Sound absorption coefficient

[chickenfingers]

Hi

Does a sound abs. coefficient of 1 for a frequency f, means that 100% of the incident sound energy at that frequency is absorbed? If not, what percentage is absorbed and what formulae/reasoning can I use to derive the percentage of absorbed energy?

cheers

Sergio

 

[mike c]

i think you got it right.

1.0 coefficient is a 100% absorption at X frequency

 

[Glenn Kuras]

correct!!

Glenn

 

[Glenn Kuras]

Guys, I just got a email from Bryan (who works for me) that he does not agree 100% with this. The way to calculate the absorption coefficient is, from "A" mount lab testing, is to take the sabins/squ footage. But Bryan, would like to add to this. Bryan????

Glenn

 

[bpape]

This is a common misconception. In theory, this should be correct but due to how measurements are taken and converted from Sabines to coefficients, it doesn't work out that way. It's actually a simple ratio rather than a percentage.

The reason it has issues is due to the edges of the samples. For instance, if I take a measurement of 80 sq ft of a product and get 80 sabines absorbtion at say 250Hz, my coefficient is 1.0 Now, if I get 90 sabines absorbtion, I get a ratio of 1.125 for the coefficient. How can we absorb more than 100%? We can't. However, the number is correct.

The problem is that when the formula is run, we divide sabines by the surface area of the face of the sample. This is the standard way to do measurements and calculations. However, if the sample is say 4" thick and all of the sides are exposed, and the samples are laid out as 2 rows of 5 2'x4' panels, then there is an additional 9.324 sq ft of surface area on the sides which is also absorbing and accounting for the additional sabines recorded - but this area is not figured into the coefficient calculation.

Does this mean that the measurements are not correct? Absolutely not. They are correct - in sabines. So, a 2'x4' panel with a coefficient over 1 is still providing the additional absorbtion.

When looking at materials, you must know how the tests were performed vs how you're going to use the materials. You should also always look at the sabines at the various frequencies to see what you're actually getting. Sabines per dollar is a good way to look at things in addition to sabines per unit area from a functionality standpoint.

Bryan

 

[Glenn Kuras]

When looking at materials, you must know how the tests were performed vs how you're going to use the materials. You should also always look at the sabines at the various frequencies to see what you're actually getting. Sabines per dollar is a good way to look at things in addition to sabines per unit area from a functionality standpoint.

This is a great point which I just want to make sure it is read. You should really go by sabines as you can have a product that has a absorption coefficient 1.0 at 80hz that is 1"x1". Looking at the sabins tells you how much absorption are getting for your dollar.

Glenn

 

[Savant]

Wow, Bryan. Thanks for that post. I would only add that even with the edges covered, the absorption coefficients measured in a reverb chamber - a.k.a. random incidence absorption coefficients or Sabine absorption coefficients - can still be greater than 1.0. Remember that the division by the area is, in essence, arbitrary. The misplaced equivalence of the Sabine absorption coefficients to the normal incidence absorption coefficients (the latter, typically measured in an impedance tube, can never be greater than 1.0) has its history in Sabine's early work to develop his now-famous equation, which is the basis for calculating absorption coefficients in the chamber method.

But, again, thanks for the post. I'm glad I'm not the only one preaching that gospel.

 

[Doc Roc]

Sorry to butt in guys. The topic is very interesting and I grasp about 60% of it. But luckily, that's about as much as I need . I have a slight tangent of a question. If you have a 2x4 sound panel made of, lets say IIG 1280 or Roxul 80. Is there any way, knowing a products absorption coefficient information at 2" and 4", to be able to tell (or come close to telling) what the coefficient info will be at 6"? Basically, They don't make the product at 6" thickness so there is not info on it, but I have made 6" panels for bass traps and would like to calculate an estimated absorption rating for the 1/3 band spectrum up to 4K. Thanks all!

 

[Savant]

Chris Whealy's Excel sheet can be used to make some predictions about the absorption for normal incidence. The challenge is the normal absorption coefficients are not easily translated to random incidence absorption coefficients, which are what are commonly published. (Like here.) A very rough guess about it can be made by using Whealy's spreadsheet to predict the normal absorption coefficients for a thickness with known random absorption coefficients. Then the ratio of the normal coefficients between two thicknesses can be used to calculate an (extremely) approximate random absorption coefficient for the thickness for which you do not have data. E.g.:

an(6-inch) / an(4-inch) ~= ar(6-inch) / ar(4-inch)

Where an represents the normal coefficients (from the Whealy spreadsheet) and ar represents the estimated (for the 6-inch) or measured (for the 4-inch) random coefficients.

Using the above Bob Golds link, I found the following for IIG 1280 at 125 Hz:

0.17 / 0.11 ~= ar(6-inch) / 1.11

ar(6-inch) = 1.71

Now, to me, this seems like a generous estimate (too high), which is what I would expect. But I think it at least represents a decent upper limit. I.e., ar(6-inch) for 125 Hz for 6" IIG 1280 is probably between 1.11 and 1.71.

HTH.

 

[vanderlae]

Acoustic Panel - Rockwool question

Hello

I am building some broadband acoustic panels 3' x 5' for a music studio which I will dot around the room and am keen to know which density of rockwool slabs would be most suitable. Rockwool is basically one of very few products available in Australia that is useful in this application.

Anyway, the supplier's technical data sheet lists several the varying coefficients depending on the thickness/density.

In short, this is what is given:

_____________________________________________________________
Nominal | Sound Absorption Coefficients
Density | Frequency (Hz)
_________250___500___1000___2000___4000 ____________________
40------|--0.56---0.90---0.88---0.85----0.88
60------|--0.70---1.19---1.04---1.14----1.06
80------|--0.91---1.19---1.20---1.07----1.05
100-----|--0.97---1.18---1.00---1.04----1.02
120-----|--0.74---1.20---1.20---0.98----1.14
----------------------------------------------------------------------

Which would you choose? I need to tame the lower frequencies of course given that these are the most difficult to treat. Pity it doesn't give a rating for <250 Hz. (I generally roll off anything below 50Hz in my mixes anyway but it's useful to know this nevertheless)

To me, the difference between the thicker 120 density and the 100 is intriguing in the lower frequencies!

And no, it didn't say which method they'd used other than "Sound absorption coefficients are based on 50mm thick measured in reverberation chamber in accordance with BS 36381987."

Thank you! I know I'm resurrecting the dead here...(post)

cheers
David

 

[Savant]

Rockwool composition tends to be fairly consistent from manufacturer to manufacturer. Why don't you compare these numbers with those of others on the Bob Golds page I linked above? Find some numbers that look similar for the bands you have and see if information at lower bands is available. This should at least give you some idea of performance.

 

[highfigh]

IIRC, the NRC is the percentage of a Sabine, so if the NRC is .7, it's 70% of 1 Sabine.

 

[Savant]

IIRC, the NRC is the percentage of a Sabine, so if the NRC is .7, it's 70% of 1 Sabine.

NRC is the average of the absorption coefficients in the 250, 500, 1000, and 2000 Hz bands. It is not a percentage of anything.

The absorption coefficients are, theoretically, percentages. However, absorption coefficients obtained from the reverberation room testing method (e.g., ASTM C423) are not percentages in terms of how they are calculated. Hence, there can be numbers that exceed 1.0, which is merely indicative of relative absorber performance. There are other threads that cover this in detail and I wrote an FAQ on it back when I was with Auralex.