A numerical framework for 3D Computational Fluid Dynamics (CFD) simulations of gasoline-fuelled burners to be adopted in catalyst preheating of vehicle exhaust systems

Air pollution from vehicular traffic is a growing concern, leading to stricter regulations on the exhaust gas emissions. In case of spark-ignition gasoline engines, the limitation of the emissions is challenging, especially during the cold start, when the already critical conditions combine with the inefficiency of the aftertreatment system because of the low temperature. A possible solution to reduce the light-off time needed to bring the catalyst to the operating temperature is represented by a gasoline-fuelled burner, which produces and drives the hot exhaust gases towards the aftertreatment system. This strategy is useful not only for traditional oil-derived gasoline, but also for carbon neutral fuels such as e- and bio-gasoline. In the present paper, a 3D-CFD modelling framework for the simulation of gasoline-fuelled burners for catalyst preheating is proposed and validated against experimental data on a prototype. The framework is mostly inherited from in-cylinder simulations of spark-ignition internal combustion engines, in order to demonstrate that a unique numerical setup can be exploited for multiple purposes with undoubtful saving of cost and time. In this regard, two well consolidated combustion models for engine simulations are tested, namely ECFM-3Z and Detailed Chemistry (DC). Alternative spray and liquid film stripping models are purposely developed in this work and presented in the paper. Validation is carried out by comparing numerical temperature and emissions at the exhaust of the burner with the experimental counterparts. Both the combustion models are able to properly match the experimental outlet temperature, with an error lower than 4 %. As for the emissions, ECFM-3Z and DC are able to predict, at least as order of magnitude, the ppm concentrations of NO and CO at the exhaust. Overall, a good agreement is obtained between simulations and experiments with both the models. Besides the proposal of a robust 3D numerical framework, the present paper aims at demonstrating the primary role of the CFD to evaluate and understand phenomena that cannot be easily investigated via experiments. In fact, prefilmer effectiveness and wall heat losses are identified by the simulations as the most critical aspects of the prototype. In particular, prefilmer and adiabatic efficiencies result equal to 78.2 % and 73.1 %, respectively. These values lead to a poor global efficiency of the burner, equal to 56.1 %, despite a combustion efficiency of 98.2 %. Finally, the combustion regime inside the investigated burner is analysed and the adopted combustion models are found to be (partly or totally) out of their applicability range. Nonetheless, similarly to in-cylinder simulations, they are still capable to provide promising results if properly tuned, as confirmed by the presented validation. The proposed numerical framework is fuel-agnostic, i.e. it can be applied to conventional, e- and bio-gasoline like fuels.