Modelling and simulation of wave transformation in porous structures using VOF based two-phase flow model

This paper represents the results of wave transformation in porous structures and hydraulic performance of a vertical porous seawall. The study was carried out using a VOF based two-phase numerical hydrodynamic model. The model was developed by coupling an ordinary porous flow model based on extended Navier–Stokes equations for porous media, and a two-phase flow model. A unique solution domain was established with proper treatment of the interface boundary between water, air and the structure. The VOF method with an improved fluid advection algorithm was used to trace the interface between water and air. The resistance to flow caused by the presence of structural material was modeled in terms of drag and inertia forces. The parameters that govern resistance to flow in a porous media were calibrated for a typical structural setup and then the computational efficacy of the model was evaluated for several wave and structural conditions other than the calibrated setup. A set of comparisons of wave properties in and around the structure showed that the model reproduced reasonably good agreement between computed results and measured data. The model was then applied to investigate wave transformation in a vertical porous structure. The role of porosity and width of a structure in reducing wave reflection and increasing energy dissipation was investigated. It is confirmed that there exists an optimum value of structure width and porosity that can maximize hydraulic performances of a porous seawall.     

Treatment of colored effluent of the textile industry in Bangladesh using zinc chloride treated indigenous activated carbons

The adsorption of colored compounds from the textile dyeing effluents of Bangladesh on granulated activated carbons produced from indigenous vegetable sources by chemical activation with zinc chloride was studied. The most important parameters in chemical activation were found be the chemical ratio of ZnCl₂ to feed (3:1), carbonization temperature (450-465 °C) and activation time (80 min). The adsorbances at 511 nm (red effluent) and 615 nm (blue effluent) were used for color estimation. It is established that at optimum temperature (50 °C), time of contact (30-40 min) and adsorbent loading (2 g l⁻¹), activated carbons developed from Segun saw-dust and water hyacinth showed substantial capability to remove coloring materials from the effluents. It is observed that adsorption of reactive dyes by all sorts of activated carbons is higher than disperse dyes. It is explained that activated carbon, because of its acidic nature, can better adsorb reactive dye particles containing large number of nitrogen sites and -SO₃Na group in their structure. The use of carbons would be economical, as saw-dust and water hyacinth are waste products and abundant in Bangladesh.     

A dynamical system approach to Heisenberg uniqueness pairs

A canonical system is a kind of first-order system of ordinary differential equations on an interval of the real line parametrized by complex numbers. It is known that any solution of a canonical system generates an entire function of the Hermite-Biehler class. In this paper, we deal with the inverse problem to recover a canonical system from a given entire function of the Hermite-Biehler class satisfying appropriate conditions. This inverse problem was solved by de Branges in 1960s. However his results are often not enough to investigate a Hamiltonian of recovered canonical system. In this paper, we present an explicit way to recover a Hamiltonian from a given exponential polynomial belonging to the Hermite-Biehler class. After that, we apply it to study distributions of roots of self-reciprocal polynomials.     

Strong stability of the embedded Markov chain in an GI/M/1 queue with negative customers

This paper is devoted to the investigation for sufficient conditions of the strong stability of the embedded Markov chain in GI/M/1 queueing system with negative customers. After perturbing the occurrence rate of the negative customers, we prove the strong stability of the considered Markov chain with respect to a convenient weight variation norm. Furthermore, we estimate the deviation of its transition operators and provide an upper bound to the approximation error. This results allow us to understand how the negative customers will affect the system’s level of performance.     

Local generalized quadratic distance metrics: application to the <Emphasis Type="Italic">k</Emphasis>-nearest neighbors classifier

Finding the set of nearest neighbors for a query point of interest appears in a variety of algorithms for machine learning and pattern recognition. Examples include k nearest neighbor classification, information retrieval, case-based reasoning, manifold learning, and nonlinear dimensionality reduction. In this work, we propose a new approach for determining a distance metric from the data for finding such neighboring points. For a query point of interest, our approach learns a generalized quadratic distance (GQD) metric based on the statistical properties in a “small” neighborhood for the point of interest. The locally learned GQD metric captures information such as the density, curvature, and the intrinsic dimensionality for the points falling in this particular neighborhood. Unfortunately, learning the GQD parameters under such a local learning mechanism is a challenging problem with a high computational overhead. To address these challenges, we estimate the GQD parameters using the minimum volume covering ellipsoid (MVCE) for a set of points. The advantage of the MVCE is two-fold. First, the MVCE together with the local learning approach approximate the functionality of a well known robust estimator for covariance matrices. Second, computing the MVCE is a convex optimization problem which, in addition to having a unique global solution, can be efficiently solved using a first order optimization algorithm. We validate our metric learning approach on a large variety of datasets and show that the proposed metric has promising results when compared with five algorithms from the literature for supervised metric learning.