"The Dreaded" Third Leaf: Fact or Fiction?

I have a theory. I actually think (am guessing!) that the third leaf effect is actually less severe than we fear under certain circumstances. I'm guessing that the more critical factor is the depth of the widest gap between the heaviest leaves, as long as the middle leaf is one of them and the wimpy third leaf is on the quiet side. In other words, I can't help but think that two heavy decoupled leaves with a wide air gap is good on its own, and that having an extra rather wimpy leaf firmly attached to the outside of one of them.would not harm anything much, and in fact might actually help a bit. Of course square footage can be maximized by a strict two leaf system, but in the case of retrofits, we sometimes must compromise. In short, when building from scratch, materials, square footage / headroom, money, and time can be saved by adhering strictly to a strict two leaf system... But when doing a retrofit/conversion of existing space, when square footage / headroom can be compromied a bit, it's probably easier and cheaper to add layers to the existing wall /ceiling surface (to become the "dreaded" middle leaf") and then build an equally massive inner leaf some distance (a foot or a bit more?) away from that, as opposed to ripping off the middle leaf and beefing the outer leaf up from the inside. I just really doubt that a third leaf would really amplify things much (if at all). I know -- resonance frequecies . . . But if the two inner leaves are doing their job, there won't be much of a frequency left to sympathetically vibrate to, right? :roll: --Keith :mrgreen: Am I on the right track with this thinking?

Sharward, I'm really glad you started this post. I hope you don't mind if I post what I wrote in the other thread so people can get an idea of a real case scenario: (quoting myself) "My next question will be 'to beef or not to beef' (the outer walls of the garage): From what I understand, the 3 leaf effect applies mostly to smaller sealed airspaces, is this correct? So if I build a 'room within a room', and have a good deal of air surrounding that room, wouldn't that negate the 3 leaf effect? For instance, if the dimensions of the garage are 10' high by 11' wide by 21' long, and the room I build is 8' high by 9' wide by 16' long, that's a difference of 1158 cubic ft. of air. Surely that volume of air can't be acting as a 'spring' can it? I'm sure that certain conditions must be met to create the 3 leaf effect- my question is whether or not this scenario would meet those conditions." Thanks

sharward wrote:

I have a theory. I actually think (am guessing!) that the third leaf effect is actually less severe than we fear under certain circumstances. I'm guessing that the more critical factor is the depth of the widest gap between the heaviest leaves, as long as the middle leaf is one of them and the wimpy third leaf is on the quiet side. In other words, I can't help but think that two heavy decoupled leaves with a wide air gap is good on its own, and that having an extra rather wimpy leaf firmly attached to the outside of one of them.would not harm anything much, and in fact might actually help a bit. Of course square footage can be maximized by a strict two leaf system, but in the case of retrofits, we sometimes must compromise. In short, when building from scratch, materials, square footage / headroom, money, and time can be saved by adhering strictly to a strict two leaf system... But when doing a retrofit/conversion of existing space, when square footage / headroom can be compromied a bit, it's probably easier and cheaper to add layers to the existing wall /ceiling surface (to become the "dreaded" middle leaf") and then build an equally massive inner leaf some distance (a foot or a bit more?) away from that, as opposed to ripping off the middle leaf and beefing the outer leaf up from the inside. I just really doubt that a third leaf would really amplify things much (if at all).

