Ensuring compliance with the future CO2 fleet emission targets presents the entire automotive industry with enormous challenges in both the short and medium term. The new DDI combustion process developed at the IVT at the Graz University of Technology combines the high efficiency of diesel engines with the CO2 saving potential of natural gas in a natural gas-diesel dual-fuel combustion process.
What will the mobility of the future look like? Despite continuing controversial discussions about combustion engines, it will not be possible to eliminate diesel and spark-ignited engines in the short term or even in the medium term. Future CO2 emission targets of 95g/km from 2020 and corporate average CO2 emissions of 75 g/km from 2025 currently being discussed in the EU, however, will require significant efficiency increases to be implemented for combustion engines.
The purpose of the research project carried out at Graz University of Technology was thus to develop a combustion process with minimum CO2 emissions. Specifically, the attempt was made to combine the high efficiency of diesel engines with the CO2 saving potential of Natural Gas (NG) in one natural gas-diesel dual-fuel combustion process.
The Dual-Fuel (DF) name is a clear indicator that this concept constitutes the combination of two different types of fuel. Unlike bivalent combustion processes, in dual-fuel operation, both fuels are combusted simultaneously in the combustion chamber. In principle, different fuel combinations would be conceivable, such as gasoline and diesel, hydrogen and diesel or gasoline and natural gas. The usual case, however, is a combination of a less ignitable and a highly ignitable fuel. (1) gives an overview of the variety of dual-fuel concepts. In the research project, the combination of natural gas and diesel was examined. In this concept, the directly injected diesel works as chemical spark plug and therefore initiates the combustion. Due to the low modification efforts, conventional natural gas-diesel DF combustion processes are realised in combination with an external natural gas mixture formation. DDI (Dual Direct Injection) has been introduced as a representative acronym for naming the new combined internal mixture formation of natural gas and diesel.
Due to the favourable H/C ratio of natural gas compared to conventional fuels, the CO2 emissions can be reduced by approximately 25 %, when assuming constant brake thermal efficiency. In the DF combustion process analysed, the CO2 reduction potential is directly linked to the natural gas energy fraction, (2). Depending on the NG energy fraction, a distinction between a substitution concept and a pilot injection concept is possible. The latter is based on the minimal injectable diesel amount, which is limited either by the necessary ignition energy or the injection system. Based on these boundary conditions, the DF concepts examined were operated specifically with minimum diesel quantities, in order to be able to exploit the CO2 reduction potential.
A holistic approach was followed during the development of the new natural gas-diesel combustion process. In addition to the comprehensive experimental tests, the dual-fuel concept was modelled and analysed numerically. An overview of the topics addressed is shown in (3). The main focus was on experimental investigations, where both application and hardware parameters were analysed in detail. The focus in the context of the numerical investigations was on visualising in-cylinder flow processes. Additional information and analyses on the listed topics are documented in [1, 2].
DIFFICULTIES WITH CONVENTIONAL DUAL-FUEL CONCEPTS
The incomplete combustion of the introduced natural gas, especially in part load operation, is one of the main challenges of conventional natural gas-diesel DF concepts . A detailed discussion of these correlations as well as additional results has already been published in . Based on existing diesel engines, the DF operation is mostly realised in combination with a sequential natural gas injection into the intake manifold.
Due to the external mixture formation, a nearly homogeneous natural gas-air mixture forms in the combustion chamber at the time of the diesel injection. As a result, the air-NG equivalence ratio (λNG) exceeds the upper flammable limit in un-throttled operation even in case of a high NG energy fraction of 85 % in the operating point 1750/5, so the flame extinguishes in areas not reached by the diesel pilot injection. As a consequence, extensive portions of the premixed air-NG mixture remain unburnt, which result in high HC emissions. The intake pressure can be continuously reduced by throttling starting from wide-open-throttle. As shown in (4), the air-NG equivalence ratio is decreased from 2.1 to 1.1.
Consequently, the HC emissions of over 23 g/kWh can be reduced to 6 g/kWh. At the same time, the reduced intake manifold pressure deteriorates the brake thermal efficiency due to the higher gas exchange losses as well as the self-ignition conditions for the diesel pilot injection, which causes the Ignition Delay (ID) to increase significantly throughout the measurement series. In order to be able to keep the combustion phasing constant, the injection timing of the diesel pilot must be continuously advanced, which deteriorates the thermodynamic conditions for self-ignition even further. This reveals a trade-off between the conditions for complete combustion of the air-NG mixture and the self-ignition conditions for the diesel pilot injection, which becomes even more significant at lower load points.
This is where the newly developed DDI concept provides a solution. Instead of the homogeneous air-NG mixture, a stratified mixture is created and thus the HC emissions are reduced without negatively affecting the self-ignition conditions of the diesel pilot injection by throttling.
