Computational tests on use of Rayleigh hypothesis in elliptical-waveguide analysis

A test model is developed for studying the validity of using the Rayleigh hypothesis in elliptical-waveguide analysis. The results show that only under certain well-behaved conditions can numerical accuracies of better than +or-0.5% be obtained and that unacceptable errors of more than +or-2% may occur when an inappropriate set of input-parameter values is fed into the software. Limits on the permissible ranges of values for these input parameters can be deduced from the empirical data generated by the test model.<
>