Keith, 1st of all - let's be real here......... if someone has to live with a 3rd leaf - then they have to live with it. An example of this would be someone who owned a condo - and wanted a studio in it........ In a condo - you own from the inside face of unpainted drywall to the unfinished face of unpainted drywall (as far as walls are concerned) SO you don't have the luxury of removing a layer of drywall - adding mass within the cavity - and then constructing a new section of framing with a newly located 2nd leaf...... OK - if that's the case - then live with it and over come the short comings with more mass on the new wall that you CAN construct within your space. BUT - other than those certain cases where you have no choice- I can't see any reason to compromise. No - to answer your question(s) directly: Does it amplify?..... the answer is yes - and at more than just the reasonant frequency......... Look in the book at page 55 (figure 4.10) and picture it properly......... If I begin with a wall - and add a wall - leaving the leaf there and add 1 layer to the inside face - that's an STC 50 vrs the STC 36 you began with...... and so your initial reaction would be - nope- that didn't amplify - it increased isolation by 14dB........... But that's not really true.......... If you take the exact same material - and move it to the new inside face of the new wall - removing it from the cavity - you have an STC 57 for using the same mass as the original wall. So if you begin with 2 sheets - W/ 2 walls - and then add a piece in between the 2 - you will increase transmissions (thus amplify) by 7dB - (STC50) which is slightly more than double the power and amplitude....... Now - if we add 2 more pieces of mass to this (removing the peice in the middle) - on the outside faces...... we will get an STC 63 (an increase for doubling the mass of 6dB - just as predicted by mass law) But - then if we take those 2 pieces and put them inside the wall cavity - we now drop to an STC 40 - for an increase in transmission through the wall (thus an amplification) of 23dB........ and that's almost a wopping 400% increase in volume (minus 1 db of course) this with exactly the same material.......... I call that amplification - how about you? Now - that having been said - at what point do we stop being concerned about this? Either at the point where we have no control (condo example above - could be apartments, etc.) OR (if we do have a choice) When the distance becomes so great that it creates usable space. Examples would be an air lock- hallway, storage, iso-booth, bathroom, etc., etc.. Now - we know that at a distance of 3' from a source - we have a reduction in amplitude of 6dB - and then at 9' another reduction of 6dB - so a total of 12dB of reduction in amplitude at a distance of 9' from the source.......... so if we assume an stc of 63 (just for the sake of conversation) and a 9 foot room- with an original source of 100dB - that would be 40dB exiting the room - and only 28 dB at the face of the opposite wall........ Now a sound 28 dB is about the same as a soft whisper......... thus not enough energy to even begin to excite the wall (heck - the average backgroung noise in a home - when it's considered quiet - is 50db). So - as I have always said - whatever you can change - make better - FIX it - to not do so is penny foolish - whatever you can't do anything about - live with it - to fret over something you can't change is just plain old foolish........ As far as hard numbers go - no one has ever bothered testing all of the different parameters we consider foolish because it's a lot of money to just throw away............ if any one is really Really curious - specd the money and get the test results........ Sincerely, Rod

sharward wrote:

as long as the middle leaf is one of them and the wimpy third leaf is on the quiet side.

That (by the way) is completely irrelevant - which side of the system an inner leaf is on - Once you develope a wall system - it alwys works exactly the same in both directions - it cannot act differently from side to side......... Rod

cjbdrm wrote:

So if I build a 'room within a room', and have a good deal of air surrounding that room, wouldn't that negate the 3 leaf effect?

AndI think I answered that question with a question - which is - why would you want to throw away valuable usable space by doing that? Was it just to save sweat ? Or did you say something about wind bracing? I forget...... Rod

Great stuff, Rod. And, cjbdrm, your question in your thread actually inspired me to start this one. Rod, I guess what I'm looking for is (and perhaps you already provided it above) what one must do to make up for the third leaf. Again, assuming one can afford to give up space. Regardless, this will be a great authoritative reference thread for third leaf questions. :-) --Keith :mrgreen:

Rod, Yes you did recommend using all available space to get a better sounding room...(the room in question is actually 12 x 18 ). If it's a matter of inches that determines the creation of a third leaf effect, I would gladly lose a little real estate...for instance coming in a foot on each side, 2 feet from the ceiling, etc. At what point(distance) do the two separate walls become two separate walls?

cjbdrm wrote:

Rod, Yes you did recommend using all available space to get a better sounding room...(the room in question is actually 12 x 18). If it's a matter of inches that determines the creation of a third leaf effect, I would gladly lose a little real estate...for instance coming in a foot on each side, 2 feet from the ceiling, etc. At what point(distance) do the two separate walls become two separate walls?

That's the point I made above - In my mind it would be 3' before you saw any noticable difference. It isn't unusual in our hotels to have walls for the grand ballrooms that are enclosed with 2 feet between wall surfaces - and you can hear about the same amount of bleed through them that the regular double walls (with only one inch of open space between studs) have. The problem here is that you don't understand the air spring itself - An elevated concrete slab- open on 2 sides to a room - with the walls built on top of it =- still has an air spring that affects it even though you would imagine it didn't. It's just the degree of compression we're lking aabout - and at low frequencies that can be a whole lot more than you might envision.......... Rod So we aren't talking inchec here- rather it's feet. Rod

rod gervais wrote:

That's the point I made above - In my mind it would be 3' before you saw any noticable difference. It isn't unusual in our hotels to have walls for the grand ballrooms that are enclosed with 2 feet between wall surfaces - and you can hear about the same amount of bleed through them that the regular double walls (with only one inch of open space between studs) have. So we aren't talking inchec here- rather it's feet. Rod

Thanks Rod. That's exactly what I needed to hear. If you're saying that flanking wasn't an issue in that comparison, then I'll certainly take your word and start obsessing about some other studio build topic... :roll: 8) Thanks again!

cjbdrm wrote:

If you're saying that flanking wasn't an issue in that comparison, then I'll certainly take your word and start obsessing about some other studio build topic...

