Thanks.Okay, dudes. I will try to get the info scanned in before I head out
Andre
Questions about using Glass Sliding Doors (CR to ISO)
-
ejbragg
- Posts: 44
- Joined: Wed Sep 08, 2004 4:46 am
- Location: Fort Smith, AR
formulas
Okay, guys, I have the book with me. I will bring it on my little vacation. Right now, I have a meeting, and I'll be pretty busy.
Before I go and paste pages up, you might decide whether or not you're really interested.
I don't have a scanner handy, so I typed up a few line of the info, pertaining only to region 3 treatment. This should give you an idea of what the book is like. Now I realize this is going to create more questions than it answers. And the truth is, you can chase down answers in these books all day and still not have a clear answer. Before you get too caught up in the idea of digging up everything you are sure to be led to, let me warn you that unless you want to compute particle velocities using Maxwell's equations, I would consider getting only what you need out of this, else I should have to paste all 430 pages of this book and you will STILL not have enough detail for a solid answer! This is why so many of the acoustic companies provide vague details. It's usually not enough, but it's hard to know where to draw that line between being helpful and merely blindsiding someone with a computational monstrosity.
The info below is fairly tame, but it's only the tip of the tiger's tail. Some of the terms such as Young's modulus and Poisson's ratio are not listed in the book, but they are common mechanical engineering terms. In fact, I'll wager they can be found on the internet.
* * * * * * * * * * * * * * * *
Panel bending stiffness per unit width:
(E*I)/b * (1-u^2), where
E = Young's Modulus
u = Poisson's ratio
I/b = (t^3)/12, the area moment of inertia per unit width b of the panel
t = panel thickness
The panel bending stiffness has a negligible effect on transmission loss at frequencies below a "critical" frequency, fc. Therefore, the theoretical transmission loss in the "mass controlled" [Region III] region can be derived by considering only the acoustic pressure and normal component of velocity acting on a small mass element of the panel.
[Here's a good place to refer back to the very first diagram that Steve linked to, above. The characteristics shown describe a panel that was designed to stop those frequencies in the middle of the diagram. What happens on the two ends is how this particular barrier breaks down in its efforts to stop these vibrations when they are too low or too high.]
The transmitted wave angle is equal to the incident wave angle; the ratio of transmitted intensity to incident intensity [at the barrier surface] is given by:
tau = 1 / (1 + ((w*ps*cos (theta) / (2*pc))^2), where
ps = mass of panel per unit surface area
w = frequency of sound wave, in radians per second
theta = angle of incidence of impinging sound wave, in radians
pc = characteristic impedance of medium (415 rayls for air at std temp & press)
The transmission for normal incidence is given by:
TL0 = 10 log(1/tau) = 10 log(1 + ((w*ps)/(2*pc))^2) dB
If the transmission coefficients for angles of incidence from 0 to 90 degrees are averaged, the random-incidence transmission loss is given by:
TL ~ TL0 - 10 log(0.23 TL0) dB, [the '~' means approximately equal to]
for values of TL0 greater than 15 dB. In many practical applications, the "field incidence" transmission loss, obtained for angles of incidence from 0 to 78 degrees, is more accurate:
TL ~ TL0 - 5 dB.
* * * * * * * * * * * * * * * *
What this means:
TL0 is the transmission loss of a certain barrier of certain material, by direct angle of incidence. So once you computer TL0, just subtract 5 from this number to get the more accurate loss.
TL0 is a formula that applies to a certain material, applying a certain frequency. Or you can graph the whole spread of freqs. BUT .... you have to know where NOT to use this calculation, because the TL given by this formula will not be accurate for those frequencies.
You'll have to have the mass of the material before you can compute TL0. In this case, it's glass. Also, it looks like you'll need the characteristic impedance of glass.
Before I go and paste pages up, you might decide whether or not you're really interested.
I don't have a scanner handy, so I typed up a few line of the info, pertaining only to region 3 treatment. This should give you an idea of what the book is like. Now I realize this is going to create more questions than it answers. And the truth is, you can chase down answers in these books all day and still not have a clear answer. Before you get too caught up in the idea of digging up everything you are sure to be led to, let me warn you that unless you want to compute particle velocities using Maxwell's equations, I would consider getting only what you need out of this, else I should have to paste all 430 pages of this book and you will STILL not have enough detail for a solid answer! This is why so many of the acoustic companies provide vague details. It's usually not enough, but it's hard to know where to draw that line between being helpful and merely blindsiding someone with a computational monstrosity.
The info below is fairly tame, but it's only the tip of the tiger's tail. Some of the terms such as Young's modulus and Poisson's ratio are not listed in the book, but they are common mechanical engineering terms. In fact, I'll wager they can be found on the internet.
