Proceedings of the Institution of Mechanical Engineers,Part D: Journal of Automobile Engineering

Diesel engine transients, in particular the rapid increase of load from idle, lead to significant excursions of operating parameters from normal, steady-state operation. A typical transient event is initiated by a sudden increase of fuelling followed by a turbo lag period. The trajectories of key parameters, such as boost pressure, air/fuel ratio, and amount of residual in the cylinder, depend on the complex interplay between the control system behaviour, actuator response, and time scales of physical processes. The end result is a significant deviation of the combustion event and emission formation. For example, it has been shown that transient spikes of soot may be responsible for up to half of total particulate emissions during a city driving cycle.

The present paper introduces an experimental technique to quantify cycle-resolved pre-combustion in-cylinder constituents during transient engine operation and uses measurements to generate insight into mechanisms that cause increased emissions during a diesel engine transient event. A 6.0-litre V-8 diesel engine is instrumented for fuel injector needle lift and injection pressure measurements, cylinder pressure, simultaneous fast measurements of CO₂ concentrations in the exhaust and in the cylinder, and fast measurements of NO and particulate spectrum. Analytical techniques are introduced to process measurements and determine cycle-resolved mass of fuel, air, and stoichiometric combustion products, thus providing insight into reasons for significant variations of emission during the transient. The technique is utilized to study engine behaviour over an FTP-72 driving cycle.

Control of performance and transient emissions from turbocharged diesel engines is an important objective for automotive manufacturers, since stringent criteria for exhaust emissions must be met. In particular, (cold) starting is of exceptional importance owing to its significant contribution to the overall emissions during a transient test cycle. In the present work, experimental tests were conducted on a turbocharged and after-cooled bus–truck diesel engine in order to investigate the engine operating behaviour and the formation mechanisms of nitric oxide, smoke, and combustion noise during cold, warm, and hot starting. With this as a target, a fully instrumented test bed was set up, using ultra-fast response analysers capable of capturing the instantaneous development of emissions and various key engine and turbocharger parameters. The experimental test pattern included a variety of starting conditions, defined by the thermal status of the engine (i.e. the coolant temperature) and its idling speed. As expected, turbocharger lag was found to be the major contributor for the pollutant emissions spikes in all cases, with the thermal status of the engine and its idling speed playing important roles in the combustion (in)stability, turbocharger response, and noise radiation.

Diesel particulate filters (DPF) are becoming a standard technology in diesel engines because of the need for compliance with forthcoming regulations regarding soot emissions. When a great degree of maturity in management of filtration and regeneration has been attained, the influence of the DPF placement on the engine performance emerges as a key issue to be properly addressed. The novelty of this work leads to the study of an unusual location of an aftertreatment device in the architecture of the turbocharged diesel engine exhaust line. The problem of the pre-turbo DPF placement is tackled comparing the engine response under full-load transient operation as opposed to the traditional DPF location downstream of the turbine. The study has been performed on the basis of a gas dynamic simulation of the engine, which has been validated with experimental data obtained under steady-state and transient conditions. The DPF response has been simulated with a model able to deal with the characteristic highly pulsating flow upstream of the turbine. Several levels of DPF soot loading have been considered to represent fully the most exigent conditions in terms of performance requirements. As a result, the main physical phenomena controlling the engine and DPF response and interaction have been identified. Placing the DPF upstream of the turbine will lead to a number of important advantages, owing to the continuous regeneration mode at which the DPF will operate, the lower pressure drop in the DPF, and the thermal energy storage in the DPF, which is very useful to mitigate ‘turbocharger lag’ during engine transient operation. These three effects have been evidenced with calculations performed using the validated model and the results have been fully analysed and discussed.

Fast-response nitrogen oxide (NO _x ) and carbon dioxide (CO₂) analysers have been used to study the effects of exhaust gas recirculation (EGR) delay within a modern diesel passenger car. These analysers have response times of the order of a few milliseconds, allowing the CO₂ analyser to be used to study the effects of the EGR transit delay during engine transients and the effect of sudden EGR valve actuation during a drive cycle. The effect of the EGR system delays on the NO _x emissions was measured with the fast-response NO _x analyser. The results indicate that the EGR delay contributes to the sharp short-duration spikes of NO _x .

Tests have also been carried out to identify the transient cylinder-to-cylinder EGR distribution by comparing the CO₂ concentrations in the inlet ports, again using the fast-response CO₂ analyser. The results of these tests show a variation between the CO₂ concentrations in the intake ports, suggesting a poor EGR distribution under certain conditions.

