Source Side to Receiving Side Performance

[SGleason]

We advocate what we like to call an "unbalanced wall", that is, the sound attenuation component is deployed under GWB on the source side only. That component could be RC, a GG sandwich, a layer of specialty drywall such as QR ( as a single layer) or maybe a layer of that odd cellulose sound board. These walls are always tested from the source side only and I have yet to find any real data where an "unbalanced wall" was tested from the "receiving side" as well.

Here is the question:

In a scenario where the use pattern dictates that the wall assembly needs to perform equally well from both sides, such as practice rooms, would you theoretically expect that "unbalanced wall" to function as a "two way" assembly? Or, conversely, would you recommend that the sound attenuation component be deployed on both sides of the wall?

I appreciate any insights.

Steve

[Soundman2020]

I'm not sure if you've seen the Wyle report from way back in 1973, Steve. It's a very comprehensive study of the theory and practice of walls meant for high isolation levels.

It's pretty long and detailed, but well worth taking the time to work your way through.

One of the conclusions they arrived at, is that for a two leaf wall you get maximum isolation when the wall has the same mass on both leaves, and for a three leaf wall you get maximum isolation when the middle leaf has twice the mass of the outer leaves, both of which have the same mass as each other, and the same gap to the middle leaf. In other words, a wall where the middle leaf is four layers of drywall and the leaves on each side of that are both 2 layers of drywall, spaced 6" away from the middle leaf.

You can download it from here: https://app.box.com/shared/jcaoavdc8g

They did not test walls with fiberboard, as far as I recall, but I don't see any reason why it would be different. You'd have to test it in a proper lab to be certain, but I would suspect that you'd get the best overall isolation if you have a sheet of Homasote on each leaf, in addition to the actual mass layer, as opposed to having Homasote on only one leaf.

Maybe Andre can shed more light on that.


- Stuart -

[SGleason]

Thanks for that reply, Stuart. I actually had the Wyle report up in front of me all day yesterday. Interesting use of PVC spacers between the studs and the GWB. I actually have some very, very old tests here in our library that make good reading as well. It is interesting to see what folks were doing back in the 1950's and 60's to experiment with sound attenuation in walls.

I would suspect that you'd get the best overall isolation if you have a sheet of Homasote on each leaf, in addition to the actual mass layer, as opposed to having Homasote on only one leaf.

With regard to that, we have a steel stud (25 ga.) test that shows no gain when Homasote is installed on the receiving side as well as the source side. I don't think the dynamics are the same, however, with wood studs.

Let me be more specific about my original question. With a two leaf assembly that has Homasote /GWB....or perhaps GWB/GG/GWB.....on just one side and (1x) GWB on the other, would you expect that wall to perform equally well from both directions?

Steve

[DanDan]

I am not familiar with some of the acronyms. But I think it worth noting the conditions in the source and receive spaces. There is a theoretical 6dB difference between a completely dead room and a completely live one.
Two rehearsal rooms..... I am going to assume Plasterboard. I would expect both to benefit from the absorption of the exposed homasote layers.
I would also expect the extra mass to be beneficial for isolation. And, damping of bending waves again caused by the homasote layer.

no gain when Homasote is installed on the receiving side as well as the source side

Something odd there I reckon. Was the receiving side already very dead?
DD

[SGleason]

I am not familiar with some of the acronyms. But I think it worth noting the conditions in the source and receive spaces. There is a theoretical 6dB difference between a completely dead room and a completely live one.
Two rehearsal rooms..... I am going to assume Plasterboard. I would expect both to benefit from the absorption of the exposed homasote layers.
I would also expect the extra mass to be beneficial for isolation. And, damping of bending waves again caused by the homasote layer.

Yes, but the question remains........would the Homasote perform as well if it was on the receiving side of the assembly as it did if it was on the source side? I definitely agree about the damping of the waves by the Homasote......but once the sound waves have started transmitting through the assembly that dynamic should change.

Sorry about the acronyms, DanDan. GWB....gypsum wall board. I think you guys use different ones.

no gain when Homasote is installed on the receiving side as well as the source side

Something odd there I reckon. Was the receiving side already very dead?
DD

Well....that data came from an E-90 sound test in a lab so the rooms were both set up as per ASTM E-413. I theorize that with steel studs the wave has already been degraded at the front end of the assembly and there is not sufficient energy left to excite the planes on the receiving side enough for the Homasote to make a difference. I also theorize that in a wood stud wall there would be more energy transmitting through the assembly so the presence of Homasote on the receiving side would make a measurable difference. I surely wish we had some actual test data on that but my only data on that is a bit more anecdotal.

