Skip to main content
To KTH's start page

Numerical Investigations of Secondary Flows in Low-Pressure Turbines

Time: Fri 2025-12-12 10.00

Location: D2, Lindstedtsvägen 5, Stockholm

Video link: https://kth-se.zoom.us/j/62659971159

Language: English

Subject area: Engineering Mechanics

Doctoral student: Victor Baconnet , Strömningsmekanik, SeRC - Swedish e-Science Research Centre

Opponent: Professor Johan Revstedt, Lunds Universitet

Supervisor: Professor Dan S. Henningson, SeRC - Swedish e-Science Research Centre, Strömningsmekanik; Docent Ardeshir Hanifi, SeRC - Swedish e-Science Research Centre, Strömningsmekanik

Export to calendar

QC 251120

Abstract

This thesis is concerned with the study of secondary flows in low-pressure turbines. Secondary flows refer to the vortical structures that arise due to the presence of the walls onto which turbine blades are attached in an engine. The losses caused by secondary flows can contribute up to a third of the total losses of a turbine. The tools used by turbomachinery designers are still not able reliably quantify the losses associated with secondary flows across a wide range of inflow parameters. Unfortunately, secondary flows are highly dependent on inflow conditions, and particularly so on the state of the incoming end-wall boundary layer. We carried out high-fidelity numerical simulations with varying inflow conditions to better understand the effect of such variations on turbine losses. Three simulations at an exit Reynolds number of 150,000 with different inflows are considered: one case with a laminar boundary layer, one case with added free-stream turbulence (FST), and one case with a turbulent boundary layer and FST. The turbulent profile is taken from experiments and is representative of an engine-like environment. First, we analyzed the effect of varying inflow conditions on the flow field around the turbine blade. Away from the wall, FST delays separation on the suction side and dampens the large scale vortex shedding at mid-span. We observed that FST did not affect the overall picture of secondary flow losses at the outflow. Outflow losses were strongly affected by changes in inflow endwall boundary layer, where two loss cores were observed for the laminar boundary layer, while only one loss core was present for the turbulent boundary layer. In general, secondary flow structure formation and development through the passage was also significantly affected by changes in boundary layer. We applied proper orthogonal decomposition (POD) to the three cases and found that any type of reduced-order modelling for secondary flows was not applicable and that a wide range of modes was needed to capture all of the energy of the flow. Finally, we computed total pressure loss contributions across POD modes and in various regions of the domain. We found that loss production is spread out across all modes. Within the endwall region, we found that the passage vortex was responsible for most of the losses across the blade passage.

urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-373044