Cylinder-to-Cylinder Variations in Power Production in a Dual Fuel Internal Combustion Engine Leveraging Late Intake Valve Closings.
Combustion variability is a common problem in advanced combustion modes. While advanced combustion modes including low temperature combustion (LTC) strategies can provide high efficiencies, the utilization of exhaust gas recirculation, fuels of differing reactivities, as well as early fuel injection timings can introduce high variability in the quantity and composition of the trapped mass consequently leading to variations in the combustion process [1,2]. Cycle-to-cycle and cylinder-to-cylinder variations of the combustion process are observed as differences in combustion phasing, heat release rates, and peak pressure, as well as the indicated mean effective pressure (IMEP). These variations in the combustion process have been found to have three primary causes 1) variation in gas motion in the cylinder during combustion; 2) variation in the amounts of fuel, air, and recycled exhaust gas trapped in the cylinder each cycle; and 3) variation in the mixing process of the composition within the cylinder . Although cycle-to-cycle variations pose significant challenges in advanced combustion engines, the scope of this work is limited to the investigation of combustion variations across cylinders.
Prior efforts have explored the driving factors of these variations and have investigated methods for improving combustion stability by strategically choosing particular operating parameters. For example, a study by Gong et. al  investigated the effects of injection and ignition timings, engine speed, and load on the cycle-by-cycle combustion variation of a direct-injection spark ignition (DISI) engine fueled with methanol. Their work showed that a specific fuel injection and ignition timing reduced combustion variability by promoting even mixture composition distribution. Additionally, an increase in engine load resulted in lower cyclic variations in IMEP and similarly, an increase in engine speed resulted in lower COV, which is explained by an increase of swirl intensity that forms a more uniform mixture concentration distribution. Selim  also investigated different parameters that affect cyclic variability and engine noise in a dual fuel engine. These parameters included the type of gaseous fuel, engine speed, load, and compression ratio, as well as fuel mass and injection timing. His findings also showed that optimal settings of the mentioned parameters can help reduce cycle-to-cycle variation and reduce combustion noise. A study on a dual fuel multi-cylinder engine by Bach et. al.  identified uneven distribution of the port-injected fuel and significant cylinder imbalances due to the common inlet plenum and the pressure pulsations, especially during valve overlap.
In this study, experimental testing on a heavy-duty multi-cylinder engine reveal that introducing late intake valve closing (LIVC) along with a dual fuel combustion technique can lead to significant cylinder-to-cylinder variations of the combustion process. LIVC is used to control trapped in-cylinder charge mass and subsequently the peak compression pressure and temperature. Additionally, dual-fuel strategy allows the control of fuel reactivity and helps improve combustion efficiency. Despite the benefits of LIVC and dual-fuel strategy, combining these two techniques resulted in efficiency losses due to the variability of the combustion process across cylinders.
Dual fuel combustion concepts were developed as a strategy to control combustion timing and extend the load limits in LTC operations . The dual fuel combustion strategy uses two fuels with different reactivities in order to adjust the overall fuel reactivity levels based on operating conditions, thereby controlling the combustion timing. Furthermore, by port-injecting the low reactivity fuel and direct injecting the high reactivity fuel, in-cylinder reactivity stratification can be improved resulting in more optimal combustion durations and acceptable pressure rise rates, a primary load-limiting factor for many HCCI-type (homogenous charge compression ignition) combustion strategies [7,8]. The benefits of the dual fuel engine in terms of performance and emission have been demonstrated on single cylinder engines [7,8], as well as multi-cylinder engines [9,10,11,12,13].
