Review and Validation of Models of Pressure Drop Across Diesel Particulate Filter and Particulate Loading Quantity 2016-01-2280

Diesel particulate filters (DPF) regeneration is required to remove accumulated particulate matter in DPF. High pressure drop across DPF triggers an active regeneration by the ECU to burn off the accumulated soot. In city driving such as in a small island as Malta, exhaust gas temperatures are not high enough for passive regenerations, and ECU active regeneration might fail due to short trips. The particulate loading quantity in DPF is beneficial information as it provides an estimate of the remaining mileage expectancy of the DPF. Many vehicles provide information on particulate filter loading quantity in the OBD data. However, since this parameter is not on the mandatory list, different manufacturers provide this loading parameter in different forms, e.g.: grams; percentage (%); remaining mileage; etc. Thus comparison of the loading quantity across different manufacturers is not straightforward. This research reviews delta-p models found in the literature and evaluates them to note their numerical performance. It was found that pressure drop models by different authors gave results that varied by more than an order of magnitude between them. The models found in the literature were then benchmarked against experimental. Two different vehicles with DPFs were assessed through the data available from the OBD socket. Experimental data obtained from the field loading of the DPF’s is presented. The real pressure drop (from OBD data) is compared to that calculated from the models in the literature. The OBD data, namely delta-p, air mass flow, soot loading, exhaust gas temperature, manifold absolute pressure, rpm and more were acquired through the implementation of a smartphone application. The comparison of the experimental data to the models showed that there was good overall agreement with the model by Konstandopoulos & Johnson which includes the Forchheimer effects and contraction/expansion inertial losses. It is noted that this is not the most recent model available in the literature but in fact dates back to 2002.It was however determined that the contraction/expansion inertial losses coefficient needed to be increased with the soot loading quantity to properly capture the curvature of the experimentally acquired delta-p versus mass flow rate curves.