Compression of Weighted Graphs

We propose to compress weighted graphs (networks), motivated by the observation that large networks of social, biological, or other relations can be complex to handle and visualize. In the process also known as graph simplication, nodes and (unweighted) edges are grouped to supernodes and superedges, respectively, to obtain a smaller graph. We propose models and algorithms for weighted graphs. The interpretation (i.e. decompression) of a compressed, weighted graph is that a pair of original nodes is connected by an edge if their supernodes are connected by one, and that the weight of an edge is approximated to be the weight of the superedge. The compression problem now consists of choosing supernodes, superedges, and superedge weights so that the approximation error is minimized while the amount of compression is maximized. In this paper, we formulate this task as the 'simple weighted graph compression problem'. We then propose a much wider class of tasks under the name of 'generalized weighted graph compression problem'. The generalized task extends the optimization to preserve longer-range connectivities between nodes, not just individual edge weights. We study the properties of these problems and propose a range of algorithms to solve them, with dierent balances between complexity and quality of the result. We evaluate the problems and algorithms experimentally on real networks. The results indicate that weighted graphs can be compressed efficiently with relatively little compression error.

Effects of temper level on the dependence of fatigue crack growth threshold and crack closure on the prior austenitic grain size

An attempt has been made to systematically investigate the effects of microstructural parameters, such as the prior austenite grain size (PAGS), in influencing the resistance to fatigue crack growth (FCG) in the near-threshold region under three different temper levels in a quenched and tempered high-strength steel. By austenitizing at various temperatures, the PAGS was varied from about 0.7 to 96 μm. The microstructures with these grain sizes were tempered at 200 °C, 400 °C, and 530 °C and tested for fatigue thresholds and crack closure. It has been found that, in general, three different trends in the dependence of both the total threshold stress intensity range, ΔK th , and the intrinsic threshold stress intensity range, ΔK eff, th , on the PAGS are observable. By considering in detail the factors such as cyclic stress-strain behavior, environmental effects on FCG, and embrittlement during tempering, the present observations could be rationalized. The strong dependence of ΔK th and ΔK eff, th on PAGS in microstructures tempered at 530 °C has been primarily attributed to cyclic softening and thereby the strong interaction of the crack tip deformation field with the grain boundary. On the other hand, a less strong dependence of ΔK th and ΔK eff, th on PAGS is suggested to be caused by the cyclic hardening behavior of lightly tempered microstructures occurring in 200 °C temper. In both microstructures, crack closure influenced near-threshold FCG (NTFCG) to a significant extent, and its magnitude was large at large grain sizes. Microstructures tempered at the intermediate temperatures failed to show a systematic variation of ΔKth and ΔKeff, th with PAGS. The mechanisms of intergranular fracture vary between grain sizes in this temper. A transition from “microstructure-sensitive” to “microstructure-insensitive” crack growth has been found to occur when the zone of cyclic deformation at the crack tip becomes more or less equal to PAGS. Detailed observations on fracture morphology and crack paths corroborate the grain size effects on fatigue thresholds and crack closure.

Compositional dependence of crystallization of As-Te glasses

Thermal crystallization studies have been carried out on bulk, semiconducting AsxTe100−x glasses of different compositions using Differential Scanning Caloritmery. AsxTe100−x glasses with x < 40, are found to exhibit one glass transition and one crystallization. On the other hand, glasses with composition 40 less-than-or-equals, slantxless-than-or-equals, slant 50 show one glass transition and two crystallization reactions. It has been found that in glasses with x greater-or-equal, slanted 40, the two crystallization reactions progressively merge with an increase in arsenic concentration. Consequently AsxTe100−x glasses with x greater-or-equal, slanted 50 show only one crystallization. The composition dependence of crystallization temperatures and activation energies for crystallization estimated by Kissinger's method, show marked deviations at a composition x = 40. These observations can be explained in terms of the changes in the local structure of the material with composition.

Anomalous behaviour in the composition dependence of the photoacoustic properties of Si---As---Te glasses

The optical bandgap and thermal diffusivity of Si10AsxTe90−x (10 ≤ x ≤ 50) and Si15AsxTe85−x (5 ≤ x ≤ 40) glasses have been measured using the photoacoustic technique. The anomalous behaviour observed in these properties at the mean coordination number left angle bracketrright-pointing angle bracket = 2.60 is interpreted by reference to the formation and development of a layered structure in these glasses.

On The Temperature-Dependence of Regioselectivity in the Amination of Bromobenzotropones

The ipso/cine ratio in the amination of 5-bromo-2,3-benzo- or 2-bromo-4,5-benzotropone shows a dependence upon the temperature at which the reaction is conducted, changing in favour of the ipso-product when the temperature is maintained high, ruling out an aryne-type mechanism. A comparison of independent mechanisms envisaged for the formation of the two isomeric products suggests a two-part reason: (i) at a higher reaction temperature, C-protonation, a step necessary for the formation of the cine-product, could be retarded when a direct internal mode is interfered with by a less efficient external one, and (ii) reketonisation by elimination of bromide, needed to form the ipso-product, is likely to have a high temperature coefficient enabling the rate of its formation to overtake that of the cine-product.