Furthermore, improvements in fuel efficiency and reductions in emissions have also been achieved by implementing late intake valve closing (LIVC) strategies. LIVC has been shown to improve fuel efficiency by reducing pumping losses . By closing the intake valves after conventional timing during the compression stroke, the effective compression ratios and consequently peak compression pressures and temperatures are reduced . The impact of LIVC on emissions has also been well studied. Nevin et al.  incorporated a hydraulically actuated variable IVC system onto a single cylinder diesel engine and this study revealed that late IVC resulted in a 90% N[O.sub.x] reduction at high speed and intermediate load with constant airflow and no EGR. With the addition of late IVC, less EGR is required to achieve the same N[O.sub.x] emissions, while CO and PM were reduced by nearly 70%. He et al.  also experimentally investigated the feasibility of using late IVC to improve the performance and emissions of a diesel engine with a fully flexible valve actuation system. They found that late IVC reduces approximately 25-50% N[O.sub.x] emissions and more than 95% PM at some operating conditions. The reduction in particulates is due to the longer ignition delay and a subsequent reduction in local fuel rich combustion zones; whereas the reduction in N[O.sub.x] is attributed to low combustion pressure and temperatures-both consequences of delaying the intake valve closing [16,17,18].
This study explores the combination of these two techniques, dual-fuel combustion and LIVC, which have been implemented on a heavy-duty engine to meet the standards of the SuperTruck program focused on demonstrating 50% brake thermal efficiency on a heavy duty engine and the establishment of a pathway to 55% brake thermal efficiency . In this work, the variations in combustion across cylinders stemming from the use of dual fuel and LIVC is investigated. LIVC operations were observed to exhibit significant cylinder-to-cylinder variations of up to 8% in peak pressure, up to 13% in IMEP, and up to 8% in combustion work (the cylinder-to-cylinder variations are calculated as the ratio of the standard deviation to the mean across the six cylinders). These variations are indicative of uneven distribution of fuel across the cylinders. This study seeks to understand the underlying causes for the observed fuel distribution problem in an effort to address this challenge on stock engine architectures.
This study leverages experimental data and a simulation model of the gas exchange process developed using GT-Power to investigate the main factors affecting the fuel distribution. In the following sections, the engine used for this study is described in Experimental Setup followed by the description of the GT-Power model in Model Development. Analysis of the experimentally observed variations in key metric combustion parameters during IVC modulation will be given prior to a discussion of the simulation study undertaken to explore the root causes for the experimentally observed trends. Concluding remarks and a discussion of future work will follow at the end.
The experimental investigation of cylinder-to-cylinder variations was conducted using data collected from a modified 2010 Navistar MaxxForce 13 heavy-duty engine, also used in [13,20-21]. The engine specifications are given in Table 1. As shown in Figure 1, the engine features an air system configured for high EGR delivery, utilizing two-stage turbocharging (with interstage cooling) and a dual-pass EGR cooler with high and low temperature stages.
The engine employs a fully flexible intake valve actuation system developed by Jacobs Vehicle Systems , which uses a lost-motion mechanism to control the length of the intake valve events. To enable dual-fuel operation, the engine has separate direct-injection and port-injection fuel delivery systems. The direct-injection system is used for the high reactivity fuel and a port fuel injection (PFI) system is used for the low reactivity fuel. The high reactivity fuel used is ultra-low sulfur diesel and the low reactivity fuel investigated is compressed natural gas (CNG). PFI timing in the experiments are prior to intake valve opening, and therefore against a closed intake valve. A set of two prototype ECUs are used to control the 12 CNG injectors (each ECU controlling a set of six injectors, one injector per cylinder), in conjunction with an OEM development ECU for the diesel injectors. Cylinder-to-cylinder balancing is not actively engaged for the study - all cylinders are receiving the same set fuel command.
For these tests, the engine cooling system and air system restrictions were tuned to be representative of the conditions encountered in production vehicles. Each cylinder was equipped with a Kistler 6125B pressure transducer which supplied pressure measurements at a resolution of 0.1[degrees]CA. Fuel flow was measured using Micro Motion CMFS010 Coriolis-type flow meters with a measurement accuracy of [+ or -] 0.1%.
While the experimental setup allows for accurate measurement of engine performance and emissions including combustion metrics, a simulation model was also developed to allow the gas exchange process to be more thoroughly investigated. The model used for the simulation studies of the cylinder-to-cylinder variations was built using GT-Power and is illustrated in Figure 2. The model is an inline 6-cylinder compression ignition engine that includes the cylinders connected to the intake manifold through the intake ports and intake runners, and the fuel injection systems connected to the cylinders (direct injection) and the intake ports (port-injection). Further details about GT-Power engine models and the other components in the model (Figure 2) can be found in .
