Combined Empirical and 1D Modeling Approach for Exhaust Aftertreatment System for Heavy Duty Diesel Engines

Abstract

The aim of this project was to develop a model for the first part of the exhaust aftertreatment system including diesel oxidation catalyst (DOC) and diesel particulate filter (DPF). The model would then be tested against test cell data to evaluate if it could be used in a virtual test rig. The model consists of two individual models one for DOC and one for DPF respectively, which were later linked together. The software used was GT-SUITE and the models were constructed as 1D single channel ow and tested against test cell data provided by Volvo Penta. The key parameters of investigation were concentration of CO, HC and NO, temperature and pressure drop. The reaction kinetics and properties of the substrate were optimized to get the best fit to data. The activation energy and the pre-exponent multiplier for the reaction rate expressions were optimized based on values for an entire cycle. A major error in the model is due to inconsistent degree of conversion at high temperatures (CO, HC). So at high temperatures the model predicted high conversion (100%) and the sensitivity of the parameters decreased, however, there were also come CO slip in this region which also caused problems in the model. The model should have been optimized towards the last part of the test data, where the conversion is lower than 100%. This was out of the scope for this thesis and therefore the model includes inaccurate optimized parameters for the low temperature region of the cycle. The lower conversion in test cell data might be due to a bend in the inlet pipe causing non-uniform inlet gas ow and non-uniform temperature and a ow of reactant were the model predicts almost full conversion. Another source of inaccuracy was the unknown initial soot loading of the DPF, causing errors in pressure drop simulation. Nevertheless, the models can give a good approximation to what happens in the DOC and DPF, especially when using PLM and NRTC cycles.