Three-dimensional simulation with a CFD tool of the airflow phenomena in single floor double-skin facade equipped with a venetian blind

Nowadays different kinds of double-skin facades are developed and used in new architectural projects. The aim of these facades is, on the one hand, to increase internal comfort and, on the other hand, to decrease energy consumption. In order to optimise the overall performance of the double-skin façades, their detailed behaviour needs to be better understood. The prediction of the airflow within the channel (between the two glazings) is very important for understanding of the double-skin facades behaviour, especially in summer conditions. A comprehensive modelling of a compact double-skin facade equipped with a venetian blind and forced ventilation is proposed here. The modelling is done using the CFD (computational fluid dynamics) approach to assess the air movement within the ventilated facade channel. Three-dimensional airflow is modelled using a homogeneous porous media representation, in order to reduce the size of the mathematical model. A parametric study is proposed here, analysing the impact of three parameters on the airflow development: slat tilt angle, blind position and air outlet position. The distance between the blind and the external glazing was found to have a major impact on the velocity profiles inside the double-skin facade channel.     

The influence of annealing on the conductivity of magnesium films

Polycrystalline magnesium films were deposited under ultrahigh vacuum by thermal evaporation onto a cooled silica substrate. During the growth process of a film a number of lattice defects are incorporated. It was found that the defect density decreases with increasing thickness. An annealing study of the electrical resistance and defect density in magnesium films was made. The results were interpreted on the basis of Vand's theory. The function F₀ expressing the law of distribution of the decay energies exhibited a maximum. For thick films there was no appreciable variation in the activation energy with thickness. In this case the evaluated activation energy E was found to be about 0.35 eV. For very thin films this energy decreases with increasing thickness.     

Control of neuronal morphogenesis by homeoproteins: consequences for the making of neuronal networks

The authors investigated the relationship between plasma lipids and the risk for cortical infarction (61 cases) and transient ischaemic attacks (TIA) (35 cases) compared with matched controls. They observed a maximal increase of total cholesterol, of very low-density lipoprotein and low-density lipoprotein (LDL), triglycerides, total apolipoprotein (Apo), B,LDL-Apo B and Apo-A1, and small size high-density lipoprotein (HDL) and large size HDL whose separation was not possible. In contrast they observed a decrease of HDL-ApoE, a distribution of LDL in a single fraction and the presence of LDL of low weight in the group with cortical infarction with or without cardiac arrhythmias. For the first time, we describe a decrease of the HDL-ApoE/total ApoE ratio. TIA differed from the former group by a low level of HDL and the lack of abnormalities of Apo-A1, distribution of small and large size HDL, and in the distribution and the weight of LDL. These data suggest that previously demonstrated differences in LDL-cholesterol and HDL-cholesterol levels between patients with ischaemic stroke and control subjects may apply to patients with cortical infarction, and that in TIA there are changes in the distribution and the weight of LDL.     

The chemistry and development of tannin-based weather- and boil-proof cold-setting and fast-setting adhesives for wood

Condensed tannins, being phenolic in nature, can be chemically modified to undergo reaction with formaldehyde in such a way that the adhesives formed harden at ambient temperature in industrially meaningful times. Resorcinol, m-hydroxyaniline, and their condensation products with formaldehyde and/or tannin resoles have been used for this purpose. The optimum conditions of application, such as paraformaldehyde amount, pH, and open and close assembly times, were used when comparing the different resins prepared. Results satisfying the requirements of international standard specifications and comparable to those of phenol/resorcinol/formaldehyde cold-setting resins were obtained.     

Efficient Binary Tomographic Reconstruction

Tomographic reconstruction of a binary image from few projections is considered. A novel heuristic algorithm is proposed, the central element of which is a nonlinear transformation ψ(p)=log(p/(1?p)) of the probability p that a pixel of the sought image be 1-valued. It consists of backprojections based on ψ(p) and iterative corrections. Application of this algorithm to a series of artificial test cases leads to exact binary reconstructions, (i.e., recovery of the binary image for each single pixel) from the knowledge of projection data over a few directions. Images up to 10⁶ pixels are reconstructed in a few seconds. A series of test cases is performed for comparison with previous methods, showing a better efficiency and reduced computation times.     

Representational power of restricted boltzmann machines and deep belief networks

Deep belief networks (DBN) are generative neural network models with many layers of hidden explanatory factors, recently introduced by Hinton, Osindero, and Teh (2006) along with a greedy layer-wise unsupervised learning algorithm. The building block of a DBN is a probabilistic model called a restricted Boltzmann machine (RBM), used to represent one layer of the model. Restricted Boltzmann machines are interesting because inference is easy in them and because they have been successfully used as building blocks for training deeper models. We first prove that adding hidden units yields strictly improved modeling power, while a second theorem shows that RBMs are universal approximators of discrete distributions. We then study the question of whether DBNs with more layers are strictly more powerful in terms of representational power. This suggests a new and less greedy criterion for training RBMs within DBNs.     

Fluorescence of Ta(2)O(5) thin films doped by kilo-electron-volt Er implantation: application to microcavities

Luminescent layers are prepared by the implantation of kilo-electron-volt Er ions into tantalum pentoxide (Ta(2)O(5)) thin films made by ion plating. The implantation fluences range from 3.3 × 10(14) to 2 × 10(15) ions/cm(2), and the energies range from 190 to 380 keV. Refractive index, extinction coefficient, and losses on guided propagation are investigated. We show that these Er-implanted layers present an absorption as low as that of the nonimplanted films. When optically pumped with an Ar(+) laser (λ = 0.488 μm) beam, implanted films show peaked fluorescence spectra centered near 1.53 and 0.532 μm. We show that the fluorescence intensity is correlated with the intensity of the pump beam in the region where Er ions are implanted. Radiation patterns of Er ions located inside a single layer or inside a Ta(2)O(5)/SiO(2) dielectric stack made by ion plating are also investigated. We show that, in any case, spontaneous emission of Er ions can be spatially controlled.     

Formal Proofs of Rounding Error Bounds

Floating-point arithmetic is a very efficient solution to perform computations in the real field. However, it induces rounding errors making results computed in floating-point differ from what would be computed with reals. Although numerical analysis gives tools to bound such differences, the proofs involved can be painful, hence error prone. We thus investigate the ability of a proof assistant like Coq to mechanically check such proofs. We demonstrate two different results involving matrices, which are pervasive among numerical algorithms, and show that a large part of the development effort can be shared between them.     

A finite element formulation to identify damage fields: the equilibrium gap method

It is proposed to determine damage parameters in two dimensions (surface of a material) or three dimensions (in the bulk of a solid) by using full‐field displacement measurements. A finite‐element approach is developed to evaluate piece‐wise constant elastic parameters modeled by an isotropic damage variable. Two sets of examples are discussed. The first series deals with mechanical fields obtained by finite element simulations to assess the performance of the approach. The second series is concerned with displacement measurements performed during a biaxial test on a composite material. Copyright © 2004 John Wiley & Sons, Ltd.