This week, acousticians at La Salle (Universitat Ramon Llull, Barcelona) described a device that sounds like science fiction and looks like a toy: a 15-centimeter rod with 29 rings of varying diameters, 3D-printed. You stick it into an ordinary ventilation duct — and the echo in the duct disappears. “As if the duct were infinitely long, with no far end for the sound to bounce back from,” explains engineer Oriol Guasch. The authors call the whole device a sonic black hole.
We have not, of course, gotten from ventilation into space. But the most interesting part is the route by which the idea traveled from astrophysics to air ducts. Over 40 years, the same mathematics of event horizons was needed twice by people with entirely different goals: some were looking for an analogue of Hawking radiation in the lab, others for a way to make a conference hall stop humming. In the end, it turned out to be closer to the latter.
Let’s start with the least appealing part — the mundane one. Sound in a duct behaves like water in a gutter: it travels while there is somewhere to go. It hits a cap, a bend, a sudden widening — and reflects back. In an enclosed volume, this is called reverberation; in ductwork, rattling, hum, and “why can I hear the bedroom from the kitchen?” An organ pipe works on precisely this principle: a closed end — a reflection node, a standing wave, a note. Music is successfully controlled echo; ventilation comes with no such promise.
The classic remedy is a silencer. Two basic types: reactive (use a labyrinth to turn the sound wave back into local turbulence) and absorptive (dissipative): a mineral-wool panel set into a break in the duct, which eats energy through friction. Both are, essentially, obstacles inserted into the flow: they take up space, cause pressure loss, their fibers settle and burn out over time, and low frequencies — the most obnoxious kind in residential buildings — are inherently hard to damp. The conference hall’s ventilation was installed in 1987, and nobody is going to let you tear it out.
Deng and Guasch, with colleagues, proposed another move: don’t damp the energy. Give the wave a path that physically does not leave it enough time to reach the end.
The idea is called a retarding structure. If the thickness of a waveguide wall decreases according to a power law, the speed of the wave inside it drops with the thickness, its wavelength compresses, its amplitude grows — and in the limit, when the thickness falls to zero, the wave stops moving altogether: it never reaches the tip and never comes back. The reflection coefficient is zero. That is an engineering black hole.
One caveat from the opening paragraphs of ScienceAlert is worth making here: the Zeno paradox they invoke is a false analogy. With Zeno, halving the distance is a counting trick; here the wave genuinely slows down, because physically it has nowhere to go — the speed in the medium has fallen faster than it can propagate. More precisely, the point beyond which the wave becomes stationary behaves for an observer like an event horizon: “the wave seems to require infinite time to reach the limiting point,” in Guasch’s words.
The paradox lies elsewhere. An ideal hole with a wall tapering to zero is mathematics, not a component. Nothing can be thinner than zero, and any residual thickness, however small, means the speed never quite reaches zero: the wave will eventually touch the end, and some of its energy will return. And at the tip, if the energy is not dissipated, a singularity rushes in — the amplitude goes to infinity. Hence the standard engineering compromise: residual thickness plus a viscoelastic layer in the zone where the wave has almost stopped. Friction for the slow one.
The standard sonic black hole in a duct has been around for some 20 years: concentric rings separated by cavities, built into the duct wall, with the rings’ inner radius following a power law. The duct is printed as a single piece along with the rings. This creates two problems, both addressed by the Barcelona work. First: retrofit. The duct in the building is already there; you cannot build a wall into it, and redoing it is expensive — in copper and steel systems, that means casting or milling. Second: frequency range. In the standard design, at certain frequencies it is not the black-hole effect itself that dominates, but the resonances of the cavities between the rings — what wave physics calls rainbow trapping: each frequency comes to a halt at its own cross-section of the structure. It is a beautiful effect — in optics, where it was first demonstrated in the late 2000s, it meant a rainbow getting stuck in chirped waveguides — but a narrowband one.
One of the authors of the current paper admits in a footnote that the only plug-in sonic black hole existing at the time of their work had been made 20 years earlier — by Mironov and Pislyakov. That is the bridge into the history, and it does not begin in Barcelona.
The basic principle of the retarding structure was described in a 1988 paper by M. A. Mironov in Sov. Phys. Acoust.: a wedge whose thickness smoothly tapers to zero does not reflect a flexural wave. Roughly a decade later, V. V. Krylov extended the idea into two dimensions — circular power-law-profile indentations in plates, an omnidirectional absorber — and made the very analogy with an optical black hole from which the name “acoustic black hole” stuck. In 2002, Mironov and Pislyakov showed that the same thing works in an acoustic duct with axial variation in wall impedance. So the “black hole in ventilation” is a quarter-century-old idea, grown from the Soviet-Russian school of acoustics; the Barcelonans did not invent it, but literally turned it inside out: rings on a central rod, like a Christmas tree — and the result was a module you can stick into any existing duct, whatever it is made of.
