Search

We present a pioneering model of the interaction between the solar wind and the surrounding interstellar medium that includes the possibility of different pressures in the directions parallel and perpendicular to the magnetic field. The outer heliosheath region is characterized by a low rate of turbulent scattering that would permit a development of pressure anisotropy. The effect is best seen on the interstellar side of the heliopause, where a narrow region is developed with an excessive perpendicular pressure resembling a plasma depletion layer typical of planetary magnetspheres. The magnitude of this effect is relatively small owing to the fact that proton-proton collisions don't permit the anisotropy to approach a stability threshold. We also show that if turbulence in the interstellar medium is weak, a much broader anisotropic boundary layer can exist, with a dominant perpendicular pressure in the southern hemisphere and a dominant parallel pressure in the north., Comment: 13 pages, 3 figures

In this paper we present the first set of 3D magnetohydrodynamic (MHD) simulations performed with the riemann geomesh code. We study the dynamics of the magnetically channeled winds of magnetic massive stars in full three dimensions using a code that is uniquely suited to spherical problems. Specifically, we perform isothermal simulations of a smooth wind on a rotating star with a tilted, initially dipolar field. We compare the mass-loss, angular momentum loss, and magnetospheric dynamics of a template star (with the properties that are reminiscent of the O4 supergiant {\zeta} Pup) over a range of rotation rates, magnetic field strengths, and magnetic tilt angles. The simulations are run up to a quasi-steady state and the results are observed to be consistent with the existing literature, showing the episodic centrifugal breakout events of the mass outflow, confined by the magnetic field loops that form the closed magnetosphere of the star. The catalogued results provide perspective on how angular-momentum loss varies for different configurations of rotation rate, magnetic field strength, and large magnetic tilt angles. In agreement with previous 2D MHD studies, we find that high magnetic confinement reduces the overall mass-loss rate, and higher rotation increases the mass-loss rate. This and future studies will be used to estimate the angular-momentum evolution, spin-down time, and mass-loss evolution of magnetic massive stars as a function of magnetic field strength, rotation rate, and dipole tilt.

The phenomenon of linear motion of conductor in a magnetic field is commonly found in electric machineries such as, electromagnetic brakes, linear induction motor, electromagnetic flowmeter etc. The design and analysis of the same requires an accurate evaluation of induced currents and the associated reaction magnetic fields. The finite element method is a generally employed numerical technique for this purpose. However, it needs stabilization techniques to provide an accurate solution. In this work, such a stabilization technique is developed for the edge elements. The stability and hence the accuracy is brought in by a suitable representation of the source term. The stability and accuracy of the proposed scheme is first shown analytically and then demonstrated with the help of 2D and 3D simulations. The proposed scheme is parameter-free and it would require a graded regular mesh along the direction of motion.

The finite element method is one of the widely employed numerical techniques in electrical engineering for the study of electric and magnetic fields. When applied to the moving conductor problems, the finite element method is known to have numerical oscillations in the solution. To resolve this, the upwinding techniques, which are developed for the transport equation are borrowed and directly employed for the magnetic induction equation. In this work, an alternative weighted residual formulation is explored for the simulation of the linear moving conductor problems. The formulation is parameter-free and the stability of the formulation is analytically studied for the 1D version of the moving conductor problem. Then the rate of convergence and the accuracy are illustrated with the help of several test cases in 1D as well as 2D. Subsequently, the stability of the formulation is demonstrated with a 3D moving conductor simulation.

Flavor is expressed through interaction of molecules via gustatory and olfactory mechanisms. Knowing the utility of flavor molecules in food and fragrances, it is valuable to add a comprehensive repository of flavor compounds characterizing their flavor profile, chemical properties, regulatory status, consumption statistics, taste/aroma threshold values, reported uses in food categories, and synthesis. FlavorDB2 (https://cosylab.iiitd.edu.in/flavordb2/) is an updated database of flavor molecules with an user-friendly interface. This repository simplifies the search for flavor molecules, their attributes and offers a range of applications including food pairing. FlavorDB2 serves as a standard repository of flavor compounds., Comment: 5 pages, 2 figures

