Approximation of a Variable Density Cloud of Points by Shrinking a Discrete Membrane

This paper describes a method to obtain a closed surface that approximates a general 3D data point set with nonuniform density. Aside from the positions of the initial data points, no other information is used. Particularly, neither the topological relations between the points nor the normal to the surface at the data points are needed. The reconstructed surface does not exactly interpolate the initial data points, but approximates them with a bounded maximum distance. The method allows one to reconstruct closed surfaces with arbitrary genus and closed surfaces with disconnected shells.     

Optimization of Robust Loss Functions for Weakly-Labeled Image Taxonomies

The recently proposed ImageNet dataset consists of several million images, each annotated with a single object category. These annotations may be imperfect, in the sense that many images contain multiple objects belonging to the label vocabulary. In other words, we have a multi-label problem but the annotations include only a single label (which is not necessarily the most prominent). Such a setting motivates the use of a robust evaluation measure, which allows for a limited number of labels to be predicted and, so long as one of the predicted labels is correct, the overall prediction should be considered correct. This is indeed the type of evaluation measure used to assess algorithm performance in a recent competition on ImageNet data. Optimizing such types of performance measures presents several hurdles even with existing structured output learning methods. Indeed, many of the current state-of-the-art methods optimize the prediction of only a single output label, ignoring this ‘structure’ altogether. In this paper, we show how to directly optimize continuous surrogates of such performance measures using structured output learning techniques with latent variables. We use the output of existing binary classifiers as input features in a new learning stage which optimizes the structured loss corresponding to the robust performance measure. We present empirical evidence that this allows us to ‘boost’ the performance of binary classification on a variety of weakly-supervised labeling problems defined on image taxonomies.     

Using AVAL-1D to simulate avalanches in the eastern Pyrenees

The two friction parameters used in the numerical avalanche dynamics program AVAL-1D, were calibrated empirically with data from observed avalanches in the Swiss Alps. The implementation of the model with these friction parameters in other regions with different characteristics can lead to considerable uncertainty if a previous calibration is not performed. However, direct calibration for a specific avalanche path is often not possible since the data available are insufficient. Therefore, we back-calculated twelve well-documented avalanche events from the Catalan Pyrenees to calibrate the friction coefficients to be used in this mountain range. The result of the study reveals that there is a good fit between recorded and simulated avalanche events using the friction parameters originally calibrated for the Swiss Alps, despite the difference in snow climate between these two mountain ranges.     

Integration of a molten carbonate fuel cell with a direct exhaust absorption chiller

A high market value exists for an integrated high-temperature fuel cell-absorption chiller product throughout the world. While high-temperature, molten carbonate fuel cells are being commercially deployed with combined heat and power (CHP) and absorption chillers are being commercially deployed with heat engines, the energy efficiency and environmental attributes of an integrated high-temperature fuel cell-absorption chiller product are singularly attractive for the emerging distributed generation (DG) combined cooling, heating, and power (CCHP) market. This study addresses the potential of cooling production by recovering and porting the thermal energy from the exhaust gas of a high-temperature fuel cell (HTFC) to a thermally activated absorption chiller. To assess the practical opportunity of serving an early DG-CCHP market, a commercially available direct fired double-effect absorption chiller is selected that closely matches the exhaust flow and temperature of a commercially available HTFC. Both components are individually modeled, and the models are then coupled to evaluate the potential of a DG-CCHP system. Simulation results show that a commercial molten carbonate fuel cell generating 300 kW of electricity can be effectively coupled with a commercial 40 refrigeration ton (RT) absorption chiller. While the match between the two “off the shelf” units is close and the simulation results are encouraging, the match is not ideal. In particular, the fuel cell exhaust gas temperature is higher than the inlet temperature specified for the chiller and the exhaust flow rate is not sufficient to achieve the potential heat recovery within the chiller heat exchanger. To address these challenges, the study evaluates two strategies: (1) blending the fuel cell exhaust gas with ambient air, and (2) mixing the fuel cell exhaust gases with a fraction of the chiller exhaust gas. Both cases are shown to be viable and result in a temperature drop and flow rate increase of the gases before the chiller inlet. The results show that no risk of cold end corrosion within the chiller heat exchanger exists. In addition, crystallization is not an issue during system operation. Accounting for the electricity and the cooling produced and disregarding the remaining thermal energy, the second strategy is preferred and yields an overall estimated efficiency of 71.7%.     

