A parallel and matrix free framework for global stability analysis of compressible flows

TL;DR

This paper introduces a parallel, matrix-free computational framework for global stability analysis of compressible flows, enabling large-scale problem solving with improved convergence techniques.

Contribution

It presents a novel parallel, matrix-free iterative framework with advanced eigenvalue selection and preconditioning methods for stability analysis of compressible flows.

Findings

Effective convergence acceleration methods tested

Spectral Cayley transformation improves eigenvalue selection

Parallel block-Jacobi preconditioning enhances scalability

Abstract

An numerical iterative framework for global modal stability analysis of compressible flows using a parallel environment is presented. The framework uses a matrix-free implementation to allow computations of large scale problems. Various methods are tested with regard to convergence acceleration of the framework. The methods consist of a spectral Cayley transformation used to select desired Eigenvalues from a large spectrum, an improved linear solver and a parallel block-Jacobi preconditioning scheme.