Making the World a Lot Quieter

Mechanical engineers have developed an “acoustic metamaterial” that can cancel 94 percent of sound

Boston University Ph. D. candidate Reza Ghaffarivardavagh poses with his research team. Reza holds the noise cancellation devices his team developed over his ears while his teammates make noise banging a pot, playing a trumpet, and a boombox in the background
Boston University mechanical engineers have created synthetic, sound-silencing structures—acoustic metamaterials—that can block 94% of sounds. Reza Ghaffarivardavagh (ENG) (front center) holds two of the open, ringlike structures over his ears while Stephan Anderson (MED) (left), Xin Zhang (ENG) (rear center), and Jacob Nikolajczyk (ENG) (right) make a racket. Photo by Cydney Scott

What sounds would you mute if you could? A pair of Boston University mechanical engineers are asking that question, with the ever-increasing din of drone propellers, airplane turbines, MRI machines, and urban noise pollution blaring in the mind’s ear.

“Today’s sound barriers are literally thick heavy walls,” says Reza Ghaffarivardavagh. Although noise-mitigating barricades, called sound baffles, can help drown out the whoosh of rush hour traffic or contain the symphony of music within concert hall walls, they are a clunky approach not well suited to situations where airflow is also critical. Imagine barricading a jet engine’s exhaust vent—the plane would never leave the ground. Instead, workers on the tarmac wear earplugs to protect their hearing from the deafening roar.

Xin Zhang and Ghaffarivardavagh were enticed by an alluring question: “Can we design a structure that can block noise but preserve air passage?”

Leaning on their mathematical prowess and the technology of 3D printing, it turns out they can. In a January 2019 Physical Review paper, the researchers argue that it’s quite possible to silence noise using an open, ringlike structure, created to mathematically perfect specifications, for cutting out sounds while maintaining airflow.

Numbers and noise control

“I’ve always been interested in acoustics,” says Ghaffarivardavagh, who finished his BSc in mechanical engineering at Sharif University of Technology in Tehran, Iran, before coming to Boston University for graduate school. Now, in Zhang’s lab, he’s close to finishing his PhD. “I like to work on something that I can hear or see the result. Something that I can have an impact on with issues we are facing nowadays.”

Having lived in other major cities before coming to Boston, Zhang and Ghaffarivardavagh have always marveled at the layered urban soundscape enveloping them. In Boston, the cacophony of the city is garbled together from airplanes flying overhead, the engines and horns of cars, trucks, and buses on the street, the rumble and screech of MBTA trolleys, and the hum of building appliances and power sources.

City life is so noisy, you have to find a way to create quiet moments, they say.

That got them dreaming up a sound baffle that wasn’t a barrier at all, but instead an open conduit. Such a feat could only be possible by developing a material with unusual and unnatural properties (known as a metamaterial), in this case with the ability to exert an isolated influence on sounds—an acoustic metamaterial.

“I’ve been working on metamaterials for more than a decade,” says Zhang, a multidisciplinary professor at the College of Engineering and the Photonics Center. “But it was Reza that gradually got me more excited about the fundamental idea of a marriage between acoustics and metamaterials. If you ask me and my colleagues, acoustic metamaterials is a relatively young direction…. It’s the future.”

Mute button, incarnated

Ghaffarivardavagh and Zhang let mathematics—a shared passion that has buoyed both of their engineering careers and made them well-suited research partners—guide them toward a workable design for what the acoustic metamaterial would look like.

“Sound is made by very tiny disturbances in the air. So, our goal is to silence those tiny vibrations,” Ghaffarivardavagh and Zhang say. “If we want the inside of a structure to be open air, then we have to keep in mind that this will be the pathway through which sound travels.”

They calculated the dimensions and specifications that the metamaterial would need to have in order to interfere with the transmitted sound waves, preventing sound—but not air—from being radiated through the open structure. The basic premise is that the metamaterial needs to be shaped in such a way that it sends incoming sounds back to where they came from, they say.

On one end of the pipe, a loudspeaker blasts a noise. On the other end, the open acoustic silencing metamaterial redirects the sound as long as it's fitted in place. Video courtesy of the Zhang lab

As a test case, they decided to create a structure that could silence sound from a loudspeaker. Based on their calculations, they modeled the physical dimensions that would most effectively silence noises. Bringing those models to life, they used 3D printing to materialize an open, noise-canceling structure made of plastic.

