5.6
Test Problems
Table 5.A shows the results from computing the
potentials exterior to a square. The results from Table 5.B
show similar results for the exterior to a sphere. The
noticeably less-accurate results in Table 5.B can be attributed
to the crude approximation that is made to the sphere.
In Table 5.C, the results of Table 5.B are repeated but
the sound pressures are given in terms of decibels and phases. This
demonstrates another important way of interpreting the results
from the subroutines and also shows that
often the numerical error appears far less significant
when it is viewed in decibel units.
In the final test problem the important Helmholtz
property of the radiation ratio is considered.
In Figures 5.1 and 5.2 the computed versus exact radiation
ratios for pulsating and oscillating spheres
across a wide wavenumber range are given
where the Helmholtz field is taken to be axisymmetric.
The results show some numerical drift from the exact solution
with increasing wavenumber. As a general rule, the accuracy of
the BEM solution of Helmholtz problems slowly deteriorates with
frequency if the same boundary element mesh is used throughout.
It may often be necessary to use finer meshes at higher wavenumbers.
Return to boundary-element-method.com/helmholtz
Return to boundary-element-method.com