Numerical simulations of acoustic cloaking in a real laboratory that deploys immersive boundary conditions

Abstract

Immersive boundary conditions (IBCs) act as a wavefield injection method that couples a physical wave propagation experiment to an arbitrary virtual domain. They allow novel, real-time applications for acoustic wave experimentation such as interactions between physical and virtual domains, broadband cloaking, and holography. The implementation of IBCs relies on actively injecting a particular wavefield on the boundary of the physical domain from a dense array of transducers. The injected wavefield must honour the real-world physical and the virtual or computational domains. To calculate the required inputs, a dense array of recording transducers inside the boundary is used from which the future wavefield arriving at the source transducers can be predicted using a discretised Kirchhoff-Helmholtz integral. Whereas IBCs are effectively exact for purely numerical applications, a physical implementation in a laboratory suffers from limitations mainly associated with spatial and temporal discretization issues and the imprint of the electrical devices such as transfer functions or radiation characteristics of transducers. We present a comprehensive numerical sensitivity study of a cloaking experiment in a 2D acoustic waveguide. This defines physical limitations of real-world IBCs, and systematic errors introduced due to subsampling of the recording and injection surfaces and to radiation characteristics of the transducers.