Trying it out in the lab, the researchers sealed the loudspeaker into one end of a PVC pipe. On the other end, the tailor-made acoustic metamaterial was fastened into the opening. With the hit of the play button, the experimental loudspeaker set-up came oh-so-quietly to life in the lab. Standing in the room, based on your sense of hearing alone, you’d never know that the loudspeaker was blasting an irritatingly high-pitched note. If, however, you peered into the PVC pipe, you would see the loudspeaker’s subwoofers thrumming away.

The metamaterial, ringing around the internal perimeter of the pipe’s mouth, worked like a mute button incarnate until the moment when Ghaffarivardavagh reached down and pulled it free. The lab suddenly echoed with the screeching of the loudspeaker’s tune.

“The moment we first placed and removed the silencer…was literally night and day,” says Jacob Nikolajczyk, who in addition to being a study coauthor and former undergraduate researcher in Zhang’s lab is a passionate vocal performer. “We had been seeing these sorts of results in our computer modeling for months—but it is one thing to see modeled sound pressure levels on a computer, and another to hear its impact yourself.”

By comparing sound levels with and without the metamaterial fastened in place, the team found that they could silence nearly all—94 percent to be exact—of the noise, making the sounds emanating from the loudspeaker imperceptible to the human ear.

A quieter world

Now that their prototype has proved so effective, the researchers have some big ideas about how their acoustic-silencing metamaterial could go to work making the real world quieter.

“Drones are a very hot topic,” Zhang says. Companies like Amazon are interested in using drones to deliver goods, she says, and “people are complaining about the potential noise.”

“The culprit is the upward-moving fan motion,” Ghaffarivardavagh says. “If we can put sound-silencing open structures beneath the drone fans, we can cancel out the sound radiating toward the ground.”

Detail photo of a ring shaped noise cancellation device built using an acoustic metamaterial both developed by Boston University engineers.
The mathematically designed, 3D-printed acoustic metamaterial is shaped in such a way that it sends incoming sounds back to where they came from, Ghaffarivardavagh and Zhang say. Inside the outer ring, a helical pattern interferes with sounds, blocking them from transmitting through the open center while preserving air’s ability to flow through. Photo by Cydney Scott

Closer to home—or the office—fans and HVAC systems could benefit from acoustic metamaterials that render them silent yet still enable hot or cold air to be circulated unencumbered throughout a building.

Ghaffarivardavagh and Zhang also point to the unsightliness of the sound barriers used today to reduce noise pollution from traffic and see room for an aesthetic upgrade. “Our structure is super lightweight, open, and beautiful. Each piece could be used as a tile or brick to scale up and build a sound-canceling, permeable wall,” they say.

The shape of acoustic-silencing metamaterials, based on their method, is also completely customizable, Ghaffarivardavagh says. The outer part doesn’t need to be a round ring shape in order to function.

“We can design the outer shape as a cube or hexagon, anything really,” he says. “When we want to create a wall, we will go to a hexagonal shape” that can fit together like an open-air honeycomb structure.

Such walls could help contain many types of noises. Even those from the intense vibrations of an MRI machine, Zhang says.

According to Stephan Anderson, a professor of radiology at BU School of Medicine and a coauthor of the study, the acoustic metamaterial could potentially be scaled “to fit inside the central bore of an MRI machine,” shielding patients from the sound during the imaging process.

Zhang says the possibilities are endless, since the noise mitigation method can be customized to suit nearly any environment: “The idea is that we can now mathematically design an object that can block the sounds of anything,” she says.

This work was supported by a Boston University Materials Innovation Grant and Dean’s Catalyst Award.



    1. Yes, or ban them all together. The low frequency noise penetrates windows and walls and is above EPA, NIOSH, OSHA and WHO’s recommended safe levels. The noise is an assault to every living creature.

    2. I am the Director of Quality for a metal stamping company. I have been researching ways to reduce the noise in the facility from the presses stamping out parts all day. I would love to challenge the team to visit my facility and apply this development into a practical application in an industrial situation. Thank you!

  1. I would appreciate a comment from one of the researchers as to whether or not their discovery might find a future application in significantly reducing infrasound.
    In particular the powerful 10 Hz generated by diesel engine locomotives at great distances. I am thinking of protecting the sleep & health inside homes where this infrasound penetrates under any condition inside or outside (even if the building is surrounded by other structures & is not in direct ligne of sight).
    Not to speak of the much less powerful harmonics which rarely annoying or health affecting.
    Thanks in advance,
    J. B.

