Haripriya Rajagopalan
(Advisor: Prof. Tim Lieuwen]

will defend a doctoral thesis entitled,

Predicting Turbulent Burning Velocity of High Flames

On

Wednesday, July 9 at 1 p.m. 
 Food Processing Technology Building, Auditorium 102
North Avenue Research Area, 640 Strong St NW 

Abstract
Hydrogen (H2) is increasingly being explored as an energy carrier to support a decarbonized energy economy. Both the direct utilization of H2 and its blending with conventional fuels such as natural gas are under active consideration to meet the stringent emissions regulations faced by the energy sector. Among the combustion parameters affected by H2, the turbulent burning velocity (ST) plays a central role in determining combustor operability limits, influencing blowoff, flashback, and combustion instabilities. The impact of H2 on ST becomes particularly complex under lean conditions and at high pressures, where traditional scaling models for ST, based largely on root-mean-square turbulent velocity fluctuations () and the unstretched laminar flame speed (SL,0), are inadequate.

This dissertation focuses on the co-development of data-driven and physics-based modeling frameworks to predict ST in H2-fueled premixed flames based on leading-point concepts. A physics-based correlation is developed for the turbulent global consumption speed ST,GC, rooted in Damko¨hler’s paradigm, of the form:

where, and SL,ref are generalized forms of the conventionally used root-mean-square turbulent velocity fluctuation () and the laminar unstretched flame speed (SL,0), respectively. Unlike traditional definitions, in this study, these parameters are determined based on leading-point scaling and data-driven fitting of large experimental databases. ST,GC measurements were acquired at a high-pressure Bunsen burner facility across different H2 fuel blends—(H2/CO, H2/CH4, H2/CO/CH4/N2)—operated at gas turbine–relevant pressures (1–20 bar) and preheat temperatures (300–450 K). The global turbulent consumption speed (ST,GC), representing the average conversion rate of premixed reactants to combustion products, is measured at the facility using OH* chemiluminescence, while the flow field turbulence characteristics are quantified using laser Doppler velocimetry (LDV). The primary focus of this work is on the ST,GC measurements of H2/CH4 fuel blends.

The ST correlation is derived by first applying feature selection techniques to identify the most relevant non-dimensional parameters governing ST,GC, based on their predictive importance across the dataset. This analysis reveals that a small subset of features - namely, the turbulence intensity /SL,max, normalized bulk flow Reynolds number (), and a time scale ratio associated with the flame leading edge (Kamax = τSL,max /τflow) - captures most of the observed variance in ST,GC. This is followed by symbolic regression, which produced an interpretable functional form capturing the nonlinear relationship between the selected features and ST,GC. We observe that the optimal scales as:

Unlike conventional Reynolds number considerations in ST,GC correlations, we hypothesize that this scaling highlights the influence of intermittent, extreme velocity fluctuation events, which occur in the tails of the velocity distribution and become more pronounced with increasing . These extreme events are hypothesized to represent flow velocity fluctuations interacting with the turbulent flame front at the leading points. Some evidence for the influence of on the tail events of turbulent velocity distributions is provided through flow characterization using LDV studies.

Meanwhile, the data clearly indicates that SL,ref closely corresponds to the maximum laminar burning velocity, SL,max, of critically stretched flames, with only a weak dependence on the Karlovitz number (Kamax). These findings provide useful insights into datasets showing the important effects of hydrogen composition and pressure on the turbulent flame speed, as well as the underlying fluid mechanic and chemical kinetic roots of these dependencies.

.  Committee

• Prof. Tim Lieuwen – School of Aerospace Engineering (advisor)
• Prof. Adam Steinberg– School of Aerospace Engineering
• Prof. P.K. Yeung – School of Aerospace Engineering
• Dr. Jackie Chen – Senior Scientist, Sandia National Labs
• Dr Debolina Dasgupta – Research Scientist, Argonne National Lab