What if material switching were slower than crack sound speed—would propagation even respond?

Your question touches on a subtle but important aspect of wave propagation: what happens when the medium itself changes on timescales comparable to or *slower* than the wave transit time?

The short answer is: yes, propagation *does* respond, but the response depends critically on the *coherence* of the disturbance relative to the medium change.

Here's the physics. When material properties switch slowly compared to sound transit time, the wave "sees" a quasi-static medium and adjusts adiabatically—like light bending gradually through a smoothly varying lens rather than bouncing off a sharp boundary. This is actually the regime where geometric acoustics works well: the wavelength remains much smaller than the scale of material variation, so the wave follows effective ray paths through the inhomogeneity [AG-2024.12-648].

By contrast, if switching were *fast* (comparable to or faster than acoustic transit), you'd get reflection, diffraction, and shock formation—the dramatic regime explored in nonlinear acoustics and plasma physics.

The practical issue is *causality*. If a material property tries to change faster than the sound speed can propagate the information, you get acausal behavior—negative energy densities, ghost modes, or instabilities. This is why the sound speed fundamentally constrains how rapidly a system can respond [AG-2024.12-648]. A slow switch avoids this pitfall because the acoustic medium has time to "communicate" the change throughout the region of interest.

One concrete analogy: seismic waves traveling through gradually layered rock (slow lithospheric changes) behave differently from those encountering a sharp fault (fast rupture). Both propagate, but the slow case looks like refraction; the fast case looks like scattering and reflection.

Unfortunately, the preprints provided focus on sound speeds in exotic matter (quark-gluon plasma, quantum field theory), not on time-dependent material switching, so I cannot cite specific experimental or numerical results directly addressing your scenario.