A 300mm. square panel of 10mm. mdf rigidly fixed around its periphery, will if we estimate a spl. of 140db., a total load of 18Newtons have a peak displacement of around .1mm.
if you double the thickness you get .0127mm. of movement.
The reference I have cited previously clearly states that for constrained layer damping to be effective one needs a materiel that displays hysteresis over over the range of stresses and consequent strains that occur in the system that you want to damp.
The point is for the types of displacements mentioned above green glue type compounds are not pushed into this region so they cannot provide constrained layer damping.
I would be very surprised if the various vibrational modes one finds in automotive panels are restricted to these types of displacements.
Rcw.
if you double the thickness you get .0127mm. of movement.
The reference I have cited previously clearly states that for constrained layer damping to be effective one needs a materiel that displays hysteresis over over the range of stresses and consequent strains that occur in the system that you want to damp.
The point is for the types of displacements mentioned above green glue type compounds are not pushed into this region so they cannot provide constrained layer damping.
I would be very surprised if the various vibrational modes one finds in automotive panels are restricted to these types of displacements.
Rcw.
Titebond Melamine Glue is designed for bonding wood, particleboard, MDF and other porous substrates to synthetic materials such as melamine, vinyl and HPL as well as metals
www.Titebond.com
www.Titebond.com
I glanced at the report and it doesn't say that. Even if it did the application is entirely different. They use a thin layer and in practice, on a loudspeaker enclosure a relatively thick layer is used. Another big difference in the reference and typical practice in other industries is that the layers on either side of the viscous layer are usually of comparable stiffness. This maximizes the shear on the inner layer and hence maximizes its effectiveness. The "hysterisis" that you talk about is basicaly the loss factor as the semi-minor axis and the stiffness as the semi-major axis. No one, not even the author, use anything but loss factor when talking about the material. One does want a maximum loss factor, with little stiffness - the author and I agree on that. But to say that the displacements are too small for the effect to take place is simply not true. As I said CLD is used in luxury cars where the materials are steel and the displacements are even smaller.
True, but it will bond wood to wood, and it remains flexible and high loss. Its what I use.
It does say that the energy dissipated is proportional to the loss modulus of the viscous material times the square of the peak displacement.
So with any given damping compound the amount of damping you get is proportional to the square of the movement.
Another reference, (that I can't locate at the moment), indicates that there is a fairly reliable relationship between the the loss that a substance has for a given strain, and its yield strength.
In the data the nitrile rubber 40A gives the best damping in the described system, and rough calculations I have done so far indicate that the scheme phoenix 358 outlined is a good one, this uses a 3mm. Sheet of a rubber as a damping layer, and this seems to be the sort of stuff that has the characteristics needed.
Rcw.
So with any given damping compound the amount of damping you get is proportional to the square of the movement.
Another reference, (that I can't locate at the moment), indicates that there is a fairly reliable relationship between the the loss that a substance has for a given strain, and its yield strength.
In the data the nitrile rubber 40A gives the best damping in the described system, and rough calculations I have done so far indicate that the scheme phoenix 358 outlined is a good one, this uses a 3mm. Sheet of a rubber as a damping layer, and this seems to be the sort of stuff that has the characteristics needed.
Rcw.
Hello All,
This is an interesting thread somewhere between Diy and commercial. Focusing on the DIY I suppose that the goal is to keep the power from the back side of the driver inside the box or at least reduce the transmission through the panel to a low value in comparison to what comes out the front. No matter how effective we are at keeping the 180 degrees out of phase sound power inside the box some will escape. The panels will vibrate at their own resonate frequencies. The box resonate frequencies are not very musical especially when added to what comes out the front of the box. When the panel resonate frequencies are near the primary frequency of the driver a beat frequency is the result. Guitar players use beat frequency to get things in tune, as do piano tuners. I prefer the concert to the tuning session.
