Search Results

On-board diagnostics (OBD) data contain valuable information including real-world measurements of vehicle powertrain parameters. These data can be used to gain a richer data-driven understanding of complex physical phenomena like emissions formation during combustion. In this study, we develop a physics-based machine learning framework to predict and analyze trends in engine-out NOx emissions from diesel and diesel-hybrid heavy-duty vehicles. This model differs from black-box machine learning models presented in previous literature because it incorporates engine combustion parameters that allow physical interpretation of the results. Based on chemical kinetics and the characteristics of diffusive combustion, NOx emissions from compression ignition engines primarily depend non-linearly on three parameters: adiabatic flame temperature, the oxygen concentration in the cylinder when the intake valves are closed, and combustion time duration.

A telemetry linkage system has been developed for piston temperature measurements in a direct-injection diesel engine. In parallel with the development of the telemetry linkage system, fast response thermocouples were installed at three piston locations - two on the bowl surface and one on the crown surface. A novel design was used to achieve electrical continuity between the piston and the connecting rod by means of a flexible steel strap pivoted on the piston skirt. The telemetry linkage system was then used to transport the electrical wires from the thermocouples to the external data acquisition system. A series of tests was run to determine the effects of location and load on piston surface temperatures. Surface temperature profiles varied substantially among the three locations, reflecting the differences in the combustion and heat flow characteristics of their surrounding regions.

A prototype chromel-alumel overlapping thin-film thermocouple (TFTC) has been developed for transient heat transfer measurements in ceramic-coated combustion chambers. The TFTC has been evaluated using various metallurgical techniques such as scanning electron microscopy, energy dispersive x-ray detection, and Auger electron spectroscopy. The sensor was calibrated against a standard thermocouple in ice, boiling water, and a furnace at 1000°C. The microstructural and chemical analysis of the thin-films showed the alumel film composition was very similar to the bulk material, while the chromel film varied slightly. An initial set of ceramic plug surface temperatures was taken while motoring and firing the engine at 1900 rpm to verify thermocouple operation. The data shows a 613 K mean temperature and a 55 K swing for the ceramic surface compared with a 493 K mean temperature and a 20 K swing for the metal surface at the same location.

Increasing functionality of electronic control units has enhanced our ability to control engine operation utilizing calibration static maps that provide the values of several controllable variables. State-of-the-art simulation-based calibration methods permit the development of these maps with respect to extensive steady-state and limited transient operation of particular driving cycles. However, each individual driving style is different and rarely meets those test conditions. An alternative approach was recently implemented that considers the derivation of these maps while the engine is running the vehicle. In this approach, a self-learning controller selects in real time the optimum values of the controllable variables for the sequences of engine operating point transitions, corresponding to the driver's driving style.

The number of control actuators available on spark-ignition engines is rapidly increasing to meet demand for improved fuel economy and reduced exhaust emissions. The added complexity greatly complicates control strategy development because there can be a wide range of potential actuator settings at each engine operating condition, and map-based actuator calibration becomes challenging as the number of control degrees of freedom expand significantly. Many engine actuators, such as variable valve actuation and flow control valves, directly influence in-cylinder combustion through changes in gas exchange, mixture preparation, and charge motion. The addition of these types of actuators makes it difficult to predict the influences of individual actuator positioning on in-cylinder combustion without substantial experimental complexity.

Variable geometry turbines (VGT) are of particular interest to advanced diesel powertrains for future conventional trucks, since they can dramatically improve system transient response to sudden changes in speed and load, characteristic of automotive applications. VGT systems are also viewed as the key enabler for the application of the EGR system for reduction of heavy-duty diesel emissions. This paper applies an artificial neural network methodology to VGT modeling in order to enable representation of the VGT characteristics for any blade (nozzle) position. Following validation of the ANN model of the baseline, fixed geometry turbine, the VGT model is integrated with the diesel engine system. The latter is linked to the driveline and the vehicle dynamics module to form a complete, high-fidelity vehicle simulation.

A comprehensive model for sprays emerging from high-pressure swirl injectors in DISI engines has been developed accounting for both primary and secondary atomization. The model considers the transient behavior of the pre-spray and the steady-state behavior of the main spray. The pre-spray modeling is based on an empirical solid cone approach with varying cone angle. The main spray modeling is based on the Liquid Instability Sheet Atomization (LISA) approach, which is extended here to include the effects of swirl. Mie Scattering, LIF, PIV and Laser Droplet Size Analyzer techniques have been used to produce a set of experimental data for model validation. Both qualitative comparisons of the evolution of the spray structure, as well as quantitative comparisons of spray tip penetration and droplet sizes have been made. It is concluded that the model compares favorably with data under atmospheric conditions.

