MSM questions, retrofit isolation, triple leaf, and other stuff/questions.

Hello,

Quick introduction: I'm a former sound designer (i.e. creator of sound effects for film and TV), current film editor and VFX guy, and currently planning some acoustic fixes to my Hollywood apartment. I've been reading quite a bit on retrofit situations, and have some specific questions regarding certain "Mass-Spring-Mass" configurations, especially MSMSM (oft referred to as triple leaf).

My questions are about half way down this post, but first some background:


The Project: Retrofit Isolation


SITUATION:

1) Apartment, built in 1964 using asbestos based wall and ceiling materials that cannot be removed or disturbed. inserting screws is okay - but scraping off old cottage cheese ceiling treatment or wholesale removal of drywall is impossible (state of California law requires completely vacating unit, and expensive removal by licensed asbestos removal crew, hazardous waste fees, etc etc. blah blah blah).

2) The unit is located on the second floor. It is below a unit occupied by a group of annoying people whom we shall call the Elephant Family Robinson who bring new meaning to the term "impact noise" It is located above and next to other people that I'd just as soon not annoy with my own sounds from my home theater.

3) The building is a typical Southern California construction: Wood frame with stucco exterior and 5/8" drywall interior. Little to no insulation. This is earthquake country, so the building is very flexible (that is, flexible wood frame, but no other damping) - but while that is good for surfing an earthquake, it is terrible for impact noise of upstairs neighbors - little kids jumping of the floor, or a slammed door seems to make the entire building shake.

4) As an interesting note, across the street is an annoying guy with a van with a very big subwoofer. Standing in front of the building in the free air, it is barely noticeable. HOWEVER, in the front bedroom the building seems to amplify it, and it becomes quite annoying. I haven't measured it but it seems the building has a resonance in the under 50 Hz area, and least in the exterior to interior path.


OBJECTIVES:

5) Objective one, to calm my nerves, and really the reason I embarked on this project is to suppress the Elephants upstairs. The impact noise of elephants, children, and screeching chairs has me yanking my hair out. And no, I'm not moving, this is rent control baby!

6) Objective two, bring the living room and bedroom to a much quieter level in general, so that either room can be used for voice over work.

7) Objective three, and in view of not annoying other neighbors (I could care less about the elephant people), I need to fit the living room for my home theater so that I can watch films (not to mention edit/mix the odd low budget job) at the proper level - meaning that -20 dBFS pink will result in 85 dB SPL in the room. The principal problem here IMO is the area under 125 Hz. I'm not using a subwoofer, and don't intend to get one - even with no sub, the bass in the system is plenty loud to shake the place.

8) Side objective: I also intend to tune the living room for the benefit of the internal acoustics, though that will not be discussed in this post.


BANDWIDTH NOTE: in this post I am mainly going to be asking/commenting about suppressing the transmission of frequencies below 250 Hz. An assumption is that taming the low end (especially in the 30-70Hz area), will result in an ancillary improvement in the upperbands, which I believe are less of an issue here.



-- -- -- -- -- -- -- -- -- --

MSM/MAM/Triple leaf stuff.


As I began researching solutions and available materials, I came across the GreenGlue site and others, which talked about the "dreaded triple leaf".. It made no sense to me they way they were explaining it (though to be fair, they greatly simplified the explanation). Sadly, I'm not the kind of guy that is satisfied with the Fisher-Price interpretation, and the "triple leaf syndrome" left me scratching my head with questions like:

"What about cavities of dissimilar resonance? Plates with different resonance points? Different forms of damping?" etc etc. My mind reeled and ultimately, a day or two ago I spent the day reading and researching on the subject.

Wow - it never occurred to me that the modal resonances of the air cavity were not related to the "spring" action/MSM resonance of the air in the cavity in an MSM configuration. Wow wow
wow. How un-intuitive can you get?

(note: for the purposes of discussion, when I say "plate" I mean 5/8" sheetrock, and "cavity" or "spring" is the air space between the plates).

My MSM assumptions before I ask questions:

• It is distance between plates, and NOT the size/volume of the air cavity that determines the resonant MSM frequency.
  
