
Oscillating sphere
A 30,000-element dual-BIE FMM model is compared with angular and frequency-domain analytical solutions from 20–1000 Hz.
Verification and accuracy
The cases cover exterior and interior domains, radiation and scattering, pressure, velocity and impedance boundary conditions, and low-to-high frequency behavior.
Primary benchmark
A monopole at (R/2, 0, 0) produces a nonuniform velocity boundary condition on a unit sphere. Benchmark models range from 588 to 14,520,000 DOF at ka=2 and 20.

Computed pressure on the sphere at ka=20.

Boundary and field-surface pressure for the same radiation case.
V9 performance
Using four of twenty threads on an Intel Core i7-13800H laptop, the 14.52-million-DOF FMM case required 580.97 seconds at ka=2 and 1,102.88 seconds at ka=20 in the current V9 benchmark workbook.
These values supersede the older 15- and 20-minute wording previously shown on this page.

Radiation and interior fields
Scattering
10,800-element FMM model with a reported maximum polar-pressure error of 0.18%.
View comparison187,500-element ACA model with a reported maximum polar-pressure error of 3.21%.
View comparisonPressure-release boundary case with a reported maximum polar-pressure error of 0.45%.
View comparisonPressure contours show the incident and scattered-field interaction around the sphere.
View contourSources and boundaries
A 10,800-element rigid cube compares frequency responses at a field point with analytical monopole and dipole solutions from 20–500 Hz.
View pressure contourA 51,600-element ACA model at ka=100 reports less than 2% maximum pressure error along the channel axis.
Watch animationAn 8,400-element FMM model at ka=30 reports 1.85% maximum pressure error against the analytical solution.
View comparisonThe HFBEM predicts far-field SPL from 20–2000 Hz using less than 1% of the reported conventional-BEM CPU time for this case.
View comparisonUse the Verification Manual and included sample models to examine inputs, analytical references, and expected results.