Dynamic insertion loss measurements of sound attenuators at low frequencies

Abstract

For decades the duct silencer industry has relied on ASTM E 477 as a standard for measuring the acoustic and aerodynamic properties of passive duct silencers. Historically, there has been significant concern regarding the accuracy and reliability of the measured dynamic insertion loss values for duct silencers at 125 Hz and below. The uncertainty at low frequencies has been blamed on several factors including: (1) inaccurate determination of sound level in reverberation chambers due to inaccurate spatial averaging (e.g., low modal density), and (2) varying acoustic output power from the electronic signal source due to a different impedance seen by the loudspeaker (with and without the test specimen in the duct). This paper presents a third reason for the low-frequency variability, which is likely to be more significant than all others combined. It will show that the problem is caused by acoustic resonances in the empty test duct, and these resonances will occur in every facility that is constructed in accordance with the test standard. The authors will present test data collected in a NVLAP accredited laboratory verifying this low-frequency effect, and also recommend specific changes to ASTM E477 which could eliminate this problem.