Abstract

In 2014, the Federal Aviation Administration (FAA) amended the Airport Improvement Program (AIP) Handbook to clarify guidance for the funding of noise mitigation projects. The clarification addressed requirements that structures eligible for sound insulation treatment not only be located within the day-night average sound level (DNL) 65-decibel (dB) noise contour, but also experience existing interior noise levels that are 45 dB or greater with the windows closed. This restated guidance placed greater emphasis on the accuracy in measuring the existing noise reduction of a structure.

Subsequent research by the FAA to develop best practices for measuring the aircraft noise reduction of building facades has resulted in the development of a new SAE Recommended Practice, ARP 6973, that provides additional details for application to practical situations. Like its predecessor, ASTM E966, the method requires the use of an outdoor loudspeaker. However, there are situations encountered in the field, such as areas with limited exterior space, where it is either difficult to actually position a loudspeaker, or to generate a uniform distribution of sound energy over the test facade.

In an attempt to overcome these limitations, an alternative method for measuring noise reduction was investigated, with a loudspeaker placed inside the house and the noise reduction measured from the inside to the outside. This approach removes restrictions on the placement of an external loudspeaker. The goal of this project was to conduct a feasibility study of such an indoor-outdoor test procedure for measuring noise reduction in buildings.

Following a series of preliminary parameter tests designed to provide the necessary details to define an indoor-to-outdoor (I-O) test procedure, field measurements were conducted in 10 rooms of three houses to demonstrate equivalence of the results with those of the standard outdoor-to-indoor (O-I) method. The differences between the O-I and I-O measured values of noise reduction for the 10 rooms tested were all well within ±1 dB, the average difference being 0.1 dB.