Measurements have been recorded of the EGR in the intake port runner of just one of the engine’s cylinders and have been compared with the exhaust port NO _x emissions. This provided some data on a cycle-by-cycle basis including the number of engine cycles required for the burned gas to complete the EGR circuit.

The paper presents a comprehensive study on engine performance improvement attributable to application of different electrically assisted turbocharger topologies. Performance of a baseline turbocharged high-speed direct-injection (HSDI) diesel engine is compared to the performance of an engine utilizing an electrically assisted turbocharger, an engine utilizing a turbocharger with an additional electrically driven compressor, and an engine utilizing an electrically split turbocharger. Analyses are performed based on a validated physically based engine and vehicle model comprising detailed models of all vehicle components, thus ensuring adequacy of results. Analyses are performed for various driving conditions, including tip-in in the fixed gears and the new European drive cycle (NEDC). Results reveal that electrically assisted turbocharger topologies improve transient response of the engine and thus driveability of the vehicle. Additionally, over a limited period of time, electrically assisted turbocharger topologies are able to improve steady-state torque output of the engine with retained fuelling, which is made possible by the availability of energy in electric storage devices. It was also revealed that the utilization of electrically split turbocharger enables considerable reduction in fuel consumption when driven according to urban drive cycles.

In fact, nowadays the optimization of engines and powertrains towards cleaner, more fuel-efficient, and safer vehicles requires virtual tools for the simulation of the whole system well in advance of the design process. Moreover, management and diagnostic issues have key roles in the exploitation of the used technologies and solutions (e.g. double-stage turbocharging, and high- and low-pressure exhaust gas recirculation). Without forgetting the importance of experimental investigations, mathematical models are very useful tools to improve both the performance and the behaviour of engines and powertrains, with applications ranging from optimization of the system layout to definition and testing of control strategies.

Starting from a brief overview of actual mathematical tools for simulation of the transient operation of diesel engines, several considerations are reported on the modelling criteria and approaches with reference to fast zero-dimensional mean-value models of engines and powertrains. An original library that the present authors and co-workers have built up in recent years for the ‘real-time’ simulation of diesel engines is described, and several applications are reported. The presented models were validated and used to simulate the transient behaviour of typical automotive diesel engines.

Biodiesel and bio-ethanol are expected to be the most applied biofuels in Europe in the short- to mid-term, especially in blended form with diesel and gasoline, respectively. There is a clear need to see what impact this blending will have on emissions and fuel consumption of current vehicles. Public information on this topic is mostly based on older technology. Within the Belgian research project BIOSES, emission and fuel consumption tests were performed on recent types of vehicles running on various biodiesel–diesel blends for diesel vehicles, and bio-ethanol–gasoline blends for gasoline and flex-fuel vehicles. The vehicles were tested on a proving ground with on-board emission measurement equipment, following different test cycles, including the European test cycle (NEDC) and a real-traffic-based cycle (MOL30). The paper shows an overview and discussion of the test results.

Significant excursions of engine variables occur during fast transients because of slow actuator responses and system dynamics. This creates adverse effects on dynamic performance and often causes emissions penalties. The challenge is particularly pronounced in engines with an increased number of actuators. In this paper, non-linear model predictive control (NMPC) is introduced to improve the dynamic response of a flexible engine system. NMPC combines advantages of both feed forward and feedback control while considering their constraints. The length of control horizon and prediction horizon are determined to achieve the dead-beat-like optimal control during transients and ensure smooth responses. The NMPC significantly improves the engine torque response and minimizes the excursions of in-cylinder variables under highly transient operation by adjusting each actuator control input simultaneously to achieve the control objectives.

This paper reviews the research activities within the subproject B1 Model Reduction for Low-Temperature Combustion Processes through CFD-Simulations and Multi-Zone Models of the Collaborative Research Centre SFB 686 – Model-Based Control of Homogenized Low-Temperature Combustion. The SFB 686 is carried out at RWTH Aachen University, Germany and Bielefeld University, Germany, and is funded by the German Research Foundation (DFG). This paper thereby summarizes the outcome of various publications by the authors, with the appropriate references given in the individual sections. Additionally, some new results are introduced.

The particular subject of this work is a dynamic simulation strategy for premixed charge compression ignition (PCCI) combustion that can be used in closed-loop control development. A detailed multi-zone chemistry model for the high-pressure part of the engine cycle is extended by a mean value gas exchange model accounting for the low-pressure part. Thus, an efficient model capable of describing PCCI combustion is sufficiently well established. In order to capture cycle-to-cycle dynamics, identified system dynamics influencing the input parameters are incorporated. For this, a Wiener model is set up that uses the combustion model as a nonlinear system representation. In this way, a dynamic nonlinear model for the representation of the controlled plant Diesel engine is created. The model is validated against transient experimental engine data.

