B. Butrylo, F. Musy, L. Nicolas, R. Perrussel, R. Scorretti and C. Vollaire
This paper presents new trends in parallel methods used to solve finite element matrix systems: standard iterative and direct solving methods first, and then domain decomposition…
Abstract
This paper presents new trends in parallel methods used to solve finite element matrix systems: standard iterative and direct solving methods first, and then domain decomposition methods. For example, the current status and properties of two prevailing programming environments (PVM and MPI) are finally given and compared when implemented together with a finite element time domain formulation.
Details
Keywords
Andrzej Jordan and Carsten Maple
Discusses a parallel algorithm for the finite‐difference time domain method. In particular, investigates electromagnetic field propagation in two and three dimensions. The…
Abstract
Discusses a parallel algorithm for the finite‐difference time domain method. In particular, investigates electromagnetic field propagation in two and three dimensions. The computational intensity of such problems necessitates the use of multiple processors to realise solutions to interesting problems in a reasonable time. Presents the parallel algorithm with examples, and uses aspects of graph theory to examine the communication overhead of the algorithm in practice. This is achieved by observing the dynamically changing adjacency matrix of the communications graph.
Details
Keywords
Jerzy Golebiowski and Robert Piotr Bycul
The purpose of this paper is to prepare procedures for determination of characteristics and parameters of DC cables on the basis of transient and steady thermal field distribution…
Abstract
Purpose
The purpose of this paper is to prepare procedures for determination of characteristics and parameters of DC cables on the basis of transient and steady thermal field distribution in their cross-sections.
Design/methodology/approach
Steady-state current rating was computed iteratively, with the use of steady thermal field distribution in the cable. The iterative process was regulated with respect to this field by changes of the mean surface temperature of the sheath of the cable. It was also controlled with respect to the unknown current rating by deviations of the temperature of the core from the maximum sustained temperature of the insulation (material zone) adjacent to the core. Heating curves were determined (in arbitrarily selected points of the cross-section of the cable) by a parallel algorithm described thoroughly in the first part of the paper. The algorithm was used for computing of transient thermal field distribution throughout the whole cross-section. Thermal time constant distributions were determined by the trapezium rule, where the upper integration limit of respective thermal field distributions was being changed.
Findings
Using the methods prepared the following characteristics/parameters of the cable were determined: steady-state current rating, spatial-time heating curves, mean thermal time constant distribution. The results were verified and turned to be in conformance with those of the IEC 287 Standard and a commercial software – Nisa v. 16. Speedup and efficiency of the parallel computations were calculated. It was concluded that the parallel computations took less time than the sequential ones.
Research limitations/implications
The specialized algorithms and software are dedicated to cylindrical DC cables.
Practical implications
The knowledge of the determined characteristics and parameters contributes to optimal exploitation of a DC cable during its use.
Originality/value
The algorithms of determination of the steady-state current rating and thermal time constant are original. The software described in the appendix has also been made by the authors.