Skip to main content

Diesel Engine Cold Start and Transient Improvement

Final Report Summary - DECOST (Diesel Engine Cold Start and Transient Improvement)

Advances in engine technologies such as high pressure common rail direct injection, multiple fuel injection strategies, reduced compression ratio and increased boost pressure have improved the higher specific power and reduce emissions. However, the reduction in compression ratio potentially affects other areas of performance notably cold start behaviour. Cold ambient temperature has a significant effect on the cranking speed, cylinder gas pressure, cylinder gas temperature, blow-by rate, evaporation rate, mixing, ignition delay, and fuel burning. Diesel engine combustion initiation relies on the temperature rise achieved by compression and the reduction in compression ratio is likely to bring the cold operation into an unstable region, possibly even leading to a total failure of the engine start. The cold startability problem is more severe for the diesel engines while using biodiesel blends at cold ambient temperatures. Currently in Europe, biodiesel (EN 14214) is blended up to 7% in diesel and it is also proposed to increase the biodiesel concentration in the near future. There is very limited research work carried out on cold startability, gaseous emissions, particle number and particle size distribution of biodiesel fuelled engines under cold start and transient conditions. This project is aimed to research new strategies for reducing the cold start difficulties associated with CRDI diesel engines under cold ambient conditions especially with respect to biodiesel and at the same time to investigate the particle matter emission characteristics under transient conditions.

A latest generation six-cylinder turbocharged common rail direct injection diesel engine supplied by Jaguar Land Rover was installed in the newly commissioned “the Sate of the Art” cold cell transient testing facility which is capable of conducting low temperature tests down to -20°C. The engine exhaust system was suitably modified to sample and measure the emissions as per European emission legislations. AMA i60, DMS 500 and Smart sampler were used to measure the gaseous, particles number and size, and particulate matter respectively. A coolant and intake air heating systems were installed in suitable locations and these can be controlled independently. Shell Global UK supplied winter diesel and biodiesel for the project and supported in analysis of fuel properties.

The baseline cold start tests were conducted using the winter diesel at different ambient temperature conditions and the test results revealed that the cold ambient temperatures increased the exhaust emissions by several times. The exhaust particulates of diameter 10-100 nm were higher at +20°C and -7°C whereas at -20°C the particulates concentration was shifted towards accumulation mode due to the incomplete combustion of fuel. The mass of accumulation mode particulates was higher than the nucleation mode particulates and the particulate masses during cold start and idle operation were increased with the drop in ambient temperatures. It is also found that particulate mass at idle was much lower than that at cold start.

Engines in passenger cars are evaluated for their emissions in a standard driving cycle (such as NEDC, FTP) on a chassis dynamometer. With the support from JLR, speed of the vehicle (km/h) for the each step of the NEDC cycle is converted into engine speed (rpm) and torque (Nm) with respect to time and input to the Puma control system of the test bed. This approach is a typical hardware-in-the-loop (HIL) system i.e. engine-in-the-loop (EIL) that simulates the testing of a passenger car on the chassis dynamometer. The coolant and air temperatures were controlled at the start of test and then the engine is tested under simulated conditions.

A series of cold start tests have been conducted considering the variables such as coolant temperature, oil temperature, biodiesel concentration in the blend and ambient temperature. A database of cold start emissions with biodiesel blends at different coolant and intake air temperatures has been created and the data generated was used for developing the Model Based Calibration (MBC) model in the MatLab platform. This model is capable of predicting the cold start emissions HC, NOx and PN as a function of intake air temperature, coolant temperature, ambient air temperature and biodiesel concentration in the blend. This model is used to optimise the coolant and air heater temperatures to achieve the minimum emissions during the cold start.

The intake air heating strategy has reduced the HC emissions significantly, for instance, heating the intake air from -7°C to 5°C reduced the HC emissions by 50%; NOx by 17% and PM by 50%. At the optimised coolant temperature of 10°C in the -7°C environment, the HC emission was reduced by HC 80%; NOx by 33% and particulate number by 15%.

The coolant and intake air heating strategies has been practised during cold start operation using rapeseed methyl ester (RME) blends also. The biodiesel blends decreased the gaseous emissions at ambient temperature or warmup test conditions. But at cold ambient temperature conditions, the higher viscosity of biodiesel blends affected the fuel spray pattern and the incomplete fuel combustion demanded higher fuel consumption than diesel. B10 and B20 blends increased the peak HC emission by 50% and 215% compared with diesel at -7°C. The peak HC emissions at -7°C were 13, 35 and 100 times higher than that of normal ambient temperatures.

The coolant and air heating strategies have been implemented for New European Driving Cycles (NEDC) and at optimised conditions tests were conducted using biodiesel blends in the +20, 0, -7°C environment. Relatively higher intake air temperatures reduced the diameter and number count of particulates and the particulates of diameter 10-23 nm was accounted for ~45% for the all intake air temperature conditions. The particulate number for the first part of NEDC was ~25% for -7°C intake air temperatures and was reduced to ~20% for the warm intake air. The particulate mass was significantly higher (~20% of NEDC) during initial stages of NEDC due to higher number of accumulation particulates at -7°C and was reduced to 12-14% by intake air heating.

Exhaust particulate matter was also collected on a PTFE filter paper using the AVL Smart sampler at different ambient temperatures and analysed for unregulated emissions. For this study, the advanced two dimensional gas chromatography technique (2D- GC) was used to quantify the components such as PAH, alkanes, alkenes, cyclehexane, benzene etc. from the exhaust. Cold ambient conditions increased the PAH emissions whereas biodiesel reduced the PAH emissions due to lower aromatics and increased oxygen in fuel. It is also found that lube oil has contributed more in formation of alkanes as compared to fuel due to higher boiling points of lube oil compounds (>C20). This study also identified the other fuel components and additives in the present and its contribution in particulate emissions.

In summary, this project demonstrated the emission reduction potential of cold start strategies such as coolant heating and air heating in reducing the cold start emissions and thereby limiting the regulatory driving cycle emissions at cold ambient temperatures. Also, biodiesel blends significantly increased the particulate emissions during the cold start compared to diesel and these emissions were controlled by cold start emission reduction strategies. An advanced control algorithm is under development in the lab with the contribution of the current research for the control of EGR and coolant water flow using an open ECU supplied by JLR. As a result of this work the Research Fellow has acquired a strong intellectual knowledge and practical skills which have allowed him to consolidate his future career.

Contact: Prof. Hongming Xu, Head of Vehicle and Engine Technology Research Centre,
School of Mechanical Engineering, University of Birmingham, B15 2TT, UK, Tel: ++44 121 414 4153; Email: h.m.xu@bham.ac.uk