Parametric and sensitivity analysis of diesel particulate filter regeneration

Particulate matter emissions from diesel-fueled cars, trucks, and buses are regulated by state and federal government agencies. Although improvements in engine performance have taken place, aftertreatment is necessary to meet the existing emissions standards. In order for a particulate emissions control system to function under real driving conditions, real-time model predictive control is needed. As a first step towards this goal, we perform a parametric study to compare two existing models of diesel particulate trap regeneration: a one-dimensional plus time model that tracks gas and solid temperatures and the particulate deposit thickness developed by Bissett (Chemical Engineering Science, 1984, 39, 1233-1244) and a one-dimensional plus time model that uses averaging theory to focus solely on the thermal evolution in the diesel particulate trap developed by Zheng and Keith (AIChE Journal, 2007, 53, 1316-1324). We use both models to predict the ignition time and ignition location within the diesel particulate trap. In this parametric study, three operating parameters are varied: the initial deposit thickness, the exhaust gas flowrate, and the exhaust gas temperature. It was found that the Zheng and Keith model agrees very well with the Bissett model under most operating conditions except when the initial deposit thickness and exhaust gas temperatures are low and the gas flow rate is high. These studies suggest the Zheng and Keith model may be appropriate for real time control of diesel particulate filter regeneration. We then perform a sensitivity analysis of the ignition time and ignition length to changes in the initial deposit thickness and gas flowrate under various conditions using the averaged model. We have found that the ignition time and ignition length are most sensitive to changes in deposit thickness when the deposit thickness and exhaust temperature is low. Also, the ignition time is relatively insensitive to changes in gas flow rate, but the ignition length is most sensitive to changes in gas flow rate at low exhaust temperatures. These studies are useful towards the ultimate development of a predictive particulate emissions control system for diesel-fueled vehicles. © 2009 The Berkeley Electronic Press. All rights reserved