The engine geometric dimensions and piping dimensions (including distances from cylinders to inlet of intake manifold) were adjusted to best match the experimental setup and the effect of the turbocharger and EGR loop were simulated by dictating the inlet conditions (pressure, temperature, and composition) that matched experimental results. This model also uses a mass flow controller at the inlet of the manifold to set the charge mass flow rate to that of the experiments. The settings of the inlet conditions, as well as the mass flow controller (a PID-like regulator), assure that the air handling performance of the engine are accurately depicted by the model.
In part of the investigation, additional fuel injectors were added directly in the intake manifold to investigate how the fuel in the manifold is distributed across cylinders. Fuel flow monitors were added at each intake valve and at the intersections of the manifold and intake runners to allow fuel distribution to be closely examined. The model is used to investigate the gas exchange process, specifically the distribution of the fuel in the intake manifold and across the cylinders. Alternatively, the accurate depiction of the combustion process (start of combustion, combustion duration, heat release rates, etc.) is outside the scope of this study. As such, a simplified combustion model is used for the study. The GT-Power model uses a DI-Jet combustion model with default calibration settings. Further details about the combustion model can be found in .
RESULTS AND DISCUSSION-EXPERIMENTAL OBSERVATION
The effect of IVC timing on combustion was investigated experimentally to determine its impact on cylinder-to-cylinder variations in a dual fuel combustion modes. The operating points investigated represent an IVC sweep from 570[degrees] after top dead center (ATDC) to 620[degrees] ATDC at an engine speed of 1200 RPM and a brake mean effective pressure (BMEP) of 16.5 bar. The engine operating condition for this study is representative of a higher load cruise condition for a class 8 on-highway truck. Combustion variations are analyzed by examining differences in IMEP, peak pressure, as well as differences in combustion and compression work stemming from changes in IVC timing.
IVC impacts key metrics including IMEP. The impact of IVC on the peak pressure level is also of particular interest as peak cylinder pressure is an important constraint at high loads, where the peak pressure constraints of the engine is 220 bar. Figure 3 provides a comparison of the cylinder-specific peak pressures with respect to IVC. Here, each line represents the peak pressure of a specific cylinder at the varying IVC timing. The bars are used to represent the cylinder-to-cylinder variation (ratio of the standard deviation to the mean), and the values for variation should be read using the secondary vertical axis on the right. This representation of combustion characteristics and cylinder-to-cylinder variation is consistent throughout Figures 3, 4, 5, 6, 9, and 11.
As illustrated in Figure 3, all cylinders except cylinder 6 exhibit drops in peak pressure as IVC is retarded; however, the rate at which the peak pressures drop across cylinders 1-5 is very different. Consequently, as displayed by the bars in Figure 3, the cylinder-to-cylinder variation in peak pressure increases significantly as IVC is retarded from 2.4% at an IVC of 570[degrees] ATDC to 7.9% at an IVC of 620[degrees] ATDC.
While peak pressure is a critical parameter, variation in cylinder work output is of primary concern. A comparison of the average IMEP across the cylinders displays similar increases in variation as IVC is retarded (Figure 4). The cylinder-to-cylinder variation in IMEP increases from 3.7% at 570 [degrees]ATDC (IVC) to 12.9% at 620 [degrees]ATDC (IVC). As can be seen from Figure 4, the cylinder-to-cylinder variation in IMEP becomes more pronounced as IVC is incrementally delayed, generating a significant gap between the cylinders with high and low IMEPs. At an IVC of 620 [degrees]ATDC, the highest IMEP is 20.4 bar (Cylinder 6) and the lowest IMEP is 13.9 bar (Cylinder 1). The gap between the IMEP of these cylinders of 6.5 bar represents nearly 40% of the average IMEP, which is very significant and directly attributed to the late IVC timing. These high variations in IMEP represent a large difference in the output across the cylinders and are detrimental to the overall engine performance.
In order to further quantify the cylinder-to-cylinder variation and explain the underlying causes for the high variations in IMEP observed at varying IVC timings, the compression and combustion work were also evaluated for each cylinder and are displayed in Figures 5 and 6 respectively. The pumping work during the intake and exhaust stroke was also calculated but is not presented here as it had minimal effect on the IMEP.
