AdBlue verursachte Ablagerungen II : Untersuchung und Modellierung der Ablagerungsbildung während der Abgasnachbehandlung durch die Einspritzung von AdBlue vor den SCR-Katalysator

Abstract

Future emission legislation will increase the pressure on European car manufacturers and suppliers to reduce nitric oxide emissions from vehicles powered by diesel engines. On the other hand, engine-out nitric oxide will rather increase due to the engine efficiency optimisation and the cut-down of CO2 emissions. As a result, there is a substantial need for effective exhaust gas aftertreatment with SCR technology using AdBlue as a reducing agent for NOx. Incomplete evaporation and decomposition of the spray may lead to an extensive liquid film formation on exhaust walls, mixer elements or the catalyst, resulting in deposit formation. In the preceding project “AdBlue Deposits I”, a fundamental understanding of the underlying mechanisms and measures for deposit formation was gained in the cooperative work between Karlsruhe Institute of Technology (KIT) and TU Wien (TUW). Extensive experimental work was carried out and a first comprehensive modelling approach was developed to predict deposit formation. In this follow-on project, a transfer of the methods to more complex scenarios was performed. The impingement of AdBlue droplets on rough, porous and coated surfaces was investigated, and improved regime maps, including a new regime transition methodology, were developed. The accumulation of liquid film at different obstacles like welding seams and the stripping of secondary droplets was studied and simulated. Experimental data on fluid viscosity, surface tension and contact angle were created and implemented in the liquid film models of the CFD simulation. A revised reaction mechanism for urea decomposition and improved film phase models for describing the gaseous, liquid and solid species were integrated into the simulations. The developed comprehensive modelling approach includes the latest models from both projects that were carried out. For a demonstration of the capabilities, the software StarCCM+ was used, but in general, the models can be implemented in every CFD software. The objective of the research project was achieved.