Differential Mass and Energy Balances in the Flame Zone From a Practical Fuel Injector in a Technology Combustor

Abstract

This paper presents further analysis of experimental results from an Air Force program conducted by researchers at Brigham Young University (BYU), Wright-Patterson Air Force Base (WPAFB), and Pratt and Whitney Aircraft Co. (P&W) (Hedman et al., 1994a, 1995). These earlier investigations of the combustion of propane in a practical burner installed in a technology combustor used: (1) digitized images from video and still film photographs to document observed flame behavior as fuel equivalence ratio was varied, (2) sets of LDA data to quantify the velocity flow fields existing in the burner, (3) CARS measurements of gas temperature to determine the temperature field in the combustion zone, and to evaluate the magnitude of peak temperature, and (4) two-dimensional PLIF images of OH radical concentrations to document the instantaneous location of the flame reaction zones. This study has used the in situ velocity and temperature measurements from the earlier study, suitably interpolated, to determine local mass and energy balances on differential volume elements throughout the flame zone. The differential mass balance was generally within about ±10 percent with some notable exceptions near regions of very high shear and mixing. The local differential energy balance has qualitatively identified the regions of the flame where the major heat release is occurring, and has provided quantitative values on the rate of energy release (up to −400 kJ/m3 s). The velocity field data have also been used to determine Lagrangian pathlines through the flame zone. The local velocity and temperature along selected pathlines have allowed temperature timelines to be determined. The temperature generally achieves its peak value, often near the adiabatic flame temperature, within about 10 ms. These temperature timelines, along with the quantitative heat release data, may provide a basis for evaluating kinetic combustion models.