ASA 126th Meeting Denver 1993 October 4-8

4pED7. Linearized least-squares data analysis.

S. L. Garrett

Phys. Dept.---Code PH/Gx, Naval Postgraduate School, Monterey, CA 93943

One of the most attractive features of many acoustic teaching laboratory experiments is their ability to produce high-quality (precision) data. This affords the possibility of also using the results of these laboratory exercises to teach advanced techniques for data analysis that exploit the ubiquity of least-squares data analysis routines which are now included in most hand-held calculators and all personal computer plotting packages. This talk will concentrate on the transformation of general two-parameter nonlinear equations into linear forms suitable for least-squares-fitting techniques. This transformation technique will be demonstrated to (i) extract the effective moving mass of the spring to correct the simple (omega)=(k/m)[sup 1/2] expression for the simple harmonic oscillator; (ii) extract the distance from the end of the pendulum string to the pendulum bob center-of-mass in the simple pendulum experiment; (iii) extract the location of the acoustic center of a projector from measurement of the free-field pressure versus separation; and (iv) extract the cutoff frequency and thermodynamic sound speed from the phase speed versus frequency measurement in a water-filled, pressure-released waveguide.