Don't be putting words in my mouth - the fact is that flanking would possibly takeover in that comparison......... I have no knowledge of the construction to make that detrmination......... flannking is always an issue seperate from individual component isolation - which is why we need to take such special care to construct in such a manner as to minimize flanking possibilities.......... Rod

Good stuff. I guess my three leaf wall between studio and my office (small 10x10 bedroom) is going to be perform worse than I had thought... :roll: It will consist of, from studio to office:

• 5/8" Type X drywall
• Green Glue at some percentage of recommended amount (probably ~75%)
• 5/8" Type X drywall
• 5/8" Type X drywall
• 15/32" OSB
• 2x4 studs, 16" o.c.
• At least 3" and possibly 6" of Thermafiber SAFB with nylon or some kind of strapping to secure firmly against OSB
• Total of about 12" of air space (including insulation)
• 1/2" drywall
• Green Glue at 75% of recommended amount
• 5/8" Type X drywall
• 5/8" Type X drywall
• 5/8" Type X drywall
• 3/8" OSB
• 2x4 studs, 16" o.c.
• Fiberglass insulation
• 1/2" drywall (third leaf)

I was hoping that my massive four layers of drywall, some Green Glue, and OSB middle leaf, plus the 12" or so of air gap, would make up for the wimpy third leaf. If it doesn't perform very well, I'm prepared to add a layer on the office side with Green Glue. --Keith :mrgreen:

Holy sheetrock Batman! I'm no scientist, but with all that mass and airspace and the 2 layers of damping(GG), any energy making it to that last layer of 1/2" drywall on the office side has got to be insignificant... Do you really think it'll make that much of a difference to go from 3 to 4 layers of drywall per side (diminishing returns)? It almost sounds like overkill...but then again, I'm the student here, not the teacher... :oops:

I havbe to believe that in this case - with the amount of mass we're talking on the 2 other faces - that the one sheet of 1/2 on the inside face is going to be insignificant to say the least........

The 4th layer of drywall was just a bonus... I intended to do "just" the three layers of 5/8", but I had three 4x10 sheets of 1/2" drywall left over from my beef-up, so rather than breaking it up into little pieces to fit into my garbage container over several weeks, I decided to slap it onto my infamous three leaf wall. And the Green Glue was left over from my outer leaf ceiling beef-up. It's my weak link -- just two layers of 5/8" with Green Glue at 75% coverage in between. I had to stick with a lightweight outer leaf ceiling so as to not overload the trusses. But enough about me. ;-) --Keith :mrgreen:

Wow Keith....what kind of STC, LF TL #'s are you going for with that build? Cheers Kris

I'm going for a low end TL of "holy crap." ;-)

I want to build a room within my finished garage. To avoid the triple leaf effect, I should: 1. remove inner layer of drywall from garage walls, use it to 'beef up' outer garage wall. 2. Build 'room within a room', using single leaf partition (drywall on only one side of 2x4 framed wall). Is this correct? Will this produce the best possible results for me? This will create a true mass-air-mass scenario, correct? Sharward, we have a lot in common buddy :)

cjbdrm wrote:

I want to build a room within my finished garage. To avoid the triple leaf effect, I should: 1. remove inner layer of drywall from garage walls, use it to 'beef up' outer garage wall. 2. Build 'room within a room', using single leaf partition (drywall on only one side of 2x4 framed wall). Is this correct? Will this produce the best possible results for me? This will create a true mass-air-mass scenario, correct?

cj, you got it........... Rod

Keith: For a moment, I felt as if I was channeling my thoughts through you... And I was hopeful.... But.... (resisting the temptation; saying to self...): ... I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything. I will not say anything.

As a crazy thought experiment...(i.e. not that anyone would ever build this), imagine if you did have a pipe between the inner and outter leaves vented to a second space (perhaps an 'infinite' space like outdoors). Now instead of just allowing sound to passively travel through the pipe (which would only reduce the effective spring of the air very slightly) imagine that a blower motor is used to actively reduce the pressure between the walls. This should serve to lower the effective spring constant of the air thereby improving the performance of the overall partition. Of course in reality the blower will create noise that will resonate in the walls themselves, and the 'infinite space' will become loud...but as a thought experiment it's kindof interesting...any dissenting views? Cheers Kris

sharward wrote:

I have a theory. I actually think (am guessing!) that the third leaf effect is actually less severe than we fear under certain circumstances. I'm guessing that the more critical factor is the depth of the widest gap between the heaviest leaves, as long as the middle leaf is one of them and the wimpy third leaf is on the quiet side. In other words, I can't help but think that two heavy decoupled leaves with a wide air gap is good on its own, and that having an extra rather wimpy leaf firmly attached to the outside of one of them.would not harm anything much, and in fact might actually help a bit. Of course square footage can be maximized by a strict two leaf system, but in the case of retrofits, we sometimes must compromise. In short, when building from scratch, materials, square footage / headroom, money, and time can be saved by adhering strictly to a strict two leaf system... But when doing a retrofit/conversion of existing space, when square footage / headroom can be compromied a bit, it's probably easier and cheaper to add layers to the existing wall /ceiling surface (to become the "dreaded" middle leaf") and then build an equally massive inner leaf some distance (a foot or a bit more?) away from that, as opposed to ripping off the middle leaf and beefing the outer leaf up from the inside. I just really doubt that a third leaf would really amplify things much (if at all). I know -- resonance frequecies . . . But if the two inner leaves are doing their job, there won't be much of a frequency left to sympathetically vibrate to, right? :roll: --Keith :mrgreen: Am I on the right track with this thinking?

Keith, you said "I'm guessing that the more critical factor is the depth of the widest gap between the heaviest leaves, as long as the middle leaf is one of them and the wimpy third leaf is on the quiet side." And there is validity to that w/o question. Please allow me to blabber about that in a moment and just offer that there is unquestioanble validity to this statement as well: -no triple leaf construction can ever offer as good of low-freq performance as a double leaf created with the same amount of space and materials. This is because by splitting the air cavity into two smaller ones and lightening the outer leaves, you will have created a system with a higher frequency of MSM and a higher frequency where decoupling starts to work for you instead of against you. But there is no certainty that any given triple leaf construction is sure to perform badly. I think that this thread in combination with the question asked earlier in the "green glue" thread (with the price boo boo) can be pretty well addressed under the subject of "turning a garage into a studio", which i hope to post tonight or tomorrow. But you will introduce a second resonance into the system, and in a way that tends to get super-imposed onto the behavior of the new decoupled system provided that the "extra" resonance is much higher in frequency than the primary MSM. for a very brief comment, imagine an enormously deep (say 10 feet) double stud wall with an extremely low MSM that was actually a quadruple leaf in that each "leaf" of the double stud wall was actually a 2x4 wall with drywall on both sides. But a 10 foot air gap in between. For the sake of simplicity in this evenings post, we'll ingore the fact that this is a room and not a cavity inside a wall and just take it as an example and not a realistic situation. In a case like this its reasonable to think of the performance of a very deep double stud wall like that as TL = TLleaf1 + TLleaf2 + TLinsulation This will hold, in theory, at frequencies above the primary "room mode" of the cavity between the leaves. With a 10 foot cavity, this will hold to rather low frequencies. For the sake of further simplicity, we will simply assume there is no absorption (even from air) in the cavity at all. What shakes out of that is basically this: Imagine that the TL of the 2x4 walls that form the "leaves"o f this double stud wall are 63 = 20 125 = 15 250 = 35 500 = 40 1000 = 45 2000 = 40 4000 = 50 just for the heck of it, imagine those values. But imagine that you didn't have the 2x4 walls on each side, but just 2 layers of drywall instead, then you might have this 63=20 125=26 250 = 32 500 = 37 1000 = 39 2000 = 32 4000 = 42 The TL of the double leaf -vs- the quadruple leaf would shake out about like this: 63 = 40 quad leaf = 40 double leaf 125 = 30 quad leaf = 52 double leaf 250 = 70 quad leaf = 64 double leaf 500 = 80 quad leaf = 74 double leaf 1000 = 90 quad leaf = 78 double leaf etc. this is just an example and nothing to be taken as more than just a simple silly example for illustrative purposes, the assumptions above are unrealistic and over-simplified. But the point is still good. The resonances of each side of the wall do still matter and degrade the performance, much as if you had a double-window and wanted to stop alot of light, but one side of the window was really transparent to say 500 nanometer wavelengths... 500 nanometers would come through more even if the other side of the window was good at 500nm. if both sides were good, except for stopping yellow light, then alot of light is stopped, except the yellow. When the MSM's of each leaf sort of "line up" the results get chaotic and impossible to really predict with simple calculations. i.e., a triple leaf formed by adding RC and drywall under an existing cieling. The MSM of the RC/drywall system is fairly close to the structural resonance of the rigid not-decoupled original ceiling. Ir811 has such a comparison and you can see that the added results (ADDING to a wall to make a triple leaf) are better than the not-decoupled-at-all double leaf. Its also deeper, and heavier. When compared to a double leaf with decuopling in the middle, its vastly worse. (thats also available in IR811) to sum, we can say that when facing a situation where you must work with a triple leaf, just the really simplified analysis above does reveal some points to be considered. 1. the MSM of the new "decoupled" part should be FAR lower in frequency than the primary low-freq resonance of the other side of the assembly. So if adding studs and airspace in front of an existing 2x4 wall, the MSM should be FAR lower than the ~125-160hz primary structural resonance of the rigid not-decoupled 2x4 wall. This ensures that the resonant sins of the original 2x4 wall are effectively decuopled from the other side (it will still be worse than if it was a dboule leaf, but this is alot of help) 2. A triple or quadruple leaf should really strive to not be symmetrical. NEVER build a 2x4 wall with drywall on both sides a little ways away from another 2x4 wall with drywall on both sides. I'll try to address this with some data i have and a better overall discussion maybe tomorrow. But this is a very good topic, i think.

