[Back]

C. Ferris Roig, M. Zöchbauer:
"Validation of a numerical spray-wall interaction model against data from literature and application to automotive SCR systems";
Report No. B13015, 2013; 81 pages.

As time goes by, legislation regarding to internal combustion engine emissions is becoming stricter. The nitric oxides present in diesel engine exhaust gases are one example of this. The most promising system to reduce its content before releasing the hot gases to the atmosphere is the Selective Catalytic Reduction (SCR).
The goal of the present work is to capture and model the main processes related with the droplet-wall interaction phenomenon that takes place inside the exhaust line when the Urea-Water-Solution (UWS) is injected. To do so, a commercial CFD code is used to compare different droplet-wall impingement models and assess their
suitability for SCR simulations.
Multiple CFD simulations have been run in a simplified geometry that pretends to reproduce the different impingement situations that can be found in SCR systems.
Three models have been compared and their suitability for the simulations of UWS injection inside the exhaust system has been assessed. These three models are available in the different versions of the CFD software Star-CD from CD-adapco: the Bai model, the Bai-ONERA model and the Senda model.
Besides that, three sensitivity studies have been carried out in order to derive a guide for future simulations in actual SCR system geometries. As a first step, the influence of the solid region discretization over the deposition boundary and the cooling of the
wall because of droplet impingement have been investigated. For this purpose, the solid region has been divided into a variable number of layers. The thickness of these layers is governed by a stretching factor.
Afterwards, the role of the injection-rate and the time step on the wall cooling due to droplet impingement has been studied. For the first, the wall temperature at which the deposition boundary is located has been evaluated for different meshes and two
injection-ratios. For the second, a solid and shell temperature comparison is carried out for three different time step values.
Finally, the chosen model has been used to simulate the impingement of a normal-tothe-wall spray. Then, the outcome of these investigations has been compared against literature data by visual means.