• Quadrupling the distance between plates will cut the resonance frequency in HALF when the sound is normal to the plate.
  
• Increasing the mass of the plates and/or damping the air spring (by placing fiberglass insulation the air cavity) will lower resonance frequency.
  
• The MSM resonance is not related to the modal resonances of the cavity.
  
• The angle of incidence of a sound entering the MSM system will affect the resonance of that system. Sound normal to the first plate will result in a lower resonance than sound that is "glancing" to the plate.
  
• If the plates are mechanically decoupled (sound clips, double studs, etc) the "air spring" will still couple the plates at the resonant MSM frequency.
  
• The issue with "triple leaf effect" is not due to 3 plates, it is that there are 3 plates and two springs (cavities) in series so you have MSMSM, wherein the two air springs are at similar (enough) resonance to multiply and thus significantly reduce the insulating effect at/around the resonance frequency. But obviously if you have 4 regular sized rooms in a row, (assume a simple plate between rooms) you have a triple leaf between rooms 1 and 4 - but the "air spings" here are sufficiently large that their resonance is far below the audible spectrum and are thus the plates are essentially decoupled (at least from the air springs) for our purposes. (This ignores the fact that this is a waste of space of course).

Okay, so now, HERE ARE MY QUESTIONS:

--> This is mostly talking about adding clips+drywall to an existing ceiling withOUT removing the old drywall.

--> 1 thru 3 b assume mounting drywall (one or two layers) on clips or RC, leaving the entire wall in place, resulting in a new 1.5" (or so) wide cavity.

1) I've read about a study that used multiple independently tuned Hemholtz resonators inside the cavity to damp the spring action. I wonder if it is possible to have enough damping of the spring with Hemholtz resonators that would fit in a very narrow space, such as 1/4" or 1" etc. I think it's easily possible to construct Hemholtz resonators from 1" galvanized pipe, or perhaps PVC pipe would do fine though the calculators I have don't seem to indicate Q - nor the reduction in terms of dB.

1b) On this line for installing RC+drywall OVER existing drywall (though probably not in my case due to asbestos): What if we used a jig saw, and make a series of 0.1" cuts (i.e. just kerf width) about every 5 or 6 inches in the existing drywall? This would turn the existing drywall and existing cavities into slat type resonators tuned at around 150 to 200 hz, which is about the resonance of a plate with a 1.5" distance to another plate. So the MSMSM becomes MS-resonator-M...

2) Is the spring a function of the compressibility of the fluid? (air)? Would the slot damping mentioned above merely provide a place for the air "to go", and would that in itself help?

3) If it is due to compressibility, what if the the spring in the newly added drywall cavity were simply vented or ported, say to another room?

3b) What if in this porting, the new cavity itself were tuned to itself be a hemholtz resonator (with a port to another room/area) - Where would the low frequency energy that was traveling through the wall go? I.e. we tune the cavity to resonate as a hemholtz resonator at the same frequency that the air spring WOULD have resonated if it were an MSMSM configuration, only now it is M(new)-Hemholzt S - MSM(old).

--> Questions 4 to 5b assume mounting the new drywall layer approximately 1/4" from the existing ceiling. 1/4" was chosen as the existing asbestos filled cottage cheese treatment is slightly less than 1/4".

4) Okay, so instead of using RC or clips, I use another method and mount new drywall 1/4" away from the existing. The mounting method used will be mechanically decoupled from the existing wall, but there will be the 1/4" air spring. According to the various calculators I've played with this seems to send the MSM resonant frequency way up to over 340 Hz (for normal sound) and as high as 3500 Hz for sound glancing at an angle 5 degrees from parallel to the wall. This calculator was NOT optimized for "triple leaf" calculations (and I'm not sure where to start on that).

4b) But I digress. The question I am asking is that if I used mechanical decoupling (something viscoelastic) and get the new drywall about 1/4" from the existing, driving the resonance of the new MSM up past 300 Hz, where mass of the new plate will be more effective, and putting this resonance in the area where there's "plenty" of TL going on already. Further, this MSM resonance is substantially higher than that of the existing wall or ceiling (around 50Hz for the ceiling).

