Real-time on-board indirect light-off temperature estimation as a detection technique of diesel oxidation catalyst effectiveness level
Abstract
The latest US emission regulations require dramatic reductions in Nitrogen Oxide (NOx) emissions. Selective Catalyst Reduction (SCR) is the current technology that achieves NOx reductions of up to 90%. It is typically mounted downstream of the existing aftertreatment system, i.e, after the Diesel Oxidation Catalyst (DOC) and Diesel Particulate Filter (DPF). Accurate prediction of input NO2:NO ratio is needed for the SCR urea injection control algorithm to reduce NOx output and NH3 slippage downstream of the SCR catalyst. Most oxidation of NO to NO2 occurs in the DOC since its main function is to oxidize emission constituents. The DOC thus determines the NO2:NO ratio as feedgas to the SCR catalyst. The prediction of NO2:NO ratio varies as the catalyst in the DOC ages or deteriorates due to poisoning. Therefore, the prediction cannot be determined by a single model. Instead, the model should take into account the correlation of DOC conversion effectiveness and the aging of the catalyst. This research project is aimed at detecting the aging level in the DOC in real time on board the vehicle by estimating light-off temperature. Initial light-off analysis was conducted on catalyzed DPF (CDPF) with the instrumentation of 12 thermocouples. An improvement on the existing light-off temperature detection equation provided by Cummins was done by adding substrate heat rate storage term. This finding was then implemented on the instrumented DOC for light-off temperature analysis with testing to simulate transient exhaust temperature. Evaluation of the equation terms showed that heat loss analysis to the surroundings is negligible while the post-fueling to the exhaust stream was found to have a significant discrepancy to the commanded quantity. A comparison analysis for the fresh and aged DOC was conducted. Clear light off temperature shifts were detected between both and the indirect estimation could roughly predict the aged DOC conversion. However, the technology is still preliminary and further work on data repeatability and consistency needs to be done before the technology can be adopted for real world application.
Degree
M.S.M.E.
Advisors
Meckl, Purdue University.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.