Session: 04-02: Flashback and Blowoff
Paper Number: 79347
79347 - Computational Fluid Dynamics Modeling of Fuel Properties Impact on Lean Blowout in the ARC-M1 Combustor
The flow and flame dynamics within liquid fueled gas turbine combustors are complex due to the interactions between the highly turbulent flow, spray dynamics and combustion. Computational tools help understand these governing processes. Predictive modeling capabilities for gas turbine combustors are a necessity for accelerating the design optimization cycle as well as improving the understanding of combustion dynamics and rare events within the combustor such as lean blow out. In liquid fueled combustors, atomization and vaporization process couple with local fuel/air mixing and the flow structures in the near-nozzle region to govern flame stabilization. This leads to higher sensitivity to the operating conditions and the fuel properties for the LBO limits. Limited studies have investigated LBO in liquid-fueled gas turbines. In general, literature is divided over whether LBO is limited by fuel physical or chemical properties and this, in turn, inhibits the ability to develop optimized fuels that satisfy LBO requirements. Most simulation studies of LBO focused on the easier problem of premixed flames and noted a strong dependence of LBO on the flow dynamics. Limited CFD studies have attempted LBO simulations in liquid-fuels combustors since including liquid fuels introduces complexities of two-phase atomization, mixing, vaporization, and partially premixed flames.
To this end, a computational fluid dynamics (CFD) model for Army Research Combustor Midsize (ARC-M1) is developed using CONVERGE to characterize the complex turbulent flow, multi-phase spray physics and hydrocarbon chemistry. The computational domain includes all the complex geometry components of the combustor including the main air inflow pipe, swirler, annular cooling air region, and dilution air holes. The gas-phase is described using an Eulerian approach and the liquid spray is modeled with discrete droplets in a Lagrangian frame. A finite-volume based compressible flow solver with a finite rate chemistry solver is used. As LBO is dependent on the unsteady fuel/air mixing processes, large-eddy simulations (LES) CFD model with fuel-dependent properties and chemistry is developed. Near nozzle spray characteristics are particularly challenging to obtain for initialization and validation of simulations. With the availability of high-quality X-ray data for the research combustor (ARC-M1), this near nozzle behavior can be characterized and used for extensive validation of the Computational Fluid Dynamics modeling approach at gas turbine conditions. Recent measurements, using Argonne’s Advanced Photon Source (APS) facility, provide a first-of-its-kind dataset for initialization and were used for the validation of near nozzle spray dynamics. The LBO study is performed by first establishing a stable flame with the corresponding liquid flow rate. The liquid flow rate is reduced in steps and the heat release rate is tracked within the combustor. Each stepped liquid flow rate is maintained for one flow through time to establish a “steady-state” flame. Beyond a certain liquid flow rate, the heat release rate shows a strong decline and this flow rate is tagged as the LBO limit for a given operating condition. The condition tested in the current study includes air temperature of 394 K and combustor pressure of 2 atm. To understand the overall impact of liquid properties, the liquid properties corresponding to Jet A and F-24 are tested, keeping all other models, inflow conditions and geometry consistent. It is observed that F-24 has a higher LBO liquid flow rate compared to Jet A. To understand the impact of individual properties, liquid properties such as density, viscosity, specific heat, surface tension, heat of vaporization, thermal conductivity and vapor pressure are changed one at a time. It was observed that a reduction in density, viscosity, heat of vaporization and specific heat w.r.t Jet A tends to increase the LBO liquid flow rate i.e. makes the flame blow-off at higher equivalence ratios. The physics governing these changes will be further explored in the paper.
Presenting Author: Debolina Dasgupta Argonne National Laboratory
Presenting Author Biography: Dr. Debolina Dasgupta (ES) is a Research Scientist in the Multi-Physics Computations section within the ES division. She will be converted to staff starting in October 2021. She has extensive experience in high-fidelity DNS, device level CFD simulations, HPC, and large-data analysis facilitating understanding and model development for turbulent combustion, thermo-acoustic instability in gas turbine engines for aviation and stationary power generation applications.
Authors:
Debolina Dasgupta Argonne National LaboratorySibendu Som Argonne National Laboratory
Eric Wood University of Illinois at Urbana-Champaign
Tonghun Lee University of Illinois at Urbana-Champaign
Eric Mayhew Army Research Laboratory
Jacob Temme Army Research Laboratory
Chol-Bum Kweon Army Research Laboratory
Computational Fluid Dynamics Modeling of Fuel Properties Impact on Lean Blowout in the ARC-M1 Combustor
Paper Type
Technical Paper Publication