After the potentials and difficulties of the conventional DF combustion process with external mixture formation had been dealt with during the first project phase, the next development was the integration of a low-pressure natural gas direct injection into the combustion chamber. Apart from the technical data of the base engine and the characteristics of the DDI concept in (5), (6) shows the adapted engine components and a detailed view of the combustion chamber.
The natural gas injector literally took a central role in the DDI combustion process. More information on the DI CNG injector from Delphi that was used is published in . In order to achieve an optimal fuel distribution and mixture preparation of the pilot injection, the injector nozzle geometry was adapted to consider the inclined mounting situation. Due to limited available space, one outlet valve per cylinder was removed in order to allow the integration of both injectors. These circumstances required extensive design adjustments. Apart from an entirely modified cylinder head, the cylinder head cover, the camshaft carrier and the outlet camshaft had to be redesigned.
INFLUENCE OF NATURAL GAS DIRECT INJECTION
Due to its importance, the influence of the start of injection of natural gas on the HC emissions is explicitly explained. Apart from the emissions, the model representation of the mixture formation is shown in (7) both for homogeneous operation as also for stratified operation. Based on a very early start of injection (SOING = 260 °CA before firing TDC), HC emissions will drop at first until they increase again in the area of BDC (SOING = 180 °CA before firing TDC). One possible explanation for this observation was found by means of the 1D engine process simulations and the 3D flow simulations.
Due to the Coanda effect, natural gas collects along the roof of the combustion chamber, especially towards the end of injection. If the injection process coincides with the time frame of Intake Valve Closing (IVC), the natural gas can flow into the inlet ports due to backflow of charge. In addition, the natural gas located at the roof of the combustion chamber leads to very lean mixtures in the squish area and the piston top land. These regions are not directly accessible for the diesel pilot injection and must be captured by the subsequent flame front. Due to the lean air-NG mixture and the long flame paths, these areas can only be converted incompletely.
A start of injection after IVC avoids this problem, but any further reduction of HC emissions is primarily achieved by means of charge stratification. In order to guarantee a supercritical pressure ratio during the entire injection time, the latest start of injection was limited to 70 °CA before TDC. The diagrams in (7) show the schematic trend of the local air-fuel equivalence ratio λlocal, starting from the diesel pilot injection through the combustion chamber for the various different injection timings. Like the DF concept with an external mixture formation, an early injection leads to a homogeneous air-NG mixture in the combustion chamber, so there is a transition from pure fuel (λlocal = 0) to a lean air-NG mixture (λlocal = λNG) at the time of diesel injection.
Due to direct injection, the development of the local air-fuel equivalence ratio can be actively influenced by means of the injection timing. By combining the central natural gas injector with a combustion chamber with a piston bowl, the natural gas can be captured in the piston bowl during the compression phase. Through charge stratification, the development of the local equivalence ratio is different in all directions. Nevertheless, the equivalence ratio can still be illustrated in general starting from the diesel pilot injection. Due to charge stratification, the major share of the fuel is within the flammable limits of diesel and natural gas. Despite this, lean zones where the flame front extinguishes still exist in the fringe area of the stratified mixture, resulting in a small share of unburnt components that remains in the combustion chamber even during lean stratified operation.
An operating strategy was derived for the DDI combustion process through the entire engine map on the basis of the obtained findings. (8) shows the three operating modes and the investigated load points. The operating strategy can be divided into the following modes and areas:
:: From medium part load to full load, the DDI combustion process is operated homogeneously and with an overall stoichiometric air-fuel equivalence ratio (λ = 1) because of the robust exhaust gas aftertreatment with a three-way catalyst.
:: A lean and stratified operation (λ > 1) can be realised in lower part load due to the internal mixture formation of NG. In a direct comparison with the DF combustion process with external natural gas mixture formation that cannot be realised in the lower area of the operating map due to the massive HC emissions, the operating area can be extended to the lower partial load area with the DDI combustion process thanks to lean stratified operation.
:: Given the amount of diesel that is required for the ignition of the air-NG mixture, idle mode and lowest part load are covered with diesel only. Stratified operation in combination with low pressure injection reaches its limits due to the limited injection timing caused by the supercritical pressure gradient required and the low quantities of natural gas.
COMPARISON OF COMBUSTION PROCESSES
In order to clearly point out the significant progress that has been achieved with the DDI concept, (9) shows the measurement results at operating point 1500/3. Among all investigated load points this is the one with the lowest load, which is therefore particularly critical for dual-fuel operation. The optimised DDI combustion process is compared to the conventional DF combustion process, the base diesel engine and a modern spark-ignited gasoline engine that is also a four-cylinder turbocharged engine with a displacement of 2.0 l.