* * * * * * * * * * * * * * * *
Panel bending stiffness per unit width:
(E*I)/b * (1-u^2), where
E = Young's Modulus
u = Poisson's ratio
I/b = (t^3)/12, the area moment of inertia per unit width b of the panel
t = panel thickness
The panel bending stiffness has a negligible effect on transmission loss at frequencies below a "critical" frequency, fc. Therefore, the theoretical transmission loss in the "mass controlled" [Region III] region can be derived by considering only the acoustic pressure and normal component of velocity acting on a small mass element of the panel.
[Here's a good place to refer back to the very first diagram that Steve linked to, above. The characteristics shown describe a panel that was designed to stop those frequencies in the middle of the diagram. What happens on the two ends is how this particular barrier breaks down in its efforts to stop these vibrations when they are too low or too high.]
The transmitted wave angle is equal to the incident wave angle; the ratio of transmitted intensity to incident intensity [at the barrier surface] is given by:
tau = 1 / (1 + ((w*ps*cos (theta) / (2*pc))^2), where
ps = mass of panel per unit surface area
w = frequency of sound wave, in radians per second
theta = angle of incidence of impinging sound wave, in radians
pc = characteristic impedance of medium (415 rayls for air at std temp & press)
The transmission for normal incidence is given by:
TL0 = 10 log(1/tau) = 10 log(1 + ((w*ps)/(2*pc))^2) dB
If the transmission coefficients for angles of incidence from 0 to 90 degrees are averaged, the random-incidence transmission loss is given by:
TL ~ TL0 - 10 log(0.23 TL0) dB, [the '~' means approximately equal to]
for values of TL0 greater than 15 dB. In many practical applications, the "field incidence" transmission loss, obtained for angles of incidence from 0 to 78 degrees, is more accurate:
TL ~ TL0 - 5 dB.
* * * * * * * * * * * * * * * *
What this means:
TL0 is the transmission loss of a certain barrier of certain material, by direct angle of incidence. So once you computer TL0, just subtract 5 from this number to get the more accurate loss.
TL0 is a formula that applies to a certain material, applying a certain frequency. Or you can graph the whole spread of freqs. BUT .... you have to know where NOT to use this calculation, because the TL given by this formula will not be accurate for those frequencies.
You'll have to have the mass of the material before you can compute TL0. In this case, it's glass. Also, it looks like you'll need the characteristic impedance of glass.
-
AVare
- Confused, but not senile yet
- Posts: 2336
- Joined: Thu Feb 05, 2004 1:56 pm
- Location: Hanilton, Ontario, Canada
-
ejbragg
- Posts: 44
- Joined: Wed Sep 08, 2004 4:46 am
- Location: Fort Smith, AR
-
fbars
- Posts: 28
- Joined: Fri Jan 16, 2004 12:16 am
- Location: Illinois
Hi all,
Last night was the first chance I had to actually measure the db level of the doors. With the meter set for slow with "C" weighting, I measured a level of 91 db in the control room. Which is WAY louder than I would ever mix at. With the doors closed, and at a distance of 2 inches from the glass, the loudest I was able to measure was 50 to 51 db. My inexpensive RS meter (yeah I know) doesn't read any lower.
As time allows, I will try to do a more accurate and detailed test. If anyone has a more scientific method let me know. I hope to gain the use of a B & K meter which has Leq, and can check it's calibration before testing.
Tom
First Bass Audio
Last night was the first chance I had to actually measure the db level of the doors. With the meter set for slow with "C" weighting, I measured a level of 91 db in the control room. Which is WAY louder than I would ever mix at. With the doors closed, and at a distance of 2 inches from the glass, the loudest I was able to measure was 50 to 51 db. My inexpensive RS meter (yeah I know) doesn't read any lower.
As time allows, I will try to do a more accurate and detailed test. If anyone has a more scientific method let me know. I hope to gain the use of a B & K meter which has Leq, and can check it's calibration before testing.
Tom
First Bass Audio
-
ejbragg
- Posts: 44
- Joined: Wed Sep 08, 2004 4:46 am
- Location: Fort Smith, AR
as promised
Sorry for the delay. Could not get away to do any scanning during my vacation. (I was just enjoying myself too much.)
I will leave this info posted for a short time, only, as I don't wish to get in trouble (if you know what I mean). I put this portion of the doc in Word format. There are 30 pages scanned at high res, shrunk to lower res jpeg. This file is about 3.2 MB, so be patient.
http://www.fountainsquarehouse.com/assets/Mech_eng.doc
Hope it helps somebody (besides just Andre)
Eric
I will leave this info posted for a short time, only, as I don't wish to get in trouble (if you know what I mean). I put this portion of the doc in Word format. There are 30 pages scanned at high res, shrunk to lower res jpeg. This file is about 3.2 MB, so be patient.
http://www.fountainsquarehouse.com/assets/Mech_eng.doc
Hope it helps somebody (besides just Andre)
Eric