Development and Application of a Virtual Soot Loading Sensor
Published May 23, 2004 by Society of Automotive Engineers of Korea in South Korea
The particulate filter is set to become a widespread system to control particulate matter (PM) emissions from diesel-powered engines. Whereas trapping of soot is easy in general - there do exist a multitude of filtering substrates, offering high efficiencies on dry soot - attention has to be paid to the question of the regeneration process including its initiation, which is still not satisfactorily solved.
The soot loading condition and thus the initiation of regeneration process is normally controlled by exhaust gas pressure loss within the filter. However, this measured value characterizes the level of soot loading only incompletely. The accumulated soot mass has to be limited in order to avoid excessive exothermal regeneration which may lead to a destroyed filter substrate. Based on this, regeneration intervals must be shortened due to safety reasons and product durability.
A possible solution and improvement represents the application of a Virtual Soot Loading Sensor. This paper reports on development and application work of this sensor, being currently still under development. The sensor claims the ability to calculate accumulated soot mass during engine operation. The achieved accuracy at this stage of development promises this to become a helpful instrument for advanced DPF regeneration strategies. The Virtual Soot Loading Sensor consists of two separate tools covering the soot loading and the regeneration, as the calculations carried out during soot loading and regeneration phases are basically different. The calculations under soot loading conditions are based on several actual engine data such as pressures, temperatures and air & fuel quantities. The second tool describes the amount of soot loss whenever an oxidation process (also partially) inside the filter takes place.
The results reported in this paper were obtained from two modern passenger car diesel engines at steady state and dynamic operation and show a good conformity between calculated and gravimetrically detected values.