The critical layer in quadratic flow boundary layers over acoustic linings

TL;DR

This paper analyzes the critical layer effects in quadratic flow boundary layers over acoustic linings, revealing dominant non-modal contributions and stabilization mechanisms using Frobenius series and Fourier transforms.

Contribution

It introduces a Frobenius series-based method to accurately compute the Greens function and identify hidden modal poles in the critical layer of flow boundary layers.

Findings

Critical layer contribution dominates pressure perturbations in thick boundary layers.

The continuous spectrum branch cut stabilizes otherwise unstable modes.

The method accurately locates modal poles behind the branch cut.

Abstract

A straight cylindrical duct is considered containing an axial mean flow that is uniform everywhere except within a boundary layer near the wall, which need not be thin. Within this boundary layer the mean flow varies parabolically. The linearized Euler equations are Fourier transformed to give the Pridmore-Brown equation, for which the Greens function is constructed using Frobenius series. Inverting the spatial Fourier transform, the critical layer contribution is given as the non-modal contribution from integrating around the continuous spectrum branch cut. This contribution is found to be the dominant downstream contribution to the pressure perturbation in certain cases, particularly for thicker boundary layers. The continuous spectrum branch cut is also found to stabilize what are otherwise convectively unstable modes by hiding them behind the branch cut. Overall, the contribution from the critical layer is found to give a neutrally stable non-modal wave with a phase velocity equal to the mean flow velocity at the source when the source is located within the sheared-flow region, and to decay algebraically along the duct as $O(x^{-5/2})$ for a source located with the uniform flow region. The Frobenius expansion, in addition to being numerically accurate close to the critical layer where other numerical methods loose accuracy, is also able to locate modal poles hidden behind the branch cut, which other methods are unable to find; this includes the stabilized hydrodynamic instability. Matlab code is provided to compute the Greens function.