Ignition and Combustion Characteristics of OME _3-5 and N-Dodecane: A Comparison Based on CFD Engine Simulations and Optical Experiments

Abstract

<div class="section abstract"><div class="htmlview paragraph">Synthetic fuels derived from renewable power sources, so-called e-fuels, will play a crucial role in achieving climate-neutral future mobility because they can be used in the existing fleets and in hard-to-decarbonize applications. In particular e-fuels that contain oxygen in their chemical structure can also burn more cleanly in terms of soot formation. For compression-ignition engines, polyoxymethylene dimethyl ethers (PODEs or OMEs) are among the most promising candidates for such oxygenated e-fuels.</div><div class="htmlview paragraph">Here, we investigated the characteristics of injection and combustion of OME<sub>3-5</sub> mixture compared to n-dodecane, a reference diesel-like fuel. Both single and multi-injection, comprising a short pilot injection, is used. Experiments were performed in a single-cylinder optically accessible Bowditch-type engine, injecting with 1500 bar pressure with a 3-hole injector (Spray B of the Engine Combustion Network). Liquid and vapor penetration were measured by imaging the spray illuminated by a pulsed light-emitting diode (LED). Ignition delay, lift-off length and flame morphology were investigated based on multi-spectral high-speed imaging of chemiluminescence. For simulations, a 3D CFD engine model was developed. The combustion simulation was performed on a 120° sector mesh onto which flow and turbulence fields from a gas exchange simulation are mapped prior to fuel injection. The model accounts for piston-ring blow-by. For the combustion of both fuels, detailed reaction mechanisms were used. In general, quite good agreement between model predictions and experimental results was achieved. In particular the consideration of blow-by losses by the CFD model produced a realistic behavior during the high-pressure cycle.</div><div class="htmlview paragraph">Both CFD simulation and optical experiments, reveal significant differences between the two fuels. For OME, the liquid phase penetrates further into the combustion chamber, the ignition delay is shorter compared to n-dodecane and the equivalence ratio of OME during combustion is significantly leaner.</div></div>

LES and RANS Spray Combustion Analysis of OME3-5 and n-Dodecane

“…The entire combustion process was shown not to exceed equivalence ratios greater than two for OME, which is, of course, in stark contrast to the combustion of n-paraffinic fuel like n-dodecane. These observations were confirmed for a single-cylinder optically accessible diesel engine in [18] using the same RANS setup and injection strategy and also recently by García-Oliver et al [19] utilizing a different simulation setup and injection strategy but the same OME fuel. However, the simulations could not reproduce the strong OH* chemiluminescence signal in the spray center axis and the peak intensity near the flame lift-off length in either the constantpressure vessel or the single-cylinder engine.…”

“…After analyzing the global combustion behavior for both fuels and operating points, the spatial distribution of the high-temperature flame morphology will be discussed in detail. The results in Figures 5 and 6 The comparison between the two fuels for OP1 (900 K) in Figure 5 demonstrates significant differences in the spatial distribution of the high-temperature reaction zone, already observed in [17,18]. For n-dodecane (Figure 5a), the highest intensity is measured and simulated in the shear layer of spray and ambient air, which are also the locations of the first ignition kernels.…”

“…Figure 2 visualizes the spray-box mesh with its refinement areas and a cell count of 966,000. The mesh and the models for the liquid and gaseous phase for the RANS calculations were validated extensively in [17,18]. The validation included a grid independence study, which was conducted for non-reacting as well as reacting conditions.…”