Date of Award


Degree Name

Master of Science


Mechanical Engineering


Ahsan Choudhuri


Gas turbines are used for power generation and aviation all over the world. Fossil fuels remain the main source of energy for the operation of gas turbines but the recent increase in environmental awareness has influenced government agencies to impose stringent limits on pollutant emissions from power generation plants. Pollutant emissions control is one of the major challenges for gas turbine designers. In this study, experimental methods and computational fluid dynamics (CFD) technologies are used to study the problems that arise when coal derived fuels are used in Gas Turbine Combustors.

Gas turbine designers need both experimental data and combustion models to design combustors that are efficient and have low emission characteristics. The trend in industry is towards less reliance in experimental methods and testing and more on numerical methods which would allow for shorter design cycles. The focus of this work is to obtain a comprehensive understanding of the instabilities that develop in gas turbine combustors and to map the operability range of a typical combustor for different fuel compositions. The modeling effort in this study will provide a critical insight on the flow-flame interaction through flow visualization techniques and the use of unsteady computational models. The outcomes from the research will provide design tools for developing gas turbine combustors fueled with gasified coal and other hydrogen containing fuels.

It is concluded in this study that when the hydrogen content of the fuel mixture is increased, for any given flow rate, the flashback propensity of the combustor increases. It was determined as well that syngas fuel derived from coals containing higher hydrogen content such as Brown and Bituminous are more prone to flashback than syngas derived from Lignite or Coke coal; this conclusion is congruent with the finding that hydrogen content increases the flashback propensity of a fuel mixture.




Received from ProQuest

File Size

63 pages

File Format


Rights Holder

Gilberto Corona