    1. As far as deep bass goes, nothing can hinder it – since it’s not based on sound waves but sound pressure – so it passes though objects, even these sound dampeners. The most effective blocking of bass is high mass material, like cork.

    2. I would imagine the solution is similar to much of audio recording: “get it at the source”. In other words, not solving the problem by building homes that can block the sound, but by reducing/eliminating the sound coming out of the engine in the first place, and perhaps this research can help there!

    3. Hi Joseph,
      The silencer design may be tuned to target different frequency ranges. However, in very low frequencies (infrasound), both the design of the structural shape and wave penetration into the structural material would be very challenging and problematic even in the context of this particular metamaterial design. However, infrasound silencing may be approached in a different manner where there is no requirement for the openness and preservation of ventilation, which were critical facets of this work. Many thanks!
      Best Regards,

      1. hi Xin, to follow up on what alot of people are curious about, what have you found to be a reasonable expectation of size vs low frequency reflection?
        for example, the specific prototypes in this picture, what frequency band are they tuned to reflect, and how large would something similar be for the lower frequency bands? is there any publicly available published data on this yet? good work, this is fantastic.

  2. Upon review, it is more like an aphysical review letter, as it’s not like those four got it out where nature could abhor it or not, unfrustrate urban bees, put it on all the campus service putt-abouts, performing arts building tunings, etc. as opposed to ASTM E 2611-09, which I choose to believe is a dronecore Boston band (…have citations for AIP always looked like that?) that stays under 600Hz in order to get a reputation or something.

    Additionally, I look forward to spring renderings in, with or on moss, at the diesel outfitters in a 10Hz loopba…okay, that’s clearly the right size…, and possibly at the clanky workers and fast recovery facilities in the gyms. We need to have a cutout for Boston Cheers to pass…

  3. This is amazing and am sure there will be number applications for this. Jet engines, MRI machines, Flyovers and roads which are near residential complexes and so on.

  4. Please consider open-sourcing the 3d files for those of us who want to experiment. My primary interest is in ventilation for soundproofed tiny homes and I’d be happy to provide real world feedback of whatever I find.

    1. Increase transmission loss of sound path (e.g., double-pane laminated windows) or decrease sound power of the source (e.g., dog is better trained).

      Nobel prize please :D

  5. Congratulations on an amazing design that will literally change the world.

    I have a home recording studio that is air tight but for two- 4” ducts pumping fresh air into the room and sucking old air out.

    The problem is I can’t use powerful enough fans to get the needed amount of air into the room due to incredible noise levels.
    This is the perfect solution and those circular units would be all I need.
    I would love to get my hands on two of them if someone from the group would message me privately please with more info…
    Thank You in advance.


  6. Now the interesting question is, how would/could this work in a real world scenario? There’s a static sound, which is being driven through a pipe.

    I’d love to hear how the material behaves when it’s not a static, eg. 400Hz sound, but music, talking etc.

  7. Talk to San Diego Airport.. they have an archaic blast fence to decrease sounds at takeoff on the single runway. All San Diegans would LOVE you if you helped reduce the noise!
    Good Luck with the new will help in millions of places! :-)

    1. Dan I’m going out on a limb but I really think all airports would benefit from this invention. I know I most definitely would.

  8. I need a Tesla covered in this material. The road noise in my current Model S is so annoying. Please Elon, cover the car and make is silent inside!

  9. Hello,

    Great work!

    A fellow field audio engineer posted a link to your article, and like him I’m interested in the possibility of the metamaterials you’ve designed being used in sound engineering for film and video production and live sound; for instance with sound isolation for headsets and earplugs, and also applying these materials to designing better directional microphones. I wonder if there are some applications for this in speaker design as well?

    I have some questions: How is sound cancelling effected when sound is coming from off axis? How does it respond to different frequencies ? How does it respond to complex acoustic situations like music and non patterned wide spectrum sounds?

    Thank you!

    I’m a member of the Norcal mixers group ( in the San Francisco Bay area, and we have many audio engineers here who I know would be willing to test out your technology and give you valuable feedback from the field.