The development and use of a simulation of an Integrated Starter Alternator (ISA) for a High Mobility Multi-purpose Wheeled Vehicle (HMMWV) is presented here. While the primary purpose of an ISA is to provide electric power for additional accessories, it can also be utilized for mild hybridization of the powertrain. In order to explore ISA's potential for improving HMMWV's fuel economy, an ISA model capable of both producing and absorbing mechanical power has been developed in Simulink. Based on the driver's power request and the State of Charge of the battery (SOC), the power management algorithm determines whether the ISA should contribute power to, or absorb power from the crankshaft. The system is also capable of capturing some of the braking energy and using it to charge the battery. The ISA model and the power management algorithm have been integrated in the Vehicle-Engine SIMulation (VESIM), a SIMULINK-based vehicle model previously developed at the University of Michigan.

This paper describes a test cell setup for concurrent running of a real engine and a vehicle system simulation, and its use for evaluating engine performance when integrated with a conventional and a hybrid electric driveline/vehicle. This engine-in-the-loop (EIL) system uses fast instruments and emission analyzers to investigate how critical in-vehicle transients affect engine system response and transient emissions. Main enablers of the work include the highly dynamic AC electric dynamometer with the accompanying computerized control system and the computationally efficient simulation of the driveline/vehicle system. The latter is developed through systematic energy-based proper modeling that tailors the virtual model to capture critical powertrain transients while running in real time. Coupling the real engine with the virtual driveline/vehicle offers a chance to easily modify vehicle parameters, and even study two different powertrain configurations.

The emerging Variable Valve Timing (VVT) technology complicates the estimation of air flow rate because both intake and exhaust valve timings significantly affect engine's gas exchange and air flow rate. In this paper, we propose to use Artificial Neural Networks (ANN) to model the air flow rate through a 2.4 liter VVT engine with independent intake and exhaust camshaft phasers. The procedure for selecting the network architecture and size is combined with the appropriate training methodology to maximize accuracy and prevent overfitting. After completing the ANN training based on a large set of dynamometer test data, the multi-layer feedforward network demonstrates the ability to represent air flow rate accurately over a wide range of operating conditions. The ANN model is implemented in a vehicle with the same 2.4 L engine using a Rapid Prototype Controller.

A regenerative braking model for a parallel Hybrid Electric Vehicle (HEV) is developed in this work. This model computes the line and pad pressures for the front and rear brakes, the amount of generator use depending on the state of deceleration (i.e. the brake pedal position), and includes a wheel lock-up avoidance algorithm. The regenerative braking model has been developed in the symbolic programming environment of MATLAB/SIMULINK/STATEFLOW for downloadability to an actual HEV's control system. The regenerative braking model has been incorporated in NREL's HEV system simulation called ADVISOR. Code modules that have been changed to implement the new regenerative model are described. Resulting outputs are compared to the baseline regenerative braking model in the parent code. The behavior of the HEV system (battery state of charge, overall fuel economy, and emissions characteristics) with the baseline and the proposed regenerative braking strategy are first compared.

To achieve full automation in self-driving vehicles, environmental perception sensing accuracy is critically important. However, ambient particles in adverse weather like foggy, rainy, or snowy conditions can significantly scatter the incident laser beam, and therefore contaminate the intensity and accuracy of light detection and ranging (LiDAR) sensors. Especially compared to the rapidity of technology development in self-driving vehicles, there is a significant lack of documented research on LiDAR systems with wavelength longer than 1 μm for application in Advanced Driver-Assistance Systems. In this work, experimental studies were performed with a state-of-the-art 1.55 μm wavelength automotive-grade LiDAR system in a controlled laboratory fog chamber. The goal of the research is to correlate laser attenuation and the optical properties of fog particles.

Calibration of automotive engines to ensure compliance with emission regulations is a critical phase in product development. Control of engine-out particulate emissions, which directly impact the environment and public health, is particularly important. Detailed physics-based models are typically used to gain a rich understanding of the complex physical phenomena that drive the soot particle formation in an engine cylinder. However, such models often fail to correctly represent the highly dynamic nature of the underlying mechanisms under transient combustion conditions. Moreover, most physics-based models were initially developed for diesel engine applications and their applicability to gasoline engines remains questionable due to differences in flame structure and fuel-wall interactions. Black-box models have been previously proposed to predict engine-out soot emissions, but their lack of physical interpretability is an unsolved drawback.

Wall wetting of injected fuel onto the intake manifold and cylinder wall causes unpredictable transient behavior of air-fuel mixing which results in a significant emission of unburned hydrocarbon (HC) emission during cold start operation. Heated exhaust gas oxygen (HEGO) sensors cannot measure the air-fuel ratio (A/F) of exhaust gas during cold start condition. Precise and fast estimation of air/fuel ratio of the exhaust gas is required to elucidate the wall wetting phenomena and subsequent HC formation. Refined A/F estimation can enable the control of fuel injection minimizing HC emissions during cold start conditions so that HC emissions can be minimized. A new estimator for A/F of the exhaust gas has been developed. The A/F estimator described in this study utilizes measured exhaust gas temperature and general engine parameters such as engine speed, airflow, coolant temperature, etc.