The dip at 1443 Hz looks like anti resonance! [oh, have I opened a can of worms?]. Here's Wiki's thought on the subject:
"Anti resonances can occur in all types of coupled oscillator systems, including mechanical, acoustical, electromagnetic, and quantum systems. They have important applications in the characterization of complicated coupled systems".
Does the acrylic/mdf/acrylic/wood glue now look like two coupled systems?
The Fc would be a positive peak on your trace.
Last edited:
This is by no means complete, so I've labelled the graphs as "preliminary". It needs a lot more work, but more feedback may ensure it's not wasted effort.
I've picked out what I believe to be "good" pulses for a few existing panels and used Equalization, filtering and Logarithmic Decrement to calculate the Damping factor at several frequencys. In figure W, I've repeated the effort for two similar pulses on the PL (Orange and Brown traces) and Woodglue panels (Green and Red traces).
They don't repeat very well do they? I'm offering these explanations:
a) Hjam picks poor pulses.
b) That's the nature of the materials. More pulses are needed to characterize them.
c) Other? Bryan, have you got another explanation? I'm guessing you've already seen this sort of behaviour.
Figure X shows the waveform and spectrum of the "PVA" Pulses in figure W, just in case it's (a).
If it's (b) , the woodglue (PVA) traces look good up to 150 hz - (DF around 0.4). Above that, it's the bad part of town. The DF of the two traces do not really agree.
The PL glue does this at about 300 hz. I've only done one pulse for Maxstretch.
I'd say that Anti-resonance seems like a good explanation for the bad parts of town. So does Hysterysis, harmonic distortion, etc.
Thoughts?
Hugh
I've picked out what I believe to be "good" pulses for a few existing panels and used Equalization, filtering and Logarithmic Decrement to calculate the Damping factor at several frequencys. In figure W, I've repeated the effort for two similar pulses on the PL (Orange and Brown traces) and Woodglue panels (Green and Red traces).
They don't repeat very well do they? I'm offering these explanations:
a) Hjam picks poor pulses.
b) That's the nature of the materials. More pulses are needed to characterize them.
c) Other? Bryan, have you got another explanation? I'm guessing you've already seen this sort of behaviour.
Figure X shows the waveform and spectrum of the "PVA" Pulses in figure W, just in case it's (a).
If it's (b) , the woodglue (PVA) traces look good up to 150 hz - (DF around 0.4). Above that, it's the bad part of town. The DF of the two traces do not really agree.
The PL glue does this at about 300 hz. I've only done one pulse for Maxstretch.
I'd say that Anti-resonance seems like a good explanation for the bad parts of town. So does Hysterysis, harmonic distortion, etc.
Thoughts?
Hugh
Attachments
John,
I messed that up even on the second try. I'll try it again...
As a guess I think that using CLD (e.g. Greenglue) to damp Acrylic/MDF/Acrylic MAY cause higher amplitudes at lower frequencies than stiff glues like PL. I really need to test this again with a better method.
Jaw is sore from eating my words.
Could the peak around 1500 - 1700 hz be Fc? I did verify that the newer AD Accelerometer can get to at least 2500 hz.
H
I'm pretty sure I realize why my Panzerholz readings were mucked up earlier. I'll revisit those later on and compare them to the Acrylic MDF mix.
Is the Saturated sand anything like the newer play sands for kids? Moon Sand would be one brand. I think they add Propylene Glycol to the sand so it molds a bit like clay. Or is this actually adding an oil?
H
There is an oil added to produce a mixture.
The Sand/Bentonite/Oil are mixed to a ratio and as a damping material, it is looking to be very cost effective and has dissipation/damping qualities.
I usually share in demonstrations only and carry out subjective assessment, this method is not really suited to the assessments being discussed within this thread, that are supported by measurements.
A Web Search will show measurements carried out on Saturated Sand Damping and some will cover the variations in methods used to produce mixtures/ratio's.
The Sand/Bentonite/Oil are mixed to a ratio and as a damping material, it is looking to be very cost effective and has dissipation/damping qualities.
I usually share in demonstrations only and carry out subjective assessment, this method is not really suited to the assessments being discussed within this thread, that are supported by measurements.
A Web Search will show measurements carried out on Saturated Sand Damping and some will cover the variations in methods used to produce mixtures/ratio's.
I wouldn't say Green Glue is used in a cld configuration. For optimum results, the thickness of both the VE layer and constraining layer have to be thin, and the elastic modulus (Young's Mod') of the constraining layer has to be low (it must bend!).
I believe Green Glue is used as a thick layer, and as such, does not work as a cld VE layer, more as an extensional damping layer.
Here's a more about the adhesives. These tests are very dependent on which hammer tap pulses are picked - Frustrating! Each graph could use more repeats, but there is a general trend, so posting them anyway.
Figure W1 (left side) has two traces from two pulses on a panel that used Wood glue (PVA). The first trace suggests good damping ( above 0.35) up to almost 500 hz. But the 2nd trace suggests good damping only up to about 150 hz.
The right side uses Lepage's PL. Surprisingly, it seems better than the Wood glue. Good damping up to over 300 hz or almost 900 hz.