    Jack Morris

  10. I too am curious about this application for subbass frequencies.

    Also, I would like to see a demo of the contraption (muffler) used on that PVC pipe where the speaker plays a sine wave sweep from 1hz to 20khz. I can’t but help thinking the sine wave chosen for the demo was one which the attenuation device was optimized for. As we know, noise naturally is comprised of a fundamental tone and overtones. Lots of noise includes white noise too. We need to hear a demo of a wide variety of tones to better grasp how it can work for the real world. I can see how something simple like that could affect tuning frequency of a port (the PVC pipe) and the pressure nodes and anti-pressure nodes. It is more simple for one sine wave, especially one that is a midrange frequency as shown in the video.

    Yes, a leaf blower would be a good test since it has a sine sweep range between idle and wide-open throttle. :)

    1. I agree the test video is only one frequency. There are many Helmholtz resonator silencer designs that are tuned for certain frequencies. The plug also would likely have very high pressure loss for duct or pipe gas flow.

    1. I thought the same thing. 94% isn’t as impressive when talking about the logarithmic nature of sound perception.

      10*log10(0.06) ≈ -12.2 dB

  11. Is that 6 complete turns in the spiral, or is it fractional, introducing a phase change in the inner part of the pipe?

  12. This is amazing and congratulation to the team!

    I am currently looking for a solution to HVAC noise problem in my house, and I am wondering if there is a way to purchase one of these. I would be a superhero to my spouse and kid if could find something that can cut down on the mechanical noise which is drowning out our living room conversations (okay really our Netflix and bad LP’s, but still).

    I have looked into fiberglass duct insulation, but it will only have a nominal impact on reducing the noise and will be a pain to install. The situation seems perfect for what you have developed. I will love to be a real-world guinea pig if you are looking for one. Please contact me if it is possible or if you would like more details.

    1. Same here! We are located next to a draft fan for a boiler system and it’s awful. Sound reduction interior windows have helped some, but not entirely. This, however, would change our lives.

  13. This is great! I do have a question about the testing apparatus; is there a reason why the PVC pipe is so long? Would the results be the same if the pipe was half the diameter and length?

  14. No doubt…grt Job. Sound has power to vibrate glass.till it its(u r device) absorbing the sound wave… convert that to something like any other form of energy.

  15. Congratulations team ! – could you tell me: is this frequency specific, or does it cover a spectrum? Is the 3D model available to be printed? Lastly I secretly wonder how such a shape may be used with photons….hmm. :)

  16. Noise pollution can have such a damaging impact on wildlife, I hope it can be applied to sources of travel that affect such things

  17. The device is effective for a narrow band of frequencies centered around the design frequency. Impulse or broadband sound would be a challenge.

  18. Please work on noise canceling floors. The downstairs neighbors are complaining even though we are very quiet. Congratulations on your enterprise!

  19. Xin Zhang,

    Congratulations on the achievement!

    It’s hard to tell from the picture what your device looks like. I’d love if you could give some more insight into the structure/make-up of the device, and how it’s particularly designed structure helps to reduce sound effectively?

    I have guesses, but, you’re the expert. Enlighten us?

  20. Could this be employed on boat engines to quiet the human sounds affecting marine life? I am thinking of the whales who can no longer communicate over long distances.

  21. Promising work! In audio, dB is used. 94% claimed attenuation is -24dB. The demo video exhibits -13dB attenuation, which is quite an achievement.
    The demo original sound is a 500Hz sine. How does the system perform with real life annoying noise, ie with a larger spectrum including frequencies below 100Hz? How much is size an issue for lower freq attenuation? thanks.

  22. Im also wondering if the test could be done with White noise .. not a single tone …

    I’m wondering if the structure is designed to reflect a narrow band of frequency’s and its 3rd harmonic or a span until the functional wavelength is less than 1/2 the width of the inner diameter of the noise cancelling device?

  23. Congratulations, I find this is very fascinating, which causes me to wonder that if sound waves are traveling vibrations of particles in media such as air, do they have mass on some level?

  24. Please translate “94%” to “dB decrease”.

    Also, are these devices reflectors or absorbers? I would conclude from the article that they are reflectors, but not 100% certain. Not even 94% certain :)

    Thanks !

  25. This would be an absolute god-send to parents who live with adult children with autism or other developmental disabilities. My son scripts constantly and loudly, and the sound travels through our ducts throughout the house. Being able to dampen that noise would improve our quality of life and make it easier for us to keep our son at home.

  26. Hello, congratulations to the research team! Brilliant new development! Could I get the STL file to print a sample for tryout? I realize the structural geometry is a unique development, but is the material a special formulation, or is it a generic resin?

  27. Thank you everyone for all of your interest, we sincerely appreciate all of the questions and comments that have been posted. We are doing our best trying to keep up and answer your queries. Thanks again!