In this discussion of materials and methods there seems to be an assumption that MDF and the like are homogeneous and that a thin layer of glue or epoxy between two layers of MDF has magic properties. Not true. Have you ever driven on highway 101 through lumber country behind a flatbed 18 wheeler with a 20 ton black bladder of magic glue on the bed? That glue is mixed with what was “waste” wood fiber, extruded then baked. Not a sandwich but a cookie. There you have it, MDF. When flexed MDF will return to the original position but with less pop. Internal friction due to shear will convert vibration into heat.
It seems that if we are intent to make a sandwich it should have some meat in the middle, how about a 1/8 inch layer of closed cell neoprene foam. Neoprene Closed Cell Foam and Specialty Foam : Water Resistant Fabric : Seattlefabrics.com and 3mm Neoprene
Anyone up to some materials testing?
DT
Just for fun!
This is an interesting thread somewhere between Diy and commercial. Focusing on the DIY I suppose that the goal is to keep the power from the back side of the driver inside the box or at least reduce the transmission through the panel to a low value in comparison to what comes out the front. No matter how effective we are at keeping the 180 degrees out of phase sound power inside the box some will escape. The panels will vibrate at their own resonate frequencies. The box resonate frequencies are not very musical especially when added to what comes out the front of the box. When the panel resonate frequencies are near the primary frequency of the driver a beat frequency is the result. Guitar players use beat frequency to get things in tune, as do piano tuners. I prefer the concert to the tuning session.
In this discussion of materials and methods there seems to be an assumption that MDF and the like are homogeneous and that a thin layer of glue or epoxy between two layers of MDF has magic properties. Not true. Have you ever driven on highway 101 through lumber country behind a flatbed 18 wheeler with a 20 ton black bladder of magic glue on the bed? That glue is mixed with what was “waste” wood fiber, extruded then baked. Not a sandwich but a cookie. There you have it, MDF. When flexed MDF will return to the original position but with less pop. Internal friction due to shear will convert vibration into heat.
It seems that if we are intent to make a sandwich it should have some meat in the middle, how about a 1/8 inch layer of closed cell neoprene foam. Neoprene Closed Cell Foam and Specialty Foam : Water Resistant Fabric : Seattlefabrics.com and 3mm Neoprene
Anyone up to some materials testing?
DT
Just for fun!
What I was attempting to point out is the storage modulus is a function of the slope of the major axis of the hysteresis curve, and the loss modulus is some fraction of the storage modulus.
With a given damping material you can increase the shear strain by increasing the thickness of the damping layer, but for small panel displacements the thickness of some of the materials being advocated rapidly becomes excessive, and it is then better to use a material that has a higher storage modulus.
This gives a scheme with the maximum thickness, (and therefore stiffness), in the two outer panels and minimizes the thickness of the constrained layer.
Rcw.
With a given damping material you can increase the shear strain by increasing the thickness of the damping layer, but for small panel displacements the thickness of some of the materials being advocated rapidly becomes excessive, and it is then better to use a material that has a higher storage modulus.
This gives a scheme with the maximum thickness, (and therefore stiffness), in the two outer panels and minimizes the thickness of the constrained layer.
Rcw.
Finding the optimum layer stiffness and loss factors is indeed a difficult task, and one that needs to be done for the specific example (hence not quoting aircraft landing gear studies). But I have found, in general, that typical cabinet materials (1/2" MDF or plywood) along with typical bonding/viscous layer agents (Melamine glue or soft polyurethane) work very well. Are they "optimum"? - probably not. Are the materials readily available to do the "optimum"? - probably not.
Hi All,
New here and was browsing the threads to see if anyone has worked with plastiboard (recycled plastic "plywood" like) sheets as a dampening layer.
Wishbone Ltd. | Recycled Plastic Sheeting
http://www.renewresources.com/pdfs/polyboard_sheetsize.pdf
This stuff is dense (between 51 and 56 lb/ft3, denser than MDF which is about 47 lb/ft3), flexible, comes in 4'x8' sheets, would be easy to work with, should glue OK and not all that pricey. The specs list a sound transmission loss of 56.5 dB over 80 to 6300 Hz. Would this material be good as the inner part of a constrained layer?
Also found this page interesting.