[DanDan]

You need to join the EU, I believe there may be a vacancy soon.....
I don't have any of the standards so I will assume your test was in two roughly equal rooms?
I would generally assume sound transmission to be symmetrical.
Sorry if I am a bit slow on the uptake here.
So, are you saying that in a test a partition of Source- H/GWB/Gap/GWB-Receiver performed no better when an extra H was added on the R side?
DD

[SGleason]

You need to join the EU, I believe there may be a vacancy soon.....

I'll talk to Management.

I would generally assume sound transmission to be symmetrical.

Me too....unless, of course the wall assembly is asymmetrical. Still wondering about unbalanced wall performance when measured from the weak side.

So, are you saying that in a test a partition of Source- H/GWB/Gap/GWB-Receiver performed no better when an extra H was added on the R side?
DD

Yes....except the Homasote is always under the GWB.

Wall assemblies as follows (from source side):

5/8" Type X GWB
1/2" Homasote 440 SoundBarrier
3-5/8" steel studs (25 ga.)
Fiberglass Insulation
5/8" Type X GWB

STC 54 (Test #NGC 2011041)


5/8" Type X GWB
1/2" Homasote 440 SoundBarrier
3-5/8" steel studs (25 ga.)
Fiberglass Insulation
1/2" Homasote 440 SoundBarrier
5/8" Type X GWB

STC 54 (Test #NGC2011032)

In fact, I just pulled the tests and the wall with Homasote on both sides is on 24" centers while the other is on 16" centers. That, plus the second layer of Homasote would seem to give that assembly an edge. Go figure, right? I have this other theory about improved cavity resonance when Homasote faces GWB in the cavity.

Steve

[DanDan]

No Mass.png

Perhaps this no change with GFR or Density continues all the way up to Homasote.
What GFR is this material? Density?

DD

[Soundman2020]

Me too....unless, of course the wall assembly is asymmetrical. Still wondering about unbalanced wall performance when measured from the weak side.

I'm not aware of any way in which a wall could be made to isolate MORE in one direction and LESS in the other, as in an "acoustic diode" for TL. Following the same electrical analogy, a wall is basically the same as the simplest RC tuned circuit: a resistor and capacitor wired in parallel. It does not matter which way the current flows: it still has the same impedance, will still be tuned to the same frequency, will still have the same reactance and resistance, etc.

Or instead of electronics, try a thought experiment. I have a solid concrete leaf next to a thin paper leaf: I can't think of any way that sound would move differently if it goes from left to right, or right to left through that wall. Each will do what it always does. There's no reason the sound should behave differently when it hits the paper "leaf", regardless of whether it first went through a bunch of empty air, or first went through the concrete. The sound would be DIFFERENT in level, sure, (it would have been attenuated by the wall) but that difference should have no effect that I can see, on how it behaves when going through the paper.

I do know that it is theoretically possible to construct an "acoustic diode" in the laboratory using certain arrangements of exotic materials, that have strange properties to start with, and by setting up materials at certain exact angles to tightly focused sound rays tuned to specific frequencies in the ultra-sonic region, but that isn't anything like the case of a wall built with typical material and angles and typical studio sounds.

Maybe I'm wrong, but I don't see any way of getting different results from different sides of the wall. If the TL is 43.7 dB going one way through, it should be 43.7 dB going the other way through.

- Stuart -

[Soundman2020]

I just pulled the tests ... That, plus the second layer of Homasote would seem to give that assembly an edge. Go figure, right?

Your tests are showing STC-numbers, not real transmission loss, so there's your problem!

STC isn't a good method for measuring studio isolation at all. Yes, it's used all over, but take a look at what the ASTM guys say about it...

Do you have the actual TL numbers for each frequency band, from those tests, or just the final STC rating? I suspect that you'll see a difference in the real numbers...


- Stuart -

[SGleason]

This is an excellent discussion....just what I need.

I'm not aware of any way in which a wall could be made to isolate MORE in one direction and LESS in the other, as in an "acoustic diode" for TL. Following the same electrical analogy, a wall is basically the same as the simplest RC tuned circuit: a resistor and capacitor wired in parallel. It does not matter which way the current flows: it still has the same impedance, will still be tuned to the same frequency, will still have the same reactance and resistance, etc.

The concept of a wall as a tuned circuit has my attention.

