Reactivity Controlled Compression Ignition (RCCI) of Gasoline- CNG Mixtures

Reactivity controlled compression ignition (RCCI) is a dual fuel combustion method that relies on the significant difference in reactivity of the fuels involved. RCCI had a low performance at high engine speed due to its high tendency on knocking and high pressure rise rate. Therefore, this study investigates the effect of the fuel stratification on the RCCI combustion and its extended to the interaction of two low reactive fuels, gasoline and compressed natural gas (CNG), in the RCCI combustion system. The investigation was experimentally performed on a single cylinder engine and constant volume chamber. The stratification was created by varying injection timing in the engine by injecting CNG at 80 and 120 before top dead center (BTDC) and varying injection gap in the constant volume chamber with the gaps between two fuel injection timingwere varied between 0 ms to 20 ms. The results in the engine experiment show that proportions of gasoline and CNG and degree of stratification of CNG were found to be effective means of combustion control within certain limits of engine load and HC and CO emissions could be significantly reduced. While in constant volume chamber it has a significant effect on the combustion phasing. Stratified mixture produces shorter combustion duration while homogeneous mixture produces longer duration.


Background
Perhaps the most graceful invention by humankind that ever had a greater impact on society, the economy, and the environment is the reciprocating internal combustion engine, in general, called the IC engine. For decades, this magnificent invention proved to play a vital role in the automobile system, used almost exclusively today. There are two types of internal combustion engines: spark ignition (SI) and compression ignition (CI). For the last decades, rapid improvements in the efficiency have been achieved on both types of IC engine.
Unfortunately, at present, there is a pressing need to develop advanced combustion engines that maximize the engine efficiency and totally mitigate the exhaust pollutants. Profound understanding of both SI and CI combustion principles has been achieved during the last decades to improve the efficiency and reduce the emissions. The conventional SI combustion, which is characterized by flame propagation in near-stoichiometric homogeneous mixtures, produces very low exhaust emissions in combination with a three-way catalytic converter but has a relatively low thermal efficiency, which is its main drawback.
CI combustion, on the other hand, that is characterized by the autoignition of a lean fuel-air mixtures, has a very high thermal efficiency; yet, it has very high soot and NO x (nitrogen oxide) emissions. Diesel engines typically produce lower carbon monoxide (CO) and unburned or partially burned hydrocarbons (HC) compared to the gasoline engines. However, NO x, which comprises nitric oxides and nitrogen dioxides, in addition to particulate matter (PM) or soot, is significant pollutant from diesel engines, which require proper control strategies as they pose adverse health and environmental impacts.
The engine technologies are advancing at a significant rate during the last decades. The engine technology development timeline is depicted in Figure 1. The engine performance was the main priority in the first era of engine developments. Technologies such as turbocharger, port fuel injection, high compression ratio, direct fuel injection, and engine lightweight material were the technologies that are focusing on increasing the engine power output to its maximum capabilities. This early era was driven by the abundant amount of fuels relative to its low demand as the automobile was still an exclusive technology.
The second era of the engine technology development was mainly driven by increasing concern about exhaust emissions and efforts in achieving low fuel consumption. In the earlier technologies, efficiency was improved and the engine downsizing was the primary target of the engine development. Homogeneous charge compression ignition (HCCI) engine is one of the promising alternatives in order to achieve these objectives.
The spark ignition engine (SI) and compression ignition engine (CI) are the established engine technologies, and each have their advantages and disadvantages. SI has a faster response and low emissions yet low efficiency, while CI offers high efficiency and low fuel consumption yet higher emissions and slower response. This makes the development of the SI and CI engine followed different approaches. Nevertheless, the main objective of an engine is mainly to achieve high performance, high efficiency with low emissions. In order to achieve this goal, HCCI was introduced in 1983. Major obstacle in HCCI is its rapid heat release rate with a very high maximum pressure, which is detrimental to engine structure. In the effort of controlling the combustion in HCCI engines, technologies such as exhaust gas recirculation (EGR) and variable valve timing (VVT) are utilized to improve the controllability of HCCI combustion, which is used in the commercial HCCI engine such as skyActive technology (Mazda) and diesOtto engine (Mercedes).
