Classification and Characterization of Heat Release Rate Traces in Low Temperature Combustion for Optimal Engine Operation

Abstract

<div class="section abstract"><div class="htmlview paragraph">Low temperature combustion (LTC) modes are among the advanced combustion technologies which offer thermal efficiencies comparable to conventional diesel combustion and produce ultra-low NOx and particulate matter (PM) emissions. However, combustion timing control, excessive pressure rise rate and high cyclic variations are the common challenges encountered by the LTC modes. These challenges can be addressed by developing model-based control framework for the LTC engine. In the current study, in-cylinder pressure data for dual-fuel LTC engine operation is analyzed for 636 different operating conditions and the heat release rate (HRR) traces are classified into three distinct classes based on their distinct shapes. These classes are named as Type-1, Type-2 and Type-3, respectively. To this end, HRR traces are analyzed for each class based on start of combustion (CA10), combustion phasing (CA50), burn duration (BD), maximum in-cylinder pressure (P_max), location of peak pressure (<i>θ</i>_Pmax), maximum in-cylinder temperature (T_max), maximum pressure rise rate (MPRR) and coefficient of variation of indicated mean effective pressure (COV_IMEP). 47.5% of the data points in Type-3 and 43.5% of the data points in Type-2 resulted in maximum in-cylinder temperature below 1500K which helps in the prevention of NOx formation. 90% of the data points in Type-1 showed COV_IMEP below 5%. All the data points in Type-3 resulted in MPRR below 8 bar/CAD. 80.3% of the data points in Type-2 resulted in indicated thermal efficiency above 35%. This analysis is used as a basis to develop machine learning classification algorithms for model-based control and optimization of LTC engine.</div></div>