Figure W2 (left side) uses the softer GE Maxstretch. Now it's looking good past 400 hz - maybe even 1100 hz.
As a reference, the right side is a piece of 8mm Panzerholtz. In this example, the "good" region is up to 550 hz. I'm sure a second pulse will show differing results.
At one point, I thought Maxstretch may somehow amplify the low frequencies with repect to PL. A set of tests with beads indicates this is not so. The same test selecting hammer pulses says it is true. Stuck on this point for now.
I did try some 1/4" Acrylic with 1/8" MDF. That looks very good, so more on that later.
Hugh
Figure W1 (left side) has two traces from two pulses on a panel that used Wood glue (PVA). The first trace suggests good damping ( above 0.35) up to almost 500 hz. But the 2nd trace suggests good damping only up to about 150 hz.
The right side uses Lepage's PL. Surprisingly, it seems better than the Wood glue. Good damping up to over 300 hz or almost 900 hz.
Figure W2 (left side) uses the softer GE Maxstretch. Now it's looking good past 400 hz - maybe even 1100 hz.
As a reference, the right side is a piece of 8mm Panzerholtz. In this example, the "good" region is up to 550 hz. I'm sure a second pulse will show differing results.
At one point, I thought Maxstretch may somehow amplify the low frequencies with repect to PL. A set of tests with beads indicates this is not so. The same test selecting hammer pulses says it is true. Stuck on this point for now.
I did try some 1/4" Acrylic with 1/8" MDF. That looks very good, so more on that later.
Hugh
Attachments
I think I'm sorted out on the hard vs softer adhesives. It seems that PL and Maxstretch respond about the same to taps from a 12V solenoid. The applied taps are a bit stronger than hammer taps. Not good to measure damping, but good for evaluating the magnitude of the spectrums.
Earlier, when I was using hammer taps, the Maxstretch needed harder taps to get "good" pulses. This made the low frequency magnitudes appear larger than those for PL. So, it seems that the softer, stretchier Maxstretch allows good damping (> 0.35) at higher frequencies. If there's any sort of penalty, I'm not seeing it yet.
Earlier, when I was using hammer taps, the Maxstretch needed harder taps to get "good" pulses. This made the low frequency magnitudes appear larger than those for PL. So, it seems that the softer, stretchier Maxstretch allows good damping (> 0.35) at higher frequencies. If there's any sort of penalty, I'm not seeing it yet.
I use clamps to to press the layers together and squeeze the VE layer down as thin as I can. It's a poor adhesive, so screws are sometimes required to maintain contact ( therefore non-CLD).
I do recall seeing a paper describing multi layer CLD. But I don't recall if it allowed different materials in the layers. So let's go with your answer and say that Acrylic/MDF/Acrylic with Greenglue, Maxstretch, etc. is something other than CLD.
There's a video showing how to apply Greenglue between layers of Gypsum board. I recall that they do squeeze the sheets together. But they put long Drywall screws through the lot. So this too would not meet the definition of CLD.
I'm using a pressure sensitive for the adhesive. This is similar
3M™ Acrylic Adhesive Transfer Tape 468 (7955MP)
These adhesives do not dry out so retain the ability to disipate vibrations as heat via shear.
Used it to make a CL sandwiched between 2 x 4mm aluminium pieces for an armboard and worked a treat. I'm also working toward a top plate using 2mm polished stainless and 4mm of aluminium. The idea is, with a suspended TT design, to have the springs and their mounting bolts fixed to the polished steel part with the motors mounted to the underlying aluminium part.
Anyway, for what its worth, hopefully of use to someone
Chris
3M™ Acrylic Adhesive Transfer Tape 468 (7955MP)
These adhesives do not dry out so retain the ability to disipate vibrations as heat via shear.
Used it to make a CL sandwiched between 2 x 4mm aluminium pieces for an armboard and worked a treat. I'm also working toward a top plate using 2mm polished stainless and 4mm of aluminium. The idea is, with a suspended TT design, to have the springs and their mounting bolts fixed to the polished steel part with the motors mounted to the underlying aluminium part.
Anyway, for what its worth, hopefully of use to someone
Chris
Hi Hugh,
I believe it could work as you suggest, but, IIRC, extensional damping gives about 7dB loss of vibrational amplitude, while cld can give up to 25 dB. (That may not be exactly the figures, but near!).
I would not think bentonite granules or sand would provide any benefit in the green glue (or else it would have been added, where B or S are a lot cheaper!).
sounds good, Chris. I have several 3M products of different formats, including 3M 2552 cld tape and 3M VHB 4929, although I haven't been able to test any of them, yet. The 2552 cld tape cost nearly £150 a roll, now!
sounds good, Chris. I have several 3M products of different formats, including 3M 2552 cld tape and 3M VHB 4929, although I haven't been able to test any of them, yet. The 2552 cld tape cost nearly £150 a roll, now!
The product I link to is £12.40 a sheet at 610 x 900mm. Its simply a thin layer of adhesive on a release liner so perhaps worth a go given the low cost
