Unicast-based multicast communication in wormhole-routed networks

Multicast communication, in which the same message is delivered from a source node to an arbitrary number of destination nodes, is being increasingly demanded in parallel computing. System supported multicast services can potentially offer improved performance, increased functionality, and simplified programming, and may in turn be used to support various higher-level operations for data movement and global process control. This paper presents efficient algorithms to implement multicast communication in wormhole-routed direct networks, in the absence of hardware multicast support, by exploiting the properties of the switching technology. Minimum-time multicast algorithms are presented for n-dimensional meshes and hypercubes that use deterministic, dimension-ordered routing of unicast messages. Both algorithms can deliver a multicast message to m-1 destinations in [log ₂ m] message passing steps, while avoiding contention among the constituent unicast messages. Performance results of implementations on a 64-node nCUBE-2 hypercube and a 168-node Symult 2010 2-D mesh are given     

Application of carbon nanostructures toward SO2 and SO3 adsorption: a comparison between pristine graphene and N-doped graphene by DFT calculations

We studied the adsorption of SO_x (x?=?2,3) molecules on the surface of pristine graphene (PG) and N-doped graphene (NDG) by density functional theory (DFT) calculations at the B3LYP/6-31G(d) level. We used Mulliken and NBO charge analysis to calculate the net charge transfer of adsorbed SO_x on pristine and defected graphene systems. Our calculations reveal much higher adsorption energy and higher net charge transfer by using NDG instead of pristine graphene. Furthermore, the density of state (DOS) graphs point to major orbital hybridization between the SO_x and NDG, while there is no evidence of hybridization by using pristine graphene. Based on our results, it is found that SO₂ and SO₃ molecules can be adsorbed on the surface of NDG physically and chemically with adsorption energies (E_ads) of ?27.5 and 65.2?kJ?mol^?1 (19.6 and 51.4?kJ?mol^?1 BSSE), respectively, while low adsorption energies were calculated in the case of using pristine graphene. So we introduced NDG as a sensitive adsorbent/sensor for detection of SO₂ and SO₃.     

On the experimental and numerical investigation of clay/epoxy nanocomposites

Due to increasing use of clay/epoxy nanocomposites in industry, investigation of mechanical properties of clay nanocomposites has become of great interest. While the stiffening mechanism of clay nanocomposites is well documented, there is still not a clear understanding about how addition of clays affect the fracture behavior of clay/epoxy nanocomposites. The main aim of this paper is to measure and explain the effect of clays on ductility reduction of these nanocomposites. First, epoxy and clay/epoxy nanocomposites with different clay weight ratio were built. Then, the damage parameters of epoxy and clay/epoxy nanocomposites were measured by variation of the elasticity modulus. Based on loading–unloading experiments, the Lemaitre damage parameters for epoxy and clay/epoxy nanocomposites were extracted. Crack initiation and propagation in dog-bone sample were simulated for epoxy and clay/epoxy nanocomposites using the eXtended Finite Element Method (XFEM). The comparison between experimental and numerical results shows that the proposed method can predict the crack initiation location and propagation path in clay/epoxy nanocomposites.     

Bioethanol Production from Saccharomyces cerevisiae through Conventional and Membrane Batch Fermentation: Experimental and Modeling Studies

Theoretical Foundations of Chemical Engineering - Kinetics for bioethanol production from glucose using Saccharomyces cerevisiae (PTCC 24860) was experimentally studied in a batch membrane...     

