Discretization effects in TLM diffusion models
โ Scribed by A. J. Wilkinson; S. H. Pulko; A. Saidane
- Publisher
- John Wiley and Sons
- Year
- 2000
- Tongue
- English
- Weight
- 167 KB
- Volume
- 13
- Category
- Article
- ISSN
- 0894-3370
No coin nor oath required. For personal study only.
โฆ Synopsis
The standard methods for bringing a TLM di!usion model to convergence are necessary but tedious. However, much time and e!ort can be saved if the initial choices of elemental size and iteration timestep are appropriate. Here we consider the general form of the solution to the di!usion equation and sample the component dynamics to determine the impulse invariant response. We then derive a convenient method for charting the discrepancy between the coe$cients of the continuous description and the impulse invariant transformation, as this discrepancy varies with discretization of space and time. The use of the curves in estimating appropriate elemental size and iteration timestep is described.
๐ SIMILAR VOLUMES
We re-examine the potential drop method to give a direct estimation of the absolute error in transmissionline matrix (TLM) diffusion modelling. In particular, the open-circuit boundary is taken into account. Promising results are presented with a structure relevant to thermal analysis of semiconduct
A new hybrid TLM-FDTD algorithm for solving di!usion problems is described. The method utilizes the transmission line model to de"ne the time step and the FDTD's leap-frog algorithm to determine the voltages and currents of the network analogue of the di!usion equation. Unlike the standard TLM metho
The TLM method, being a time domain method, cannot deal directly with frequency dispersive media. A way for modelling such media, for instance, dielectric dispersive ones, is proposed, starting from the causality relationship between "eld vectors D and E, which is discretized over the same space}tim
We show that the general boundary condition DaWax + a@ = 0 (D is the diffusion coefficient and a is a ## constant) in TLM diffusion modelling can be expressed accurately in terms of a voltage reflection coefficient p = ( A x -a A t ) / ( A x + aAt), where Ax is the spatial resolution and At is the