Large Eddy Simulations of the Combustor Turbine Interface: Study of the Potential and Clocking Effects

Abstract

With the generalization of Lean Burn combustors, the flow field entering the turbine tends to feature higher levels of swirl, turbulence, while different hot streak patterns often emerge if compared to the previous generation of combustion chambers. In this context, the combustor-turbine interactions and more specifically the transport of hot streaks through the turbine need to be further analysed to gain engine performance and improved turbine life. Considering this new context, a non-reactive axial combustor simulator representative of a Lean Burn architecture, together with a 1.5 high pressure turbine stage is developed within the European project FACTOR. The interaction between these two modules is numerically investigated by performing Large Eddy Simulation (LES) of the combustor simulator equipped with two Nozzle Guide Vanes (NGVs). By using such an integrated approach, the combustor-turbine interface (plane 40) disappears, allowing: (i) more realistic inlet conditions to the turbine by suppressing all the assumptions associated with averaged profiles; (ii) to account for the potential effect of the vanes on the chamber. Note that if compared to classical approaches, the use of time-resolved LES has the advantage of well predicting the combustor mean flow and turbulence, resulting in more realistic flow properties at the turbine inlet as confirmed by previous works on this configuration. This paper focuses first on two LESs of the combustor-turbine specific configurations: i.e., two clocking positions of the hot streak relative to the NGVs. Significant changes on the thermal field around the vanes are highlighted. When the hot streak is injected in front of a vane leading edge, it considerably heats up the pressure side compared to the adjacent vane although the temperature field is quite uniform at the NGV exit because of the enhanced mixing of the hot streak. On the other hand, when the hot streak is injected in the passage between two adjacent vanes, it remains away from the vane walls preventing them from heating up. The hot streak however crosses the vane passage without being significantly distorted resulting in a more heterogeneous flow field at the rotor inlet. Second, the potential effect induced by the presence of the vanes is investigated by comparing the flow field inside the chamber with and without NGVs. It is found that the potential effect does not alter temperature patterns while a significant radial and azimuthal mass flow redistribution is observed up to about 25% axial chord length upstream of the vanes. The turbulence level is affected by the presence of vanes up to plane 40 when the hot streak is aligned with the passage.

Computational Study of Combustor–Turbine Interactions

“…These effects are related to each other in a complex manner. Although there are a number of experimental and numerical studies [7][8][9][10][11][12] which have sought to ascertain the combustor-turbine interactions, we still have not achieved a complete understanding of how flow fields (including the hot-streaks) are affected by the effect of the presence of the stator and clocking. This study is a first step toward the ultimate objective of tightly-coupled unsteady simulations of combustor-turbine interactions for a realistic combustor and a turbine geometry.…”

“…For instance, in Case 1, the middle vane (V2) is located between the fuel nozzles, but the leading edge (i.e., stagnation point) of the last vane (V3) is aligned with the fuel nozzle (N2). Recently, Koupper et al 9 conducted a Large Eddy Simulation to investigate the combustor-turbine interactions for a Lean Burn chamber and was able to show that the different clocking indeed results in the different migrations of the hot-streaks around the vanes. Also, the presence of the two vanes at the combustor exit alters the radial and azimuthal mass flow distribution and the turbulence level.…”

Hot Streak Evolution in an Axial HP Turbine Stage

“…These features might significantly alter the blade surface temperature, with noteworthy implications on the rotor cooling effectiveness [2]. In addition to detailed investigations on the aerothermal features of the flow released by combustors ( [8], among others), recent studies have considered more realistic configurations, combining the hot streak with a local streamwise vorticity [9] and studying the potential for clocking between the burners and the stator blades [10]. The residual hot streak entering the rotor induces even more complex features within the rotor blade row, including the generation of further vorticity cores which are pushed towards the endwalls [11], altering the wall temperature in these regions and triggering the development of novel cooling techniques [12].…”

“…The device was allowed to reach 390 K in the core of the hot streak, as measured on a traverse placed one vane axial chord upstream of the vane leading edge. This temperature peak corresponds to an increase of 20% of the main stream temperature, which is a realistic representation of hot-streak-induced temperature perturbation; just to set a general context, [9,12] documented a temperature ratio of ~1.09 while [8,10] imposed a ratio greater than 1.5. The hot streaks were injected at 70% of the span in the stream-wise direction, with the aim of minimizing the injector blockage and of limiting the jet interaction with the vane secondary flows.…”

Parameterised Model of 2D Combustor Exit Flow Conditions for High-Pressure Turbine Simulations

“…The typical numerical approach is to run two separate Reynolds-averaged Navier-Stokes (RANS) simulations of the combustor and turbine, and to exchange flow conditions at a defined interface (referred to as combustor turbine interface), i.e., combustor outlet conditions are used as turbine inlet conditions. Novel approaches make use of unsteady, time step-wise coupling of the combustor and turbine codes (Vagnoli et al [7]) or aim towards an integrated simulation of combustor and turbine together with RANS as done by Klapdor [1], Raynaud et al [8], or by using large eddy simulations (LES) as per Koupper et al [9]. Turbine designers are interested in how the approaching flow field affects the HPT's aerodynamics and cooling performance.…”