Frederic Laville
Jean Nicolas
Dept. of Mech. Eng., Sherbrooke Univ., Sherbrooke, PQ J1K 2R1, Canada
The standardization of sound power determination using sound intensity measurements is a much more difficult task than in the case of pressure measurements. The difficulty comes from the fact that both the precision of each intensity measurement and of the resulting sound power depend strongly on the acoustic field. This effect has been analyzed with a computer simulation that has a broader scope than previous ones that were restricted to the evaluation of the sampling error for the source field only. Because the intensity estimated with two microphones is computed for sources represented by monopole arrays located inside as well as outside the measurement surface, the effects of instrumentation error and background noise field are included. The discrete point measurement standards proposed by ISO and ANSI are critically examined with the computer simulation. Among other results, it is shown how F[sub 3] as defined in the ISO standard is a good indicator of the phase mismatch error but should not be used in conjunction with F[sub 2] as suggested in the standard at this approach leads to variable error limits. [Work supported by IRSST.]