On the other hand, in an effort to find the best method for controlling HCCI combustion, the HCCI has evolved into various types of controlled autoignition-based engine (CAI) such as premixed charge compression ignition (PCCI) in 1995 [1] and reactivity controlled compression ignition (RCCI) engine in 2011 [2]. These methods are proven to be able to improve the operating range of CAI engine. However, there are no established methods that are proven to be effective in controlling CAI engine. Figure 2 depicts the differences in each combustion system control strategy. The CI system (Figure 2 (a)) creates a very high pressure inside the combustion chamber, and the fuel is directly injected to combust. SI system (Figure 2b)), on the other hand, forms a homogeneous mixture either by direct injection or by port injection before the mixture is ignited by a spark, whereas the HCCI engine (Figure 2 (c)) produces a homogeneous mixture by injecting the fuel during the intake stroke, and the mixture is autoignited due to the compression. HCCI combustion combines the best features of gasoline and diesel engines to produce diesel-like power and efficiency while maintaining gasoline-like soot free emissions within certain operating limits.
The engine performance was limited to the part load conditions and controlling the combustion process was very problematic due to the autoignition being highly dependent on the  temperature, pressure, and mixture composition inside the combustion chamber. In its development, HCCI was limited by the narrow operating range and the unpredictable combustion delay and behavior. The maximum pressure generated was very high but produced a low mean effective pressure due to the short combustion duration. Many methods and possibilities were proposed to control HCCI engines. In this process, the paradigm of creating an autoignition process from homogeneous charge is shifting to the method of controlling autoignition. The first development is the premixed charge compression ignition (PCCI). This concept was introduced by Aoyama et al. [1]. Gasoline was subjected into diesel-like environment with high compression ratio 17.4:1. A port injection method was used to create the premixed charge in the combustion chamber where the fuel is injected very close to the intake valve closing time as shown in Figure 3.
The combination starts with gasoline and diesel as the low-reactive fuel and high-reactive fuel, respectively. It was found that RCCI combustion was able to operate in a wide range of engine loads with near-zero NO x and soot emissions, accepted pressure rise rate and high indicated efficiency. However, RCCI still could not achieve high load operation with power outputs comparable to CI engine. The combustion behavior of RCCI is somewhat still unpredictable. Further investigation on the important parameters in RCCI combustion control will improve the understanding of the combustion process of RCCI that leads to better control of the process and better engine output.
The premixed charge compression ignition (PCCI), reactivity charge compression ignition (RCCI), and spark-assisted HCCI are some of the established methods of controlling the autoignition process in an engine. All of these methods are categorized as CAI engines.
Regardless of the limitations of the controlled autoignition (CAI)-based combustion system, CAI offers high efficiency [4], low fuel consumption [5], and low emission [6], which are the main aims of future engine development. The attributes that differentiate SI, CI, HCCI, PCCI, and RCCI are shown in Table 1.
These earlier works are the basis of the controlled autoignition engine concept. The primary focus of the CAI combustion concepts is identifying the relevant influencing parameters as well as control parameters of this system to widen the operating range and improve the efficiencies. The need for a thorough understanding of the CAI combustion process initiates further discussion on the method to control its combustion. The next stage is to determine the engine parameters that have a direct effect on the combustion. As these steps are carefully defined, high efficiency, low fuel consumption, and low emissions internal combustion engines are achievable ( Table 2). Many researches have been done in the effort of controlling CAI combustion-based system. Agarwal et al. [7] summarize the various types of combustion method and some method in controlling the combustion process. There are various areas that require improvement in order to reshape the combustion and improve the efficiency while still having a low emission. Development of control on ignition timing [8], method in slowing down the heat release rate at high load [9,10] and development of intake and exhaust manifold for multicylinder engine are among the few area of improvement that have been identified in the area of CAI combustion system. Focusing on the RCCI combustion control method, Li et al. [11] categorized the control by two main categories, fuel and engine management. The fuel management includes two fuel strategies [12,13] and single fuel strategy with additives, while the engine management include fuel ratio [14], injection strategy [15][16][17][18], EGR rate [19], compression ratio [16], bowl geometry [20,21], stability control, and utilization of two injectors.