Maybe I'm wrong, but I don't see any way of getting different results from different sides of the wall. If the TL is 43.7 dB going one way through, it should be 43.7 dB going the other way through.

But, Stuart, if the sound wave enters the cavity stronger (greater amplitude) from one side than it does from another, wouldn't that affect the cavity resonance? I think it would affect the low frequency resonance. I do have some data that has caused this question to nag at me.

Five years ago, when I was very green at all this, I was asked to test a 2x6 wood stud wall with Homasote on just one side. We had a 2x4 wall that tested pretty well and all expectations were that the 2x6 wall would at least match those 2x4 numbers and hopefully exceed them slightly. Instead, the numbers were appreciably lower. The lab technician was at a loss, as was I.........until, that is, I had a chance to review the whole test with the photos. They had tested the wall from the non-Homasote side. The photos are absolutely conclusive of that as are my distinct memories. The 2x4 test was conducted from the Homasote side. The 2x6 test document was written up as being conducted from the Homasote side as well but that is absolutely not the case.

Here are the low frequency TL numbers for those two tests

2x6 wall - Homasote on receiving side
100 Hz - 25 dB
125 Hz - 25 dB
160 Hz - 25 dB
200 Hz - 32 dB
250 Hz - 41 dB
315 Hz - 41 dB

2x4 wall - Homasote on source side
100 Hz - 21 dB
125 Hz - 25 dB
160 Hz - 30 dB
200 Hz - 35 dB
250 Hz - 43 dB
315 Hz - 46 dB

Both assemblies were on 16" centers and both assemblies had full fiberglass insulation. Unless there is some other dynamic that makes a 2x6 wood wall perform poorly, i have to draw the conclusion that the position of the Homasote was what caused this differential, that is, having the sound attenuation component on the source side of the wall rather than on the receiving side leaves the transmitting sound wave with less amplitude therefor causing diminished cavity resonance. I apologize for that somewhat simplistic approach but that the best I've got at this point.

I appreciate this dialog greatly.

Steve

[SGleason]

Do you have the actual TL numbers for each frequency band, from those tests, or just the final STC rating?

One Side.pdf

Both Sides.pdf

Steve

[DanDan]

The 6" deep partition should perform better than the 4". There is something amiss here.
Have you tests of the same or similar structures with no Homasote?

if the sound wave enters the cavity stronger (greater amplitude) from one side than it does from another, wouldn't that affect the cavity resonance? I think it would affect the low frequency resonance.

How? Amplitude doesn't change frequency.

I think inserting Homasote in there is loaded with variables.
Is it attached to the GWB or touching it with an unknown amount of pressure and intimacy.
It would probably need to be glued to it in order to provide the same amount of bending wave damping as a full fill of fibre.
Speaking of which, you referred to the void being filled with fibre. How filled? There is room for a lot of variation there.

DD

[SGleason]

The 6" deep partition should perform better than the 4". There is something amiss here.
Have you tests of the same or similar structures with no Homasote?

I agree.....the test direction seems to be the only obvious variable. We have no 6" base data tests....there do not seem to be any anywhere, which is odd.

How? Amplitude doesn't change frequency.

But doesn't amplitude affect the strength of the cavity resonance? I have a huge conch shell here in my office. If I listen to it I can hear what I would describe as the soft resonance in the chambers within caused by ambient air movement. If I blow in it from the hole near its apex, which I occasionally do, it gets very loud.

I think inserting Homasote in there is loaded with variables.

The Homasote is screwed to the studs. The GWB is attached to the Homasote with screws only that do not go into the studs.....only screwed to the Homasote. I believe that that intimate contact creates a mass/spring relationship between the GWB and the Homasote.

Thanks for engaging in this, DanDan.

Steve

[Soundman2020]

But doesn't amplitude affect the strength of the cavity resonance?

A wall is a resonant system. It's not really the cavity that resonates: a wall is not a Helmholtz device, such as the shell you mentioned. It's a mass-spring-mass device. The air is the spring here, and the masses on either side are the resonant objects. The air in the cavity compresses and expands due to the movement of the mass panels, yes, but the air is the spring, and the panels are the "resonance". Thus, the wall acts as a system, not as individual parts. Yes, there might be some cavity resonance going on (notably, standing waves inside the cavity), but those should be damped by the insulation. So basically the wall acts as two masses linked by a spring, and it doesn't really matter which one you "trigger": the system will still end up resonating in the same way. Even if the two masses are different, the end result will still be the same, since they are linked by the spring and nothing else.

As Dan said: something doesn't add up in the results you mentioned.

- Stuart -