Adding Insulation Panels to Door Assemblies

#1

This post is intended as a supplement to pp. 109-111 in Rod Gervais "Home Recording Studio - Build It Like the Pros" book. If there is any copyright infringement please let me know and I'll amend the post accordingly.

It's an intriguing little section describing some acoustic laboratory measurements made by Brian Ravnaas;

Rod Gervais wrote:During the testing process for Green Glue, Brian Ravnaas decided (at some point) while taking a wall assembly apart to do a simple test of one-sided assembly. He wanted to find the effect that insulation played in regards to the TL value for something like a door panel.

So they removed the drywall from one side only and experimented with a few different configurations. Figure 5.13 is a chart with the results from those tests.

I wanted to do some analysis of those results so transcribed them into an Excel spreadsheet. I've attached it in case anyone else might find it useful. As a bonus I was able to get a graph somewhat clearer than the one in the book. There are some very worthwhile gains evident in the midrange!

I also attempted to figure out the mass of the single leaf panel in those measurements. It is two layers of plasterboard but the thickness is not specified. The tests in the original table are numbered OL-061214 through to OL-061218. Stuart has some tests from this series in his post about Green Glue;

https://digistar.cl/Forum/viewtopic.php?t=77

Those tests are OL-061222 & OL-061223, and the plasterboard in those is 5/8" gypsum board, two layers giving a density of 24 kg/m2.

By way of real world comparison, a typical UK FD30 rated fire door is 27k/m2. The mass law gives a 1dB theoretical TL improvement on the results above. Not too shabby at all.

Cheers!
Jennifer

#2

Thanks Jennifer! I don't recall seeing that graph before in Rods' book. Maybe I have an earlier edition. But it is very interesting, and completely demonstrates the point completely filling the wall cavity produces a very large increase in isolation: partial fill is not very good at all. There are very good reasons for that in acosutic theory, but its always good to have a simple graph that demonstrates it so clearly. I once posted something similar on Gearslutz, showing that there's a difference of up to about 16 dB in isolation between not insulating and fully insulating a well-built studio wall: someone over there whose should know much better posting a scathing response, saying it was all garbage, and that insulation only had a minor effect, despite all the evidence I posted... including some from that person's very own website! Hard to fathom the hatred that seems to go around, sometimes. But anyway, your graph is very clear indication that the text books and equations actually are correct, and that theory does a good job predicting reality, despite the unfounded opinions of self-appointed "experts"!

If there is any copyright infringement please let me know and I'll amend the post accordingly.

That's probably fin, and falls under "fair use". Besides, Rod seems to be fairly cool with that: as long as you attribute the info to him, he seems to be OK with it, from what I've seen in the past.

- Stuart -

#3

a single panel will be limited by mass law, and the higher frequencies will be attenuated by the absorption of the insulation - which besides burning off that energy, also damps the panel which improves the transition (somewhat flattened) on coincident frequency of the panel.

#4

Thanks chaps, very illuminating.

So if I understand correctly, the mass law by itself would only give a 5 dB TL increase with each doubling of frequency. This does indeed seem to be the case for the midrange frequencies with the untreated panel, apart from the coincidence frequency dip.

With full coverage of R19 insulation this gets closer to 10 dB in the midrange. Nice. Probably about as close as we'll ever get to a free lunch in this lark, other than the "air gap behind absorber" effect.

There's an additional quite from Brian Ravnaas you may enjoy Stuart;

Brian Ravnaas wrote:The results speak for themselves. Covering a single leaf wall (like a door) in absorbing material results in a profound improvement in transmission loss, beginning at the frequency where the absorbing material in question becomes very effective. Covering the entire surface is distinctly more effective than covering just a portion of the surface at higher frequencies...
We anticipate that the frequency at which the absorbing material layer will start to be meaningfully effective will be, as is true for sound absorption, higher with thicker material, etc. So 3" of 703 should work better and lower than 3" of R13, 4" of 703 should be effective to a lower frequency, and more effective overall than 2" of 703.

(Perhaps the higher word in bold is a typo and was meant to be lower?)

So assuming the door is not at a first reflection point or similar, I've been wondering if 705 would be better than 703 in this case? It seems like the insulation is operating in a similar way to how it would if one were attempting to reduce sound transmission from a machine room, for example, and not concerned about anything being reflected back into the noisy room.

BTW I'd love to read Brian Ravnaas "document zero" on this, but can't find it anywhere.

Cheers,
Jennifer

#5

Correction:

In the opening post I stated that "a typical UK FD30 rated fire door is 27k/m2". This was based on a standard domestic sized door weighing 45kg. I've since discovered that although some FD30 fire doors weigh 45kg, many, perhaps most, weigh 30kg. This gives a surface density of 18kg/m2, a bit less than the 24kg/m2 of the single leaf panel mentioned in Rod's book.

Cheers,
Jennifer

#6

the thing is, fire rating vs mass depends on the door construction - many fire rated doors are metal filled with foam vs a solid material (wood or more likely some chemically infused wood product). so mass can vary - in the end, 90% of the issue is the seals and damping the door. and two lightweight back to back doors with good seals and damped, on separate decoupled walls, will give a significant level of isolation at least on par with a super door in the same situation. however, i'm a fan of single doors where possible because i find busy room to room movement is hindered by multiple doors... :-)

#7

Yes, and the single superdoor is also essential, when like in my place it is only possible to have a single door. The fire doors in the house all have a ceramic core in a wooden frame. The outside panels are (I think) MDF or a similar material.

All that being said, there is something uniquely satisfying about building and installing a single superdoor that works as you hoped it would. I think partly to do with the combination of building and installing something that is very heavy and awkward, yet must be built to very precise tolerances, be able to move, and at the same time look good

#8

if you film the super door movement and play it in slow motion, you just need some epic bank vault music to accompany it ;-)

Soundman2020 - Studio Design Forum