As displayed by the bars in Figure 5, the cylinder-to-cylinder variation in compression work is not significantly affected by changes in IVC timing. All cylinders exhibit similar drops in the work input as IVC is retarded and the cylinder-to-cylinder variation in the compression work is maintained at 2-3%.
In contrast, the combustion work displayed in Figure 6 illustrates a significant increase in cylinder-to-cylinder variation as IVC is retarded. As can be seen from the figure, cylinders 1 and 2 exhibit significant drops in combustion work whereas the other cylinders have relatively similar combustion work output throughout the IVC sweep. At late IVC timings, the combustion work across cylinders is significantly different whereas the compression work is relatively similar. As the IMEP is the representation of the net work output (compression, combustion, and pumping), the comparison of Figures 5 and 6 suggest that the variations in IMEP observed in Figure 4 are almost entirely due to variations in the combustion work.
Considering the large difference in combustion work between the cylinders, the most plausible explanation for the cause of such variation is a disproportionate distribution of the fuel across cylinders. Although the amount of fuel directly injected (high-reactivity) and port-injected (low-reactivity) into each cylinders is identical for these tests, the timing of IVC seems to have a significant effect on the amount of port-injected fuel that remains in each cylinder for combustion.
In the cases where IVC occurs after bottom dead center (BDC-540[degrees]ATDC), the intake valve remains open during the beginning of the compression stroke allowing some of the trapped mass, which includes the port-injected fuel, to flow back out of the cylinders into the intake manifold. As such, the longer the intake valves remain open during the compression stroke, the more fuel is expelled from the cylinders into the intake manifold. Although the fuel is drawn back into the cylinders during the intake stroke, the amount of fuel going into each cylinder is now largely affected by the flow in the intake manifold. This explains the larger variations in IMEP (and combustion work) observed at late IVCs and furthermore explains the higher IMEPs observed in the cylinders furthest away from the intake port. As can be seen from Figure 3, as IVC is retarded, the rate of change in IMEP of each cylinder correlates exactly with the cylinder's position in the inline arrangement. At the latest IVC timings, the cylinders furthest away from the inlet of the intake manifold exhibit the highest IMEP (Cyl 6>Cyl 5>Cyl 4>Cyl 3>Cyl 2>Cyl 1).
Due to the inline arrangement of the cylinders and the position of the intake inlet at the front of the engine, the flow in the intake manifold is inherently from cylinder 1 to cylinder 6. Thus, as fuel escapes from each cylinder into the intake manifold during the compression stroke, most of the fuel is pushed towards the cylinders furthest away from the intake port leading to an uneven distribution of the fuel in the intake manifold that favors the cylinders furthest away from the intake port. Consequently, the cylinders' position in the inline arrangement is indicative of the relative amount of fuel available for combustion as well as the relative IMEP.
The disproportionate distribution of the port-injected fuel in the intake manifold explains the increasing cylinder-to-cylinder variations in IMEP (as well as peak pressure and combustion work) as IVC is retarded and the observed correlation between the inline arrangement of the cylinders and the relative IMEP levels between the cylinders. The main cause of variation is the increasing amount of fuel that is dissipated from the cylinders into the intake manifold as the intake valve is opened longer in the compression stroke. In the next sections, simulation studies from a GT-Power model allow a closer look of the fuel flow at the valve and throughout the intake system. The factors causing this uneven distribution of the port-injected fuel is further studied using the simulation model.
RESULTS AND DISCUSSION-SIMULATION STUDIES
In this section, the cylinder-to-cylinder variations are investigated by taking a closer look at the flow of the port-injected fuel flow using a simulation model developed with GT-Power. The simulation study will provide a better understanding of the fuel distribution across the cylinders by simulating the effects of IVC, intake runner length, and load on the fuel distribution. The design of the intake system, specifically the length of the intake runner, as well as the effective compression ratio determined by the IVC timing were found to have the most significant effect on the fuel distribution across the cylinders.
