|Good fuel models are required for good predictions. Left: modeling using a reduced n-heptane model (34 chemical types) vs. data. Right: mofe accurate n-heptane model (174 chemical types) vs. data. Source: Reaction Design. Click to enlarge.
Reaction Design is introducing the first volume of the industry’s most well-validated available Model Fuel Library, the result of seven years of research and validation under the Model Fuels Consortium (earlier post). MFC members included Toyota, GE Energy, VW, Suzuki, Petrobras and Conoco. The MFC is ending its work in December.
The Model Fuel Library is a subscription-based library which includes more than 40 fully validated, self-consistent components that can be used to simulate fuel effects in virtually all types of automotive and aircraft engines, as well as engines used for electric power generation. The components can be combined to model a large variety of new or existing fuel blends.
Fuel properties affect a number of combustion and engine properties, including ignition delay; knocking tendency; flame speeds; pollutant emissions; sooting tendency and particle size distributions; and density, viscosity and heating value. However, real fuels, with their hundreds of components, are too complex to simulate directly. The MFC uses one or two molecules to represent each significant chemical class, and builds detailed chemistry models for each molecule. The MFC applied this concept of surrogate fuels to develop both fossil-based and bio-based fuel models.
Major liquid fuel classes represented in the library include hydrogen; n-alkanes; iso-alkanes; 1-ring aromatics; 2-ring aromatics; cycloalkanes/naphthenes; olefins; oxygenated fuels; and soot precursors and emissions pathways.
The Model Fuel Library provides a set of accurate models for real-fuel components that enables engine designers to develop low-emission, high-efficiency engines in a more timely and cost-effective manner. Further, subscribers will have the ability to target a fuel model for specific application and simulation goals such as ignition, flame propagation, or emissions of NOx, CO and soot/particulate matter (PM) emission rates.
Volume 1 of the library, available now, includes:
Improved fundamental models for C0 to C4 hydrocarbons, enabling accurate prediction of flame propagation, auto-ignition and emissions.
Diesel fuel components, including n-heptane, iso-octane, decalin and alpha-methyl naphthalene (AMN).
Gasoline components, including toluene, propylbenzene, ethyl benzene and xylene.
Additive components such as ETBE.
Alternative fuel components, including ethanol, butanol, bio-diesel components and methyl esthers.
Jet fuel critical components: toluene and decalin.
Soot pre-cursor PAH mechanisms.
Volume 2, with availability planned for December 2013, will include a detailed soot kinetics model with detailed surface kinetics; a global soot (PM) model; updated large n-alkane for improved negative temperature coefficient predictions; and a biofuels sub-mechanism for tetrahydrofuran (THF).
Volume 3, planned for December 2014, will include a reduced (lumped) soot model for CFD and new bio-diesel components.
We set out to build a library of components that, with the right set of software tools, could be mixed and matched or plugged and played or could be trusted in terms of the way the models are built and validated. This included fundamental experiments, particularly in area of soot.
This [past] week we reported on some groundbreaking work that USC did to identify some of the behaviors of particles as they grow and coagulate and oxidize in the engine. We used that to validate our modeling in that area. We spent time talking about some of the ways to make things more elegant and even more accurate—but everybody forgets that 7 years ago we were talking about not knowing where models came from, what kind of validation they had, and so on.
This was quite an accomplishment, I think. It moved the industry forward. Part of the industry doesn’t know that yet. How do we now take this pre-competitive work—funded by this group and exclusive to this group—and now make it available?—Bernie Rosenthal, CEO, Reaction Design
The staged roll-out of the library results from the original nature of the MFC; i.e., companies paid and joined to have a period of exclusivity in use of the data and tools—two years on the chemical model side. The current commercial offering thus offers the library as it was two years ago, Rosenthal said. The improvements and the PM/soot modeling work will be progressively rolled out as the exclusivity periods end.
While the Model Fuel Library can be used with many commercial simulation packages, most have limited kinetics capabilities and may not be able to take full advantage of the library, the company suggests. Reaction Design’s own ENERGICO, FORTÉ and CHEMKIN-PRO products include the most advanced kinetic solvers in the industry and can therefore maximize the utility of the Model Fuel Library components. Using the CHEMKIN-PRO Reaction Workbench, fuel component models from the Model Fuel Library can be blended to create surrogate fuel representations that match the behavior of commercial fuels. Reaction Workbench can then be used to intelligently reduce the fuel model complexity in line with the simulation goals and the package to be used.
