Zoran S. Filipi

Effects of refinery stream gasoline property variation on the auto-ignition quality of a fuel and homogeneous charge compression ignition combustion

The combination of in-cylinder thermal environment and fuel ignition properties plays a critical role in the homogeneous charge compression ignition engine combustion process. The properties of fuels available in the automotive market vary considerably and display different auto-ignition behaviors for the same intake charge conditions. Thus, in order for homogeneous charge compression ignition (HCCI) technology to become practically viable, it is necessary to characterize the impact of differences in fuel properties as a source of ignition/combustion variability. To quantify the differences, 15 gasolines composed of blends made from refinery streams were investigated in a single-cylinder homogeneous charge compression ignition engine. The properties of the refinery stream blends were varied according to research octane number, sensitivity (S = research octane number − motor octane number) and volumetric contents of aromatics and olefins. Nine fuels contained 10% ethanol by volume, and six more were blended with 20% ethanol. Pure ethanol (E100) and an un-oxygenated baseline fuel (RD3-87) were included too. For each fuel, a sweep of intake temperature at a consistent load and engine speed was conducted, and the combustion phasing given by the crank angle of 50% mass fraction burned was tracked to assess the sensitivity of auto-ignition to fuel chemical kinetics. The experimental results provided a wealth of information for predicting the HCCI combustion phasing from the given properties of a fuel. In this study, the original octane index correlation proposed by Kalghatgi based solely on fuel research octane number and motor octane number was found to be insufficient for characterizing homogeneous charge compression ignition combustion of refinery stream fuels. A new correlation was developed for estimation of auto-ignition properties of practical fuels in the typical HCCI engine. Fuel composition, captured by terms indicating the fraction of aromatics, olefins, saturates and ethanol, was added to generate the following formula: O I JKZ = RON − K ′ · S + κ · ( Aromatic s 2 ) ( Olefins + Saturates ) + ε · ( Aromatics · Ethanol ) . The results indicate a significantly improved estimation of combustion phasing for gasoline fuels of varying chemical composition under low-temperature combustion conditions. Quantitative findings of this investigation and the new octane index correlation can be used for designing robust HCCI control strategies, capable of handling the wide spectrum of fuel chemical compositions found in pump gasoline.

The impact of a magnesium zirconate thermal barrier coating on homogeneous charge compression ignition operational variability and the formation of combustion chamber deposits

The accumulation and burn-off of combustion chamber deposits create uncontrolled shifting of the homogeneous charge compression ignition operability range. This combustion chamber deposit–created operational variability places increased control burden on a multi-mode engine. However, the operational variability can be mitigated by manipulating combustion chamber deposit accumulation. A magnesium zirconate thermal barrier coating was applied to the piston of a homogeneous charge compression ignition engine in an effort to reduce combustion chamber deposit accumulation through elevated piston surface temperatures. While reduced combustion chamber deposit thicknesses were observed on the magnesium zirconate piston periphery, combustion chamber deposit accumulation in the bowl region increased relative to aluminum piston operation. Additionally, combustion chamber deposit thicknesses on the aluminum cylinder head were reduced during operation with the magnesium zirconate coated piston. Chamber-wide alterations to combustion chamber deposit accumulation taken together with the increased burn duration and hydrocarbon emissions measured during operation with the magnesium zirconate piston indicate significant interaction between the directly injected fuel spray and thermal barrier coating porosity. The porosity and surface roughness of the magnesium zirconate thermal barrier coating are speculated to create fuel pooling/absorption within the piston bowl, increasing combustion chamber deposit accumulation in the bowl and leaning the remaining fuel–air charge. The charge leaning lengthens the magnesium zirconate burn duration and reduces cylinder head combustion chamber deposit accumulation. Furthermore, hydrocarbon emissions were increased during magnesium zirconate operation due to late desorption and subsequent incomplete burning of fuel from piston bowl and the presence of incombustibly lean areas in the remaining cylinder charge.