This chapter introduces two low reactive fuels, gasoline and CNG, in an RCCI combustion system in order to increase the limit of RCCI engine operation. It introduces a method that has different principals compared to RCCI by introducing combination of low reactive fuels rather than combination of high-and low-reactive fuel. Furthermore, this chapter also introduces the method of RCCI combustion control by varying the stratification by varying the injection timing and gap between two fuel injectors. Two approaches were done to investigate the behavior of gasoline-CNG mixtures in the RCCI combustion system. First approach is experimental testing on a single cylinder engine that was converted to dual fuel engine with gasoline injected at intake port, while CNG is directly injected into the combustion chamber. The stratification was done by varying the injection timing of CNG, while gasoline is kept homogeneous.
The second approach is the combustion testing in a constant volume chamber with both fuels that are directly injected into the combustion chamber. In this setup, the stratifications level in chamber is done with varying the injection gap between two fuel injections.

Test procedure and equipment
The test procedure and equipment used in both approaches are elaborated in this section.

Engine testing and equipment
The engine used for this experimental study houses the fuel system of direct injection of gaseous fuel and has compression ratio of 14. This engine is a single cylinder water-cooled engine coupled to an electric dynamometer that can be used for starting the engine and measuring the brake torque produced by the engine. Figure 4 shows a schematic drawing of the engine.
An electric heater is provided to heat the lubricant oil to help warm up the engine faster. A separate control unit controls the operation of the pumps and the temperature of the oil and water by temperature controllers. The control unit also controls the operation of the dynamometer, which also serves as the starter motor and the engine can be motored at a wide range of speeds. There are standard features of safety included in the control unit such as emergency switch, automatic shut down upon the excessive rise in the oil and/or coolant temperatures, or any abnormal conditions of electrical power supply ( Figure 5).
A commercially available gasoline port fuel injector was used and its specifications can be found in Table 3. This injector has low flow rates and was selected to match the requirement of injecting very low quantities of gasoline to operate in HCCI mode with ultralean mixtures. The injector comes calibrated in the factory to inject and precisely meter the volume against the specified injection duration.  The change in air flow rate and volumetric efficiency at different engine speeds were taken into account, and the gasoline injection durations were adjusted so as to operate with constant equivalence ratios of gasoline from 0.20 to 0.26.  The CNG quantity was varied and limited by the combustion stability of the mixture by limiting the maximum coefficient variation (CoV) of the combustion to 10%. Thus, the readings were obtained at different speeds, different gasoline flow rates, and different injection timings with various quantities of CNG injected ( Table 5 and Figure 6).   Figure 7 depicted the injection timing, and intake and exhaust valve timing of the engine. It shows that injection timing 300, 240, and 180 BTDC take place before the intake valve close, while 120 and 80 BTDC take place after intake valve close.

Constant volume combustion chamber testing and equipment
Following section describes the detailed experimental setup and equipment used on the constant volume chamber. Constant volume combustion was used as the primary method for characterizing the interaction of the parameters such as fuel compositions, lambda, and mixing ratio. The equipment and control system were designed, manufactured, and calibrated to accommodate the experimental works. The primary data in this experimental setup are combustion images and pressure trace of which acquired by Schlieren and direct measurement method, respectively.
The experiments were carried out in a constant volume combustion chamber with diameter and length 80 and 100 mm, respectively. A 100 W cartridge heater was placed in the middle of the chamber to increase in-cylinder gas temperature with a maximum temperature of 820 C. There were two window access planes to the chamber in order to facilitate the Schlieren image visualization method (Figures 8-10).

Gasoline-CNG mixtures in RCCI combustion system
The gasoline-CNG mixtures performance and combustion in RCCI combustion system from both methods are elaborated in this section. It explained the parameters affecting the combustion of low-reactive fuels and method in controlling the combustion process.

Gasoline-CNG combustion behavior in RCCI combustion-based engine
The degree of stratification of CNG in the total mixture was found to have significant effects on the maximum load in terms of the IMEP attainable and φTotal. The degree of stratification is determined by the injection timing with 300 BTDC representing the homogeneous mixture, while 120 BTDC represents the stratified mixture. The 300 BTDC has a very early injection timing, and the fuel is injected during intake valve is open. It is, therefore, the fuel that has sufficient time to be completely mixed with air and create a homogeneous mixture. While on the other hand, for 120 BTDC, the fuel is injected after intake valve closed and mixing time between fuel and air is very short and does not allow a complete mixing to take place. Figure 11 shows that at 300 and 240 BTDC injection higher total equivalence ratios could be operated. But with 180 and 120 BTDC, the maximum operable φTotal was limited with reduced IMEP at a given φTotal when compared to the 300 and 240 BTDC cases.
