Development of a Rapid Compression Machine for Screening Alternative Fuel for Gas Turbines.

The reduction of NOx, soot and other emissions in aero or industrial gas turbines are engage by the concept of new combustion system such as the lean premixed pre-vaporized system. Which brings with it several engineering solutions such as combustion instabilities, flashback and autoignition. An experimental test rig named Sheffield Rapid Compression Machine (Shef-RCM) is designed to investigate the autoignition chemistry of alternative fuels relevant to gas turbine plant and to handle high boiling point long chain hydrocarbon fuels. The Shef-RCM incorporates a hydraulic stopping mechanism, piston release mechanism, an optimal crevice piston design and a reactor chamber is designed which utilises the direct test chamber method for easy admittance of fuels. The machine is pneumatically driven and hydraulically stopped. The novelty in the design of the Shef-RCM is the introduction of a piston release mechanism (brake) pneumatically operated use to hold the reactor piston in position at its bottom dead centre. A computational fluid dynamics study on crevice piston was undertaken to produce the best-optimized crevice piston head design that will suppress the roll-up vortex to enhances the homogeneity of the temperature field of the reactor chamber. The simulation used a 2-Dimensional computational moving mesh axisymmetric in the commercial code of Ansys fluent. The model adopted for this calculation was the laminar flow. Appropriate choice of the model parameter was taken into consideration during the simulation to reduce errors caused by a poor mesh quality. These parametric studies examine the time step size and the mesh density, which was been deemed necessary for running the simulation to handle errors of negative cell volume. The parameters maintained for the model are a constant stroke length of 142 mm with a volume clearance height of 17 mm. Further optimisation of a 282 mm3 crevice volume on the width resulted to five different crevice widths of 3 mm, 5 mm, 7mm, 9 mm and 12 mm respectively. The widths of the piston head crevice of 5mm gave a better result regarding the peak pressure profile and maintained a homogeneous temperature field at the end of the TDC at post compression time of about 40ms. Performance characterization of the Shef-RCM, using inert gases, N-Heptane and Jet A-1 showed that the experimental data obtained was highly reproducible and repeatable. The machine is vibration free, allows for fast compression, less than 35 ms, an obtainable compressed gas pressure of 22 bar. The estimation of the compressed gas temperature at the top dead centre using numerical modelling was 698 K; the heat loss implemented in the model used an effective volume approach, which showed a perfect match for the model with experiment. Ignition delay time measurement for Jet A-1 are reported for low to intermediate temperatures regime (734 ≤ TC ≥ 815)K, compressed gas pressure, PC = 6 and 10 bar and equivalent ratios, ф= 0.5, 0.75 and 1.0 in air. Jet A-1 exhibited Arrhenius behaviour at 6 bar and 10 bar except for some suspected traces of NTC at ф = 0.5,which needed to be fully established. No evident of Negative Temperature Coefficients (NTC) behaviour at a higher pressure of 10 bar. The kinetic modelling conducted for Jet A-1 used Ranzi et al.[1] model with Dooley et al. [2] and Aachen[3] surrogate mixture. At a compressed pressure of 6 bar, ф= 0.75, the model predicted a shorter ignition delay time and displayed a two stage ignition delay time for Jet A-1 fuel, and the model was in agreement with the experiment. The Shef-RCM facility has also been used to measure the combustion behaviour of Banner-Solvent at low to intermediate temperature regime (718 ≤ TC ≥ 916) K at compressed pressure, PC = of 6 and 10 bar, and equivalence ratios, ф= 0.5, 0.75 and 1.0 in air. Various diluent mixtures were carried out to alter the end gas temperature, it was found that ignition delay within the temperature range of 718 – 916 K exhibited NTC behaviour at lean condition and stoichiometric. Banner-Solvent reacts faster compare to Jet A-1, and this showed some trend of Negative Temperature Coefficient behaviour at a compressed gas pressure of 6 bar. Experimental measurement of the ignition delay response of UCO-HEFA at low to intermediate temperature regime (680 ≤ TC ≥ 777) K at compressed gas pressure, PC = 6 and 10 bar, and equivalence ratios, ф= 0.5, 0.75 and 1.0 in air was studied. The effects of temperature, pressure, and equivalence ratio and oxygen concentration on the ignition delay time was investigated. The overall reactivity of the three fuels showed that Banner-Solvent had showed a higher reactivity than Jet A-1 and UCO-HEFA at 10 bar compressed gas pressure. At 6 bar compressed gas pressure, UCO-HEFA showed some signs of NTC behaviour. The uncertainty for the three fuels was considered and this was seen to be within the limits compared in literature. The global correlation for Jet A-1 and UCO-HEFA were derived for both fuels.