M Sa M Sb M

Sa resonance 300 Hz, Sb resonance 44 Hz. Now, nearly 3 octaves apart, i'd think the two springs would not contribute to each other - would they even damp each other?

5) So here's a proposed configuration for the home theater ceiling:

Mass one (new): 5/8" drywall+GG+5/8" Drywall OR one sheet quietrock, then 1/4" airspace (leaving old cottage cheese in place), Mass one decoupled from Mass two using an isolating connector, Mass two (old) is 5/8" drywall (circa 1964), 9.5" air space (no insulation, 9.5" floor joists on 16" centers), then Mass Three (the floor) which I believe is some sort of thin concrete, probably poured over plywood.

Dw+GG+Dw - 1/4" air - Dw - 9.5" joist/Air - Ply/concrete


5b) Is my thinking here for this configuration flawed?


MORE GENERAL MSM questions (centered mostly on damping the spring):

6) So, for the very low frequency TL figures, the consensus is to increase the spacing between plates, and damp the spring using fiberglas batts, yes? So then in any case, we can assume that in MSM configurations the low frequency TL can be aided by damping the spring.

6b) So, the reason that the DISTANCE between plates is important and the total cavity volume is not (for MSM resonances) is that the fluid (air) becomes substantially more viscous, coupling the plates with that viscous fluid - yes? So it might be helpful to visualize that the cavity between the plates is filled with a liquid like hydraulic fluid (just for visualization purposes). Is this a good analogy?

7) Is an analogy that, at resonance, it's like tapping on a full aquarium? Or hitting a steel box (or 55 gal drum) filled with oil with a hammer ?

8) If this is the case, I wonder about if it is possible to convert the normal force to a sheering force.

9) And as I mentioned earlier - what about porting? Let's say we have an MSM wherein the resonance is 81Hz - so we add a port, or a number of ports. This should help to decouple the air spring from the two masses, yes?

10) What about bubble wrap? If the bubbles in a layer of bubble wrap were filled with carbon dioxide, they would have a different resonance point than the air - cold this dampen the spring enough to make a difference?



Thanks for muddling through my post Any comments or thoughts on these subjects are much appreciated as I attempt to understand the nuances of MSM and MSMSM configurations.


My general intention with the project:

WALLS:

In addition to the ceiling (the biggest issue), I intend to affix a single layer of 5/8" sheetrock, using GreenGlue (or other viscoelastic adhesive) to the existing walls. (so no MSMSM issue here, just damping with the GG solution).

I wonder though if this is enough to tame the exterior sound issues like to guy with the van and subwoofer.


FLOOR:

And then there is the floor. Float it on 2x2s? But won't THAT create another small distance MSMSM situation?

What about just laying down Acoustiblok under the carpet? I don't care for MLV that much, so is acoustiblok really useful? Substantially better than MLV?

What about "Quietwood" from the Quietrok people? I imagine it's pricey, but I do want to keep the increased floor thickness under an inch.

OR how about 1/2" MDF, with Acoustiblok between it and the existing floor? then foam/carpet on top. The doors will still work, but will I see realistic improvements?


And... bleh, now I have a headache. again thanks for reading, and any comments, thoughts, suggestions, etc. are very appreciated!!


Cheers,

Andy

Hello all, thanks for the replies thus far.

BigMouthinDc, and Dennis: just to reiterate, the asbestos cannot be removed. Federal and California laws dictate restrictions on how asbestos can be removed and disposed of. In the present case it is far far too expensive.

There is nothing wrong with undisturbed asbestos - covering g it up and encapsulating it is just fine.

It is relavent only for the ceiling, the primary source of the impact noise, as I only ever intended to use drywall+greenglue against the existing walls.

For the ceiling most I could consider would be to scrape off the cottage cheese Treatment in specific areas so that drywall could be affixed to the ceiling with green glue. But without decoupling with clips/rc, I'm not sure it's enough to damp the impact noise.


3finger: I'll make some diagrams and post.