The IMEP results show agreement with investigation from Genchi G and Pipitone E [22] where the increased composition of CNG produces higher IMEP. With the highest degree of stratification, although the maximum load was limited, there was no significant drop in the IMEP and the trend was similar to 300 and 240 BTDC conditions. The corresponding values of indicated thermal efficiencies are shown in Figure 12. The maximum load was observed to be limited by knocking when CNG was injected at 300 BTDC, and for the other cases, increasing CNG injection rate led to unstable operation or misfire. From Figure 13, it can be seen that the ignition timing could be altered by changing the timing of injection of CNG at a given load. The ignition timing was determined by identifying the start of heat released rate and mass fraction burned derived from the pressure data where the 0% points before the continuous propagation of the mass fraction burned is determined as the start of ignition of the analyzed combustion cycles.  When the rate of CNG injection was increased, the ignition timing was delayed due to the higher octane number of CNG. Also, higher degrees of stratification resulted in higher increments in the delay of ignition timing as the CNG injection rate was increased. The slope of the curves was steeper when the injection timing was delayed. For a given increase in the CNG injection rate, the increase in delay in ignition timing was higher when the degree of stratification was increased. That is both the injection rate and the degree of stratification of CNG had significant effects on the ignition timing when operated with φg = 0.20. However, the maximum total equivalence ratio was less than that obtained with CNG injection at 300 and 240 BTDC.
It was found that the combustion duration was reduced when CNG injection rate was increased at 300, 240, and 80 BTDC. When CNG was injected at 180 and 120 BTDC, the combustion duration was marginally affected and it initially decreased up to certain values of CNG injection rate and then it increased again. Figures 15-18 show the rate of heat release and pressure rise at various injection timings. Increasing the rate of CNG injection at 300 BTDC was limited by knocking as shown in Figure 14. But with later injection timings, with φg = 0.20, any increase in CNG injection rate resulted in a delayed autoignition and reduced peak pressure. Therefore, increasing CNG injection rate beyond certain levels led to misfire or no fire, thereby defining the maximum load limit.
As shown in Figure 15, when CNG injection rate was increased, it resulted in delayed ignition timing. Up to φTotal = 0.33, the resultant peak pressure increased, and with a further increase in CNG injection rate, it decreased. Also, above φ Total = 0.33, the delay in ignition timing was more significant and resulted in decreased peak pressures. As will be discussed later in this section, combustion efficiency of both the fuels increased and CH4 emissions decreased with an increase in φTotal above 0.33 as shown in Figures 23 and 29. Therefore, it can be concluded that, above φTotal = 0.33, the peak pressure was reduced due to delayed ignition, and the combustion was more complete with increase in injection rate at 240 BTDC. That is, increasing φg above 0.33 resulted in reduced peak pressures without a decrease in thermal efficiency as shown in Figure 12. Heat release rates increased with an increase in CNG injection rate up to φTotal = 0.40 above which it reduced again. Above φTotal = 0.42, increasing CNG injection rate resulted in misfire or no fire, and both gasoline and CNG combustion was quenched.
With CNG injection at 180 BTDC, increase in CNG injection rate resulted in a more significant delay in ignition timing. There was a marginal increase in peak pressure when φTotal was increased to 0.26 above which it reduced again. Thermal efficiency and combustion efficiency increased primarily due to the remarkable increase in the completeness of CNG combustion as suggested by CH4 emissions as shown in Figure 29. Heat release rate increased with an increase in CNG injection rate as shown in Figure 17. However, increasing CNG injection rate above φTotal = 0.26 resulted in decreased overall combustion efficiency and high CH4 emissions as shown in Figures 23 and 29. This suggests that the degree of stratification created at 180 BTDC injection results in deterioration in combustion and leads to decrease in thermal efficiency as shown in Figure 12. Similar, trends were observed with CNG injection at 120 BTDC when φTotal was increased above 0.24 as shown in Figure 16.
