Discussion of impacts of measurement error in architectural acoustics performance specifications

Abstract

Customers for acoustic products need manufacturers and professionals to provide performance specifications to allow informed design decisions. Acoustic professionals rely upon standard tests and special labs to quantify products. Data published by the ASTM for E90 shows that an STC rating for a single construction can have a variability range of up to 7 points. Being logarithmic, this equates to a sound pressure range of more than 2×, larger than 100%. The log nature of many acoustical performance metrics hides this from most. This large variability causes challenges for the architectural acoustics community. The primary negative impact: customers do not trust the acoustical ratings manufacturers provide. This is not an issue of the manufacturers providing inaccurate information, it is an issue of under-reported and incomplete understanding of the uncertainty in the metrics themselves. For perspective, NRC measured the Young’s Modulus of OSB material to have a mean of 6.8 × 109 and a standard deviation of 1.5 × 108 (Ref: NRC IR-766, Table 22). This yields a measurement error of 3.8%. It is simple to account for this in the design process. If the metric has an uncertainty as large as the acoustic metrics, the standard paradigm for accounting for error not only breaks down, it incentivizes bad actors: cherry picking, lab shopping, etc. More open acknowledgement of the measurement error in architectural acoustic metrics would increase the confidence customers have in them and would allow designers to make more informed decisions.