Membranes vs VPRs

#1

Hi all,

Firstly, I just want to say thanks to all the folks who contribute to this forum. It's a wealth of knowledge. Thank you.

OK, I've got a few theoretical questions that the pro designers among us may be unwilling to publicly divulge their expertise on. That's fine - I'll try my luck, and maybe others can chime in.

Imagine you have two membrane traps that attenuate the same frequency: one trap has a heavier mass membrane and a shallower cabinet, one trap has lighter mass membrane and a deeper cabinet. If room dimensions and practical considerations are not an issue, which trap would perform better? By performance I mean start to attenuate quicker/quicker to achieve oscillation and damping.

There's a few aspects to this equation that I can't get my head around:
1. A heavier mass membrane will take longer to oscillate, but...
2. Stiffer membrane + shallower cabinet will increase the stiffness of the spring (air)
3. A shallow cabinet reduces amount of absorption you can use, affecting spring stiffness + damping
4. Maybe proximity to wall boundary, where modal energy is highest, affects a membrane's speed to oscillation

That said (and without doing any real testing), I suspect that a thinner/lighter membrane performs better - I say this because, after reading many textbooks and scouring online resources, I've observed that designers generally don't prefer to use the heaviest mass possible on the shallowest cabinet possible. This leads me to my final question...

VPRs!
Given all I've written above, is the sole purpose of a VPR to add mass without adding too much stiffness to the membrane? What performs better: a 10kg/m2 membrane or a 5kg/m2 steel sheet sandwiched between 2.5kg/m2 MLV? This is assuming that all other parameters are adjusted so that both traps attenuate the same frequency.

Hope I explained myself clearly,
Aureliano

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

As a sort-of-related aside, I find it interesting that sealed membrane traps are generally not used in FTB or non-environment rooms. Instead there are sufficient layers of damping material (with membrane used as a constraint layer) to attenuate reflections and resonances. There is no 'waiting' for a sealed membrane trap to build up to resonation before it can start attenuating. Then there's SBIR, which cannot be efficiently dealt with by resonation: the reflection needs to be attenuated in one passage through the treatment. I wonder to what extent this logic of 'speed' can be applied to sealed membrane traps. If you have to use a sealed trap because of room restraints, how do you get the quickest resonation/oscillation? But what about dampening? What's the peak of the bell-curve when it comes to factoring in: quickness to oscillation vs quickness to damping? More damping (insulation material) would slow the speed of oscillation/resonation, which affects attenuation.

All the calculators I've ever used just show the absorption co-efficient. Is there anyway to calculate (or theorise) response in the time domain? Is this even an important issue? Perhaps the differences are negligible.

#2

welcome!

a VPR "membrane" is a piece of sheet steel (mostly) which has mass and stiffness. it works because the steel will more quickly create the resonant modes (steel - speed of sound transmission) and as it's thin surface (mostly) is damped quickly from the bound foam backing, it can be highly absorptive at the frequencies where it has those modal responses.

so like any other absorption membrane, to absorber low frequencies, it has to be large. say 1m x 2m (quite heavy even at 2mm thick), and still has to be placed on mode points in the space (corners work nicely here). why large? LF sounds are large :-)

small stuff might do some bits of attenuation, but most of it goes around.

so using layers of insulation and MLV (for example) can cover much wider areas (which means more attenuation even if "inefficient") and tortured pathways to/from hard boundaries are the trick.

#3

I learned something new ...

Thanks gullfo.

#4

Thanks for replying Glenn. This is so cool. The speed of sound transmission in steel is a factor I haven't come across before (with regard to VPR design), and it makes sense.

Would you say there's a certain threshold of thickness and/or kg/m2 of a membrane like MLV that ought not to be crossed? And from that point switch over to a VPR-style design?

I'm imagining a situation where you have, say, only 30cm of space on the rear wall, but a 3mm 5kg/m2 membrane on a 30cm deep sealed cabinet would not be sufficient to reach a specific modal frequency. Perhaps using anything thicker/heavier with a MLV-style membrane is inefficient, and its better try VPR.

#5

i have done a couple of designs which used VPR -- but the units were large and heavy (~500Kg each) and were somewhat needed in very large places to kill deep modal issues. but in general, tortured path layers using 1lb/ft2 or even 2lb/ft2 MLV often quells the necessary LF response anomalies.

#6

gullfo wrote:so using layers of insulation and MLV (for example) can cover much wider areas (which means more attenuation even if "inefficient") and tortured pathways to/from hard boundaries are the trick.

Are you able to explain why most designers use a denser layer of insulation material first, in front of less dense material? (I say first as in the side closest to the speaker)

I've never managed to find a solid answer to this question. I guess it has something to do with surface impedance affecting flow rate.

When I try using the multi-layer absorber calculator in AcousticModelling.com, using a denser material first gives a worse absorption coefficient. I'm not sure why.

#7

the denser stuff can help with supporting HMF and HF on purely absorptive surfaces. but yes, denser (and remember large coverage) helps with attenuation across a large range because sound is made up of particles colliding in patterns we see as waves. and we also know that dense boundaries (like walls) will reflect the energy - so not only do the boundary constructs matter, but the patch back from the reflected energy matters, i might say more as the reason for the issues we are trying to solve are mostly due to hard boundaries...

#8

gullfo wrote:Source of the postthe patch back from the reflected energy matters, i might say more as the reason for the issues we are trying to solve are mostly due to hard boundaries...

This is a really helpful way to think about it. Thank you.

I've been thinking a lot about tortured pathways + large coverage since you first replied Glenn, and the pros and cons of different applications of them. I remember seeing somewhere that you've used a 60/40 ratio to stagger membranes between layers of insulation material. That seems like a great approach to improve the performance of panels and deep absorbers. In wall designs though, it seems like 100% coverage is the norm - like in Newall rooms. Then there's waveguides. Brandt uses waveguides in corners, which to me makes sense, as waveguides will be more broadband than a sealed membrane resonator, and since all modes terminate in corners, having the most tortured, non-sealed path possible there would be optimal. Then, for modal issues, use sealed membranes in the centre of the related walls.

Another example of a tortured path would be in the screenshot I've attached. It's a wall of a certain well-known designer. The arrows (not drawn by me) point to 'constraint membrane' [text written in non-cropped version]. It looks as though there are two different layers of insulation material, of the same thickness, followed by a wooden frame (I guess) which would suspend a membrane. This pattern is then repeated, maybe three times. I don't know for sure if it really is a membrane, its just a photo I found online, but it's reasonable to expect it is having read many of Thomas' posts online.

It's impossible to know what is going on from a picture, and each design is different and specific to the room, etc etc. I'm just trying to think through the logic of this approach. I might be wrong but I imagine the membrane is held against the insulation; so: two layers of insulation > air gap > membrane (repeat). The air gap is another layer of resistance/impedance change, and keeping space in front of the membrane stops it from being over-damped.

The size of this supposed air gap made me think though,
- Could you put a membrane either side of the air gap? Both membranes touching insulation, facing each other with air in-between. Would that increase performance?
- Could the 2nd iteration of the pattern in the photo have a heavier membrane? Makes sense to me: the further through the treatment you go, the lower (in terms of targeted frequency) you could try to absorb/attenuate.

#9

besides tortured path you also have impedance shifts and mechanical movement to convert energy. depending - a membrane which is unconstrained or full damped by contact _could_ be more efficient. but most times that usage would be to target something specific. but as a general rule: broadband first. i'm a fan of wave guides as well as slat resonantors, PoA (plain old absorption), tortured path, and membranes. and in most of my designs there are most of those... :-)

Soundman2020 - Studio Design Forum