In the last part, the effects of fuel injection system parameters on the performance and emissions characteristics of an RCCI engine are discussed. The injection system parameters include Premixed Ratio (PR), injection pressure, Start of Injection (SOI) timing and spray angle. The CFD model is then used to suggest an injection strategy capable of achieving optimized RCCI engine operation.

Abstract : Low Temperature Combustion (LTC) is a combustion strategy that burns fuel at lower temperatures and leaner mixtures in order to achieve high efficiency and near zero NOx emissions. Since the combustion happens at lower temperatures it inhibits the formation of NOx and soot emissions. One such strategy is Reactivity Controlled Compression Ignition (RCCI). One characteristic of RCCI combustion and LTC com- bustion in general is short burn durations which leads to high Pressure Rise Rates (PRR). This limits the operation of these engines to lower loads as at high loads, the Maximum Pressure Rise Rate (MPRR) hinders the use of this combustion strategy. This thesis focuses on the development of a model based controller that can control the Crank Angle for 50% mass fraction burn (CA50) and Indicated Mean Effective Pressure (IMEP) of an RCCI engine while limiting the MPRR to a pre determined limit. A Control Oriented Model (COM) is developed to predict the MPRR in an RCCI engine. This COM is then validated against experimental data. A statistical analysis of the experimental data is conducted to understand the accuracy of the COM. The results show that the COM is able to predict the MPRR with reasonable accuracy in steady state and transient conditions. Also, the COM is able to capture the trends during transient operation. This COM is then included in an existing cycle by cycle dynamic RCCI engine model and used to develop a Linear Parameter Varying (LPV) representation of an RCCI engine using Data Driven Modeling (DDM) approach with Support Vector Machines (SVM). This LPV representation is then used along with a Model Predictive Controller (MPC) to control the CA50 and IMEP of the RCCI engine model while limiting the MPRR. The controller was able to track the desired CA50 and IMEP with a mean error of 0.9 CAD and 4.7 KPa respectively while maintaining the MPRR below 5.8 bar/CAD.

Premixed Compression Ignition (PCI) strategies are promising methods to achieve low engine out NOx and soot emissions and high efficiency. However, PCI strategies have failed to see widespread implementation due to difficulties controlling the heat release rate and lack of an adequate combustion phasing control mechanism. In this research, a dual fuel reactivity controlled compression ignition (RCCI) concept is proposed to address these issues. In the RCCI strategy, two fuels with different auto ignition characteristics are blended inside the combustion chamber. Combustion phasing is controlled by the relative ratios of these two fuels and the combustion duration is controlled by spatial stratification between the two fuels. The study has three primary sections. The first section highlights the development of the RCCI strategy using computational fluid dynamics (CFD) modeling. The second section uses CFD modeling and metal engine experiments to evaluate the performance and emissions characteristics of RCCI combustion. The metal engine experiments confirm that RCCI operation is possible over a wide range of conditions with near zero levels of NOx and soot emissions. Additionally, it is found that RCCI is able to achieve very high indicated efficiency (greater than 50%) by lowering heat transfer losses and improving the control over the combustion phasing and burn duration. The third section uses optical engine experiments to validate model predictions and provide a fundamental explanation for the processes controlling RCCI combustion. The results of the optical engine experiments clarify the mechanisms controlling the RCCI energy release. Chemiluminescence imaging shows that RCCI features a reaction zone that appears to grow by the appearance of small auto ignition pockets. The fuel tracer fluorescence imaging shows that the ignition locations correspond to the regions with the lowest primary reference fuel (PRF) number and highest equivalence ratio. The rate of reaction zone growth is then controlled by the level of stratification in equivalence ratio and PRF number. Kinetics modeling based on the fuel tracer fluorescence imaging shows that the PRF number has the largest effect on the rate of reaction zone growth.

Abstract : Reactivity controlled compression ignition (RCCI) is a combustion strategy that offers high fuel conversion efficiency and near zero emissions of NOx and soot which can help in improving fuel economy in mobile and stationary internal combustion engine (ICE) applications and at the same time lower engine-out emissions. One of the main challenges associated with RCCI combustion is the difficulty in simultaneously controlling combustion phasing, engine load, and cyclic variability during transient engine operations. This thesis focuses on developing model based controllers for cycle-to-cycle combustion phasing and load control during transient operations. A control oriented model (COM) is developed by using mean value models to predict start of combustion (SOC) and crank angle of 50% mass fraction burn (CA50). The COM is validated using transient data from an experimental RCCI engine. The validation results show that the COM is able to capture the experimental trends in CA50 and indicated mean effective pressure (IMEP). The COM is then used to develop a linear quadratic integral (LQI) controller and model predictive controllers (MPC). Premixed ratio (PR) and start of injection (SOI) are the control variables used to control CA50, while the total fuel quantity (FQ) is the engine variable used to control load. The selection between PR and SOI is done using a sensitivity based algorithm. Experimental validation results for reference tracking using LQI and MPC show that the desired CA50 and IMEP can be attained in a single cycle during step-up and step-down transients and yield an average error of less than 1.6 crank angle degrees (CAD) in the CA50 and less than 35 kPa in the IMEP. This thesis presents the first study in the literature to design and implement LQI and MPC combustion controllers for RCCI engines.

