Variable Granularity Access Tracking Scheme for Improving the Performance of Software Transactional Memory

Software transactional memory (STM) has been proposed as a promising programming paradigm for shared memory multi-threaded programs as an alternative to conventional lock based synchronization primitives. Typical STM implementations employ a conflict detection scheme, which works with uniform access granularity, tracking shared data accesses either at word/cache line or at object level. It is well known that a single fixed access tracking granularity cannot meet the conflicting goals of reducing false conflicts without impacting concurrency adversely. A fine grained granularity while improving concurrency can have an adverse impact on performance due to lock aliasing, lock validation overheads, and additional cache pressure. On the other hand, a coarse grained granularity can impact performance due to reduced concurrency. Thus, in general, a fixed or uniform granularity access tracking (UGAT) scheme is application-unaware and rarely matches the access patterns of individual application or parts of an application, leading to sub-optimal performance for different parts of the application(s). In order to mitigate the disadvantages associated with UGAT scheme, we propose a Variable Granularity Access Tracking (VGAT) scheme in this paper. We propose a compiler based approach wherein the compiler uses inter-procedural whole program static analysis to select the access tracking granularity for different shared data structures of the application based on the application's data access pattern. We describe our prototype VGAT scheme, using TL2 as our STM implementation. Our experimental results reveal that VGAT-STM scheme can improve the application performance of STAMP benchmarks from 1.87% to up to 21.2%.

Modeling of seepage flow through layered solis

A two-dimensional finite difference model, which solves mixed type of Richards' equation, whose non-linearity is dealt with modified Picard's iteration and strongly implicit procedure to solve the resulting equations, is presented. Modeling of seepage flow through heterogeneous soils, which is common in the field is addressed in the present study. The present model can be applied to both unsaturated and saturated soils and can handle very dry initial condition and steep wetting fronts. The model is validated by comparing experimental results reported in the literature. Newness of this two dimensional model is its application on layered soils with transient seepage face development, which has not been reported in the literature. Application of the two dimensional model for studying unconfined drainage due to sudden drop of water table at seepage face in layered soils is demonstrated. In the present work different sizes of rectangular flow domain with different types of layering are chosen. Sensitivity of seepage height due to problem dimension of layered system is studied. The effect of aspect ratio on seepage face development in case of the flow through layered soil media is demonstrated. The model is also applied to random heterogeneous soils in which the randomness of the model parameters is generated using the turning band technique. The results are discussed in terms of phreatic surface and seepage height development and also flux across the seepage face. Such accurate modeling of seepage face development and quantification of flux moving across the seepage face becomes important while modeling transport problems in variably saturated media.

Arbitrary Lagrangian-Eulerian finite-element method for computation of two-phase flows with soluble surfactants

A finite-element scheme based on a coupled arbitrary Lagrangian-Eulerian and Lagrangian approach is developed for the computation of interface flows with soluble surfactants. The numerical scheme is designed to solve the time-dependent Navier-Stokes equations and an evolution equation for the surfactant concentration in the bulk phase, and simultaneously, an evolution equation for the surfactant concentration on the interface. Second-order isoparametric finite elements on moving meshes and second-order isoparametric surface finite elements are used to solve these equations. The interface-resolved moving meshes allow the accurate incorporation of surface forces, Marangoni forces and jumps in the material parameters. The lower-dimensional finite-element meshes for solving the surface evolution equation are part of the interface-resolved moving meshes. The numerical scheme is validated for problems with known analytical solutions. A number of computations to study the influence of the surfactants in 3D-axisymmetric rising bubbles have been performed. The proposed scheme shows excellent conservation of fluid mass and of the total mass of the surfactant. (C) 2012 Elsevier Inc. All rights reserved.

