Combustion characteristics of butanol isomers in multiphase droplet configurations
Abstract This study reports results of experiments on the isolated droplet burning characteristics of butanol isomers ( n -, iso -, sec -, and tert -) under standard atmosphere conditions in an environment that promotes spherical combustion. The data are compared with predictions from a detailed numerical model (DNM) that incorporates complex combustion chemistry, radiative heat transfer, temperature dependent variable fluid properties, and unsteady gas and liquid transport. Computational predictions are generated using the high temperature kinetic models of Sarathy et al. (2012) and Merchant et al. (2013). The experiments were performed in a free-fall facility to reduce the effects of buoyancy and produce spherical droplet flames. Motion of single droplets with diameters ranged from 0.52 mm to 0.56 mm was eliminated by tethering them to two small-diameter SiC filaments (∼14 µm diameter). In all the experiments, minimal sooting was observed, offering the opportunity for direct comparison of the experimental measurements with DNM predictions that neglect soot kinetics. The experimental data showed that the burning rates of iso - and sec -butanol are very close to that of n -butanol, differing only in flame structure. The flame stand-off ratios (FSR) for n -butanol flames are smaller than those for the isomers, while tert -butanol flames exhibited the largest FSR. DNM predictions based upon the kinetic model of Sarathy et al. over-predict the droplet burning rates and FSRs of all the isomers except n -butanol. Predictions using a kinetic model based upon the work of Merchant et al. agree much better with the experimental data, though relatively higher discrepancies are evident for tert -butanol simulation results. Further analyses of the predictions using the two kinetic models and their differences are discussed. It is found that the disparity in transport coefficients for isomer specific species for Sarathy et al. model fosters deviation in computational predictions against these newly acquired droplet combustion data presented in this study.