Attenuator Reduction Values

The sound reduction that can be achieved by a sound attenuator, is heavily dependent on the sound field inside a duct. Different noise sources will excite different sound fields, especially above ' cut-off' frequency. 'Cut-off' frequency is the first frequency the sound wave changes character from being only plane wave modes to also become cross modes, and various combinations of these modes are established depending on size and shape of the duct.

The sound reduction that can be achieved by a sound attenuator is limited by the nature of the duct arrangement/configuration. Measurements taken on site often vary from those taken in laboratories, and can be affected by:
  • Flanking noise
  • Background noise
  • Resonance noise
  • Vibration noise
  • Reflective surfaces
  • Etc.
These factors are not always taken into consideration, but can result in significantly different measurements.

Laboratory measurements often vary from site measurements. The most commonly accepted international standard for testing ducted sound attenuators is ISO 7235. Testing is done indirectly by measuring the noise level both with and without the attenuator. The difference in noise level between the two measurements represents the actual attenuator reduction value, or insertion loss value. Site measurements, on the other hand, are normally made directly, by measuring the improvement after the introduction of the attenuators.

ISO 7235 is 'ideal' in that all 'disturbing' factors (flanking noise, background noise etc.) that may influence the measurement are checked and corrected. Laboratory testing will give more exact theoretical reduction values for the respective attenuator, but does not necessary give the correct values for the actual site arrangement.

A loud speaker is used as noise source for testing of sound attenuators according to ISO 7235. Other noise sources, such as ventilation fans, generators or compressors will propagate very different sound fields, and therefore yield different attenuator reduction values. Because of these reasons, it is often difficult to evaluate the expected resulting noise levels by purely considering the data quoted in suppliers' catalogs. One must also consider the site arrangement and dynamic conditions.