Speedskater: room in a room was my first idea... But I don't want to lose the room height which is already just 8'.

So let's assume there is no space to float an Inteior room.


The project parameters are:

1) existing ceiling and walls must remain.

2) carpet can be removed.

3) any addition to the floor but not interfere with doors when carpet is installed.

4) The principal issue from above is impact noise. ( diplomaacy here has failed)

5) The principal issue from below is the downstairs neighboor playing music too loud (diplomacy has worked here)

6) the principal issue with walls that face the exterior is reduction of subwoofer/saturday night boom box noise

7) the principal issue with inteior walls is reducing the noise I transmit when I watch Apocolypse Now! at the proper level....


All that said, I was hoping to discuss MSM theory a little more in depth, this information about the project was really more for background.

I am very interested in the causes of msm resonace, and I assebit is due to air becoming very viscous at the resonance point??


Thanks again for the replies....


Andy

LOL - yes, I AM considering that, but have not found a place as well situated as this one.

And I LOATH moving.



I'm not saying it is "theory", I am just asking about the underlying reasons that it acts the way it does - my assumption is that air becomes very viscous at low frequencies, and loses its compressibility, especially at the MSM resonance.

This is something I want to explore and discuss, and it is the main reason for my post.




I understand this, however, it is not my question. I want to discuss the underlying mechanics of MSM and MSMSM configurations.

For the purpose of discussion, we have to assume that the existing ceiling cannot be removed, and so the point is moot.

It would cost approximately $20,000 to remove the existing ceiling because of the asbestos. To do it legally It's over $3600 just to scrape the popcorn asbestos treatment, not to mention the drywall.

For that kind of money I could just use multiple layers of QuietRock THX.


QUOTE=Dennis Erskine;18337803]
Eliminating the option of removing existing materials frankly means you'll need to lower your expectations or increase your costs. [/quote]

And this is where I am looking to determine the value of doing any isolation here at all.

In other-words, what can I really expect, given the constraints. Assume that I want the ceiling and the floor to compliment the wall treatment.

The wall treatment is going to be no more than a single sheet of 5/8" sheetrock affixed to the existing wall using greenglue (or other viscoelastic attachment method, such as walldamp).

So, I'd like the ceiling and and floor to perform similarly to the wall, given the constraints.

QUOTE=Dennis Erskine;18337803]
And, btw, your named Mass Loaded Vinyl product will do little more than assist in increasing your costs.[/quote]

I mentioned the wrong product - I was actually thinking of "walldamp" and "IsoDeck" from ASC.

The idea is using 5/8" QuietWood floor over the existing floor, or 5/8" MDF, over the existing floor using WallDamp or IsoDeck between the existing floor and the MDF.



As mentioned, material removal is not possible.



On the "triple leaf" MSMSM issue, I notice that WallDamp (being 1/6" thick) leaves a 1/6" air gap between plates.

The makers of green glue suggest leaving "blank space" to make small air gaps there as well.


At 1/16" between two sheets of 5/8" sheetrock, the MSM resonance is about 600 Hz.

So, it seems that one can use damping and very small airspaces, so my question is, at what point does it really start to fail?

Clips+hat channel over existing would give us an MSM resonance of about 150Hz - right in a trouble area where we need all the isolation we can get.

On the other hand, having an MSM resonance at 600+ Hz still seems to give good transmission loss curves, at least in the tests I've read at various manufacture's sites.



And I have been reading more about using multiple independently tuned helmholtz resonators inside an air cavity to damp the air spring of that cavity.

According to NASA, "substantial improvements" can be made with this technique.

http://adsabs.harvard.edu/abs/1988JSV...124..367M

Though I haven't seen anyone talking about this in terms of room/studio/theater acoustics - so I think it's worth looking into, even as a way to improve LF TL in the under 125Hz area for typical double leaf configurations.


Cheers

Andy

First: Hi Terry! Thanks for chiming in! Your comments on MSM are the kind of information I'm looking for.

And, Dennis, thanks for sticking with me - I know I can be annoying, especially when I'm trying to learn something new (MAM in this case) - your comments and experience are very helpful.

