Lean Blowout Detection Techniques for Partially Premixed Flames in a Dump Combustor - Novel Combustion Concepts for Sustainable Energy Development - page 223

out to characterize and predict fl flame behavior near LBO. It is observed that for

premixed flames, LBO is preceded by a high level of intermittent behavior char-

acterized by frequent and irregular extinctions and reignitions. This behavior is
absent in lower level of air

fl

fuel premixing.
For all levels of premixing, a shift of color from reddish to bluish as equivalence
ratio reduces toward LBO. A metric has been developed for early detection of
imminent blowout. The ratio of red to blue intensities, integrated over all points,
showed a monotonical variation with change in equivalence ratio. This behavior is
consistent over a large range of degree of premixing of different fuels and

-

ow rate.
The characteristics can be measured using a standard CCD/CMOS sensor and used
as a detection metric. The method is cheaper, computationally easier, and easy to
implement in real time.
The time series of optical data obtained from PMT has been used to develop
another metric for early LBO detection. The time series is

fl

rst converted to a
symbolic time series, by partitioning the data. Two methods of partitioning have
been discussed here. As the flame shifts toward LBO, the symbolic time series

shows a deviation from the normal operating condition (stoichiometric or near
stoichiometric). The deviation can be quanti

fl

ed using a norm called anomaly
measure. This anomaly measure can be used as a metric for LBO detection. Both
partitioning methods showed satisfactory result regarding LBO detection.

Acknowledgment The authors sincerely acknowledge the help extended by Suvojit Ghosh,
Rakesh Prasad Sahu, Tanoy Paul, Souvick Chatterjee, Arkadeep Kumar, and Kamalika chatterjee.
The contribution and collaboration from Prof. Asok Ray, Pennsylvania State University, is
gratefully acknowledged. We acknowledge DRDO, AICTE, and CSIR for their financial support
toward this work.

References

Ateshkadi A, McDonell VG, Samuelsen GS (2000) Lean blowout model for a spray-fired swirl-
stabilized combustor. Proc Combust Inst 28:1281 1288

Barnes JC, Mellor AM (1998) Effects of unmixedness in piloted-lean premixed gas turbine
combustors. J Propul Power 14:967

-

973
Barnum BM, Bell RC (1993) Flame failure detection method. US Patent-5235802, filed 17 Aug
1993
Bukkapatnam S, Kumara S, Lakhtakia A (1995) Analysis of sensor signal shows turning on a lathe
exhibits low-dimensional chaos. Phys Rev E 52:2375

-

2387
Bukkapatnam S, Kumara S, Lakhtakia A (2000) Fractal estimation of

-

fl flank wear in turning. ASME

94
Chakraborty S, Gupta S, Ray A, Mukhopadhyay A (2008) Dynamic data-driven detection and
estimation of faults in thermal pulse combustors. Proc Inst Mech Eng: J Aerosp Eng 222:1097 -
1108
Chao YC, Chang YL, Wu CY, Cheng TS (2000) An experimental investigation of the blowout
process of a jet flame. Proc Combust Inst 28:335 - 342
Chaudhari RR (2011) Investigation of thermoacoustic instabilities and lean blowout in a model gas
turbine combustor. PhD thesis, Jadavpur University

J Dyn Syst Measur Control 122:89

-