Efficient computation of global resolvent modes
arXiv:2008.10904 · doi:10.1017/jfm.2021.364
Abstract
Resolvent analysis of the linearized Navier-Stokes equations provides useful insight into the dynamics of transitional and turbulent flows and can provide a model for the dominant coherent structures within the flow, particularly for flows with large gain separation. Direct computation of force and response modes using a singular value decomposition of the full resolvent matrix is feasible only for simple problems; despite recent progress, the cost of resolvent analysis for complex flows remains considerable. In this paper, we propose a new matrix-free method for computing resolvent modes based on integration of the linearized equations and the corresponding adjoint system in the time domain. Our approach achieves an order of magnitude speedup when compared to previous matrix-free time stepping methods by enabling all frequencies of interest to be computed simultaneously. Two different methods are presented: one based on analysis of the transient response, providing leading modes with fine frequency discretization; and another based on the steady-state response to a periodic forcing, providing optimal and suboptimal modes for a discrete set of frequencies. The methods are validated using a linearized Ginzburg-Landau equation and applied to the three dimensional flow around a parabolic body.
22 pages, 20 figures
References in corpus (1)
Cited by in corpus (11)
- Data-driven resolvent analysis
- Efficient global resolvent analysis via the one-way Navier-Stokes equations. Part 1. Forced response
- Resolvent-based tools for optimal estimation and control via the Wiener-Hopf formalism
- Nozzle dynamics and wavepackets in turbulent jets
- Improved convergence of the spectral proper orthogonal decomposition through time shifting
- Towards real-time reconstruction of velocity fluctuations in turbulent channel flow
- Wavelet-based resolvent analysis of non-stationary flows
- Self-consistent model for active control of wind turbine wakes
- Resolvent4py: a parallel Python package for analysis, model reduction and control of large-scale linear systems
- BROADCAST: A high-order compressible CFD toolbox for stability and sensitivity using Algorithmic Differentiation
- Linear model reduction using spectral proper orthogonal decomposition