Reactivity of LaNi1−y Co y O3−δ Perovskite Systems in the Deep Oxidation of Toluene

In the present work we have evaluated the oxidation of toluene over different lanthanum perovskites with a general composition of LaNi_1?y Co _y O_3?δ. These catalysts, prepared by a spray pyrolysis method, have been characterised by XRD, BET and FE-SEM techniques. Additional experiments of temperature programmed desorption of O₂, reduction in H₂ and X-ray absorption spectroscopy were also performed in order to identify the main surface oxygen species and the reducibility of the different perovskites. The catalytic behaviour toward the oxidation of toluene (as a model for VOCs compounds) was evaluated in the range 100–600 °C, detecting a total conversion for all the samples below 400 °C and higher activities for the cobalt-containing perovskites. The catalytic behaviour of these samples is consistent with a suprafacial mechanism, with the α-type oxygen playing an active role in the oxidation reaction.     

Ultrathin PECVD epitaxial Si solar cells on glass via low‐temperature transfer process

Fabrication of high‐quality ultrathin monocrystalline silicon layers and their transfer to low‐cost substrates are key steps for flexible electronics and photovoltaics. In this work, we demonstrate a low‐temperature and low‐cost process for ultrathin silicon solar cells. By using standard plasma‐enhanced chemical vapor deposition (PECVD), we grow high‐quality epitaxial silicon layers (epi‐PECVD) from SiH₄/H₂ gas mixtures at 175 °C. Using secondary ion mass spectrometry and transmission electron microscopy, we show that the porosity of the epi‐PECVD/crystalline silicon interface can be tuned by controlling the hydrogen accumulation there. Moreover, we demonstrate that 13–14% porosity is a threshold above which the interface becomes fragile and can easily be cleaved. Taking advantage of the H‐rich interface fragility, we demonstrate the transfer of large areas (∽10 cm²) ultrathin epi‐PECVD layers (0.5–5.5 µm) onto glass substrates by anodic bonding and moderate annealing (275–350 °C). The structural properties of transferred layers are assessed, and the first PECVD epitaxial silicon solar cells transferred on glass are characterized. Copyright © 2016 John Wiley & Sons, Ltd.     

Effect of different thermal treatments on the mechanical performance of poly(L‐lactic acid) based eco‐composites

PLLA‐based eco‐composites reinforced with kenaf fiber and rice straw and containing red or yellow pigments have been studied. The mechanical behavior of the composites was tested by DMTA at two different annealing temperatures (65°C and 85°C) and times (15 min and 120 min) as well as at two preparation conditions: vacuum drying and long time at room temperature. A decrease of microhardness was observed during the water absorption tests. Moreover, the rice straw‐based composites absorbed more water than the kenaf‐ones. Generally, the dyed NFs composites presented better water resistance than undyed ones. The pigments improved the adhesion and led to better mechanical performance. The natural fibers favored the cold crystallization process of PLLA and shifted the cold crystallization peak temperature to lower values, as it was confirmed by DSC measurements. The values of tensile storage modulus obtained after different preparation condition were strongly affected by the process of physical ageing. According to, tan δ parameter, the samples stored at room temperature for a long time showed the highest amorphous content. The PLLA eco‐composite reinforced with kenaf fibers, dyed with the red pigment, and annealed at 85°C for 2 h displays the best mechanical properties. © 2009 Wiley Periodicals, Inc. J Appl Polym Sci, 2010     

Improvement of mechanical performance of epoxy resins filled with cobalt and chromium powders

Epoxy/ powder metal composites have interesting electrical properties, becoming conductors above the percolation threshold. To complete this study, mechanical investigations have been carried out to show the influence of the fillers on the mechanical performance of these composites. In this framework, different epoxy/metallic powders (Cobalt, Chromium) composites were prepared. Scanning Electron Microscopy showed that the dispersion of the metallic fillers in the matrix is almost homogeneous. The dynamic mechanical thermal analysis (DMTA) measurements showed the dependence of the viscoelastic parameters with the frequency, temperature, nature, and content of fillers. The main relaxations observed are the primary α relaxation (associated to the glass transition, T_g) and a secondary β relaxation. A second DMTA run on the same samples showed a slight increase of the T_g. It clearly showed that the used metallic fillers improve the mechanical properties of the obtained composites. © 2010 Wiley Periodicals, Inc. J Appl Polym Sci, 2010