On the dynamic contact angle in simulation of impinging droplets with sharp interface methods

Effects of dynamic contact angle models on the flow dynamics of an impinging droplet in sharp interface simulations are presented in this article. In the considered finite element scheme, the free surface is tracked using the arbitrary Lagrangian-Eulerian approach. The contact angle is incorporated into the model by replacing the curvature with the Laplace-Beltrami operator and integration by parts. Further, the Navier-slip with friction boundary condition is used to avoid stress singularities at the contact line. Our study demonstrates that the contact angle models have almost no influence on the flow dynamics of the non-wetting droplets. In computations of the wetting and partially wetting droplets, different contact angle models induce different flow dynamics, especially during recoiling. It is shown that a large value for the slip number has to be used in computations of the wetting and partially wetting droplets in order to reduce the effects of the contact angle models. Among all models, the equilibrium model is simple and easy to implement. Further, the equilibrium model also incorporates the contact angle hysteresis. Thus, the equilibrium contact angle model is preferred in sharp interface numerical schemes.

Penetrative phototactic bioconvection in an isotropic scattering suspension

Phototaxis is a directed swimming response dependent upon the light intensity sensed by micro-organisms. Positive (negative) phototaxis denotes the motion directed towards (away from) the source of light. Using the phototaxis model of Ghorai, Panda, and Hill ''Bioconvection in a suspension of isotropically scattering phototactic algae,'' Phys. Fluids 22, 071901 (2010)], we investigate two-dimensional phototactic bioconvection in an absorbing and isotropic scattering suspension in the nonlinear regime. The suspension is confined by a rigid bottom boundary, and stress-free top and lateral boundaries. The governing equations for phototactic bioconvection consist of Navier-Stokes equations for an incompressible fluid coupled with a conservation equation for micro-organisms and the radiative transfer equation for light transport. The governing system is solved efficiently using a semi-implicit second-order accurate conservative finite-difference method. The radiative transfer equation is solved by the finite volume method using a suitable step scheme. The resulting bioconvective patterns differ qualitatively from those found by Ghorai and Hill ''Penetrative phototactic bioconvection,'' Phys. Fluids 17, 074101 (2005)] at a higher critical wavelength due to the effects of scattering. The solutions show transition from steady state to periodic oscillations as the governing parameters are varied. Also, we notice the accumulation of micro-organisms in two horizontal layers at two different depths via their mean swimming orientation profile for some governing parameters at a higher scattering albedo. (C) 2013 AIP Publishing LLC.

MODELING AND SIMULATION OF HYDRODYNAMIC INTERACTION OF DNA IN A MICRO-FLUIDIC CHANNEL

The motion of DNA (in the bulk solution) and the non-Newtonian effective fluid behavior are considered separately and self-consistently with the fluid motion satisfying the no-slip boundary condition on the surface of the confining geometry in the presence of channel pressure gradients. A different approach has been developed to model DNA in the micro-channel. In this study the DNA is assumed as an elastic chain with its characteristic Young's modulus, Poisson's ratio and density. The force which results from the fluid dynamic pressure, viscous forces and electromotive forces is applied to the elastic chain in a coupled manner. The velocity fields in the micro-channel are influenced by the transport properties. Simulations are carried out for the DNAs attached to the micro-fluidic wall. Numerical solutions based on a coupled multiphysics finite element scheme are presented. The modeling scheme is derived based on mass conservation including biomolecular mass, momentum balance including stress due to Coulomb force field and DNA-fluid interaction, and charge transport associated to DNA and other ionic complexes in the fluid. Variation in the velocity field for the non-Newtonian flow and the deformation of the DNA strand which results from the fluid-structure interaction are first studied considering a single DNA strand. Motion of the effective center of mass is analyzed considering various straight and coil geometries. Effects of DNA statistical parameters (geometry and spatial distribution of DNAs along the channel) on the effective flow behavior are analyzed. In particular, the dynamics of different DNA physical properties such as radius of gyration, end-to-end length etc. which are obtained from various different models (Kratky-Porod, Gaussian bead-spring etc.) are correlated to the nature of interaction and physical properties under the same background fluid environment.

