Heat2Control: modeling of Split-injection Spray-A Case from ECN with FGM in OpenFOAM

Diesel like combustion is mostly preferred for the internal combustion engines due to its efficiency. However, it has several drawbacks at which emissions are one of them. Reducing emissions down while keeping the efficiency high is a tough task to cope with. In this respect, multi-phase injection studies are performed in order to control the combustion since diesel combustion is mostly a mixing controlled process. In this project, nominal condition for the double injection Spray-A case from Engine Combustion Network (ECN) is modelled with Flamelet-Generated-Manifold (FGM) in OpenFOAM computational fluid dynamics (CFD) tool with Reynolds Averaged Navier-Stokes (RANS) turbulence model to validate the solver code that will be used for engine applications later on. FGM is an efficient chemistry reduction method, which holds good accuracy while keeping the computational power lower compared to the other reduction methods for chemical kinetics. The solver code for OpenFOAM is from Lib-ICE library (Lucchini, 2013). FGM method is implemented into the OpenFOAM solver since the original solver utilizes Tabulation of Dynamic Adaptive Chemistry (TDAC) method for chemical kinetics. TDAC solves transport equations for chemical kinetics online; therefore it’s limited to chemical reaction mechanisms with 50-species and 100-reactions whilst FGM method is suitable for larger reaction mechanisms since the chemistry is decoupled from the flow field. FGM look-up table stores sources terms for transport equations and thermophysical variables that are required for transport equations with respect to user-defined controlling variables. Besides, selected species can be stored in the look-up table to be retrieved with respect to the controlling variables. Furthermore, thermophysical properties are updated with the virtual-fuel approach. In this respect, a linear system is solved to evaluate virtual species mass fractions by satisfying the conservation of energy, mixture properties, and elemental masses. This approach improves the computational time significantly, 30 times faster for each CFD simulation of 3 [ms], compared to the original OpenFOAM solver code with TDAC and provides comparable results. the chemistry effects on ignition and lift-off length for both injections that are computed in CFD simulations with respect to the different chemical reaction mechanism, which are for n-dodecane surrogate fuel. The smallest mechanism (Yao, 2015) in terms of the number of species gives comparable results for ignition delays and lift-off lengths with respect to the experimental results, (Skeen, 2015).