Application of COMSOL to Acoustic Imaging

K. Mcilhany, and J.C Hernandez
U.S. Naval Academy, Annapolis, MD, USA
Published in 2010

Acoustic Imaging of hand movement is being studied with COMSOL and Matlab. A hardware implementation is being pursued that will be an array of 16x16 ultra-sonic speakers placed in a grid facing a similar grid of 16x16 microphones, operating at 40kHz. COMSOL is used to repeatedly calculate the diffraction pattern from a small scattering center, approximately 1.0cm in diameter. In conjunction with a hardware setup, COMSOL was instrumental in the design phase of the project, helping to establish running parameters such as wavelength, timing resolution as well as detector placement. The goal of the project is to collect wavefront arrival times which is then correlated to positions of the scattering centers. A neural network is employed to map the arrival time vector to the collection of scattering center positions. A human hand can be approximated as roughly 30, 1cm marble sized balls. The final position of the scattering centers enables software to reconstruct the hand position, making the device effectively an acoustic imager of hand movement, to be used for control systems.