    Best Regards,

    Xin Zhang

    1. Would love to 3D print hexigon versions to interlock as a wall panel. I’d rear mount 1/8” 1lb/sqft mass loaded vinyl to handle the lower frequencies. Could offer before/after data using DIRAC room correction analysis (impulse+frequency response measure and correction)

  28. I am curious about the potential application of this concept to firearms suppressors (silencers). Assuming that nothing could be done about the crack caused by the supersonic expansion of gases, would the metamaterial be able to attenuate the rest of the report?

  29. The challenge in HVAC applications is balancing pressure drop with acoustic performance. Silencers have been the go to solution for 30 years which are costly and typically add about .25″wc PD for a good selection. If the metamaterial offers lees press PD and cost than silencers, you have a winner.

  30. Can you please clarify a few things?

    1) Does this silencer only work at a single frequency baked into the geometry? If so, is there any hope of extending on the principles to make a broad band silencer? Alternatively, could it be made to be deformed in a predicable way that would allow tuning it to a range of frequencies?

    2) Does the attenuation depend on the incoming sound wave being a plane wave? Another way of asking this, would it work if it was installed as the outlet to a plenum?

  31. Hello,Ms Zhang, did u test the gyroid filling structure for the noise decrease device? With Cura 3.6, can select gyroid filling, i prefer with 5% fill density. The gyroid is super light and can reflect most noise movement face.

  32. From what I can gather this is a tuned, narrowband, passive attenuation device and so would offer very limited suppression of broadband noise such as a jet engine or a lawnmower. Likewise for an impulse such as a dog bark. It would be nice to see a curve of attenuation vs, frequency.

  33. At first blush, this structure appears to redirect a fraction of the impinging sound through a more lengthy channel, such that it emerges from the channel in anti-phase with the original, unmodified, un-delayed sound, in such a proportion as to most closely result in mutual cancellation. As such, it will only work in a narrow range around a specific frequency, and odd multiples of that frequency. So it would be useful for reducing the sound of a constant-frequency source emanating from a restricted aperture, such as the sound of a furnace blower through a heating duct. But it will not significantly reduce sound of arbitrary or variable frequencies, such as human voice sounds, music, variable engine frequencies, or other common disturbances.

  34. Please help people who are living with hyperacusis. Auditory hyperacusis causes every sound to be magnified. People who have this condition need to wear 30dB or 35dB earplugs or shooting range headphones just to leave their houses. Forty percent of people with Lyme disease experience some, or suffer from chronic auditory hyperacusis. And now there are way too many people with Lyme disease (& additional tick-borne illnesses).

  35. Is it mandatory that the center space be unobstructed , or could a screen or mesh be placed over the center area to slightly limit the air flow but still produce the desired results of sound reduction?

  36. First of all, thank you so much for your valuable comments and suggestions. Much appreciated!

    Herein, we have summarized several of the points that have received the most questions and we sincerely encourage you to read the full paper at the link below for further details:

    1. Acoustic metamaterials refer to structures in which the effective acoustic properties are dominated by their geometrical shape/pattern rather than their composite materials.

    2. The main focus of our study is to present a silencer design methodology that features a dominant open area to preserve efficient ventilation. Indeed, it is demonstrated in this paper that an ultra-open silencer (+90% openness) may be designed within the presented framework.

    3. Acoustic silencers are commonly in-duct based bulky structures which result in a considerable pressure drop in the case of forced ventilation. Our silencer features a large central opening to ensure efficient ventilation and its overall dimension is deep subwavelength. Moreover, it may be leveraged as a stand-alone structure such as the housing of a fan or propeller.

    4. The silencer performance is based on the Fano-like interference which arises from the destructive interference between the continuum and resonating states. Consequently, the silencing performance is frequency selective and occurs when an acoustic wave in the helical channels of the periphery reaches its resonant state. Furthermore, silencing emerges for a targeted frequency range as well as its higher harmonics (both odd and even) which may make this design particularly suitable for tonal noise suppression. For variable speed fans and propellers, multiband and tunable versions of this methodology may be adopted.

    5. The 94% reduction in energy is approximately equivalent to the 13dB loss which is strictly limited by the fabrication quality in our experiment. More advanced fabrication methods can significantly improve dB loss at the targeted frequency.

    Thanks again!

    Xin Zhang
    Boston University
    Tel.: 617-358-2702

Post Your Comment