It can be seen that lean and stratified operation with the DDI concept is able to reduce engine-out HC emissions by a factor of 5, compared to the conventional DF concept. The DDI concept can reach engine-out pollutant emission figures on the level of a modern spark-ignited gasoline engine. The effective brake thermal efficiency was increased by 0.9 %-points due to the faster and earlier combustion compared to the diesel engine and even by 4.5 %-points compared to the spark-ignited engine. This results in CO2 savings of between 22 % and 29 % at the investigated operating point. The even and fast conversion of the DDI concept is very clearly visible when the heat release rate is considered. This fact indicates that the premixed combustion of the air-NG mixture through the propagation of a flame front clearly dominates. The heat-release rate is also significantly faster than it is in the spark-ignited engine due to the higher ignition energy of the diesel pilot injection.
With regard to lean operation, the calibration of the DDI concept was set in such a manner that the engine-out NOx emissions achieved are equal to those of the diesel engine. The great challenge of the DDI combustion process, however, is the aftertreatment of the remaining HC emissions that consist almost entirely of methane. The conversion in an oxidation catalyst requires exhaust gas temperatures of at least 450 °C . Due to the high compression ratio and the high charge dilution in part load operation caused by EGR and excess air, the exhaust gas temperature of the DDI concept is, however, only in the range of 250 and 300 °C. This means that the remaining methane emissions cannot be converted using currently available catalysts. At the moment this issue is standing in the way of further realisation for the DDI concept.
A new natural gas-diesel dual-fuel combustion process for passenger car engines has been developed. By contrast to conventional dual-fuel concepts, direct injection of natural gas has here been integrated into the DDI concept. This allows the realisation of charge stratification of the air-natural gas mixture, considerably reducing engine-out HC emissions in part load operation. Even in the critical lower load areas, the engine-out HC emissions were reduced to the level of a modern spark-ignited gasoline engine. By this approach, the operating area of the DF combustion process has been significantly extended towards lower loads. Due to the high brake thermal efficiency of the DDI concept and the use of natural gas, CO2 savings between 20 and 30 % can be achieved over the engine map compared to diesel and spark-ignited engines.
The investigations also showed, however, that due to the low exhaust gas temperatures in lean stratified operation, the remaining engine-out HC emissions mainly consisting of methane cannot be converted with currently available oxidation catalysts. The realisation of the DDI concept thus requires a further reduction of engine-out HC emissions through optimisation of the combustion process and ensuring exhaust gas aftertreatment in all operating conditions.
 Sprenger, F.: Entwicklung eines Erdgas-Diesel Dual-Fuel-Brennverfahrens zur signifikanten CO2-Reduktion bei Pkw-Motoren. Graz, University of Technology, doctoral thesis, 2017
 Fasching, P.: Natural Gas as Fuel for Monovalent and Dual Fuel Combustion Engines – an Experimental and Numerical Study. Graz, University of Technology, doctoral thesis, 2017
 Garcia, P.; Tunestal, P.: Experimental Investigation on CNG-Diesel Combustion Modes under Highly Diluted Conditions on a Light Duty Diesel Engine with Focus on Injection Strategy. In: SAE International Journal Engines 8 (2015), No. 5, pp. 2177–2187
 Fasching, P.; Sprenger, F.; Granitz, C.: A holistic investigation of natural gas–diesel dual fuel combustion with dual direct injection for passenger car applications. In: Automotive Engine Technology 2 (2017), No. 1, pp. 79–95
 Fasching, P.; Sprenger, F.; Preuhs, J. F.; Hoffmann, G.; Piock, W. F.: The challenges of natural gas direct injection and its application to a natural gas-diesel dual fuel concept. 16th conference “The Working Process of the Combustion Engine”, Graz, 2017
 Bank, R.; Etzien, U.; Buchholz, B.; Harndorf, H.: Methane Catalysts at an Upstream Turbine Position. In: MTZindustrial 5 (2015), No. 1, pp. 14–21
DR TECHN. FLORIAN SPRENGER is a former University Assistant at the Institute of Internal Combustion Engines and Thermodynamics at the Graz University of Technology (Austria).
DR TECHN. PAUL FASCHING is a former Scientific Staff Member at the Institute of Internal Combustion Engines and Thermodynamics at the Graz University of Technology (Austria).
DIPL.-ING. CHRISTINA GRANITZ is Scientific Staff Member at the Forschungsgesellschaft für Verbrennungskraftmaschinen und Thermodynamik mbH in Graz (Austria).
UNIV.-PROF. DR HELMUT EICHLSEDER is Head of the Institute of Internal Combustion Engines and Thermodynamics at the Graz University of Technology (Austria).
The authors would like to thank the Austrian Research Promotion Agency for its financial support for the project.