The system will be down for regular maintenance from 8:00-10:00am PDT on April 3rd, 2024.

Show simple item record

dc.contributor.advisorNovosselov, Igor V
dc.contributor.authorGupta, Saurabh
dc.date.accessioned2018-07-31T21:15:42Z
dc.date.available2018-07-31T21:15:42Z
dc.date.submitted2018
dc.identifier.otherGupta_washington_0250O_18933.pdf
dc.identifier.urihttp://hdl.handle.net/1773/42471
dc.descriptionThesis (Master's)--University of Washington, 2018
dc.description.abstractGas turbine engines are usually operated at lean equivalence ratios (typically about 0.45 to 0.60) in order to achieve better fuel efficiency and to limit NOx emissions, but this increases the risk of the occurrence of a lean flame blowout (LBO). LBO can cause critical safety concerns for aero-based gas turbine engines while for land-based gas turbines, predominantly used for power generation, such a phenomenon can result in expensive and time-consuming shutdown and restart procedures. Previous research shows that the proximity to such a blowout condition in a premixed combustor can be predicted using the combustion species data obtained from a real-time Chemical Reactor Network (RT-CRN) model. The main advantage of this novel technique is that unlike most of other LBO prediction methods which require significant hardware modifications for monitoring of optical or acoustic parameters of the system, this technique uses computational results based on the combustor temperature only and not requiring any additional combustor modifications. This thesis develops a generic approach for a controlling LBO in a combustor based on the RT-CRN prediction methodology. All calculations shown here are based on experiments conducted in a laboratory single-jet stirred reactor (JSR) operated at atmospheric pressure on methane fuel, which is designed to represent the flame zone of practical combustors. This approach, however, can be easily extrapolated to other systems, contingent to the availability of a working CRN model for the system and a detailed analysis of the OH-radical behavior across the elements thereof. The algorithm utilizes a 3-element CRN design for the JSR, developed and validated by Kaluri [1]. This design employs a series of three Perfectly Stirred Reactors (PSRs) to model the flame, post-flame and recirculation regions of the JSR respectively. The full GRI 3.0 chemical kinetic mechanism is used for calculating the concentrations of the combustion species in the CRN code. The proposed methodology is validated by experiments conducted on the JSR apparatus. For all these validation experiments, the air flow in varied as the independent variable and the fuel flow control signal is actuated based on the output of the control algorithm. Two independent sets of experiments are conducted by increasing the system airflow (i) as a step function and (ii) as a monotonically increasing function. The results are examined to confirm the functionality of the devised LBO prevention scheme in terms of its ability to identify and prevent an incipient blowout and to stabilize the system once such an event has been averted.
dc.format.mimetypeapplication/pdf
dc.language.isoen_US
dc.rightsnone
dc.subjectChemical Reactor Network
dc.subjectCombustion Control
dc.subjectJet-stirred Reactor
dc.subjectLean Flame Blowout
dc.subjectMechanical engineering
dc.subject.otherMechanical engineering
dc.titlePrevention of Lean Flame Blowout using a Real-Time Chemical Reactor Network
dc.typeThesis
dc.embargo.termsOpen Access


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record