Conductor moving in magnetic field is quite common in electrical equipment. The numerical simulation of such problem is vital in their design and analysis of electrical equipment. The Galerkin finite element method (GFEM) is a commonly employed simulation tool, nonetheless, due to its inherent numerical instability at higher velocities, the GFEM requires upwinding techniques to handle moving conductor problems. The Streamline Upwinding/Petrov-Galerkin (SU/PG) scheme is a widely acknowledged upwinding technique, despite its error-peaking at the transverse boundary. This error at the transverse-boundary, is found to be leading to non-physical solutions. Several remedies have been suggested in the allied fluid dynamics literature, which employs non-linear, iterative techniques. The present work attempts to address this issue, by retaining the computational efficiency of the GFEM. By suitable analysis, it is shown that the source of the problem can be attributed to the Coulomb's gauge. Therefore, to solve the problem, the Coulomb's gauge is taken out from the formulation and the associated weak form is derived. The effectiveness of this technique is demonstrated with pertinent numerical results., Comment: 8 pages, 15 figures

Methods proposed in the literature for zero-shot learning (ZSL) are typically suitable for offline learning and cannot continually learn from sequential streaming data. The sequential data comes in the form of tasks during training. Recently, a few attempts have been made to handle this issue and develop continual ZSL (CZSL) methods. However, these CZSL methods require clear task-boundary information between the tasks during training, which is not practically possible. This paper proposes a task-free (i.e., task-agnostic) CZSL method, which does not require any task information during continual learning. The proposed task-free CZSL method employs a variational autoencoder (VAE) for performing ZSL. To develop the CZSL method, we combine the concept of experience replay with knowledge distillation and regularization. Here, knowledge distillation is performed using the training sample's dark knowledge, which essentially helps overcome the catastrophic forgetting issue. Further, it is enabled for task-free learning using short-term memory. Finally, a classifier is trained on the synthetic features generated at the latent space of the VAE. Moreover, the experiments are conducted in a challenging and practical ZSL setup, i.e., generalized ZSL (GZSL). These experiments are conducted for two kinds of single-head continual learning settings: (i) mild setting-: task-boundary is known only during training but not during testing; (ii) strict setting-: task-boundary is not known at training, as well as testing. Experimental results on five benchmark datasets exhibit the validity of the approach for CZSL.

Zero-shot learning is a new paradigm to classify objects from classes that are not available at training time. Zero-shot learning (ZSL) methods have attracted considerable attention in recent years because of their ability to classify unseen/novel class examples. Most of the existing approaches on ZSL works when all the samples from seen classes are available to train the model, which does not suit real life. In this paper, we tackle this hindrance by developing a generative replay-based continual ZSL (GRCZSL). The proposed method endows traditional ZSL to learn from streaming data and acquire new knowledge without forgetting the previous tasks' gained experience. We handle catastrophic forgetting in GRCZSL by replaying the synthetic samples of seen classes, which have appeared in the earlier tasks. These synthetic samples are synthesized using the trained conditional variational autoencoder (VAE) over the immediate past task. Moreover, we only require the current and immediate previous VAE at any time for training and testing. The proposed GRZSL method is developed for a single-head setting of continual learning, simulating a real-world problem setting. In this setting, task identity is given during training but unavailable during testing. GRCZSL performance is evaluated on five benchmark datasets for the generalized setup of ZSL with fixed and dynamic (incremental class) settings of continual learning. The existing class setting presented recently in the literature is not suitable for a class-incremental setting. Therefore, this paper proposes a new setting to address this issue. Experimental results show that the proposed method significantly outperforms the baseline and the state-of-the-art method and makes it more suitable for real-world applications.

Recently, zero-shot learning (ZSL) emerged as an exciting topic and attracted a lot of attention. ZSL aims to classify unseen classes by transferring the knowledge from seen classes to unseen classes based on the class description. Despite showing promising performance, ZSL approaches assume that the training samples from all seen classes are available during the training, which is practically not feasible. To address this issue, we propose a more generalized and practical setup for ZSL, i.e., continual ZSL (CZSL), where classes arrive sequentially in the form of a task and it actively learns from the changing environment by leveraging the past experience. Further, to enhance the reliability, we develop CZSL for a single head continual learning setting where task identity is revealed during the training process but not during the testing. To avoid catastrophic forgetting and intransigence, we use knowledge distillation and storing and replay the few samples from previous tasks using a small episodic memory. We develop baselines and evaluate generalized CZSL on five ZSL benchmark datasets for two different settings of continual learning: with and without class incremental. Moreover, CZSL is developed for two types of variational autoencoders, which generates two types of features for classification: (i) generated features at output space and (ii) generated discriminative features at the latent space. The experimental results clearly indicate the single head CZSL is more generalizable and suitable for practical applications., Comment: Zero-shot Learning, Continual Learning, Incremental Learning

