Multidimensional Modeling and Validation of Dual-Fuel Combustion in a Large Bore Medium Speed Diesel Engine

Abstract

High fidelity, three-dimensional CFD was used to model the flow, fuel injection, combustion, and emissions in a large bore medium speed diesel engine with different levels of natural gas substitution. Detailed chemical kinetics was used to model the complex combustion behavior of the premixed natural gas, ignited via a diesel spray. The numerical predictions were compared against measured multiple cycle pressure data, to understand the possible factors affecting cyclic variation in experimental data. Under conditions with high natural gas substitution rates, diesel was injected much earlier than firing-TDC. This additional mixing time allowed the active radicals from diesel dissociation to initiate combustion from the cylinder wall and propagate inwards. 0%, 60%, and 93% natural gas substitution rates (by energy) were tested in this study to develop computational capabilities needed to accurately model and understand the underlying physics. Several innovative computational methods such as adaptive mesh refinement (which automatically refines and coarsens the mesh based on the existing solution parameters), and multi-zoning (which groups chemically similar cells together to reduce combustion calculation time) were utilized to obtain accurate predictions at a lower computational cost. Important engine emissions such as NOx, CO, unburnt HC, and soot were predicted numerically and compared against measured engine data.

Towards a Predictive Simulation of Turbulent Combustion?—An Assessment for Large Internal Combustion Engines

“…In [107] the authors used a combination of coherent flame and surface wrinkling model and declared that in spite of capturing some physical trends, further improvements of the mixing controlled combustion are still needed. Research work was carried out with a special focus on the extended chemistry involved due to the two fuels in [108][109][110]. Eder et al showed results with an ECFM-3Z model and empiric correlations for ignition delay and laminar flame speed [111] that reasonably correlated with test data.…”

Numerical Investigation of the Turbulent Flame Propagation in Dual Fuel Engines by Means of Large Eddy Simulation

“…The flame front thickness in a lean natural gas-air mixture under enginelike conditions is usually in the range of a few hundred micrometers [13], which raises the need for computational cells with a size one order of magnitude below the size that is usually used to resolve the combustion chamber to ensure an appropriate resolution of the flame front. To overcome the resulting overwhelming computational effort, some authors applied an adaptive mesh refinement to increase the local resolution in the vicinity of the flame front to generate reasonable results within certain limits [14][15][16][17][18]. However, the predicted burn rates are very sensitive to the selection of the reaction mechanism and sometimes need some additional adaption of single reaction rates [17].…”

“…To overcome the resulting overwhelming computational effort, some authors applied an adaptive mesh refinement to increase the local resolution in the vicinity of the flame front to generate reasonable results within certain limits [14][15][16][17][18]. However, the predicted burn rates are very sensitive to the selection of the reaction mechanism and sometimes need some additional adaption of single reaction rates [17].…”

Cycle to Cycle Variation Study in a Dual Fuel Operated Engine

“…Therefore, a RANS simulation would not realize to a DNS simulation with sufficient grid resolution. It has been shown previously that grid resolutions in the order of 0.25 mm is sufficient to model turbulent combustion in diesel engines [11], gasoline engines [12], gasoline and diesel dual-fuel engines [13], and in natural gas and diesel dual-fuel engines [14]. Using 20 consecutive cycles, Scarcelli and Sevik [12] showed how RANS simulations can be effectively used to represent cyclic variability for EGR dilute and non-dilute gasoline engines.…”

“…A recent study [14] showed 45-65% speedup using METIS for a GDI simulation using the same numerical tool. …”