Loudspeaker construction
He makes a few good points about where the sound energy inside a speaker enclosure goes : it can be converted into heat in the stuffing or the walls or it has to exit the enclosure, through the speaker cone or the walls.
I'm thinking a thin layer of HDF (hardboard), 3/8" plastiboard and then MDF.
Last edited:
You mean Green Glue, right? Have you tested Green Glue? I'm sure the difference in Q between Green Glue + wood panel - vs - Tightbond Melamine glue & wood panel is much greater with the GG. Green Glue should be much superior at decoupling the two panels it is used between.
It should also be noted that decoupling may be better for speaker boxes than constrained layer damping. CLD is really preferred if using metals, although it could make sheet steel a more viable material for making a speaker box out of.
http://www.earsc.com/pdfs/engineering/understandingdamping.pdf
http://www.earsc.com/pdfs/engineering/understandingdamping.pdf
Thanks for the link ... interesting article. There is a forum where decoupling for drivers was discussed.
RCW: I for one, follow your hysteresis point. While the material remains in the elastic regionn of deformation, energy is conserved, damping is zero to minimal. When stressed into plastic deformation, energy is converted to heat, partly. Since many polymers are essentially plastic in nature a laminated polymer panel makes a great deal of sense, above those which rely on fibrous tension for their rigidity.
Bringing this old thread back to life..
Surely the damping materials in common use for audio like sheet butyl rubber, sorbothane or green glue are in hysteresis pretty much all the time since they do not have any 'spring'. Take a sheet of 1mm thick sorbothane rubber and see if it returns the energy put in during displacement; it does not, right? Pretty much all energy disappears as deformation. Perhaps this is not the case at audio frequencies though?
Regarding materials of twice the thickness having eight times the stiffness this may be true for static loads, but not at resonance. All materials will have resonance regardless of thickness / stiffness. Read the BBC paper where they test 9mm ply against 18mm ply and find equally high amplitude resonances in each only at higher frequency in the 18mm sample.
I would certainly like to know how to calculate the optimal damping layer thickness for CLD. Is it related to the base and constraining layer thickness, or only stiffness? If we took two layers of 18mm ply I wonder what the optimal damping layer thickness is? I've seen research that shows the damping will increase with a thicker damping layer until all of a sudden it gives hardly any damping at all. In the case I remember going from 9mm to 10mm damping layer gave a difference between best and worst damping performance.
Surely the damping materials in common use for audio like sheet butyl rubber, sorbothane or green glue are in hysteresis pretty much all the time since they do not have any 'spring'. Take a sheet of 1mm thick sorbothane rubber and see if it returns the energy put in during displacement; it does not, right? Pretty much all energy disappears as deformation. Perhaps this is not the case at audio frequencies though?
Regarding materials of twice the thickness having eight times the stiffness this may be true for static loads, but not at resonance. All materials will have resonance regardless of thickness / stiffness. Read the BBC paper where they test 9mm ply against 18mm ply and find equally high amplitude resonances in each only at higher frequency in the 18mm sample.
I would certainly like to know how to calculate the optimal damping layer thickness for CLD. Is it related to the base and constraining layer thickness, or only stiffness? If we took two layers of 18mm ply I wonder what the optimal damping layer thickness is? I've seen research that shows the damping will increase with a thicker damping layer until all of a sudden it gives hardly any damping at all. In the case I remember going from 9mm to 10mm damping layer gave a difference between best and worst damping performance.
Last edited:
Sorry, but this does not seem feasible. I would expect the damping to increase to a maximum with increased thickness and then decrease as the two layers get so far apart that they decouple. But a sudden shift in damping does not seem reasonable.
This is where I read it - http://maxwellsci.com/print/rjaset/v4-3130-3136.pdf
Have a look at Case Study 2.
What you say seems to be the case but the shift is very sudden.
Any comment on whether base and constraining layer thickness is important, or only stiffness with regards to damping layer thickness?
Have a look at Case Study 2.
What you say seems to be the case but the shift is very sudden.
Any comment on whether base and constraining layer thickness is important, or only stiffness with regards to damping layer thickness?
Last edited: