With advanced engines pushing the limits of fuel efficiency, rapid development and improvement of engines increasingly rely on insights from simulations. Reliable simulations require fuel models that consist of a fuel surrogate and its kinetic mechanism. As complexity and sources of fuels vary, a good surrogate needs to be tailored for the specific test fuel. A simple surrogate, typically consisting of 1 to 3 components, can match a single property of the real fuel, such as ignition quality or average molecular weight. More complex surrogates with 4 to 7 components can capture many properties simultaneously. While simple surrogates are good for estimating ignition in engines they require some compensation for the mismatch of the fuels’s physical properties. Complex surrogates can be used to directly represent real fuels in both laboratory experiments and simulations.

We have developed a surrogate blending methodology to identify surrogates with a desired degree of complexity. This involves methods that estimate properties for fuel blends, including ignition quality, sooting propensity, distillation curve, as well as other physical and chemical properties that are important to combustion behavior in simulations. We have assembled and developed a rich library of over 60 fuel components from which we can formulate surrogates to represent most gasoline, diesel, gaseous fuels, renewable fuels, and several additives. The components cover a carbon number range from 1 to 20, and chemical classes including linear and branched alkanes, olefins, aromatics with one and two rings, alcohols, esters, and ethers. As part of the library, we have assembled self-consistent and detailed reaction mechanisms for all the components. The mechanisms also include comprehensive NOx creation and destruction pathways, molecular weight growth kinetics leading to the formation of polycyclic aromatic hydrocarbons (PAH), and a detailed soot-surface mechanism. The mechanisms have been validated extensively using over 500 published sets of experimental kinetics data from a wide range of facilities and diagnostic methods. Over the past decade, the validation suite has been used to improve the kinetics database such that good predictions and agreement to data are achieved for the fuel components and fuel-component blends, within experimental uncertainties. This effectively eliminates the need to tune specific rate parameters when employing the kinetics mechanisms in combustion simulations.

For engine simulations, the master mechanisms have been reduced using a combination of available reduction methods while strictly controlling the error tolerances for targeted predictions. These include several directed relation graph (DRG) based methods and sensitivity analysis. Iteratively using these reduction methods has resulted in small mechanisms for efficiently incorporating the validated kinetics into computational fluid dynamics (CFD) applications. The surrogate formulation methodology, the comprehensive fuel library, and mechanism reduction strategies suggested in this work allow the use of CFD to explore design concepts and fuel effects in engines with reliable predictions.

This content is only available via PDF.
You do not currently have access to this content.