Double Wall Weirdness!!!!

Well... I surprised myself last night while showing some friends the ongoing studio building. I was showing them the sheetrocking job and the double walls (9 foot walls 3 1/2 air gap between them) and was explaining the reasons for spacing and totally decoupling ceilings etc for reasons sound wont transfer..... well... I banged on the inside wall of my drum room.. and it vibrated the outer wall paralleled to it in the main room! WHY would THAT happen. It is plain to see the airgap all the way clear into the attic. The insulation is clearly not touching from one wall cavity to another. Can it be the impact sound waves are THAT strong to make the opposite wall facing (nearly a foot away) vibrate? Man! It was an eye opener. I was only knocking about as loud as you would on someones front door... What are your guys (and gals) thoughts? Tom

Are you using RC on the walls :?: if not, I was thinking that the reason might be the fact that your insulation is not pressed against the back of your gypsum.. :shock: If knocking the walls does that vibration, what happens when you actually play drums in that room? can you send us some pictures of your construction, preferrably some details of how you did it. best andreas

Andreas. I have not actually set up anything in there yet since I am still building and sheetrocking. I still have to put the second layer on. Basically the construction is a totally free standing shell of a building (with a 10 foot wall height on a concrete slab. Inside walls and rooms are also independantly freestanding. Walls are (9 ft tall) 2 x 4 wood studs on 16" centers, insulated with r-13 fiberglass (ceilings with r-30).All walls are a minimum of 2 inches to 3.5 inches seperated from adjoining rooms and outer walls, drywall on inside wall faces only I have one layer of 5/8ths drywall up (glued and screwed) with second layer ready to be put up after taping and bedding( and acoustic caulking). The ceilings are totally independant from adjoining rooms, resting solely on each own room's walls. I tried poking more loose packed insulation between the existing walls insulation to force the insulation up against the drywall and perhaps do some dampening, but it did not change anything. Thanks for your input. Tom

Flanking then perhaps? :idea: :? I'm pretty much finshed with my walls but I have a concrete wall separating my LR and CR so I can't really tell if knocking my walls would make my other walls vibrate. AND I'm not really fimiliar with anything else than the metric system so I can't tell you if your walls are big/thick enough... mine are 90mm studs with 95mm insulation pressed against one layer of 13mm gypsum with straps. If i knock on one of my walls there's really just a low thud, nothing more. I probably will have to put up a second layer of gypsum or something heavy if I'd like to get better STC. I guess you'd get better result after that second layer too. I'll do a search and see if you posted anything else about your construction that may give me some clues. edit: found this [/url]http://www.johnlsayers.com/phpBB2/viewtopic.php?t=5751&highlight=[url]but progress seemed to stop around may?![/url]

tmix wrote:

It is plain to see the airgap all the way clear into the attic. The insulation is clearly not touching from one wall cavity to another.

Forgetting (for a moment) your initial problem - what you describe (above) poses a different problem. You need a firestop installed at the top of that wall assembly - otherwise you are creating a chimney that will spread fire like crazy. This is a code requirement in your state - and it exists for a very good reason.

Can it be the impact sound waves are THAT strong to make the opposite wall facing (nearly a foot away) vibrate?

Well picture that you only have one layer of drywall over a 9' tall wall by whatever length it is. This is (for all intent and purpose) a very large drum head. Not only does a drum head transmit sound - it actually amplifies it, and the 2nd head amplifies it even more. There are a number of things that come into play with sound isolation - mass - air transfer between spaces - and (for lower frequencies) stiffness. Right now you have a floppy wall - and it wont behave well with direct impact noises. As you add more sheets of mass to the assembly - this will becomes lees and lees of an issue. Be careful to make certain that your assemblies do not connect at places like top and bottom plates - this will increase flanking problems. You would be amazed at the number of projects that i have inspected recently where this is an issue. Understand too that impact noise is very different from waves transmitted through the air - We construct walls that are very effective for air waves - but suck when it comes to impact noise. To really check your assembly - you should be using sound waves - not tapping. Hopefully your clients will not be banging on the walls when recording. Sincerely, Rod

Thanks Rod! I thought about the fact that hitting the wall might be different from sound waves. I do have fire stop blocking in each individual walls. How do you put a stop at the top of the walls without bridging the 2 walls together. It would seem to cause flanking horribly. I have gone back and filled the air gap with loosely filled insulation to dampen things. I have tried to be very careful to meet all code requirements of my city. I have had many inspections at every stage from the city inspectors. None of them seemed concerned in the slightest regard to the wall builds. I know that does not releive me of any responsibilty to be correct, I would have figured them to tell me if I missed something by accident though. Tom

Rod's book covers fire blocking in great detail. (I will actually have some questios for him when I get into my framing -- I'm holding off until then though so that I have some photos for reference. ;-)) --Keith :mrgreen:

rod gervais wrote:

There are a number of things that come into play with sound isolation - mass - air transfer between spaces - and (for lower frequencies) stiffness. Right now you have a floppy wall - and it wont behave well with direct impact noises. As you add more sheets of mass to the assembly - this will becomes lees and lees of an issue. Rod

