What is it about?
FDTD acoustic simulation works by solving the acoustic wave equations at every cell of a 3D grid containing the scattering object of interest. However if the outer edges of the entire grid are not treated in a special way, outgoing waves would reflect back inside the computation domain. A "perfectly matched layer" was previously proposed to absorb these outgoing waves, but in the case of 3D acoustics the value of the "loss factor" was not previously specified. Here we show how to set that loss parameter to an optimum value, to ensure that outgoing waves are maximally absorbed.
Featured Image
Why is it important?
In computer simulation of acoustics, it is critical to ensure that the waves calculated inside the computation domain originate either from the sound source or from reflections off the object of interest. Any reflections that may bounce back from the edges of the computation domain are purely artefactual and must be minimized, otherwise they would contaminate the signals of interest and make it difficult to interpret the simulation results.
Perspectives
I hope that this paper will help anyone who tries to use FDTD acoustic simulation, as it seems that prior to this work, people would set the value of the loss factor either arbitrarily or after a great deal of trial and error.
Parham Mokhtari
National Institute of Information and Communications Technology
Read the Original
This page is a summary of: Optimum Loss Factor for a Perfectly Matched Layer in Finite-Difference Time-Domain Acoustic Simulation, IEEE Transactions on Audio Speech and Language Processing, July 2010, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/tasl.2009.2035036.
You can read the full text:
Contributors
The following have contributed to this page
