When an imperfectly expanded supersonic jet exits a nozzle, the mismatch between exhaust pressure and ambient pressure creates standing shock cells — a repeating pattern of compressions and expansions downstream. Kelvin-Helmholtz instabilities grow along the jet's shear layer, producing waves that propagate downstream. When these waves interact with the shock cells, they generate upstream-propagating acoustic waves. If the acoustic waves reach the nozzle lip and trigger new instabilities at the right phase, the loop closes and the jet screams — a sustained, high-amplitude oscillation audible across an airfield.
Farghadan, Beekman, Nogueira, Edgington-Mitchell, and Towne show that the linear resonance model of screech — waves going around the loop, amplifying each pass — is insufficient. Using resolvent analysis and a novel bilinear formulation, they demonstrate that the screech mode's nonlinear self-interaction redistributes energy to other frequencies. The screech doesn't just sustain itself through linear feedback. It also drives fluctuations at harmonics and subharmonics through triadic interactions, where three different frequency components exchange energy.
Note (2026-11-13; read against the paper's full text): the paper does not say that the linear resonance picture is insufficient to sustain screech. Its first two steps are linear: the least damped global eigenmode of the linearised operator matches the experimentally observed screech frequency, and the leading resolvent mode agrees closely with the leading mode from proper orthogonal decomposition of the experimental data at the screech frequency (its figure 8), which it takes as confirming that the resolvent mode represents screech. What it says linear resolvent analysis cannot capture is the energy transfer to the mean flow and to harmonics of the screech frequency; the harmonic resolvent and the bilinear formulation are used for that, and it says the forcing from the screech mode's self-interaction, together with triadic interactions, is 'sufficient to explain' the redistribution of energy to other frequencies and the associated experimental observations. The essay's claim that the sustained tone is a pump that generates new frequencies is its reading of that result, and it names subharmonics, which I found only in cited titles and not in the paper's own results. On the loop itself, the paper cites work (Gojon et al., 2018; Edgington-Mitchell et al., 2018) arguing that the upstream-travelling component that closes the loop is better described as a guided jet mode than as a free-field acoustic wave, which the essay's opening describes as upstream acoustic waves.
The through-claim is about what self-sustaining means. A linear feedback loop maintains itself by amplifying one signal through one cycle. The jet screech maintains itself and simultaneously creates other oscillations through nonlinear coupling. The sustained tone is not just a closed loop at one frequency — it is a pump that drives energy into frequencies that were not part of the original instability. Self-sustaining oscillation in a nonlinear system is not just persistence. It is generation. The screech creates what wasn't there.