Advanced diesel engine architectures employing flexible valve trains enable emissions reductions and fuel economy improvements. Flexibility in the valve train allows engine designers to optimize the gas exchange process in a manner similar to how common rail fuel injection systems enable optimization of the fuel injection process. Modulating valve timings directly impacts the volumetric efficiency of the engine since it directly controls how much mass is trapped in the cylinders. In fact, it will be shown that the control authority of valve timing modulation over volumetric efficiency, that is, the range of volumetric efficiencies achievable due to modulation of the valve timing, is three times larger than the range achievable by modulation of other engine actuators such as the exhaust gas recirculation valve or the variable geometry turbocharger. Traditional empirical or regression-based models for volumetric efficiency, while suitable for conventional valve trains, are therefore challenged by flexible valve trains. The added complexity and additional empirical data needed for wide valve timing ranges limit the usefulness of these methods. A simple physically based volumetric efficiency model was developed to address these challenges. The model captures the major physical processes occurring over the intake stroke, and is applicable to both conventional and flexible intake valve trains. The model inputs include temperature and pressure in both the intake and exhaust manifolds, intake and exhaust valve event timings, engine cylinder bore, stroke, connecting rod lengths, engine speed, and effective compression ratio. The model is physically based, requires no regression tuning parameters, is generalizable to other engine platforms, and has been experimentally validated using an advanced multi-cylinder diesel engine equipped with a fully flexible variable intake valve actuation system. Experimental data were collected over a wide range of the operating space of the engine and augmented with air handling actuator and intake valve timing sweeps to maximize the range of conditions used to thoroughly experimentally validate the model for a total of 286 operating conditions. The physically based volumetric efficiency model will be shown to predict the experimentally calculated volumetric efficiency to within 5 per cent for all cases with a root mean square error of less than 2.5 per cent for the entire dataset. The physical model developed differs from previous physical modelling work through the novel application of effective compression ratio, incorporation of no tuning parameters, and extensive validation on a unique engine test bed with fully flexible intake valve actuation.

This paper describes a simple, analytical, controloriented and physically-based model for prediction of combustion timing during PCCI combustion. The model includes direct dependence on in-cylinder temperature, in-cylinder pressure, and the total in-cylinder O 2 mass fraction. It is extensively validated with experimental PCCI data from a multi-cylinder engine with almost exclusively stock hardware (stock pistons, injectors/nozzles, turbocharger, etc.) and variable valve actuation. The results show that across a wide range of input conditions the model predicts the start of combustion (SOC) within ±2 • CA of the experimental values for all but one of the 119 data points. The experimental SOC ranges from as early as −19.3 • CA to as late as +0.6 • CA by heavily exercising the control authority over SOC provided by SOI ecm , IVC/ECR modulation (ECR ranges from 12:1 to 18:1), and the engine's air-handling system. This PCCI combustion timing model can be coupled with a gas exchange model for control algorithm design.

Advanced combustion strategies including premixed charge compression ignition, homogeneous charge compression ignition, and lifted flame combustion are promising approaches for meeting increasingly stringent emissions regulations and improving fuel efficiency in next generation powertrains. Variable valve actuation and closed-loop control promise to play a key role in the promotion and control of these advanced combustion modes. For example, modulation of intake valve closure timing dictates the effective compression ratio and influences the total amount of charge trapped inside the cylinder, and in so doing allows manipulation of the in-cylinder reactant concentrations and temperature prior to and during the combustion process. The effort described here uses data from, and an experimentallyvalidated simulation model for, a multi-cylinder engine with variable geometry turbocharging, cooled exhaust gas recirculation, and fully flexible variable valve actuation. This effort's intent is to determine the control authority over the gas exchange process and effective compression ratio when intake valve closure timing modulation is included on a modern turbocharged diesel engine, as well as to lay the groundwork for closed-loop control design for the promotion and control of advanced combustion modes. The engine testbed at Purdue provides a unique opportunity to pursue these objectives for turbocharged engines with exhaust gas recirculation, as it is the only such engine system in academia outfitted with multi-cylinder fully-flexible valve actuation. A method to estimate in-cylinder temperature at top dead centre is also described. Candidate control architectures for both steady state and transient operation are introduced.

Modulation of the effective compression ratio, a measure of the amount of compression of in-cylinder gases above intake manifold conditions, is a key enabler of advanced combustion strategies aimed at reducing emissions while maintaining efficiency, and is directly influenced by modulation of intake valve closing time. To date, the effective compression ratio has most commonly been calculated from in-cylinder pressure data, requiring reliable in-cylinder pressure sensors. These sensors are generally not found on production engines, and thus a method is needed to determine effective compression ratio without in-cylinder pressure data. The work presented here outlines an estimation scheme that combines a high-gain observer with a physically-based volumetric efficiency model to estimate effective compression ratio using only information available from stock engine sensors, including manifold pressures and temperatures and air flows. The estimation scheme is compared to experimental engine data from a unique multi-cylinder diesel engine test bed with flexible intake valve actuation. The effective compression ratio estimator was tested transiently at five engine operating points and demonstrates convergence within three engine cycles after a transient event has occurred, and exhibits steady-state errors of less than 3%.

No abstract