Comparison of filter smoke number and elemental carbon mass from partially premixed low temperature combustion in a direct-injection diesel engine

Partially premixed low temperature combustion (LTC) is an established advanced engine strategy that enables the simultaneous reduction of soot and _NOx emissions in diesel engines. Measuring extremely low levels of soot emissions achievable with LTC modes using a filter smoke meter requires large sample volumes and repeated measurements to achieve the desired data precision and accuracy. Even taking such measures, doubt exists as to whether filter smoke number (FSN) accurately represents the actual smoke emissions emitted from such low soot conditions. The use of alternative fuels such as biodiesel also compounds efforts to accurately report soot emissions since the reflectivity of high levels of organic matter found on the particulate matter collected may result in erroneous readings from the optical detector. Using FSN, it is desired to report mass emissions of soot using empirical correlations derived for use with petroleum diesel fuels and conventional modes of combustion. The work presented in this paper compares the experimental results of well known formulas for calculating the mass of soot using FSN and the elemental carbon mass using thermal optical analysis (TOA) over a range of operating conditions and fuels from a four-cylinder direct-injection passenger car diesel engine. The data show that the mass of soot emitted by the engine can be accurately predicted with the smoke meter method utilizing a 3000 ml sample volume over a range of FSN from 0.02 to 1.5. Soot mass exhaust concentration calculated from FSN using the best of the literature expressions and that from TOA taken over all conditions correlated linearly with a slope of 0.99 and R² value of 0.94. A primary implication of the work is that the level of confidence in reporting the soot mass based on FSN for low soot formation regimes such as LTC is improved for both petroleum diesel and biodiesel fuels.