When CNG injection was retarded to 80 BTDC, increase in injection rate resulted in significant delay in ignition; however, there was less noticeable effect on peak pressures up to φTotal = 0.28. Increasing φTotal above resulted in a more significant delay in ignition and peak pressure, and heat release rates increased. Thermal efficiency and combustion efficiency increased primarily due to remarkable increase in completeness of CNG combustion as suggested by CH4 emissions as shown in Figure 29.
As shown in Figure 19, increasing CNG injection rate at 240 BTDC resulted in delayed ignition. At φTotal = 0.28 and 0.33, there was a slight increase in burning rate at the last stage of combustion compared to combustion with pure gasoline. At 180 and 120 BTDC, there was no significant effect on the burning rate of the fuels due to increase in CNG injection rate, but it caused a significant delay in ignition as shown in Figures 20 and 21. Similar results were obtained with CNG injection at 80 BTDC; however, at φTotal = 0.28 and 0.33, the combustion was slower at the initial stages and was faster at latter stages as shown in Figure 22. As shown in Figure 23, with an increase in φTotal by CNG injection at 300, 240, and 80 BTDC, combustion efficiency increased. The highest increment was obtained with CNG injection at 80 BTDC for a given increase in φTotal due to mixture stratification. However, CNG injection at 180 and 120 BTDC, combustion efficiency increased initially but decreased again and was below 80% for all φTotal.
The exhaust gas temperature was observed to increase as the CNG injection rate was increased as shown in Figure 24. The increase in exhaust gas temperature with increasing CNG injection rates at 180 and 120 BTDC was less than that observed with increasing CNG injection rates at 300, 240, and 80 BTDC. When the fuels were homogeneously mixed, it resulted in higher exhaust gas temperatures due to rapid burning. Similarly, when CNG was highly stratified, it also led to higher exhaust gas temperatures. Figure 25 shows the indicated specific NO x (ISNO x ) emissions. The NO x emissions were marginally affected and were around the same levels for all test conditions. However, different trends were observed at different injection timings and CNG injection rates.
Increasing the CNG injection rate resulted in a drastic increase in the NO 2 /NO x ratio up to a certain point and then it decreased. As shown in Figure 26, the ratio of NO 2 /NO x almost doubled when the CNG injection rate was increased to around φTotal = 0.33 before decreasing again. That is, up to a certain value of CNG injection rate, CNG reduced the combustion temperature and led to formation of higher amounts of NO 2 .  The indicated specific CO (ISCO) emissions were reduced significantly as the mixture was enriched with CNG by direct injection at all injection timings as shown in Figure 27. However, the reduction obtained was the highest when CNG was injected at 300 and 240 BTDC. Any increase in CNG injection rate at later injection timings resulted in less reductions in CO emissions. The lowest reduction was obtained at the injection timing of 80 BTDC, as the high degree of stratification of CNG limited the availability and distribution of oxygen and temperature differences in the CNG and air particles.
The HC emissions were found to be significantly affected by the degree of stratification of CNG as shown in Figure 28. The highest reduction in HC emissions was obtained with CNG  injection at 80 BTDC. Higher degrees of stratification of CNG resulted in more complete combustion. Figure 29 shows the mass ratio of flow rates of CH4 in the exhaust emissions and mass flow rate of CNG injected into the cylinder. At a given constant gasoline equivalence ratio of φg = 0.20, CNG direct injection at 80 BTDC resulted in the least emission of CH4. Therefore, the combustion of CNG was more complete when it was stratified. CNG injection at 300 and 240 BTDC resulted in moderate levels of CH4 emissions, and highest values were obtained with CNG injection at 180 BTDC. This was due to the turbulence created and mixing conditions in the cylinder when the piston changed its direction at 180 BTDC.

Gasoline-CNG combustion behavior in constant volume combustion chamber
The effect of injection gap on the gasoline-compressed natural gas mixture (GCNG) mixture combustion is discussed below. The injection gap alteration gave direct impact on the mixture distribution inside the combustion chamber. There are five injection gaps tested, 0, 5, 10, 15, and 20 ms. These injection gaps are expected to be able to give direct control to the mixture distribution inside the chamber.
