Effect of Gaussian wake amplitude on wake-induced transition for a T106A low pressure turbine cascade

TL;DR

This study uses simulations to explore how Gaussian wake amplitude affects flow transition on a turbine blade, showing increased wake amplitude delays separation and reduces drag, with detailed analysis of flow dynamics and energy transfer.

Contribution

It provides new insights into the influence of wake amplitude on flow transition and separation control in low-pressure turbines through detailed numerical analysis.

Findings

Higher wake amplitude delays flow separation

Increased wake amplitude reduces skin friction drag

Wake-induced transition features include puffs, streaks, and turbulent spots

Abstract

The wake-induced transition on the suction surface of a T106A low-pressure turbine (LPT) blade is investigated through a series of implicit large eddy simulations, solving the two-dimensional (2D) compressible Navier-Stokes equations (NSE). The impact of the incoming Gaussian wake amplitude on the blade's profile loss and associated boundary layer parameters is examined, revealing a 50\% reduction in skin friction drag at the highest amplitude. The results indicate that increasing wake amplitude leads to delayed separation and earlier reattachment, resulting in reduced separated flow. The vorticity and enstrophy dynamics during the transition process under varying wake amplitudes reveal characteristic features of wake-induced transition, such as puffs, streaks, and turbulent spots. The periodic passing of wakes induces intermittent "calmed regions", which suppress flow separation and improve profile loss at low Reynolds numbers (Re), typically found in LPTs. The energy budget, accounting for both translational and rotational energy via the turbulent kinetic energy (TKE) and compressible enstrophy transport equation (CETE), respectively, shows trends with increasing wake amplitude. The relative contribution to TKE production and the roles of baroclinicity, compressibility, and viscous terms are explained.