Hi Rod: I was reading the above paragraph about 'floppy versus stiff', and was wondering where 'limp' might come in? I'm not sure, but does 'floppy' means resilient?... or just that it is easy to set in motion? A material say like lead or MLV is supposed to be limp, and not stiff nor resilient --- is that one of the things that makes it so good? I guess floppy would never be desirable? But what about combinations of materials, some being stiff, some being limp? thanx for any clarifications (and btw welcome back!), K

I think the easiest way to think about this is what with only one layer of sheetrock, you've built a bunch of "drum heads." There is SOME give in any surface mounted around the perimeter. By using more than one layer, you are creating a thicker drumhead which is much harder to set in motion (and will probably not resonate as long. As an example, say you have two saw horses set 8' apart and you put an 8' 1x6 board across them. Not only will it sag in the middle but it will also bounce pretty good if hit it in the middle. Now, replace the 1x6 with a 2x6. It won't sag as much and it won't bounce as easily. Keep making the board thicker and the bouncing gets less and less. This is the same thing that is happening with your drywall at this stage. I hope that clears up the "science" a little. len

I think there are some frequencies where floppy walls perform better though. That's why Steve (knightfly) often recommends double framed walls be studded 24" o.c. rather than 16" o.c. It's yet another example of how this stuff just refuses to be simple and intuitive! :? --Keith :mrgreen:

I think the principle is that on 24" centers you have a bigger "drum head" and it acts like a panel resonator at a specific frequency. Putting the fiberglass behind it damps this a little but also spreads the reponse to cover a wider range. Since LF is the hardest to get rid of, having your walls "help" a little can't be a bad thing. len

len-morgan wrote:

I think the principle is that on 24" centers you have a bigger "drum head" and it acts like a panel resonator at a specific frequency. Putting the fiberglass behind it damps this a little but also spreads the reponse to cover a wider range. Since LF is the hardest to get rid of, having your walls "help" a little can't be a bad thing. len

I'm not sure I'm understanding this... If the 24" OC gives a bigger 'drum head', then that would mean a lower frequency, right? But if lower frequencies are harder to get rid of, why would making a panel resonate at a lower frequency be better? K

You pose a good question. This is what makes all of this acoustic stuff confusing at times to thinking people - it doesn't seem to be logical. The explanation might be that by making the "drum heads" big enough (mine for example are 24" x 8 ft), the frequency is low enough to be out of the range of human hearing. After you've got a couple of layers of dry wall up though, this becomes much stiffer and harder to set in motion (with sound, not a fist). If you add the insulation pressed up behind it to dampen the vibration, it shouldn't be too hard to control. len

tmix wrote:

well... I banged on the inside wall of my drum room.. and it vibrated the outer wall paralleled to it in the main room! WHY would THAT happen. It is plain to see the airgap all the way clear into the attic. The insulation is clearly not touching from one wall cavity to another. Can it be the impact sound waves are THAT strong to make the opposite wall facing (nearly a foot away) vibrate? Man! It was an eye opener.

Just as Rod and others said, but in a bit different words and some additions. If one speaks about decoupling, one standard refers to visible connections and visible decouplers (elastic things). One often refers to the Mass-Air-Mass resonance in analogy with the NRC/IRC papers. The general notion however is the Mass-Spring-Mass resonance. This means, even when all visible connections are decoupled by visible devices, which are never perfect but work to a certain degree, you still have the air which is an invisible decoupler. In those low frequencies, below cavity resonances, this air layer act as a spring. You could as well picture this air layer as a matrix of helical steel wire springs. Hence this cavity works as a pressure room where you compress/decompress the air. Don't see it as forceful sound waves. Sound wave behavior comes into the picture in a double leaf wall where the cavity modes come into the picture (hence wave length versus cavity width) By using your fist, you compress the air. In fact it works the same as TL, only you use much more force. And you trigger mainly the resonance frequency, where this mass on the other side on this air spring starts dancing. A sound power level of 120 dB equals 1 Watt. Your power is much larger than that. You really can see air in this cavity as the invisible counterpart of traditional visible decouplers as neoprene or whatever. This mass spring resonance of a double leaf wall is calculated with EXACT the same basic formula as a decoupler (not speaking about corrections for mounting etc., just the MSM of the panels). You only don't recognize it directly since the dynamic stiffness of air is already integrated in those constants. But if you go to the source formula, you'll notice that a double leaf wall is calculated with the same formula as the resonance frequency for a helical spring underneath a machine. Only the spring constants (related to type of spring), and masses are adjusted. This is also valid for a panel trap. Only underneath a machine and in a panel trap one simplifies the formulas assuming one of both masses has infinite mass (hence no influence) and for drywalls both masses are light. One can directly improve on the traditional published resonance formula for panel traps by substituting it by the formula for double leaf walls, since then also the backing of the trap, which can vary from a brick wall (+/- infinite mass behavior in calculation) to a MDF panel, is taken into account. One can easely check that. Use the formula for double leaf walls, then make one of the masses in the formula stupidly high, and you come to about the same result as the panel trap formula. Hence I even wondered already why they don't publish those formulas directly that way, rather than standard assuming that a panel trap has an infinit or very heavy backing. (I assume some writers don't know the physical background of the formulas themselves). One could easily say: this is the general formula, and this is the stylized version when the backing doesn't matter (very heavy), but still then the general formula works correctly. Eric