Here is some work out of NC State. It's a new way to make a Schroeder diffusor that is 1/10th the thickness of the classic diffusor. That would be amazing if it works, and it looks easy to build.
Phys. Rev. X 7, 021034 (2017) - Ultrathin Acoustic Metasurface-Based Schroeder Diffuser
in PDF: https://journals.aps.org/prx/pdf/10.1103/PhysRevX.7.021034
Supplemental: https://journals.aps.org/prx/supple....7.021034/Supporting_Information2zyf_js-1.pdf
Basically it's a flat sheet containing a a series of small chambers, similar to a honeycomb but with square chambers. Each chamber has a hole in its face. The size of the hole determines the working frequency. You could imagine a sheet of pegboard where there is a chamber behind each hole - only the holes are of different sizes.
What I can't figure out from the papers is; are the chambers all the same size? I think so. Also, how to determine the size of the chambers and the size of the holes. The depth isn't difficult. The pattern of holes is well laid out in the docs.
Looks like a sheet about 50mm or 2" thick would work down into the 340 Hz range. Any thoughts?
Phys. Rev. X 7, 021034 (2017) - Ultrathin Acoustic Metasurface-Based Schroeder Diffuser
in PDF: https://journals.aps.org/prx/pdf/10.1103/PhysRevX.7.021034
Supplemental: https://journals.aps.org/prx/supple....7.021034/Supporting_Information2zyf_js-1.pdf
Basically it's a flat sheet containing a a series of small chambers, similar to a honeycomb but with square chambers. Each chamber has a hole in its face. The size of the hole determines the working frequency. You could imagine a sheet of pegboard where there is a chamber behind each hole - only the holes are of different sizes.
What I can't figure out from the papers is; are the chambers all the same size? I think so. Also, how to determine the size of the chambers and the size of the holes. The depth isn't difficult. The pattern of holes is well laid out in the docs.
Looks like a sheet about 50mm or 2" thick would work down into the 340 Hz range. Any thoughts?
Uh, not NC State, but Nanjing University.
You need to look up Helmholtz resonators.
Helmholtz Resonance
The panel they are using uses the coupling and phase shift of them to get the effect. It is all about the size of the holes to vary the frequencies. That way the chambers can all be the same size.
You need to look up Helmholtz resonators.
Helmholtz Resonance
The panel they are using uses the coupling and phase shift of them to get the effect. It is all about the size of the holes to vary the frequencies. That way the chambers can all be the same size.
One of the guys is at NC State, the others are at Nanjing. I'll be working at Duke next month, so may be able to talk to the guy at State. Not sure how principle he is in the research. I emailed him.
Yes on the phase shift. The chambers look like Hermholtz resonators, but are not, as they say the holes are too large to function as the resonator. They mention that in the paper.
Yes on the phase shift. The chambers look like Hermholtz resonators, but are not, as they say the holes are too large to function as the resonator. They mention that in the paper.
This would be certainly interesting for anyone with access to a CNC router as a 3-sheet operation. 1/4" backing plate, something like 3/4" plywood sheet with the pockets cut out, then, a front face of 1/4" ply with the holes bored out. Wouldn't want to do it by hand, to say the least.
It's a fixed cavity depth of lambda/20. Promise it's a Helmholtz resonator, although maybe not the "common" definition, but Helmholtz equation is for waves of all shapes and sizes.
(edit: just read through and the article says exactly that)
lambda/20 means .678" cavity at 1000 Hz! With a pocket width requirement of lambda/2, it's gonna need some area to work, however.
It's a fixed cavity depth of lambda/20. Promise it's a Helmholtz resonator, although maybe not the "common" definition, but Helmholtz equation is for waves of all shapes and sizes.
(edit: just read through and the article says exactly that)lambda/20 means .678" cavity at 1000 Hz! With a pocket width requirement of lambda/2, it's gonna need some area to work, however.
Last edited:
Hmmm, I was thinking of this passage, which I might have misinterpreted.
"Although the unit cell is Helmholtz resonator (HR)-like, its cavity width and neck width are much larger than those of the classical HRs with respect to λ0. Consequently, the well-established analytical theory for classical HRs (e.g., the lumped model) is not valid anymore and must be revisited."
