gypsum/drywall: glued vs screwed

[Brian Dayton]

as a follow up to my thought posted above, i tested 3 small "panels" via knock stimulation and pink noise stimulation, FFT's the results and submit them here. the magnitude information should be totally discarded, i just wanted to observe the frequency of various modes.

test specimen were 18" x 20" 2x4 wood frame with double drywall (5/8") on both sides. they were bonded with

1) nothing
2) a flexible adhesive
3) ordinary drywall adhesive

this was a 15 min test, pic quality is modest, but it seems that the frequencies do correlate.

in all graphs the choppy response is accelerometer measured data for a 10 second pink noise pulse. the smooth response is the FFT of about 1 second of a single impact/decay measurement.

perhaps this is worth something in future tests.

[Brian Dayton]

please find attached the data on various commercially available adhesives and calculated impact on the critical frequency. location of critical frequency was determined via modulus, which was determined via a resonance test.

i decline to divulge the make/product of commercial damping glue mentioned within, as i am not affiliated with the maker and have no desire to counter any possible published specs that they may have.

enjoy, and by all means comment as you wish

this file and the material it contains are copyrighted to me, see last page.

i have retained all of the files taken during testing for future reference if anyone has an interest in pulling additional data from them.

PLEASE SEE NEW DOCUMENT IN NEW THREAD AND DISCARD OLD VERSION, IT IS WELL WORTH THE ADDITIONAL DOWNLOADING TRAUMA, THANKS

Brian

[Brian Dayton]

the paper mentioned above is now located in another thread, forward on this board.

some clarifications are in order, and i'm back to make simpler a copule of points after a note from a friend that i got a little long-winded above.  Brevity in type is a work-in-progress for this guy...  :)


first, gluing rigid members between the two leaves of a wall does not represent damping.  Damping is the rate of energy dissipation in a system - literally, that is the definition - and gluing the two sides of a double leaf wall together should be considered stiffening, not damping.

second, in data published by the NRC, stud spacing as close as 12" on center had a very significant change in the location of the MSM resonance, but no effect on the coincidence location at all.

third, adding insulation to a cavity will generally have no effect to a slight RAISING of the coincidence location (refer to IR-761 for a handful of examples where this slight or no effect is illustrated).  The resonance above shifts down considerably when insulation is added.

fourth, adding insulation to the cavity has a menial effect on TL below the 600-1000hz dips, and had a notable impact on performance above the dip.  this is classic behavior of a double leaf wall.  It behaves approximately lke a single panel below the resonance, then de-couples (and insulation has an effect) above it.

fith, coincidence is a wave phenomenon, and increasing stud spacing is associated with an increase in the dpeth of the coincidence dip in many cases, as is making more resilient the wall (the bending waves can travel, being built-up by the waves in air, farther before they are interfered with by the studs).  the faint dip at 2000hz may represent a coincidence dip, or it may be off the screen in this case, i haven't a clue, if the data went higher we could be more certain of where the dip lies.

finally, and a bit off topic, note that the minima around the MSM of these systems was the same for both the 5" deep cavity and the 1" deep cavity.  

if we apply mass-air-mass resonance theory, it should have moved down by more than one ocatave.  why this does not shift should be explained as well.  i refer to curve 1 in the top graph and both curves in the second (Staggered stud) example.  

finally^2, it is interesting to note that the staggered stud construction - very much deeper, more decoupled, and with far more insulating material - exhibits poorer transmission loss than the thin sample.  more to think about, and more questions


this seems to be, perhaps, the largets misunderstanding among people - what damping is.  it's energy being removed from a system, that's it, and that's all.  stiffening is something else.

[Bob]

I'm unable to download that PDF.

[Brian Dayton]

i tried and got it by right clicking and selecting "save target as", but clicking directly failed for me as well

[Bob]

Thank's Brian. Right click then save target as works for me too.