i have to mostly avoid studiotips because i'm really tempted to start a debate about the coincidence phenomenon, which will keep me preoccupied i'm sure as those gents will put up alot to think about. now i'm going to have to avoid this forum too lest i wind up triple-leaf-pondering myself into working until midnight. :lol: :cry: j/k but its on my mind so i'll blabber away for a few minutes. Not long ago for a specific big project we tinkered with 1/2" board on each side of styrofoam. That panel exhibits an amazingly severe MSM at around 800hz (the foam acts like a really stiff air spring between the layers). Adding even 4 layers of drywall to each side didn't bring the results up very high at all, so it was determined that studs + an air space should be used. 2" steel studs with double drywall/GG and then again with GG on the "other" side, glued to the foam panel. The first graph shows the results of this triple leaf. With the panel left to resonate on one side of the wall, its resonance (the resonance of the "third" leaf) is suprimposed on the behavior of the other wall and looks like a dip at 800hz. This disspears when adding to the other side as well. (this is also getting precariously close to the flanking limit of the laboratory, which obscures some details). In this case the presence of the "third" leaf and associated resonance aren't a big deal because they are so high in frequency, where the wall is so good, that the dip doesn't really make a difference in the STC or A-weighted isolation calculations. cjbdrm asks about adding in front of a 2x4 wall to make a triple leaf. This is potentially more problematic as the resonance of the "third" leaf is the rseonance of the 2x4 wall, which tends to be both fairly severe and fairly low in frequency. for 16" OC studs its typically 125hz for 1/2" drywall 160 for 5/8" and for 24" Oc studs its lower in frequency. But its at a frequency where you don't necessarily want it to be. If you successfully damp the resonance of the 2x4 wall and then add your new decouple leaf, that will hlelp mitigate this issue. It will also add weight (if you add glue and drywall to the existing 2x4 wall before adding the next leaf) which will help lower the primary MSM of the final assembly. From there, your results are mostly defined by -how deep the new cavity in front of the old wall is -how much weight is on the other side (the new side) of the new air cavity as these will define the location of the MSM. w/o question if i was to use GG in an assembly like this as cj inquired about, but only wanted to use it in one place, i would opt to put it on the existing wall before adding the new leaf in front of it to mitigate the "third leaf"s resonance. I would also go for quite a bit of weight on the "new" leaf, and as much air space as i could spare. One can calculate the MSM of the "new" leaf by treating the weight of the side of the decoupled assembly as the entire weight, studs an all, of the existing 2x4 wall plus whatever you add to it, and the weight of the other side as the drywall/;ayers on the wall (plus studs). If you improved both sides of the existing wall with drywall before adding the new leaf, you'd have about 50 kg/m^2 (assuming 5/8" drywall) for that side. If the other side was double drywall, with a 4.5" air cavity, the MSM could be expected to fall around 34hz, and with a third layer of drywall on the "new" side of the same assembly, you would have an MSM of around 30hz. The low MSM combined with the very high weight (for a stud wall) would predict good results. With more air space, say 8", between the old wall and the new wall with the same layers as above, the MSM would be extremely low at around 22-23 hz and triple leaf or not with that low of a primary MSM and that much weight, it'll be a great wall (especially if the original walls resonance behavior is tamed a bit. But, if you had the same # of layers, same amount of space, but a double leaf, then you'd have an MSM of around 16hz, and better low-end isolation. Always a comparable double-leaf will have lower MSM and better low-end isolation. Anyway, if that was too winding and rambling, just throw some eggs and tomatos my way, but i hope it helps.

Brian, Thanks for your input.