The effect of injection gaps is shown in Figure 30. It shows two mixture compositions, 50 and 90% GCNG composition. The injection gap gives different effect between the two compositions. In 50% GCNG, longer injection gap gives higher combustion efficiency, maximum pressure, total heat released (THR) and shorter delay. Furthermore, it also shows longer duration for  all combustion stages. While on the contrary, longer injection gap reduces the combustion efficiency, maximum pressure, THR and longer combustion delay for 90% GCNG. However, the trends of the combustion duration are similar, longer duration for longer injection gap. Figure 31 confirmed the variation of injection gap effect to the combustion process of the GCNG mixture. The turning point is shown between 70 and 80% GCNG mixture compositions. For all the mixture above 80% shows decreasing combustion efficiency with the increase in injection gaps which is contrary with the mixtures below 70% that show an increment of combustion efficiency with the increase of injection gaps. These differences may cause by the mixture distribution inside the chamber.  The mixture distribution inside the chamber for GCNG mixture 30 and 90% with 0 and 20 ms injection gaps are depicted in Figure 32. In the figure, highly stratified mixture for 30% GCNG mixture with 0 ms injection gap. The stratification is marked by darker color on the bottom of the chamber that indicates high density fluid (gasoline). The image shows that most of the gasoline was collected at the bottom of the chamber because of the momentum of CNG injection that prevents the gasoline from reaching the top side of the chamber. 20 ms injection gap, on the other hand, shows better fuel mixing shown by fairly similar image intensity throughout the chamber.
The injection gaps for 30% GCNG mixture improve the mixing rate thus increase the combustion performance. Furthermore, the gasoline fuel is mainly accumulating at the bottom side     which also has average low temperature compared to the top one. It makes the vaporization rate of the gasoline took longer time which also elongates the combustion delay as shown in Figure 33.
The injection gaps at 90% GCNG mixture, on the other hand, have similar liquid fuel distribution as in Figure 32 where both injection gaps show concentrated fuel distribution at the top of the chamber. In spite of the similarity, the 0 ms injection gap shows the higher intensity of the liquid fuel (darker region) on the top side of the chamber compared to 20 ms injection gap. It shows that 0 ms injection gap have higher gasoline population compared to 20 ms injection gap thus the longer time required for vaporization process. This is the main reason for the lower combustion output as well as longer combustion delay for 0 ms compared to 20 ms injection gap.
The combustion sequence for GCNG at 0 and 20 ms injection gaps is depicted in Figure 34.   lower intensity compared to 100% gasoline, while 0 ms injection gap shows yellow color. It can be assumed that the blue color is the product from the same reaction that generates hydroxyl peroxide and increases the combustion output of the mixture.
In 100% gasoline combustion, the blue flame occurs because of the homogeneous mixture that creates multipoint combustion behind the flame front that significantly increases the combustion output. A similar process occurs in the 20 ms injection gaps, mixture homogeneity is achieved with this mixture as the effect of CNG injection shown by longer combustion delay relative to the start of injection as depicted in Figure 35.
Injection gaps proved to have a direct influence on the fuel distribution inside the chamber, thus affecting the combustion characteristics of the mixture. The combustion process in the CVC is mostly affected by the characteristics of the fuel distribution inside the chamber at the time the combustion occurs. The injection gaps, in this case, highly affect the mixture distribution inside the chamber where longer gap promotes the mixing and create a mixture that is more homogeneous.

Conclusions
1. It was found that more amount of gasoline was needed to ignite stratified CNG. Increasing the degree of CNG stratification resulted in delayed ignition and reduced peak pressures. The combustion efficiency of overall mixture and that of CNG were higher at high degree of CNG stratification. When the degree of CNG stratification was increased, significant reduction in HC emissions could be achieved even at low gasoline equivalence ratios. However, NO x and CO emissions increased especially at low engine loads and high speeds. Maximum load attainable at low engine speeds could be extended by increasing the degree of CNG stratification.

2.
Injection gap capable of controlling the stratification level in the CVC shown by longer injection gap produces relatively homogeneous mixture compared to the short injection gaps. The results show that the fuel stratification level has a less significant effect on the combustion stages although it has a significant effect on the combustion phasing. Stratified mixture produces shorter combustion duration especially on the main and final combustion stages, while homogeneous mixture produces longer duration. This trend is applicable for both GCNG and DCNG mixtures.