The findings from the simulation studies are discussed in 2 sections. First, the main aspects affecting the flow of the port-injected fuel are illustrated. These include the backflow of fuel during the compression stroke from each cylinder and the uneven redistribution of the fuel reaching the intake manifold. Second, the impact of the intake runner length and IVC timing are investigated by observing the resulting effect of changes in the effective compression ratio and the relative volume of the intake manifold on the fuel distribution across cylinders.
Fuel back/low and re-distribution
In the experimental studies, the cylinder-to-cylinder variations in both IMEP and peak pressure were observed to increase as IVC is retarded. The large differences in IMEP and peak pressure observed across cylinders (38% and 24% respectively at 610[degrees] ATDC IVC) were best attributed to uneven fuel distribution across cylinders. In this section, a GT-Power model is used to identify how the fuel flow in the intake system varies at different IVC timings and pinpoint the driving causes of uneven fuel distribution.
As suggested in the experimental studies, a larger variation in fuel distribution is observed at late IVC timings due to the larger amount of trapped mass and fuel that is pushed back into the manifold the longer the valves remain (partially) open in the compression stroke. The GT-Power model is used to demonstrate this effect by monitoring the fuel flow at the valves and at the intersection of the runners and manifold at different IVC timings; two of which are illustrated in Figure 7. In Figure 7, the fuel flow at the intake valve (representing the fuel going in and out of the cylinder) as well as the fuel flow at the manifold/runner intersection (representing the amount of fuel coming from and going into the intake manifold) are plotted alongside the intake valve lift profile and the port fuel injection profile for two operating conditions with varying IVC timing. These plots represent the fuel flow across the runner and valve of cylinder 1, but are representative of the effect of IVC timings across all cylinders.
Figure 7 clearly demonstrates that backflow of fuel occurs during the initial stages of the compression stroke (540-660[degrees] ATDC) during which the intake valve is still partially open. As illustrated in the figure, as the intake valve is kept open longer in the compression stroke, a greater amount of fuel is pushed out of the cylinders ultimately reaching the intake manifold. In the upped subfigure of Figure 7, the intake valve closure is at 570 [degrees]ATDC, and fuel predominantly flows into the intake valve as shown by the blue line. Flow into and out of the intake manifold is shown by the red line and is quite minimal in this case. However, when a late IVC timing is applied, as in Figure 7b, there is significant backflow of fuel at the intake valves during the early stages of the compression stroke. The fuel flow between cylinders and the manifold also become more significant at late IVC timings.
The figure also illustrates the backflow of fuel right when the intake valve is opened (~340-360[degrees] ATDC). This region represents the backflow of the hot burned gas from the cylinder into the intake port right when the valve is opened, which pushes a portion of the injected fuel to the manifold. In the late IVC operations in which large cylinder-to-cylinder variation is observed (>3%), the backflow of fuel in this region represents only a small fraction of the total amount of fuel that reaches the intake manifold and is thus not further emphasized in the rest of the analysis.
At the late IVC timings, more fuel reaches the intake manifold. As expected, the fuel that is pushed into the manifold is eventually redistributed across the cylinders. In the ideal case, the fuel reaching the intake manifold is evenly redistributed across cylinders; and, the higher amount of fuel backflow at the late IVC timings would have no additional effect in increasing cylinder-to-cylinder variation. However, as observed experimentally, later IVC operations result in higher variation; thus, this is suggestive of uneven redistribution of the fuel across cylinders.
In order to study the actual redistribution of the fuel reaching the intake manifold, six additional fuel injectors were added to the model directly in the intake manifold across each cylinders' runner (see Figure 2) and were individually operated to analyze the amount of fuel that reaches each cylinder from each fuel injection point. As previously discussed in the experimental studies, the effect of the dominating direction of the flow in the intake manifold appeared to be quite significant. The distribution of the fuel injected in the intake manifold is illustrated in Figure 8 for two operating loads with two different charge mass flow rates (in which the air-fuel ratio is maintained constant).