Now:



Okay, understood (finally).

Now, as to the ability of air to be an effective coupler (spring) as a certain frequency, here's a question (well, a couple):

I very curious about the Q of an MSM configuration - is the Q essentially constant regardless of the resonance frequency?

And does the TL in dB at peak resonance vary relative to the distance between plates? Or total air cavity?

I.e. do larger air cavities not only lower the resonant frequency, but ALSO reduce the detrimental effect on TL?


I might play with some FEA software this weekend to explore this.




Okay, that had me confused - it was due to playing with an MSMresonance calculator from Eric Desart, which can be found here:

http://forum.studiotips.com/viewtopic.php?f=27&t=1414




Yea, I wondered about that.

I found the more recent study I was actually thinking of, and he indicated using a plurality of resonators, at various frequencies around the MSM resonance.

http://www.empa.ch/plugin/template/empa/*/48491




The front door is my only real door concern - I just examined it and I should be able to add an inch to 1.5". Maybe 2" if I eliminate carpet there and instead use thin tile over the door swing area.

By Acousik Mat, do you mean "QuietZone" acoustic mat?

And can green glue be used between flooring materials?

What about acoustic mat + 1/2" MDF + GG + 1/2"MDF ?

And here's a crazy thought - obv. drywall can't be used as a floor surface, or an underlayment, but what about as an internal layer? The *compression* strength of drywall is good, it's just bad in terms of tensile strength.

BUT, if it was: Existing Floor + viscoelastic + drywall + viscoelastic + MDF or Particleboard then the top MDF layer would spread the load over the drywall adequately, and the drywall would be a lowest-cost way to increase mass.





Any idea of the depth of the GG air gaps? I've never worked with the material, so when you sandwich two gypsum panels together with 2 tubes of GG, what is the resulting surface distance of the inside faces of the panels?



Okay, I think that makes sense, so really what we are saying is the AIR between plates in terms of MSM resonance is resonating at the air's "coincidence frequency", not to be confused with the modal resonances? But I though in terms of musical instruments, we were talking about modal, and not coincidence, resonances?




Does GG damp the "air spring"? Or just damp the solid mechanical connections between plates and studs, etc.?



I guess I was imagining making resonators out of galvanized pipe which were long and thin, though like I said I have only read a small amount about this.


HERE'S A THOUGHT: Mount the Helmholtz resonators OUTSIDE of the air cavity (i.e. in the room) with the open port connecting into the cavity. This would allow for tuning the resonators to the wall AFTER the wall was constructed.

Note that this is not just for fixing triple leaf issues, but perhaps could greatly improve typical double leaf designs?

Even GG walls with: Dw + GG + Dw + RC + Stud + Dw + GG + Dw

Only have 20 dB TL at 80 Hz. Could resonators damp the air spring here and improve LF TL even further?




The added mass of the pressure vessel could only help...

(of course then I'd be stuck trying to quiet the vacuum pump...)




Right... so, damping the fluid - been trying to consider that aspect. What about some sort of large bubble wrap sheet in the middle of the air cavity, wherein the bubbles were filled with Carbon Dioxide? That would create a layer with significantly different "springyness"?


Something else to model....


Cheers,

Andy

I guess I wasn't thinking GG was useful in flooring, though thanks, I've now looked at their site with that in mind.

What do you think of this configuration:

ExistingFloor + GG + 1/2" drywall + GG + 1/2" MDF + carpetPad + carpet

OR

ExistingFloor + rubberMat (or cork?) + 1/2" drywall + GG + 1/2" MDF + carpetPad + carpet





How about a broadband TL of 10 dB? Reasonable?



My thought on the electrical outlets was to remove only enough drywall around an outlet to access it, move it forward (to be flush with the new wall surface), and cover the box with something like metacaulk or QuietPutty.

For HVAC, this building does not have central HVAC. The advantage is that I have control on how I deal with this and I'm considering some ideas - though the decisions on the ceiling/walls/floor will drive that.




Okay, then how about rubber mat? Cork? GG?



Cheers!

Andy