Full-rate full-diversity finite feedback space-time schemes with minimum feedback and transmission duration

A Finite Feedback Scheme (FFS) for a quasi-static MIMO block fading channel with finite N-ary delay-free noise-free feedback consists of N Space-Time Block Codes (STBCs) at the transmitter, one corresponding to each possible value of feedback, and a function at the receiver that generates N-ary feedback. A number of FFSs are available in the literature that provably attain full-diversity. However, there is no known full-diversity criterion that universally applies to all FFSs. In this paper a universal necessary condition for any FFS to achieve full-diversity is given, and based on this criterion the notion of Feedback-Transmission duration optimal (FT-optimal) FFSs is introduced, which are schemes that use minimum amount of feedback N for the given transmission duration T, and minimum T for the given N to achieve full-diversity. When there is no feedback (N = 1) an FT-optimal scheme consists of a single STBC, and the proposed condition reduces to the well known necessary and sufficient condition for an STBC to achieve full-diversity. Also, a sufficient criterion for full-diversity is given for FFSs in which the component STBC yielding the largest minimum Euclidean distance is chosen, using which full-rate (N-t complex symbols per channel use) full-diversity FT-optimal schemes are constructed for all N-t > 1. These are the first full-rate full-diversity FFSs reported in the literature for T < N-t. Simulation results show that the new schemes have the best error performance among all known FFSs.

A POSTERIORI ERROR CONTROL OF DISCONTINUOUS GALERKIN METHODS FOR ELLIPTIC OBSTACLE PROBLEMS

In this article, we derive an a posteriori error estimator for various discontinuous Galerkin (DG) methods that are proposed in (Wang, Han and Cheng, SIAM J. Numer. Anal., 48: 708-733, 2010) for an elliptic obstacle problem. Using a key property of DG methods, we perform the analysis in a general framework. The error estimator we have obtained for DG methods is comparable with the estimator for the conforming Galerkin (CG) finite element method. In the analysis, we construct a non-linear smoothing function mapping DG finite element space to CG finite element space and use it as a key tool. The error estimator consists of a discrete Lagrange multiplier associated with the obstacle constraint. It is shown for non-over-penalized DG methods that the discrete Lagrange multiplier is uniformly stable on non-uniform meshes. Finally, numerical results demonstrating the performance of the error estimator are presented.

Response of vertically loaded pile in clay: a probabilistic study

This study presents the response of a vertically loaded pile in undrained clay considering spatially distributed undrained shear strength. The probabilistic study is performed considering undrained shear strength as random variable and the analysis is conducted using random field theory. The inherent soil variability is considered as source of variability and the field is modeled as two dimensional non-Gaussian homogeneous random field. Random field is simulated using Cholesky decomposition technique within the finite difference program and Monte Carlo simulation approach is considered for the probabilistic analysis. The influence of variance and spatial correlation of undrained shear strength on the ultimate capacity as summation of ultimate skin friction and end bearing resistance of pile are examined. It is observed that the coefficient of variation and spatial correlation distance are the most important parameters that affect the pile ultimate capacity.

Binaural Signal Processing Motivated Generalized Analytic Signal Construction and AM-FM Demodulation

Binaural hearing studies show that the auditory system uses the phase-difference information in the auditory stimuli for localization of a sound source. Motivated by this finding, we present a method for demodulation of amplitude-modulated-frequency-modulated (AM-FM) signals using a ignal and its arbitrary phase-shifted version. The demodulation is achieved using two allpass filters, whose impulse responses are related through the fractional Hilbert transform (FrHT). The allpass filters are obtained by cosine-modulation of a zero-phase flat-top prototype halfband lowpass filter. The outputs of the filters are combined to construct an analytic signal (AS) from which the AM and FM are estimated. We show that, under certain assumptions on the signal and the filter structures, the AM and FM can be obtained exactly. The AM-FM calculations are based on the quasi-eigenfunction approximation. We then extend the concept to the demodulation of multicomponent signals using uniform and non-uniform cosine-modulated filterbank (FB) structures consisting of flat bandpass filters, including the uniform cosine-modulated, equivalent rectangular bandwidth (ERB), and constant-Q filterbanks. We validate the theoretical calculations by considering application on synthesized AM-FM signals and compare the performance in presence of noise with three other multiband demodulation techniques, namely, the Teager-energy-based approach, the Gabor's AS approach, and the linear transduction filter approach. We also show demodulation results for real signals.