Dependence of the large-scale vortex instability on latitude, stratification and domain size

In an earlier study, we reported on the excitation of large-scale vortices in Cartesian hydrodynamical convection models subject to rapid enough rotation. In that study, the conditions for the onset of the instability were investigated in terms of the Reynolds (Re) and Coriolis (Co) numbers in models located at the stellar North pole. In this study, we extend our investigation to varying domain sizes, increasing stratification, and place the box at different latitudes. The effect of the increasing box size is to increase the sizes of the generated structures, so that the principal vortex always fills roughly half of the computational domain. The instability becomes stronger in the sense that the temperature anomaly and change in the radial velocity are observed to be enhanced. The model with the smallest box size is found to be stable against the instability, suggesting that a sufficient scale separation between the convective eddies and the scale of the domain is required for the instability to work. The instability can be seen upto the colatitude of 30 degrees, above which value the flow becomes dominated by other types of mean flows. The instability can also be seen in a model with larger stratification. Unlike the weakly stratified cases, the temperature anomaly caused by the vortex structures is seen to depend on depth.

Pulsewidth analysis in a mode locked internally frequency doubled Ti: sapphire laser

Theoretical analysis of internal frequency doubling in actively mode locked broadband solid state lasers is presented. The analysis is used to study the dependence of mode locked pulsewidth on the second harmonic conversion efficiency, the modulation depth, and the tuning element bandwidth in an AM mode locked Ti: sapphire laser. The results are presented in the form of graphs.  

Temperature dependence of hydrogen evolution reaction on nickel oxide electrode in sealed nickel/cadmium cells

The impedance of sealed nickel/cadmium cells around a cell e.m.f. of 0.0 V was measured at five different temperatures between � 10 and +30 °C. The results show that the behaviour is similar at all temperatures. Based on the experimental results, the relation between charge-transfer resistance (Rct) and temperature (T) has been established for the Volmer reaction. Further, the value of cathodic transfer coefficient (?) has been estimated.

Size dependence of the bulk modulus of semiconductor nanocrystals from first-principles calculations

The variation in the bulk modulus of semiconductor nanoparticles has been studied within first-principles electronic-structure calculations using the local density approximation (LDA) for the exchange correlation. Quantum Monte Carlo calculations carried out for a silicon nanocrystal Si87H76 provided reasonable agreement with the LDA results. An enhancement was observed in the bulk modulus as the size of the nanoparticle was decreased, with modest enhancements being predicted for the largest nanoparticles studied here, a size just accessible in experiments. To access larger sizes, we fit our calculated bulk moduli to the same empirical law for all materials, the asymptote of which is the bulk value of the modulus. This was found to be within 2-10% of the independently calculated value. The origin of the enhancement has been discussed in terms of Cohen's empirical law M.L. Cohen, Phys. Rev. B 32, 7988 (1985)] as well as other possible scenarios.

Dependence of magnetism on GdFeO3 distortion in the t(2g) system ARuO(3) (A = Sr, Ca)

We have examined the stability of the ferromagnetic (FM) state in CaRuO3 and SrRuO3 as a function of the GdFeO3 distortion. Model calculations predict the dependence of the FM transition temperature (T-c) on the rotation angle theta to vary as cos(2)(2 theta) for e(g)-electron systems. However, here, we find an initial increase and then the expected decrease. Furthermore, a much faster decrease is found than predicted for e(g)-electron systems. Considering the specific case of CaRuO3, a larger deviation of the Ru-O-Ru angle from 180 degrees in CaRuO3 as compared to SrRuO3 should result in a more reduced bandwidth, thereby making the former more correlated. The absence of long-range magnetic order in the more correlated CaRuO3 is traced to the strong collapse of various exchange interaction strengths that arises primarily from the volume reduction and increased distortion of the RuO6 octahedra network that accompanies the presence of a smaller ion at the A site.

Improved mechanical properties of polymer nanocomposites incorporating graphene-like BN: Dependence on the number of BN layers

The mechanical properties of composites of polymethylmethacrylate (PMMA) with two-dimensional graphene-like boron nitride (BN) have been investigated to explore the dependence of the properties on the number of BN layers. This study demonstrates that significantly improved mechanical properties are exhibited by the composite with the fewest number of BN layers. Thus, with incorporation of three BN layers, the hardness and elastic modulus of the composite showed an increase of 125% and 130%, respectively, relative to pure PMMA. (C) 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.