The objective of this study is to discern and characterize the factors contributing to the variability in Particulate Matter (PM) mass emissions from current model year diesel and natural gas engines observed during in-use chassis dynamometer testing and compare it with an alternate method of mass measurement involving instantaneous particle size distribution and number count. The study involves the analysis of engine and chassis dynamometer data collected from different engine technologies and chassis dynamometer test cycles in order compare the variability in gravimetric PM mass as well as the PM mass estimated by number-concentration. PM and oxides of nitrogen (NOx) have been the two most stringently regulated emissions constituents from heavy-duty diesel engines. Current US-EPA 2010 PM regulations, set at 0.01 g/bhp-hr, have forced manufacturers to implement the use of diesel particulate filters (DPF) in order to comply with the regulations. The use of DPF's has resulted in PM mass decreasing by orders of magnitude since 2004, hence laboratory measurement techniques and instrumentation to accurately quantify the true mass of PM emitted from such engines have been posed with a challenge. Also, the widely gained acceptance of heavy-duty natural gas engines, characterized by their low soot combustion, has inevitably resulted in significant measurement variability due to the high volatile organic content in the exhaust. Particulate matter mass comparisons are performed between PM sampled using the regulated "CFR 1065 Methodology" and PM number count measurements performed with similar exhaust dilution conditions as employed for gravimetric PM sampling. Particle size distribution and number concentration measurements were performed using the Engine Exhaust Particle Sizer ® (EEPS) spectrometer for transient engine operation and Scanning Mobility Particle Sizer ® (SMPS) spectrometer for steady-state modal tests performed on an engine dynamometer. Additionally, PM mass comparisons are extended to an in-use study in order to compare Not-To-Exceed (NTE) PM emission limits. Results from this study showed that an "effective-density" based conversion technique correlated well with gravimetric filter mass for pre-2010 engine technologies without aftertreatment systems, in particular DPFs for diesel fueled vehicles. Gravimetric PM measurements from a natural gas engine resulted in a standard deviation of 3.1 mg/bhp-hr in comparison to mass calculated through particle size distributions, which resulted in a standard deviation of 0.36 mg/bhp-hr. The use of particle size based measurements for in-use PM monitoring resulted in better resolution of PM mass during short, valid NTE windows.

Dual-layered Multi-Objective Genetic Algorithms (D-MOGA): A Robust Solution for Modern Engine Development and Calibrations Pragalath Thiruvengadam Padmavathy Heavy-duty (HD) diesel engines are the primary propulsion systems used within the freight transportation sector and are subjected to stringent emissions regulations. The primary objective of this study is to develop a robust calibration technique for HD engine optimization in order to meet current and future regulated emissions standards during certification cycles and vocational activities such as drayage operations. Recently, California-Air Resources Board (C-ARB) has also shown interests in controlling off-certification cycle emissions from vehicles operating in the state of California by funding projects such as the Ultra-Low NOx study by Sharp et. al [1]. Moreover, there is a major push for the complex real-world driving emissions testing protocol as the confirmatory and certification testing procedure in Europe and Asia through the United Nations-Economic Commission for Europe (UN-ECE) and International Organization for Standardization (ISO). This calls for more advanced and innovative approaches to optimize engine operation to meet the regulated certification levels. A robust engine calibration technique was developed using dual-layered multi-objective genetic algorithms (D-MOGA) to determine necessary engine control parameter settings. The study focused on reducing fuel consumption and lowering oxides of nitrogen (NOx) emissions, while simultaneously increasing exhaust temperatures for thermal management of exhaust aftertreatment system. The study also focused on using D-MOGA to develop a calibration routine that simultaneously calibrates engine control parameters for transient certification cycles and