Neuro-fuzzy model tree approach to virtual sensing of transient diesel soot and NOx emissions

Diesel engine combustion and emission formation are highly nonlinear and thus create a challenge related to engine diagnostics and engine control with emission feedback. This article describes the development of neuro-fuzzy models for prediction of transient NOX and soot emission from a diesel engine. The modeling techniques are motivated by the idea of divide and conquer the input–output space. The complex problem is divided into multiple simpler subproblems, which are then identified using simpler class of models. This article explores two different choices of local models, specifically polynomial and neural networks. The modeling technique is augmented with input relevance algorithm to select the most relevant input regressors. Two algorithms, namely, orthogonal least square and automatic relevance determination, are introduced. The models are data driven, and an advanced experimental setup incorporating a medium duty diesel engine and fast emission analyzers for soot and NOX is used to generate training data. The choice of local models and input relevance algorithm is validated with instantaneous emission recorded during transient schedules different from those used in development. High prediction accuracy, both qualitatively and quantitatively, is demonstrated with low computational cost.

Naturalistic drive cycle synthesis for pickup trucks.

PROBLEM Future pick-up trucks are meeting much stricter fuel economy and exhaust emission standards. Design tradeoffs will have to be carefully evaluated to satisfy consumer expectations within the regulatory and cost constraints. Boundary conditions will obviously be critical for decision making: thus, the understanding of how customers are driving in naturalistic settings is indispensable. Federal driving schedules, while critical for certification, do not capture the richness of naturalistic cycles, particularly the aggressive maneuvers that often shape consumer perception of performance. While there are databases with large number of drive cycles, applying all of them directly in the design process is impractical. Therefore, representative drive cycles that capture the essence of the naturalistic driving should be synthesized from naturalistic driving data. METHOD Naturalistic drive cycles are firstly categorized by investigating their micro-trip components, defined as driving activities between successive stops. Micro-trips are expected to characterize underlying local traffic conditions, and separate different driving patterns. Next, the transitions from one vehicle state to another vehicle state in each cycle category are captured with Transition Probability Matrix (TPM). Candidate drive cycles can subsequently be synthesized using Markov Chain based on TPMs for each category. Finally, representative synthetic drive cycles are selected through assessment of significant cycle metrics to identify the ones with smallest errors. SUMMARY This paper provides a framework for synthesis of representative drive cycles from naturalistic driving data, which can subsequently be used for efficient optimization of design or control of pick-up truck powertrains. IMPACT ON INDUSTRY Manufacturers will benefit from representative drive cycles in several aspects, including quick assessments of vehicle performance and energy consumption in simulations, component sizing and design, optimization of control strategies, and vehicle testing under real-world conditions. This is in contrast to using federal certification test cycles, which were never intended to capture pickup truck segment.

Control-oriented model-based ignition timing prediction for high-degrees-of-freedom spark ignition engines

The pressure to improve automotive fuel economy and emissions has driven the introduction of more complex spark ignition engines. As the number of control actuators increases, traditional ignition timing calibration and control methods become restrictive, creating a need for new feedforward approaches to manage transient operation. This research was conducted to determine whether a control-oriented turbulent flame entrainment model could be developed to predict the ignition timing of an engine with a large number of control actuators. A physics-based approach is used to capture the influence of additional control actuators on the complex interactions affecting the ignition timing. Each actuator is characterized by its influence on the fundamental combustion parameters, such as the residual gas fraction and the turbulence intensity. Experimental results are used to generate semiempirical input and combustion models that are capable of running in real time within an engine controller. The model is used to predict the combustion duration, from the spark to 50% mass fraction burned, at every crank angle position within a reasonable ignition timing window for each engine operating point. With minimal engine mapping, the model was capable of predicting the spark timing to within several degrees of ideal values, demonstrating the feasibility of this approach for use in high-degrees-of-freedom spark ignition engines.

Real-Time Transient Soot and NO

Diesel engine combustion and emission formation is highly nonlinear and thus creates a challenge related to engine diagnostics and engine control with emission feedback. This paper presents a novel methodology to address the challenge and develop virtual sensing models for engine exhaust emission. These models are capable of predicting transient emissions accurately and are computationally efficient for control and optimization studies. The emission models developed in this paper belong to the family of hierarchical models, namely “neuro-fuzzy model tree”. The approach is based on divide-and-conquer strategy i.e. to divide a complex problem into multiple simpler subproblems, which can then be identified using simpler class of models. Advanced experimental setup incorporating a medium duty diesel engine is used to generate training data. Fast emission analyzers for soot and NOX provide instantaneous engine-out emissions. Finally, the Engine-In-the-Loop is used to validate the models for predicting transient particulate mass and NOX .Copyright