We started working on acoustic streaming because it is a source of loss of performance in thermoacoustic devices. It was through this 10-year research work that the Parisian women encountered Steven Garrett and enjoyed all our exchanges. We investigated the physical phenomena responsible for the deformation of Rayleigh streaming patterns in a standing-wave guide at high acoustic levels, focusing on the appearance of reverse streaming cells near the acoustic velocity antinodes observed in both experiments and numerical simulations. Thermal effects were first put aside by considering isentropic acoustic propagation, showing that inertia alone modifies streaming cells at a high acoustic level, but does not create any extra cells since the acoustic field is not modified by the streaming field. When the interaction between streaming and acoustic fields is allowed, small modifications in the acoustic velocity generate extra streaming cells, indicating non-linear interactions. Finally, when taking into account heat transfer within the fluid and the solid walls, transverse mean temperature gradients are generated at high acoustic levels, possibly inducing reverse streaming cells. An analytical criterion for the transition in streaming patterns was established based on the intrinsic coupling between thermal effects and acoustic streaming, and then validated with numerical and experimental results.
Baltean-Carlès et al. (Tue,) studied this question.