Figure 8 illustrates the portion of fuel that reaches each cylinder based on the injection point. For example, the area enclosed in red indicates that 29.4% of the fuel injected at point 3 (man-fs-3 in Figure 2) went into cylinder 5. The rest of the boxes should be read accordingly. As anticipated, the distribution of the fuel from the intake manifold is largely disproportionate across cylinders. The cylinders furthest away from the inlet of the manifold (cylinders 5 and 6) are favored, while very minimal or basically no fuel from the intake manifold reaches the ones closest to the inlet (cylinders 1 and 2). This is a direct consequence of the dominating direction of the flow in the intake manifold (from cylinder 1 towards cylinder 6), which limits the flow of fuel in the opposing direction.
Figure 8. Distribution of fuel in intake manifold across cylinders at two operating loads. Values below 5% have been omitted. Fuel Injection Point (*) (a) Mass flow rate: 100g/s Fuel reaching Cyl 1 1 2 3 4 5 6 Cyl 6 14.0% 37.0% 21.1% 45.1% 61.1% 52.0% Cyl 5 15.5% 31.6% 29.4% 40.8% 29.2% 42.2% Cyl 4 19.4% 16.5% 47.4% 11.0% 9.1% Cyl 3 48.9% 14.0% Cyl 2 Fuel Injection Point (*) (b) Mass flow rate: 400g/s Fuel reaching Cyl 1 1 2 3 4 5 6 Cyl 6 13.3% 42.5% 15.0% 34.5% 81.1% 61.5% Cyl 5 14.9% 32.6% 28.5% 60.1% 16.9% 36.3% Cyl 4 52.4% 12.6% 55.8% 5.2% Cyl 3 18.3% 11.4% Cyl 2 52.4% (*) Fuel injection points correspond to the flow volumes labelled man-fs-# in Figure 2. -The area enclosed in red indicates 29.4% of the fuel injected at point 3 (man-fs-3 in Figure 2) went into cylinder 5. -Note: There is no fuel reaching cylinder 1 from the intake manifold. Note: Table made from bar graph.
While similar trends are seen at both operation points, some variations are observed due to the different pressures in the intake manifold. However, the effect of load is relatively minimal considering the large difference between the two chosen points (4x difference) and the minor difference in the fuel distribution. The effect of the direction of the flow in the manifold is much more pronounced and plays a more vital role. The flow of fuel is generally away from the inlet of the intake manifold (for example, fuel pushed back out of cylinder 3, is mainly redistributed between cylinders 4-6); and as such, the cylinders furthest away from the inlet of the intake manifold receive more fuel and this poor fuel distribution is the main source for the observed cylinder-to-cylinder variations.
In summary, the variation in the amount of trapped fuel across cylinders is more significant at late IVC timings in part due to the larger amount of the total injected fuel that is ultimately pushed into the intake manifold. Furthermore, due to the dominating direction of the flow in the intake manifold, the fuel that reaches the manifold is unevenly redistributed across the cylinders favoring the cylinders furthest away from the inlet of the manifold. The next section evaluates the relative roles of IVC timing and the design of the intake system on the fuel distribution. As will be shown, one additional parameters that has significant effect on the fuel distribution is the length of the intake runner. Both the impact of IVC timing and runner length on fuel distribution are investigated in the next section.
Effect of Intake Runner Length and IVC Timing
One possible method of addressing the uneven fuel distribution at late IVC timings is by introducing a new intake system which features longer intake runners. Such an implementation would reduce the amount of fuel that is pushed into the manifold and would decrease the variations in the net amount of fuel going into each cylinders. As the length of the intake runner and the IVC timing both play a vital role in the distribution of the port-injected fuel across the cylinders, their relative impact on fuel flow are analyzed and discussed in this section.
In Figure 9, the results from a GT-Power simulation displaying the net amount of port-injected fuel reaching each cylinder as a function of the intake runner length are illustrated. As shown, the cylinder-to-cylinder variation drops significantly as the runner length is increased. The IVC is set at 610[degrees] ATDC throughout the sweep of intake runner length displayed in Figure 9 and the cylinder-to-cylinder variations in the trapped port-fuel amount ranges from 31% at 50mm runner length to 5% at 530mm. Although the drop in variation is significant, this decrease is achieved at the cost of increasing the length of the intake runner by a considerable amount (which may not in most cases be practical).
