TY - JOUR
T1 - A dynamic self-adaptive turbulence modelling approach for swirling reactive flow in a dual-swirl gas turbine combustor
AU - Xia, Zhaoyang
AU - Chen, Tao
AU - Han, Xingsi
AU - Zhao, Majie
AU - Shi, Baolu
N1 - Publisher Copyright:
© 2026 Elsevier Masson SAS.
PY - 2027/1
Y1 - 2027/1
N2 - Swirling flames are the dominant configuration in modern gas turbine combustors, yet accurate and efficient numerical prediction of high-Reynolds number turbulent swirling reactive flows remains a long-standing challenge. Traditional large eddy simulation (LES) suffers from prohibitive computational cost for industrial applications due to its stringent near-wall grid resolution requirements. An improved dynamic turbulence modelling approach, named self-adaptive turbulence eddy simulation (SATES), featuring a dynamically formulated mode transition coefficient CH that significantly enhances near-wall modeling performance, is proposed in this work. Comprehensive numerical simulations of a well-documented dual-swirl gas turbine model combustor (GTMC) are conducted under both isothermal and reactive conditions, with comparisons to two LES models (with and without wall damping) and experimental measurements. The flamelet generated manifold (FGM) approach based on the detailed GRI-Mech 3.0 mechanism is employed for turbulent combustion modelling. The results demonstrate that the dynamic SATES method accurately reproduces the large-scale recirculation zones, flow split ratio (relative error < 9%), and the characteristic precessing vortex core (PVC) frequency (relative error < 2%) under isothermal conditions. In contrast, LES with wall damping predicts an unphysical Coandă jet due to inaccurate inflow split ratio estimation. For reactive flows, the SATES-FGM model achieves excellent agreement with experimental measurements of velocity, temperature, and major species distributions, and exhibits significantly lower grid sensitivity than LES. This study reveals that accurate prediction of the swirler flow split ratio is critical for reliable downstream flow and flame simulations, and demonstrates the promising potential of the dynamic SATES method for efficient and high-fidelity numerical simulations of industrial gas turbine combustors.
AB - Swirling flames are the dominant configuration in modern gas turbine combustors, yet accurate and efficient numerical prediction of high-Reynolds number turbulent swirling reactive flows remains a long-standing challenge. Traditional large eddy simulation (LES) suffers from prohibitive computational cost for industrial applications due to its stringent near-wall grid resolution requirements. An improved dynamic turbulence modelling approach, named self-adaptive turbulence eddy simulation (SATES), featuring a dynamically formulated mode transition coefficient CH that significantly enhances near-wall modeling performance, is proposed in this work. Comprehensive numerical simulations of a well-documented dual-swirl gas turbine model combustor (GTMC) are conducted under both isothermal and reactive conditions, with comparisons to two LES models (with and without wall damping) and experimental measurements. The flamelet generated manifold (FGM) approach based on the detailed GRI-Mech 3.0 mechanism is employed for turbulent combustion modelling. The results demonstrate that the dynamic SATES method accurately reproduces the large-scale recirculation zones, flow split ratio (relative error < 9%), and the characteristic precessing vortex core (PVC) frequency (relative error < 2%) under isothermal conditions. In contrast, LES with wall damping predicts an unphysical Coandă jet due to inaccurate inflow split ratio estimation. For reactive flows, the SATES-FGM model achieves excellent agreement with experimental measurements of velocity, temperature, and major species distributions, and exhibits significantly lower grid sensitivity than LES. This study reveals that accurate prediction of the swirler flow split ratio is critical for reliable downstream flow and flame simulations, and demonstrates the promising potential of the dynamic SATES method for efficient and high-fidelity numerical simulations of industrial gas turbine combustors.
KW - Dual-swirl gas turbine combustor
KW - Dynamic self-adaptive turbulence eddy simulation
KW - Flamelet generated manifold
KW - Large eddy simulation
KW - Swirling flame
UR - https://www.scopus.com/pages/publications/105048935807
U2 - 10.1016/j.ast.2026.113617
DO - 10.1016/j.ast.2026.113617
M3 - Article
AN - SCOPUS:105048935807
SN - 1270-9638
VL - 180
JO - Aerospace Science and Technology
JF - Aerospace Science and Technology
M1 - 113617
ER -