Multistability and Noise-Induced Transitions in Dispersively-Coupled Nonlinear Nanomechanical Modes

TL;DR

This paper investigates the complex noise-driven dynamics of two dispersively coupled nonlinear nanomechanical modes, revealing multistability, phase transition-like behavior, and triple points, with implications for advanced nanomechanical applications.

Contribution

It introduces a detailed analysis of noise-induced transitions and multistability in coupled nonlinear nanomechanical modes, highlighting phase diagram analogies and novel transition phenomena.

Findings

Identification of four stable states in the coupled system.

Observation of five transition curves separating stable regions.

Discovery of two triple points where three states coexist.

Abstract

We study the noisy dynamics of two coupled bistable modes of a nanomechanical beam. When de-coupled, each driven mode obeys the Duffing equation of motion, with a well-defined bistable region in the frequency domain. When both modes are driven, intermodal dispersive coupling emerges due to the amplitude dependence of the modal frequencies and leads to coupled states of the two modes. We map out the dynamics of the system by sweeping the drive frequencies of both modes in the presence of added noise. The system then samples all accessible states at each combination of frequencies, with the probability of each stable state being proportional to its occupancy time at steady state. In the frequency domain, the system exhibits four stable regions -- one for each coupled state -- which are separated by five curves. These curves are reminiscent of coexistence curves in an equilibrium phase diagram: each curve is defined by robust inter-state transitions, with equal probabilities of finding the system in the two contiguous states. Remarkably, the curves intersect in two triple points, where the system now transitions between three distinct contiguous states. A physical analogy can be made between this nonequilibrium system and a multi-phase thermodynamic system, with possible applications in computing, precision sensing, and signal processing.