Development of a hybrid energy storage sizing algorithm associated with the evaluation of power management in different driving cycles

In this paper, a hybrid energy storage sizing algorithm for electric vehicles is developed to achieve a semi-optimum cost-effective design. Using the developed algorithm, a driving cycle is divided into its micro-trips and the power and energy demands in each micro-trip are determined. The battery size is estimated because the battery fulfills the power demands. Moreover, the ultra-capacitor (UC) energy (or the number of UC modules) is assessed because the UC delivers the maximum energy demands of the different micro-trips of a driving cycle. Finally, a design factor, which shows the power of the hybrid energy storage control strategy, is utilized to evaluate the newly designed control strategies. Using the developed algorithm, energy-saving loss, driver satisfaction criteria, and battery life criteria are calculated using a feed-forward dynamic modeling software program and are utilized for comparison among different energy storage candidates. This procedure is applied to the hybrid energy storage sizing of a series hybrid electric city bus in Manhattan and to the Tehran driving cycle. Results show that a higher aggressive driving cycle (Manhattan) requires more expensive energy storage system and more sophisticated energy management strategy.     

Evaluation of a new solid boundary implementation in the lattice Boltzmann method for porous media considering permeability and apparent slip

The accuracy of solid wall treatment in the lattice Boltzmann method (LBM) simulation of porous structures affects different hydraulic parameters including integral properties, such as permeability, or local phenomena, such as apparent slip. Based on an analysis of the advantages and disadvantages of the current methods, a new technique is introduced for exact boundary extraction from binary representation. Using this technique, the LBM model can simultaneously benefit from the advantages of existing approaches, i.e. the real micro-/nanostructure obtained with X-ray computed tomography, and a reduction in the resolution requirement. To evaluate the technique, permeability and slip length on the solid walls are investigated for a porous gas diffusion layer. The results show acceptable accuracy improvement balanced with computational costs.     

Fluid-Film Lubrication Regimes Revisited

Dimensionless film-thickness equations are provided for four fluid-film lubrication regimes found for nonconformal surfaces. Side-leakage effects are present in the equations. These regimes are isoviscous-rigid; piezoviscous-rigid, isoviscous-elastic, or soft EHL; and piezoviscous-elastic, or hard EHL. The influence or lack of influence of elastic and viscous effects is a factor that distinguishes these regimes. The film-thickness equation for the respected regimes come from earlier theoretical studies. Results are presented as a map of the lubrication regimes, with film thickness contours on a log-log grid of the viscosity and elasticity parameters for four values of the ellipticity parameter.     

Reduction of the computational time and noise filtration in the IHCP by using the proper orthogonal decomposition (POD) method

In this paper a comprehensive investigation is performed on the application of the proper orthogonal decomposition (POD) in solving the inverse heat conduction problems (IHCP's). Inverse heat conduction problems are those in which the boundary or the initial condition, or the physical properties are unknown; instead temperature measurements at some locations within the body are available. Two approaches may be taken to solve these mathematically ill-posed problems. These approaches are classified according to whether the problem is modeled using a PDE or a matrix form. In this paper a new approach, namely the proper orthogonal decomposition (POD) method is used for solving the IHCP. In this approach, based on the dynamics of the problem on hand, two courses of action may be taken: In the first course the governing PDE is reduced to a system of ODE's which reduces the computational time considerably. In the second approach the ill-conditioned matrix is modified using the singular value decomposition (SVD) which reduces the destructive effects of the random noise of the temperature data. Although both courses of action introduce a bias error into the inverse solution, the considerable reduction in the variance error in the solution makes the method to be quite attractive. In this paper these points are demonstrated through solutions obtained for two standard IHCP's. The results obtained by the methods suggested in this paper are compared with those obtained by the well-known conjugate gradient and Tikhonov regularization methods.     

Adaptive Wormhole Routing in Hypercube Multicomputers

Both adaptive unicast routing and efficient multicast communication have been shown to be important to the performance of distributed-memory multiprocessors, or multicomputers. In this paper, we propose a uniform adaptive routing strategy for wormhole-routed hypercube networks that accommodates both unicast and multicast communication. Based on a node labeling method, the resultant routing algorithms are shown to be deadlock-free without requiring virtual channels. We present an optimal ordering algorithm that minimizes the traffic generated under the proposed paradigm. A greedy algorithm with less time complexity is also proposed.