|M.Sc Student||Omari Ahmad|
|Subject||Laminar Burning Velocity Measurement and Combustion|
Characteristics of Alcohol Steam Reforming
|Department||Department of Energy||Supervisors||ASSOCIATE PROF. Leonid Tartakovsky|
|PROFESSOR EMERITUS Michael Shapiro|
|Full Thesis text|
Utilizing exhaust gas heat emitted from internal combustion engines (ICE) for on-board alcohol reforming process is a promising way to produce hydrogen-rich syngas in addition to recovering part of the otherwise totally wasted exhaust gas energy. Feeding the engine with these gases contributes to a higher heat release rate, higher knock resistance and wider lean operating capabilities, all of which result in overall efficiency increase and emission mitigation. However, the on-board alcohol reforming process is subjected to changing exhaust gas heat and temperature affecting the reformate composition and hence their combustion characteristics, which in turn affects the exhaust gas heat. To optimize the latter closed loop reformer-ICE behaviour as well as to set light on the optimal reforming process (catalyst, alcohol-to-steam ratio), the ability to predict the change in heat release rate resulting from the change in reformate composition is considered a key factor.
To provide such data, this research investigates in the laminar burning velocity and the effects of cellularity on burning velocity for a wide range of reformates. An additional goal was to investigate in different methods for measuring the laminar burning velocity using a spherical combustion vessel. In particular, the Schlieren imaging technique along with the stretched unconfined flame model as well as the pressure monitoring technique along with the confined flame model were both used and the conditions for consistency where investigated.
A novel semi-spherical combustion bomb was designed satisfying the conditions for both measuring methods and hence allowed their application on the same experiment. The following gaseous fuel mixture was used to simulate the reformates of a general alcohol reforming process; αCOβCH4(1-α-β)CO2(3-α-4β)H2, where α and β are the CO-selectivity and CH4 selectivity of the reforming process respectively. The burning velocity was mapped as a function of α and β within the range: 0<α<1 , 0<β<0.4 . Separate maps were made for the following air-excess factors; λ=1, 1.3, 1.6 and 1.9 . The results showed maximal burning velocities up to 140cm/sec for mixtures simulating either an ethanol or methanol steam reforming process with zero CH4 and CO selectivity (α=β=0). Increasing the CO selectivity from α=0 to α=1, which is accompanied by a reduction in both the CO2 and H2 fractions, showed only a slight decrease in laminar burning velocity. In contrast, increasing the CH4 selectivity from β=0 to β=0.4 almost halved the laminar burning velocity of the mixture. Flame cellularity was found to accelerate the flame propagation and thus contributing to a higher heat release rate. The effect of cellularity was quantified by an apparent cellular burning velocity which exceeded the laminar one by up to 90%. Cellularity is found to occur earlier and hence its effect was stronger for mixtures with higher hydrogen content (α & β → 0) and at higher air-excess factors. As to the comparative investigation of the above mentioned burning velocity measuring methods, it was observed that consistency in the measured laminar burning velocity value existed only when an accurate relation between the pressure and the burned mass fraction is made.