"Although the unit cell is Helmholtz resonator (HR)-like, its cavity width and neck width are much larger than those of the classical HRs with respect to λ0. Consequently, the well-established analytical theory for classical HRs (e.g., the lumped model) is not valid anymore and must be revisited."
All good Pano! When I think of Helmholtz, I go back to the original equation, rather than many of the shorthand notations. That's all I think that very quote is discussing, namely these don't behave like standard/common resonators due to their size.
End of the day, that's semantics versus being able to make an effective diffuser. Part of me wonders if you couldn't make these out of acrylic sheet and double them up as ceiling lights. Might be too many $$'s though for the raw material.
End of the day, that's semantics versus being able to make an effective diffuser. Part of me wonders if you couldn't make these out of acrylic sheet and double them up as ceiling lights. Might be too many $$'s though for the raw material.
LOL, not a bad idea! I can see it.
I was thinking of a grid sandwiched between to thin sheets of plywood or Masonite. Drill holes of calculated sizes. Not difficult, just labor intensive. I did hear back from one of the authors, he's putting me in touch with one of the other authors in Nanjing to help me with the dimensions. He says they are in the paper, but I don't see them.
I was thinking of a grid sandwiched between to thin sheets of plywood or Masonite. Drill holes of calculated sizes. Not difficult, just labor intensive. I did hear back from one of the authors, he's putting me in touch with one of the other authors in Nanjing to help me with the dimensions. He says they are in the paper, but I don't see them.
Looks similar in principle to the one discussed in Magic metamaterial diffusers -- almost!
I looked at doing that one on a 3D printer, but it didn't seem very practical and looked like it would have some structural resonance complications. This one might be more feasible (though thicker for a given low frequency limit), but better done with thin material, a table saw, and a drill. Assuming that the holes could be round rather than square, that is.
I looked at doing that one on a 3D printer, but it didn't seem very practical and looked like it would have some structural resonance complications. This one might be more feasible (though thicker for a given low frequency limit), but better done with thin material, a table saw, and a drill. Assuming that the holes could be round rather than square, that is.
A question I have about these kind of diffusors is how well a design 'scales'. For a Schroeder or a step diffusor, only linear dimensions (depths, widths) are involved, so you can scale frequency range by scaling all x, y and z dimensions together. Ignoring nonlinearities of air, surface losses, etc.
But these metamaterial things seem to depend also on the volumes inside the resonators, and of course if you scale x and y and z, the volume scales as x*y*z. What happens with the operating frequency if you scale? And does the structure stay optimized?
But these metamaterial things seem to depend also on the volumes inside the resonators, and of course if you scale x and y and z, the volume scales as x*y*z. What happens with the operating frequency if you scale? And does the structure stay optimized?
I believe the easiest type diffusers to build involve vertical slats only.
There are plenty of designs building them this way. I recall a five or 6 inch deep diffuser of this type working down to +- 1000Hz.
So the one above would be groundbreaking…
Also common is the typical wood 1-1/2” x 1-1/2” ballister diffuser. This diffuser both vertically and horizontally, but is more difficult to build.
There are plenty of designs building them this way. I recall a five or 6 inch deep diffuser of this type working down to +- 1000Hz.
So the one above would be groundbreaking…
Also common is the typical wood 1-1/2” x 1-1/2” ballister diffuser. This diffuser both vertically and horizontally, but is more difficult to build.
Good points, Bill. In the supplement they say:
"The results suggest that the MSD can be scaled up/down to an arbitrary wavelength/frequency with thickness of λ0/20"
That was for a simulated depth of 5cm
Your remark that the volume of the well is x*y*z is important. I don't know what the relation to well volume and phase shift scattering would be.
"The results suggest that the MSD can be scaled up/down to an arbitrary wavelength/frequency with thickness of λ0/20"
That was for a simulated depth of 5cm
Your remark that the volume of the well is x*y*z is important. I don't know what the relation to well volume and phase shift scattering would be.
Somehow I am not seeing them. Must be a blind spot. Can you point them out to me? Thanks.
That would be amazing. Wonder if it would be rigid enough? There a mention of wall thickness, let me see if I can find it. And would the holes have to be square?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.