Dissertation Title: Turbulent combustion modeling based on reconstructed tabulation parameter space
Date: 2025/11/14 – 2025/11/14
Dissertation Title: Turbulent combustion modeling based on reconstructed tabulation parameter space
Speaker: Junyi He, Ph.D. candidate at SJTU Global College
Time: November 14th from 3:00 -5:00 p.m., 2025 (Beijing Time)
Location: Room 414A, Longbin Building
Abstract
Turbulent combustion in industrial systems, such as gas turbines, rocket engines, and internal combustion engines, involves numerous physicochemical processes, including turbulence-chemistry interaction, flame-wall interaction, radiation, and, for spray systems, droplet break-up and evaporation, multiphase flow interaction, and spray flame interaction. With respect to these complexities, numerical simulations play an irreplaceable role in gaining a deeper grasp of the underlying mechanisms for the purposes of fundamental research and design improvement. Among various combustion models, methods based on reduced tabulation parameter space stand out as they decouple the flow simulation from chemistry, thus greatly decreasing the computational complexity.
This study aims to reconstruct and improve the current combustion modeling framework using tabulation methods from three aspects. Firstly, concerning the arbitrariness in the selection of assumed PDFs to consider the turbulence effects, an enhanced filtered turbulent flame model (FTFM-C), which directly stores the mapping relations between filtered quantities, is developed and validated against the Sydney bluff-body flame HM1 and Sydney swirl flame SMH1. The results reveal higher numerical stability and improved performance on species concentration predictions compared to its former version. The characteristic of better capturing the flame surface position compared to conventional flamelet methods is still retained. FTFM-C also proves particularly effective in resolving critical flow features in the swirl flame, including upstream recirculation zones and vortex breakdown regions.
Secondly, spray flames are distinguished by the coexistence of multiple combustion regimes, a feature absent in purely gaseous combustion. The similarity mapping for the spray flamelet progress variable model (SMFPV), which efficiently considers the spray flame interaction by introducing the evaporation source terms without direct calculation of droplets, overlooks the existence of the premixed combustion regime that arises from small, rapidly evaporating droplets. By adjusting the spatial range of the evaporation source and including the mixture fraction scalar dissipation rate as an additional entry parameter, the new model (multi-SMFPV) can capture the multiple regimes encountered in spray combustion. Comprehensive validation on several laminar counterflow configurations demonstrates its improved accuracy over existing approaches, achieving close agreement with detailed chemistry benchmarks. With regard to the turbulent combustion case, multi-SMFPV is also implemented to simulate the Sydney piloted turbulent ethanol spray flame and shows satisfactory performance compared to the literature results.
Furthermore, the SMFPV framework is advanced for complex multi-component heavy fuels such as kerosene and gasoline, using the methodology of splitting the evaporation zones of each component and changing the heating-evaporation sequence. The extended model (MC-SMFPV) captures essential physics of the droplet heating of heavy fuel and the preferential evaporation effect of multi-component fuels, as verified through n-dodecane and several multi-component fuel laminar spray flame cases, showing better performance than conventional flamelet methods, which neglect the spray flame interaction and the preferential evaporation effect. This development significantly expands the applicability of flamelet methods while maintaining the inherent advantage of parameter space reduction. A lean direct injection combustor is also simulated, and comparable performance with the more cumbersome traditional spray flamelet model is obtained.
By systematically reconstructing the tabulation parameter space, this work advances turbulent combustion modeling for both gaseous and spray flames, balancing accuracy with computational efficiency.
Biography
Junyi He is a Ph.D. candidate in Shanghai Jiao Tong University Global College, supervised by Professor Lipo Wang. His research focuses on turbulent spray combustion modeling. His dissertation, “Turbulent combustion modeling based on reconstructed tabulation parameter space”, investigates efficient combustion models for simulating multi-regime multi-component fuel spray flames.