The accuracy of the counterflow, twin-flame technique for the determin
ation of laminar flame speeds was examined analytically, numerically a
nd experimentally. The analysis was conducted by using multiple-expans
ion, large activation energy asymptotics, while the numerical simulati
on incorporated detailed chemistry and transport. In both approaches t
he solutions were obtained in a finite domain and with plug flow bound
ary conditions in order to better simulate the actual experiments. Res
ults show that linear extrapolation of the minimum velocity to zero st
retch overestimates the true laminar flame speed. This overestimate, h
owever, can be reduced by using larger nozzle separation distances. Th
e theoretical results were further confirmed by experimental measureme
nts for methane/air flames with various stoichiometries and nozzle sep
aration distances. The numerical and experimental results indicate tha
t for atmospheric methane/air flames, nozzle separation distances in e
xcess of about 2 cm yield laminar flame speeds obtained by linear extr
apolation accurate to within the uncertainty range of the experiment.
The results obtained herein thus provide further support for the viabi
lity of the counterflow technique, when the influence of the nozzle se
paration distance is properly accounted for. The viability of an alter
nate technique for the determination of laminar flame speeds, based on
the variation of how velocity at a constant temperature near the upst
ream boundary of the flame with stretch, was also theoretically invest
igated. (C) 1997 by The Combustion Institute.