Tailoring the second mode of Euler-Bernoulli beams: an analytical approach

In this paper, we study the inverse mode shape problem for an Euler-Bernoulli beam, using an analytical approach. The mass and stiffness variations are determined for a beam, having various boundary conditions, which has a prescribed polynomial second mode shape with an internal node. It is found that physically feasible rectangular cross-section beams which satisfy the inverse problem exist for a variety of boundary conditions. The effect of the location of the internal node on the mass and stiffness variations and on the deflection of the beam is studied. The derived functions are used to verify the p-version finite element code, for the cantilever boundary condition. The paper also presents the bounds on the location of the internal node, for a valid mass and stiffness variation, for any given boundary condition. The derived property variations, corresponding to a given mode shape and boundary condition, also provides a simple closed-form solution for a class of non-uniform Euler-Bernoulli beams. These closed-form solutions can also be used to check optimization algorithms proposed for modal tailoring.

Optimal error estimate using mesh equidistribution technique for singularly perturbed system of reaction-diffusion boundary-value problems

In this article, we study the problem of determining an appropriate grading of meshes for a system of coupled singularly perturbed reaction-diffusion problems having diffusion parameters with different magnitudes. The central difference scheme is used to discretize the problem on adaptively generated mesh where the mesh equation is derived using an equidistribution principle. An a priori monitor function is obtained from the error estimate. A suitable a posteriori analogue of this monitor function is also derived for the mesh construction which will lead to an optimal second-order parameter uniform convergence. We present the results of numerical experiments for linear and semilinear reaction-diffusion systems to support the effectiveness of our preferred monitor function obtained from theoretical analysis. (C) 2014 Elsevier Inc. All rights reserved.

Adaptive mesh generation for singularly perturbed fourth-order ordinary differential equations

In this paper, we consider a singularly perturbed boundary-value problem for fourth-order ordinary differential equation (ODE) whose highest-order derivative is multiplied by a small perturbation parameter. To solve this ODE, we transform the differential equation into a coupled system of two singularly perturbed ODEs. The classical central difference scheme is used to discretize the system of ODEs on a nonuniform mesh which is generated by equidistribution of a positive monitor function. We have shown that the proposed technique provides first-order accuracy independent of the perturbation parameter. Numerical experiments are provided to validate the theoretical results.

On the superposition principle in interference experiments

The superposition principle is usually incorrectly applied in interference experiments. This has recently been investigated through numerics based on Finite Difference Time Domain (FDTD) methods as well as the Feynman path integral formalism. In the current work, we have derived an analytic formula for the Sorkin parameter which can be used to determine the deviation from the application of the principle. We have found excellent agreement between the analytic distribution and those that have been earlier estimated by numerical integration as well as resource intensive FDTD simulations. The analytic handle would be useful for comparing theory with future experiments. It is applicable both to physics based on classical wave equations as well as the non-relativistic Schrodinger equation.

MOMENT ANALYSIS OF A SCROLL EXPANDER USED IN AN ORGANIC RANKINE CYCLE

Recent studies on small-scale power generation with the organic Rankine cycle suggest superior performance of positive displacement type of expanders compared to turbines. Scroll expanders in particular achieve high isentropic efficiencies due to lower leakage and frictional losses. Performance of scroll machines may be enhanced by the use of non-circular involute curves in place of the circular involutes resulting non-uniform wall thickness. In this paper, a detailed moment analysis is performed for such an expander having volumetric expansion ratio of 5 using thermodynamic models proposed earlier by one of the present authors. The working fluid considered in the power cycle is R-245fa with scroll inlet temperature of 125 degrees C for a gross power output of similar to 3.5 kW. The model developed in this paper is verified with an air scroll compressor available in the literature and then applied to an expander Prediction of small variation of moment with scroll motion recommends use of scroll expander without a flywheel over other positive displacement type of expanders, e.g. reciprocating, where a flywheel is an essential component.