AE 5090. Graduate Aerospace Engineering Colloquium: Alternative Fuels for Gas-Turbines in Propulsion and Power: Chemistry / Transport Effects and Modeling

Friday, February 15, 2019
3:00 pm to 3:50 pm
Floor/Room #: 


Alternative Fuels for Gas-Turbines in Propulsion and Power: Chemistry / Transport Effects and Modeling

Dr. Santosh Shanbhogue
Research Scientist
Mechanical Engineering
​Massachusetts Institute of Technology

3:00-3:50 am, Friday, 2/15/2019
Higgins Labs 202


Ten years ago, most power-plants were powered by coal and all civilian flight propelled by petroleum derived Jet-A/Jet A-1 fuel. Today, anxieties about climate change and the abundant discoveries of natural gas around the world is increasingly shifting us away from traditional fuels, particularly for gas-turbines. In an attempt to reduce their carbon footprint, gas-turbines for flight are now burning Jet A blended with biomass derived fuel. For power generation, land based gas-turbines now burn natural gas or syngas mixtures. In the near future, they will likely shift to oxy-combustion. This natural leads to the question –what is the impact of fuel variability on the combustion process in gas-turbines? Strangely, despite its ubiquity, there is no “octane number” for rating fuels for gas-turbine combustion. As a consequence, operability problems like combustion instability, flashback and blowout have restricted gas-turbines to a narrow margin of fuel-flexibility. To gain a fundamental understanding of the impact of fuels on combustion dynamics, this talk we will review data from a set of experiments carried out in two different combustors. One is a 2D backward step combustor at Re=6500 and the other an axisymmetric swirl stabilized combustor at Re=20,000, operated with different fuel blends – C3H8/H2/Air and CH4/H2/Air. We will see that parameters derived from a numerical twin-flame stagnation point flow can scale a variety of phenomena related to flow and flame dynamics. In the backward step combustor, PIV measurements were performed to determine the size of the recirculation zone at a fixed Reynolds number and upstream velocity for different C3H8/H2/air blends. We find that the size of the recirculation zone length varies by the addition of even small amounts of hydrogen, at a fixed flame temperature. We find that the length of the recirculation zone can be correlated with the extinction strain rate (κext) of the mixture. For swirling CH4/H2/air flames, we find that as the equivalence (Φ) ratio is increased for a given CH4/H2 blend, the flame transitions across four shapes – from a columnar flame close to lean blowoff, to a bubble flame, to an IRZ (inner recirculation zone) stabilized flame and finally to a ORZ (outer recirculation zone) stabilized flame close to stoichiometry. Increasing the proportion of H2 lowers the Φ at which these transitions occur. When plotted as a function of the corresponding κext, all transitions from one shape to another occur at the same critical value of κext. When the swirl combustor is operated with a long exhaust such that thermoacoustic oscillations are excited, transitions from one acoustic frequency to another also occur at the same κext. The observation that the extinction strain rate can be used to scale data under different conditions highlights the role of local flame response in various phenomena and points to a possible universality of the strained flamelet concept over a range of Reynolds numbers. This opens up entire new possibilities of chemistry-based combustion instability and blowoff/flashback control methods, both active and passive, in gas-turbines.

About the Speaker

Dr. Santosh Shanbhogue is currently a Research Scientist at MIT. He got his B. Tech degree from IIT Madras in 2003 and MS & PhD degrees from Georgia Institute of Technology in 2005 and 2008 respectively. His research interests are in the area of unsteady combustion, optical diagnostics and fuel processing. He has authored over fifty journal and conference papers; holds one patent and a couple more pending; and his work has been cited more than a thousand times.