Gearslutz.com - View Single Post - Owen Corning 703 or subistitue in Canada?
View Single Post
Old 23rd April 2008   #6
avare
Lives for gear
 
avare's Avatar
 
Joined: May 2004
Location: Hamilton, On Canada
Posts: 4,445

Quote:
Originally Posted by danbronson View Post
Ack! Okay, I read through the thread. You weren't kidding, it's not intuitive.

So correct me if I'm wrong, but the thicker the absorber and air gap, the lower the frequencies you can absorb BUT as those factors change, the ideal absorber changes as well. Looks like for 6" thick with a 6" air gap behind, something with a gas flow resistivity of 6,000 rayls/m is ideal. That's not much at all! In fact, it seems to me that fluffy fiberglass might be pretty close to that.

Any idea how I can calculate the gas flow resistivity of different Roxul products to determine the ideal one?
My bolding of the correct answer. Having to get into the technical stuff, the theoretical goal is for the impedance at the surface of the absorber to match the acoustic impedance of air (412 Rayls). Fortunately the effect of surface impedance is relatively minor with the resistivities of the materials we use for absorption. This ignoring that the material has to be where the kinetic energy (velocity) is high.

As you have already figured out, getting data on the resistivity of various materials is very difficult. We use material density as an indicator of resistivity. Looking at OC data on absorption of various densities of materials, the trends that might be overlooked by someone not familiar with the physics involved would overlook. As an example how at 125 Hz 4" 703 is more efficient than 705. The trend between density and resistivity also varies between fiberglass and mineral wool, just to make things even more counter intuitive.

Coming back to what you asked about 6/6 absorbers, yes fluffy fiberglass is close to optimum theoretically. The reason I say theoretically, is that when working professionally in acoustics, I never want to recommend anything without test data to confirm what I recommend unless it can be interpolated from current test data or knowledgeably extrapolated.

I hope many people read this thread and get an idea why a question that seems simple to answer is not.

Andre
avare is offline  
Reply With Quote