Why the inversion works became clear from the authors’ own numerical models. In the standard hole, absorption rests on rainbow trapping associated with cavity resonances — within a certain range, that is the main mechanism. In the plug-in Christmas tree, another effect dominates: axial Fabry–Pérot resonances along the rod — and it works across a much broader band. With the same free cross-sectional area, the plug-in hole reflects less than the built-in one. A three-stage modeling strategy — from a continuous waveguide to a discrete model with a finite number of rings — matched the impedance-tube results: a prototype with 29 rings, 15 centimeters long, printed on a 3D printer, tested with a loudspeaker and microphones. The echo has not been eliminated completely, but the simulations have been confirmed.
One side detail worth keeping in mind: the Barcelona Christmas tree is not a lone sortie, but one link in a long series. In the Journal of Sound and Vibration, the same group examined how to get rid of rainbow trapping altogether — by splitting the hole into small units — and proposed two- and three-dimensional versions with domes; another paper covers a piezoelectric version that uses the slowdown in reverse: pressure near the end of the structure intensifies, and a piezoelectric plate draws energy from it (8 microwatts, conversion close to 100% at optimal load). The same geometry works as both absorber and source.
While engineers ran sound through rings, physicists spent 40 years running phonons through superfluid liquids with the same mathematics — and achieved a kind of success worth knowing alongside the Barcelona insert as a counterpoint.
In 1981, William Unruh published a short paper in Physical Review Letters with a telling question mark in its title: “Experimental Black-Hole Evaporation?” The point was this: wave equations in a moving fluid take the same form as field equations in curved spacetime. If the fluid flows faster than the speed of sound, phonons cannot cross the boundary where the regime changes — the horizon. And if there is a horizon, then, according to Hawking, there should also be evaporation: spontaneous creation of pairs right at the horizon from vacuum fluctuations. There is nowhere to test this with a real black hole; with a sonic one, there is.
It took nearly 30 years. In 2009, at the Technion, a rubidium Bose–Einstein condensate was used to create the first laboratory analogue: a supersonic flow, a measured horizon, measured surface gravity and temperature. In 2014, Jeff Steinhauer observed self-amplifying Hawking radiation in an “analogue black-hole laser”; in 2016, quantum radiation together with entanglement; in 2019, its temperature (a paper in Nature); and in 2021, stationary spontaneous radiation. Superradiance — extracting energy from a rotating horizon, in other words a laboratory test of the process by which black holes ought to have energy stolen from their rotation — was tested on rotating analogues: the first laboratory verification dates to 2010, and the effect in a vortex water flow was demonstrated in a paper by Torres et al. in Nature Physics (2017).
Why should any of us care? The formal answer is the black-hole information problem and attempts to reconcile general relativity with quantum mechanics. The indirect one: the fact that horizons have been reproduced in condensates, water, ion rings, slow light is cited in support of the idea of emergent gravity — that spacetime itself is not fundamental, but assembled from something smaller, the way sound is assembled from molecular vibrations.
And this is where the two lines converge. The Barcelona Christmas tree is emphatically NOT a physical analogue: ScienceAlert stresses this — the device exploits a property of a black hole rather than reproducing gravity. But the move itself — “make a waveguide in which the speed falls to zero, and you get a perfect absorber” — is exactly the same one Unruh came up with for physics, and Mironov described in 1988, apparently without knowing of Unruh. Two communities, 40 years of separate development, the same patch of equations. The engineering hole turned out not to be a side illustration of the physical one, but its full-fledged twin sister, simply with a different client.
For the Christmas tree to make it into a real air duct, it has to get through a list of routine troubles that Guasch himself names: particles clogging the cavities; temperature; pressure; airflow. In ventilation, all of these are present at once and in motion — unlike in a laboratory impedance tube, where a microphone sits in the pipe and the air does not flow. An ideal hole tapering to zero is mathematics; a practical hole lives with residual thickness and friction that must be tuned.
The longer-term plans are two- and three-dimensional structures and room acoustics: if an absorber works in the cross-section of a duct, why not turn the same logic into a module that kills reflections in a room — where today you need diffusers, absorptive panels, and carefully calculated angles. Given how much low-frequency hum there is in modern buildings with open ceilings and glass facades, that would be an unexpectedly broad market for a rod with rings.
Hawking radiation in a rubidium condensate and echo in a ventilation shaft are separated by 40 years and the entire substance of physics departments. Between them lies the same formula: if the speed of a wave falls to zero, the wave does not return. Whether Mironov in 1988 or Unruh in 1981 was the first to describe that patch of equations is irrelevant to the engineering result; what matters is that the elegance proved reusable in both directions. Physicists spent a generation making sure that an analogue black hole evaporates. Engineers learned something else over those same decades — that real black holes stand in every building, only they are called ventilation. The answers to questions about the structure of the universe and about a humming fan turned out to be in the same toolbox, and someone in Barcelona simply remembered it and printed the necessary part in one evening.