Thursday, November 1, 2012

Acoustical Corrections

Once again, I have been asked to explain how to correct sound data for size, in this case, the length of a linear slot diffuser. The basic equation is that sound output is proportional to 10*Log (A), where A is the area of the sound generating source. In practice, this means that for doubling the area, and at the same time doubling the flow rate, the measured (and reported) sound level would be expected to increase by 10* Log(2*A), which is 3 dB. That is the rule; add 3dB for doubling the area of a sound generating source.

This comes to play when looking at the sound traveling down a duct. If the duct is split and feeds two separate spaces, any sound in the duct would be expected to be 3dB lower at it enters the space. If, of course, the two ducts exit into the same space, the sounds would recombine, thus adding the 3dB back. This effect is the only sound parameter that is independent of frequency.

For continuous slot diffusers, we provide a correction table for length, for lengths up to 10 feet. Someone wanted to know the correction for longer lengths. Applying the above formula to the base data, which is based on a 4 foot length, going to 8 feet would increase the sound by 3dB. It would go up by another 3dB at 16 feet (all assuming the flow rate per foot is kept constant). The problem is that by the time we get to 16 feet, the observer is so far away from the added length that one can no longer hear the sound being generated there. So we stop the published correction at 10 feet.

The posted sound data for continuous slots assumes that the supply plenum isn’t adding any sound. In practice, however, it is likely that there is some noise added by the supply plenum. As a rule, the larger the plenum, the less noise it will create. A “step sided” plenum (wider than the opening at the top of the diffuser) will generate less noise than one that is only as wide as the opening. A taller plenum is quieter than a shorter one, and allows for a round, rather than an oval, inlet. An oval inlet has less area than a round one with the same perimeter. As a “rule of thumb”, add 1dB for every 100 fpm velocity above 800 inlet velocity into the plenum.

Finally, insulating the inside of a plenum decreases the interior volume, raising velocities in the plenum, and negating the sound absorption of the insulation. This is one reason that plenums are insulated with thin insulation, because increasing the thickness of the insulation will likely result in more sound generation. In practice, spaces with a return air plenum are seldom faced with condensation on the plenum, as the space (and therefore plenum) dew point is almost always higher than the supply air temperature in the plenum. If the space has ducted returns, I recommend field installed external insulation on all exposed surfaces, especially in humid climates.

Authored by: Dan Int-Hout, Chief Engineer Krueger