This paper examines the cold-start-up events of spark and compression ignition engines in the context of transient operation and emissions control. When the temperatures of the oil, coolant and engine block are equal or close to the ambient temperature, start-up can be difficult to achieve without significant excess levels of exhaust emissions and fuel consumption. In general, the lower the ambient temperature, the more significant these problems are. The physical processes responsible for this phenomenon are briefly discussed. Excess emission factors from a pool of Euro 4 and Euro 5 petrol vehicles and three Euro 5 diesel vehicles are presented, as tested over the Urban Drive Cycle at 24 °C and at −7 °C. A full modal emissions analysis was also conducted at 24 °C and at −7 °C on one petrol vehicle and one diesel vehicle over the New European Drive Cycle. The most problematic emissions are identified for both engine types, and recommendations are made to legislators regarding cold-start emissions.

This paper presents the results from the analysis of an experimental investigation with the aim of providing an insight into the cyclic thermal shock phenomena occurring in the internal cylinder wall surfaces of a direct injection (DI), air-cooled diesel engine during the initial stage of a transient operation. The mechanism of cyclic heat transfer is investigated during engine transient events, viz. after a sudden change in engine speed and/or load. The experimental installation allowed both long- and short-term signal types to be recorded on a common time reference base during the transient event. Processing of experimental data was accomplished using a modified version of one-dimensional heat conduction theory with Fourier analysis, capable of catering for the special characteristics of transient engine operation. Based on this model, the evolution of local surface heat flux during a transient event was calculated. Two engine transient events are examined, which present a key difference in the way the load and speed changes are imposed on each one. During the analysis of experimental results the most important parameters characterizing thermal shock, such as the heat wave velocity and length of penetration, are quantified for each event, providing a comprehensive insight into the causes and consequences of this dangerous phenomenon. The results, in addition, confirm the theoretical predictions for the development of the thermal field during an engine transient event, as presented by the authors in previous work. Each thermal transient event is characterized by two distinct phases, that is the ‘thermodynamic’ and the ‘structural’ one, which are appropriately configured and analysed. From the results it is revealed that in the case of a severe variation, in the first 20 cycles after the beginning of the transient event, the wall surface temperature and heat flux amplitude on the cylinder head was almost three times higher than that observed in the ‘normal’ temperature oscillations occurring during steady-state operation.

In order to identify some of the special combustion and emission formation phenomena that occur in a turbocharged heavy-duty diesel engine during transient operation, the transient strategy of a production engine has been characterized at four different engine speeds. From each transient some points have been selected for further investigation by recreating these load points as steady-state points in a single-cylinder engine. This allows the emissions to be measured with a high degree of accuracy. An endoscope which makes it possible to evaluate flame temperatures was used in both engines. An empirically derived method of calculating nitric oxide (NO) formation from a combination of measured flame temperature, calculated gas temperature, and heat release rate has been developed and applied. This provides an increased understanding of combustion and emission formation phenomena during transient operation. An optical engine was also used to provide a full combustion chamber view for some of the operating points, and a specially developed software was used to calculate temperature distributions based on high-speed camera colour information. The NO formation formula was applied on these images, which resulted in spatially resolved NO formation distributions.

The effect of recirculating engine-out gases into the intake manifold on the cold start of direct-injection diesel engines is investigated. Two types of recirculation are examined. The first is low-pressure recirculation of engine-out gases into the intake manifold where their rate is controlled by a gas recirculation (GR) valve installed between the exhaust and intake manifold. The second is high-pressure recirculation by restricting the flow of the engine-out gases using a butterfly (BF) valve, installed in the exhaust system after the turbocharger to increase the back pressure and the rate of recirculated gases. Since there is no combustion during cranking, these gases contain evaporated hydrocarbons and partial oxidation products, mostly formaldehyde (HCHO). Experimental investigations on a four-cylinder direct-injection diesel engine indicated that high rates of cranking gas recirculation (CGR) increase the ignition delay and lengthen the cranking period. Since higher concentrations of hydrocarbons (HC) are expected to enhance the autoignition process, it is suspected that the recirculated HCHO would have an opposite effect. These opposing effects are investigated using the ChemKin diesel cycle simulation model. The model results demonstrated the effect of HCHO on slowing the autoignition and combustion reactions.