In some ways, Rosenthal noted, Reaction Design’s chemistry solvers were meant to make advanced simulation more accessible outside of a high-performance computing environment.
The whole reason that we believe that engines have not progressed more quickly in terms of their ability to predict performance and predict emissions has been as a result of the inability of the software tools—up to this point—to really handle the amount of computation necessary, particularly on the chemistry side.
That drives a need for more and more and faster computers. Our claim to fame—having the fastest chemistry solvers—obviates the need for HPC in some ways. But what we have found is that the more you tell people, the more they want to know, and so you have some really sophisticated work in spray modeling, and turbulent flow, and now sophisticated chemistry and turbulence interactions. All this driving need for computing up and up and up.
Early on, the chemistry portion of FORTÉ was scalable across clusters of computers. We’re already working to spread it across the HPC environment. The same thing is happening on the flow side, the spray side...really taking advantage of the supercomputers.—Bernie Rosenthal
The Model Fuels Consortium. Reaction Design developed the Model Fuels Consortium in 2005 to address the emerging challenges experienced by the automotive and fuel industry. The 20 Consortium members come from energy companies and engine manufacturers, with technical guidance from academic advisors.
The goal was to enable the design of cleaner-burning, more-efficient engines and fuels by accelerating the development of software tools and databases to streamline and bolster these advances.
MFC charter members representing Chevron, Dow Chemical Company, Conoco-Phillips, Institute Français du Pétrol (now IFP Energies nouvelles), PSA Peugeot Citroen, Mitsubishi Motors, Nissan and Toyota were later joined by representatives of Cummins Engine, Ford, General Electric Energy, General Motors, Honda, Mazda, Oak Ridge National Laboratories, Petrobras, Saudi-Aramco, Volkswagen, all of whom joined with Reaction Design engineers to develop, validate and apply fuel models and simulation methods that enable improvements to engine and fuel design.
With the MFC coming to an end in Dec. 2012, Reaction Design is continuing the research and assembly of the most complete and accurate fuel mechanisms available today through the Model Fuel Library subscription service. Several packages to the Model Fuels Library are available starting at $200,000; information about additional pricing configurations is available upon request.
Throughout its seven-year tenure, the Consortium conducted two phases of research. At its launch, MFC members organized with the goal of establishing a practical methodology and the associated software and models needed to improve accuracy of engine simulations. The MFC pioneered the use of “surrogate modeling” where complex fuel chemistry could be represented by a reduced number of well characterized molecular models and reactions selected to accurately simulate specific behaviors like ignition delay or pollutant emission production.
This addressed the fact that hundreds of molecules are involved in thousands of chemical reactions during the combustion process for large categories of fuels: carbon-based fuels such as diesel, gasoline, jet fuel, and natural gas; and alternative fuels such as ethanol and bio-diesel. Surrogate modeling necessitated the creation of a well-validated database of “Model Fuel” components as well as innovative new engineering analysis tools.
|The goal of MFC II was to create software models and tools to predict soot particle size and number. Source: Reaction Design. Click to enlarge.
In 2008, Reaction Design announced the launch of the MFC-II in response to tightening emissions standards and research linking respiratory conditions to harmful microscopic soot particles. As such, the MFC-II was formed with a goal to create software models and tools that could allow engine designers to predict soot particle size and number. After nearly four years of research and development, MFC-II members can now use Reaction Design’s software tools such ENERGICO and FORTÉ, and the updated MFC fuel models to accurately simulate the formation, agglomeration and oxidation of soot particles, enabling designers to create cleaner-burning engines and better respond to anticipated changes in fuel compositions.
The ability to predict soot particle sizes and reduce particulate emissions to meet regulations can save months of development time, cost and most importantly, help address growing global air quality concerns. The MFC has provided a unique opportunity for fuel and auto leaders to closely collaborate with one another for a common goal: the continued development of tools and models to solve fuel efficiency and pollution challenges we all shared.—Charlie Westbrook, senior scientist at Lawrence Livermore Laboratory and chief technical advisor to the MFC