The drop in variation as the runner length is increased is anticipated and can be explained using Figure 10. As shown in the figure, although the amount of fuel that is pushed back out of the cylinders remains relatively constant with varying runner length, the fuel reaching the intake manifold drops almost linearly as the runner length is increased. An increase in runner length represents a larger volume in the intake system pertaining to each individual cylinder before the common intake manifold; thus, the longer the intake runner length, the smaller the amount of fuel escaping into the intake manifold and consequently smaller variations in fuel distribution across cylinders. The secondary x-axis in Figure 10 represents the ratio between the runner volume and the cylinder displacement volume. The increase in runner length from 50 mm to 530 mm represents an increase of 603 mL in the runner volume, whereas the cylinder volume swept between 570 [degrees]ATDC to 610 [degrees]ATDC is 526 mL.
Similarly, changes in IVC timings significantly affect the cylinder-to-cylinder variations in fuel distribution. Figure 11 illustrates the predicted effect of IVC timing on the fuel distribution from the simulation model. As previously seen in Figures 5, 6, and 8, late IVC timings are well associated with increases in the cylinder-to-cylinder variation of the combustion process. Figure 11 specifically shows the increasing difference in trapped fuel as IVC is retarded. The sweep of IVC timings in Figure 11 represent an operating condition in which the length of the intake runners is fixed at 290mm. The increase in cylinder-to-cylinder variation observed in Figure 11 follows the same trends as cylinder-to-cylinder variations in peak pressure and IMEP observed in the experimental studies (Figures 3 and 4). The variations in fueling, as displayed in Figure 11, range from 2 % at 570 ATDC IVC to 19 % at 610 ATDC IVC. These variation levels are higher than the variations in IMEP and peak pressure observed experimentally, and this should be expected. IMEP is affected by the amount of both fuels so the impact of the high variations in the amount of port-injected fuel will be reduced by the evenly distributed direct-injected fuel. Furthermore, the peak pressure across each of the cylinders is simultaneously affected by the changes in the effective compression ratio stemming from the changes in IVC timings. Thus, the variations in peak pressure are not as accentuated as that of the trapped fuel. The consistent trends between fuel variations and the difference in peak pressure, IMEP, and combustion work further affirm that the increase in variation in the combustion process at late IVC timings is due to uneven fuel distribution.
The increasing variation in fuel distribution as IVC is retarded can be explained considering IVC's effect on the effective compression ratio (calculated as the ratio of the volume at IVC to that at TDC). Figure 12 demonstrates the linear relationship between the effective compression ratio and the relative amount of fuel pushed back out of the cylinders. In Figure 12, the Y-axis represents the fraction of the amount of fuel that is pushed back out of each cylinder relative to the total amount of fuel entering the cylinder. As the effective compression ratio is reduced (by retarding IVC), a larger portion of the fuel entering the cylinder is pushed back out. As previously seen in Figure 8, the fuel pushed back into the manifold is unevenly distributed across the cylinders. Thus, as IVC is retarded, cylinder-to-cylinder variations in the fuel distribution are amplified.
In this study, the variation in power production across cylinders on an engine leveraging dual fuel combustion and late IVC is investigated. Despite the benefits of dual fuel combustion in extending the load limitation of low temperature combustion modes as well as enabling the control of the combustion timing, and the benefits of late IVC timing in reducing pumping losses, their simultaneous use leads to uneven fuel distribution across cylinders which negatively impacts the engine's overall performance and efficiency.
In late IVC timing operations, the intake valves are maintained partially opened during the compression stroke. Consequently, a portion of the trapped mass and some of the port-injected fuel in pushed back out into the intake manifold during the period in the compression stroke in which the valves are still open. The fuel that is pushed into the intake manifold is unevenly distributed across the cylinders due to the dominating direction of the flow that restrict the flow towards the cylinders closest to the inlet of the intake manifold. The uneven fuel distribution results in varying IMEP, peak pressure, and combustion work across cylinders and is an undesirable feature of the dual-fuel and late IVC operation.
