckle measurement sensor

In this paper, we present an experimental approach to track coarsening process of foam using a computer optical mouse as a dynamic laser speckle measurement sensor. The dynamics of foam coarsening and rearrangement events cause changes in the intensity of laser speckle backscattered from the foam. A strong negative correlation between the average speed of the cursor and the evolution of bubble diameter was found. We used microscopic images to demonstrate that decrease in speed is related to increase in bubble size. The proposed set-up is not very expensive, is highly portable and can be used in laboratory measurements of dynamics in other kinds of opaque materials.     

Polarization effects in liquid metals and charge stabilized colloidal dispersions

The induced polarization contributions to the effective interionic potentials are derived for liquid metals, liquid alloys and charge stabilized colloidal dispersions. These contributions have been obtained within a perturbative scheme that includes up to three-body terms. The formalism is illustrated by calculating the static structure factor S(k) of liquid Ga near melting. It is found that the main effects due to the induced polarization in S(k) are first lowering the height of the principal diffraction peak, and second sharpening the subsidiary peak present in liquid Ga.