There is a great need in several areas of astrophysics and space-physics to carry out high order of accuracy, divergence-free MHD simulations on spherical meshes. This requires us to pay careful attention to the interplay between mesh quality and numerical algorithms. Methods have been designed that fundamentally integrate high order isoparametric mappings with the other high accuracy algorithms that are needed for divergence-free MHD simulations on geodesic meshes. The goal of this paper is to document such algorithms that are implemented in the geodesic mesh version of the RIEMANN code. The fluid variables are reconstructed using a special kind of WENO-AO algorithm that integrates the mesh geometry into the reconstruction process from the ground-up. A novel divergence-free reconstruction strategy for the magnetic field that performs efficiently at all orders, even on isoparametrically mapped meshes, is then presented. The MHD equations are evolved in space and time using a novel ADER predictor algorithm that is efficiently adapted to the isoparametrically mapped geometry. The application of one-dimensional and multidimensional Riemann solvers at suitable locations on the mesh then provides the corrector step. The corrector step for the magnetic field uses a Yee-type staggering of magnetic fields. This results in a scheme with divergence-free update for the magnetic field. The use of ADER enables a one-step update which only requires one messaging operation per complete timestep. This is very beneficial for parallel processing. Several accuracy tests are presented as are stringent test problems. PetaScale performance is also demonstrated on the largest available supercomputers., Comment: Accepted for publication in MNRAS

For the simulation of rectilinearly moving conductors across a magnetic field, the Galer-kin finite element method (GFEM) is generally employed. The inherent instability of GFEM is very often addressed by employing Streamline upwinding/Petrov-Galerkin (SU/PG) scheme. However, the SU/PG solution is known to suffer from distortion at the boundary transverse to the velocity and the remedial measures suggested in fluid dynamics literature are computationally demanding. Therefore, simple alternative schemes are essential. In an earlier effort, instead of conventional finite-difference based approach, the numerical instability was analyzed using the Z-transform. By employing the concept of pole-zero cancellation, stability of the numerical solution was achieved by a simple restatement of the input magnetic flux in terms of associated vector potential. This approach, however, is restricted for input fields, which vary only along the direction of the velocity. To overcome this, the present work proposes a novel approach in which the input field is restated as a weighted elemental average. The stability of the proposed scheme is proven analytically for both 1D and 2D cases. The error bound for the small oscillations remnant at intermittent Peclet numbers is also deduced. Using suitable numerical simulations, all the theoretical deductions are verified., Comment: appears in IET Science, Measurement \& Technology, 2016

The magnetic flow meter is one of the best possible choice for the measurement of flow rate of liquid metals in fast breeder reactors. Due to the associated complexities in the measuring environment, theoretical evaluation of their sensitivity is always preferred. In order to consider the 3D nature of the problem and the general flow patterns, numerical field computational approach is inevitable. When classical Galerkin's finite element formulation is employed for the solution, it is known to introduce numerical oscillations at high flow rates. The magnetic field produced by the flow induced currents circulate within the fluid and forms the source of this numerical problem. To overcome this, modified methods like stream-line upwind Petrov-Galerkin schemes are generally suggested in the allied areas like fluid dynamics, in which a similar dominance of advective (curl or circulation) component occurs over diffusion (divergence) component. After a careful analysis of the numerical instability through a reduced one dimensional problem, an elegant stable approach is devised. In this scheme, a pole-zero cancellation approach is adopted. The proposed scheme is shown to be absolutely stable. However, at lower flow rates numerical results exhibits small oscillation, which can be controlled by reducing the element size. The source of stability at higher flow rates, as well as, oscillations at lower flow rates are analysed using analytical solution of the associated difference equation. Finally the proposed approach is applied to the original flow meter problem and the solution is shown to be stable., Comment: IET Science, Measurement & Technology, 2016

We report on the observation of exchange bias and memory effect in double perovskite Sr$_2$FeCoO$_6$. Antiphase boundaries between the ferromagnetic and antiferromagnetic regions in the disordered glassy phase is assumed as responsible for the observed effect which reflects in the cooling field dependence and temperature evolution of exchange bias field and in training effect. The spin glass phase itself is characterized through memory, ageing and magnetic relaxation experiments. The spin glass transition temperature, $T_g$, versus $H_{dc}^{2/3}$ follows the Almeida-Thouless line yielding a freezing temperature, $T_f$ = 73 K. Time-dependent magnetic relaxation studies reveal the magnetization dynamics of the underlying glassy phase in this double perovskite., Comment: Accepted for publication in Applied Physics Letters