A Distributed Routing Method for Dynamic Multicasting

Single point, sender based control does not scale well for multicast delivery. For applications, such as group video or teleconferencing a low total cost multicast tree is required. In this article we present a destination driven algorithm to minimize the total tree cost of multicast tree in a dynamic situation for the whole session duration. In this heuristic approach we considered the staying duration of participants are available at the time of joining. The performance of our algorithm is analyzed through extensive simulation and evaluated against several other existing dynamic multicast routing and also against one well known near optimum heuristic algorithm used for solving Steiner tree problem. We have further tested our algorithm using erroneous information given by the joining participants. Simulation results show that its performance does not degrade that much even when the range of error is considerably high, which proves the robustness of our algorithm.     

Studies on turbulent momentum, heat and species transport during binary alloy solidification in a top-cooled rectangular cavity

A two-dimensional transient fixed-grid enthalpy-based numerical method is developed to analyze the effects of turbulent transport during a binary alloy solidification process. Turbulence effects are introduced through standard k-ε equations, where coefficients are appropriately modified to account for phase-change. Microscopically-consistent estimates are made regarding temperature-solute coupling in a non-equilibrium solidification situation. The model is tested against laboratory experiments performed using an NH₄Cl-H₂O system in a rectangular cavity cooled and solidified from the top. Particular emphasis is laid on studying the interaction between Rayleigh-Benard type convection and directional solidification in the presence of turbulent transport. Numerical predictions are subsequently compared with experimental results regarding flow patterns, interface growth and evolution of the temperature field, and the agreement is found to be good.     

Resource contention in shared-memory multiprocessors: A parameterized performance degradation model

A standard metric conventionally employed to compare the performance of different multiprocessor systems is speedup. Although providing a measure of the improvement in execution speed achievable on a system, this metric does not yield any insight into the factors responsible for limiting the potential improvement in speed. This paper studies the performance degradation in shared-memory multiprocessors as a result of contention for shared-memory resources. A replicate workload framework with a flexible mechanism for workload specification is proposed for measuring performance. Two normalized performance metrics—efficiency and overhead factor—are introduced to quantify the factors limiting performance and facilitate comparison across architectures. Finally, the proposed model is employed to measure and compare the performance of three contemporary shared-memory systems, with special emphasis on the newly released BBN Butterfly-II (TC2000), currently undergoing Beta test.     

Numerical models of creep and boundary sliding mechanisms in single-phase, dual-phase, and fully lamellar titanium aluminide

Finite element simulations of the high-temperature behavior of single-phase γ, dual-phase α₂+γ, and fully lamellar (FL) α₂+γTiAl intermetallic alloy microstructures have been performed. Nonlinear viscous primary creep deformation is modeled in each phase based on published creep data. Models were also developed that incorporate grain boundary and lath boundary sliding in addition to the dislocation creep flow within each phase. Overall strain rates are compared to gain an understanding of the relative influence each of these localized deformation mechanisms has on the creep strength of the microstructures considered. Facet stress enhancement factors were also determined for the transverse grain facets in each model to examine the relative susceptibility to creep damage. The results indicate that a mechanism for unrestricted sliding of γ lath boundaries theorized by Hazzledine and co-workers leads to unrealistically high strain rates. However, the results also suggest that the greater creep strength observed experimentally for the lamellar microstructure is primarily due to inhibited former grain boundary sliding (GBS) in this microstructure compared to relatively unimpeded GBS in the equiaxed microstructures. The serrated nature of the former grain boundaries generally observed for lamellar TiAl alloys is consistent with this finding.