One path to acoustic cloaking

A sound wave bends around an obstacle so cleanly that, on the other side, the water is undisturbed — no echo, no shadow, no scattering — as if the obstacle were never there. That image sounds like science fiction. Cummer and Schurig showed that the physics allows it, and the path runs through a mathematical move borrowed wholesale from optics. To understand where acoustic cloaking comes from, you have to start with light. In 2006, Pendry and colleagues discovered something remarkable about Maxwell's equations — the equations governing how electromagnetic fields behave. Those equations are form invariant under coordinate transformations. In plain language, that means if you mathematically remap space — stretch it, compress it, or bend it — the equations describing light keep their exact mathematical shape. The material parameters, permittivity and permeability, absorb the transformation. The equations themselves don't change. That property turns out to be enormously powerful. If you design a coordinate transformation that compresses a finite region of space into a shell — pushing all the physics into a ring around an empty core — you can then reinterpret that transformation as a prescription for material properties in ordinary flat space. If you engineer a material whose permittivity and permeability vary in just the right way through that shell, light propagates as if space had been bent around the core. As Cummer and Schurig put it, this approach yields a general method for rendering arbitrarily sized and shaped objects electromagnetically invisible, with theoretically perfect performance regardless of wavelength, object size, or shape. The object inside the shell is invisible. Not approximately invisible. Exactly invisible, in principle. The obvious question followed immediately: can you do the same thing for sound? Milton and colleagues answered that question, and their answer was largely no. For a general elastic medium, the equations of motion are not form invariant under coordinate transformations. A coordinate remap in a generic elastic solid generates modified equations that involve additional structure — including an anisotropic mass described by a second-rank tensor — and those modified equations are not simply the original elastodynamic equations with different parameter values. Milton and colleagues discussed conceptually how one might engineer artificial elastic media to satisfy those modified equations, but no practical scheme for a transformation-based acoustic cloak emerged from the analysis. The wall was real. The word that matters, though, is "general." Cummer and Schurig spotted the loophole. For a two-dimensional, z-invariant geometry — a fluid with no shear modulus — the linearized acoustic equations become mathematically identical in form to a single polarization of Maxwell's equations. Not analogous. Identical, via a precise variable exchange that also preserves boundary conditions. Pressure maps to one electromagnetic field component, and the two fluid velocity components map to the others. Bulk modulus maps to one material parameter, and density maps to another. Crucially, the density is allowed to be anisotropic — different in different directions — and that anisotropy is what carries the weight of the transformation. The boundary conditions match too. Acoustic pressure and normal velocity at an interface correspond exactly to the tangential electric and magnetic field components at an electromagnetic interface. Even the power flow lines up: the Poynting vector in electromagnetics — the cross product of the electric and magnetic fields — equals the acoustic power flux, which is pressure times velocity. This is not a vague structural resemblance. It's a term-by-term correspondence, and it's exact. To confirm the theory actually works, Cummer and Schurig ran time-harmonic numerical simulations. Picture a rigid cylinder sitting at the center of a flat two-dimensional plane — one point five wavelengths in diameter. Surrounding it is a concentric cloaking shell, three-quarters of a wavelength thick. A Gaussian acoustic beam is launched toward this object. The computational domain is bounded by perfectly matched layers to absorb outgoing waves. The excitation frequency, bulk modulus, and background fluid density are all normalized to unity, and the simulation grid uses twenty points per wavelength. Inside the cloaking shell, both the bulk modulus and the mass density vary with position. The bulk modulus is inhomogeneous throughout the shell. The density is both inhomogeneous and anisotropic — in cylindrical coordinates, the density tensor is diagonal, with different values in the radial and azimuthal directions. When mapped onto the Cartesian grid for computation, those off-diagonal terms appear explicitly, and the simulation includes them. The results come in three panels. Without any scatterer, a clean Gaussian beam propagates undisturbed. With the rigid cylinder present but uncloaked, the beam scatters strongly in all directions and casts a pronounced shadow downstream. With the cloaking shell in place, the wave is smoothly steered around the cylinder and rejoins on the far side, restoring wavefronts that closely match the empty-space baseline. Scattering is dramatically reduced in all directions. The shadow is gone. Acoustically, the object is effectively invisible. Some residual scattering is visible, and Cummer and Schurig are straightforward about why: numerical artifacts from approximating circular geometry on a Cartesian grid, and the relatively coarse twenty-point-per-wavelength resolution. These are computational limitations, not failures of the underlying physics. The theory is exact. Why does this only work in two dimensions? The mismatch in higher dimensions is not a minor algebraic detail. It's a genuine structural problem. Electromagnetic waves in three dimensions carry polarization — the Maxwell fields are vectorial, with two independent polarizations. A fluid acoustic field is described by a scalar pressure and a velocity vector. As Cummer and Schurig put it directly, we need three field variable components for correspondence with each electromagnetic polarization, and a fluid provides only four total with the velocity vector and the scalar pressure. The field-component counts simply don't align in three dimensions. You can't map the full three-dimensional Maxwell system onto a fluid acoustic system and preserve form invariance under coordinate transformations. In two dimensions, one polarization of the electromagnetic field behaves like a scalar, and the exact isomorphism appears. Cummer and Schurig show the mapping explicitly for transverse electric polarization, provided the coordinate transformation doesn't mix the in-plane directions with the out-of-plane one. Satisfy that constraint, and the two-dimensional acoustic problem and the two-dimensional electromagnetic problem are the same problem. That's the entire foundation of the cloak. The constraint is real, and it's why a perfect general three-dimensional acoustic cloak is ruled out by this analysis, and by Milton and colleagues' results. This isn't a temporary engineering limitation waiting on better materials. It's a statement about the mathematical structure of the equations themselves. On experimental feasibility, Cummer and Schurig are candid. Engineering a shell with continuously varying bulk modulus and spatially anisotropic mass density elements may be a challenge, but it is experimentally feasible in principle. They point to two conceptual routes. Milton and colleagues described a construction using spring-loaded masses that would produce anisotropic effective density — the acoustic analogue of self-resonant electromagnetic metamaterial elements, though of limited bandwidth. A reduced cloak is also possible, directly analogous to the reduced electromagnetic cloak demonstrated by Schurig and colleagues. In that version, the wavenumber through the shell is correct but the impedance is not, making the cloak ideal only in the short-wavelength limit. The payoff is flexibility — you can shift the required anisotropy from mass density to bulk modulus, or eliminate the inhomogeneity of one material component entirely. Cummer and Schurig also point to experimental progress already underway in acoustic metamaterials: theoretical work showing a fluid loaded with particles can realize almost any isotropic effective mass and bulk modulus, and experiments demonstrating that a one-dimensional acoustic waveguide loaded with resonators can yield an effective bulk modulus of essentially any value, including negative values. The deeper finding here, though, is the equivalence itself. The acoustic cloak is its most dramatic demonstration, but the mapping between two-dimensional fluid acoustics — with anisotropy — and a single polarization of Maxwell's equations opens a broader path. Wherever transformation optics has produced a useful electromagnetic device, a two-dimensional acoustic analogue becomes conceivable: lensing, waveguiding, and field concentrators. Cummer and Schurig found the door. The cloak is just what's immediately on the other side of it. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.