Abstract : Low Temperature Combustion (LTC) has got widespread attention over the past two decades in the field of Automotive Research and Development due to it's potential for achieving higher efficiencies with near-zero engine out NOx and soot emissions. Among all the LTC strategies Reactivity controlled compression ignition (RCCI) has shown the most promising results due to it's precise control over combustion phasing and heat release rate. However, RCCI being a dual-fuel stratified combustion, precise control over the injection timing of direct injected fuel and in-cylinder fuel reactivity of the mixture needs to be controlled effectively in order to achieve gross indicated thermal efficiencies as high as around 60%. This thesis focuses on developing real-time, model-based controller for controlling combustion phasing of an RCCI Engine. Optimum combustion phasing can be achieved by varying mixture reactivity and injection timing of higher reactive fuel. An experimental study was performed to study the effects of these variables on combustion phasing. Next,a mean-value and dynamic control-oriented model (COM) was developed to predict combustion phasing during steady-state and transient operating conditions. The validation results have shown that the COM was able to capture the experimental trends with minimal error. Next, for implementing in real time, a PI controller was developed using the COM to track the desired combustion phasing by adjusting duel-fuel premixed ratio and start of injection timing. The PI controller is then implemented on the engine plant. The validation results proved that the designed controller can follow the desired combustion phasing with an average error of 2 crank angle degrees and rise time of 3 engine cycles.

A computational investigation of methods to extend the upper limit of power output of reactivity controlled compression ignition (RCCI) engines was performed. The study utilized two approaches. The first approach is to increase the engine speed while maintaining a medium load. The second approach is to operate at higher loads without changing the engine speed. Iso-octane and n-heptane were used to represent the low-reactivity fuel and high-reactivity fuel, respectively. A light-duty diesel engine was modeled for the high speed dual-fuel RCCI combustion study. With high-speed operation several benefits were identified. Firstly, the peak pressure rise rates (PPRR), both crank angle-based and time-based, were reduced compared to those with low-speed operation. Secondly, at high speed the NO formation residence time became short, leading to reduced NOx emissions. Lastly, a frictional penalty analysis of high-speed operation using the Chen-Flynn model was conducted, which showed only 0.5 bar FMEP increase compared to that at low-speed. These findings indicate that high-speed RCCI is a very promising path for high-power output operation. For the high-load operation study use of dual direct-injectors was explored in order to direct-inject both fuels. Analysis of the optimum injection strategy revealed two main physical mechanisms enabling high-load operation with dual direct-injectors. The first exploited local evaporative cooling from the iso-octane injection, which delayed the iso-octane ignition. The second mechanism was related to the shorter chemical residence time of the iso-octane due to its late delivery into the cylinder. It was also noted that n-heptane's role as an ignition source could not be achieved with just iso-octane. Finally, the co-axial injector location assumption was removed by using an actual dual-injector layout. Unlike results with the co-axial injector design, the actual dual-injector layout exhibited soot and CO emission problems. In order to attempt to accommodate off-center injector locations, various injector hole patterns were tested. Although these unconventional injector hole patterns improved the emissions, it is concluded that the development of a co-axial dual-fuel injector is imperative in order to achieve clean RCCI combustion at high load.

An experimental investigation of the pragmatic limits of Reactivity Controlled Compression Ignition (RCCI) engine efficiency was performed. The study utilized engine experiments combined with zero-dimensional modeling. Initially, simulations were used to suggest conditions of high engine efficiency with RCCI. Preliminary simulations suggested that high efficiency could be obtained by using a very dilute charge with a high compression ratio. Moreover, the preliminary simulations further suggested that with simultaneous 50% reductions in heat transfer and incomplete combustion, 60% gross thermal efficiency may be achievable with RCCI. Following the initial simulations, experiments to investigate the combustion process, fuel effects, and methods to reduce heat transfer and incomplete combustion reduction were conducted. The results demonstrated that the engine cycle and combustion process are linked, and if high efficiency is to be had, then the combustion event must be tailored to the initial cycle conditions. It was found that reductions to engine heat transfer are a key enabler to increasing engine efficiency. In addition, it was found that the piston oil jet gallery cooling in RCCI may be unnecessary, as it had a negative impact on efficiency. Without piston oil gallery cooling, it was found that RCCI was nearly adiabatic, achieving 95% of the theoretical maximum cycle efficiency (air standard Otto cycle efficiency).