Inferring laminar burning properties from spherical expanding flames: the pitfalls of an established approach

The procedure of deriving laminar burning parameters from high-speed, high-resolution shadowgraph recordings of spherical expanding flames allows to obtain experimental data in lifelike conditions of high pressure and temperature. Straightforward in principle, relatively simple to implement, the approach has been applied in several labs worldwide, to the extent of achieving a "common practice" status, sometimes resembling a codified protocol. Simple as it may look on paper, yet the technique hides a number of pitfalls, which can impair the end result. After the introduction of solid-state high speed cameras (CCD, then CMOS), gathering high-speed videos of single combustion events became a breeze: this, along with the possibility of easy stockpiling of data, lend to underrate data analysis: sort of trading the accuracy of measurement for the repeatability. Another critical phase is dealing with the stretch, which affects any real flame: measured data must be processed to obtain the unstretched flame speed and, ultimately, the laminar burning velocity. The relationship between flame speed and stretch will be discussed, being a key factor for the deconvolution of experimental, stretch-affected data. In the present work a critical discussion is proposed, from experimental data acquisition and processing to stretch analysis: the underlying hypotheses of each step will be used as the guidance to a "good" rather than "common" practice. Reference will be made to a specific test case: the combustion of CH4 in air at P0 = 6 bar, spanning the whole flammability range.