In this study, both experimental data and simulation studies are leveraged to develop a better understanding of the underlying causes of uneven fuel distribution and the associated effects on the combustion process. Through experimental testing, the cylinder-to-cylinder variations observed in IMEP, peak pressure, and combustion work were specifically quantified. The cylinder-to-cylinder variations (calculated as the ratio of standard deviation to the mean) in these parameters at the latest IVC timing (610[degrees] ATDC) were as follows: IMEP-13%, peak pressure-8%, and combustion work-8%. The difference between the maximum and minimum values were even more significant: IMEP-38%, peak pressure-24%, and combustion work-21%. The large variations represent very different and unbalanced combustion process across cylinders, and is best explained by the uneven distribution of the port-injected fuel.
The simulation model, developed using GT-Power, allowed the analysis of the fuel flow at different locations in the engine intake system, which would not be practical from an experimental standpoint. The analysis of the fuel flow revealed that, indeed, a significant maldistribution of the port-inject fuel results from late IVC operations.
This study provides insight into one of the challenges associated with combustion modes that leverage a port-injection strategy along with late IVC operations. The study shows that the cylinder-to-cylinder variations in the combustion process significantly increase mainly due to the uneven distribution of the port-injected fuel. Although, an upstream throttle-body injection or a venturi mixer at the inlet of the manifold would result in more even fuel distribution across cylinder; due to manifold dynamics, there is no guarantee of even fuel mass delivery between cylinders. Additionally, single-point upstream injection would eliminate one of the control variables used to optimize and balance other effects influencing cylinder-to-cylinder variation in dual-fuel combustion.
This work identifies the main parameters that affect the fuel distribution and identifies trends between the operating condition and the fuel flow. Future work will look into leveraging these trends in order to build control strategies that can be implemented to mitigate the cylinder-to-cylinder variation in the combustion process that stems from the uneven fuel distribution.
[1.] Bittle, J., Zheng, J., Xue, X. and Jacobs, T. 2012. Cylinder-to-cylinder variation sources in diesel low temperature combustion and the influence they have on emissions. Proceedings of the Central States Section of the Combustion Institute, Park City, Utah.
[2.] Selim, Mohamed Y.e. "Sensitivity of Dual Fuel Engine Combustion and Knocking Limits to Gaseous Fuel Composition." Energy Conversion and Management 45.3 (2004): 411-25, 2004.
[3.] Heywood, John B. "Causes of Cycle-by-Cycle and Cylinder-to-Cylinder Variations." Internal Combustion Engine Fundamentals. New York: McGraw-Hill, 1988. 419-426.
[4.] Gong, Changming, Huang Kuo, Chen Yulin, Jia Jinglong, Su Yan, and Liu Xunjun. "Cycle-by-cycle Combustion Variation in a DISI Engine Fueled with Methanol." Fuel 90.8 (2011): 2817-819.
[5.] Bach, F., Clemens H., Uwe W., Ulrich S., and Christina S. "Low Temperature Gasoline Combustion With Diesel Micro-Pilot Injection in a Six-Cylinder Heavy Duty Engine." ASME 2012 Internal Combustion Engine Division Fall Technical Conference (2012)
[6.] Ickes, A., Hanson, R., and Wallner, T., "Impact of Effective Compression Ratio on Gasoline-Diesel Dual-Fuel Combustion in a Heavy-Duty Engine Using Variable Valve Actuation," SAE Technical Paper 2015-01-1796, 2015, doi:10.4271/2015-01-1796.
[7.] Kokjohn, S., Hanson, R., Splitter, D., and Reitz, R., "Experiments and Modeling of Dual-Fuel HCCI and PCCI Combustion Using In-Cylinder Fuel Blending," SAE Int. J. Engines 2(2):24-39, 2010, doi: 10.4271/2009-01-2647.
[8.] Hanson, R., Kokjohn, S., Splitter, D., and Reitz, R., "An Experimental Investigation of Fuel Reactivity Controlled PCCI Combustion in a Heavy-Duty Engine," SAE Int. J. Engines 3(1):700-716, 2010, doi:10.4271/2010-01-0864.
[9.] Dempsey, A., Curran, S., Storey, J., Eibl, M. et al., "Particulate Matter Characterization of Reactivity Controlled Compression Ignition (RCCI) on a Light Duty Engine," SAE Technical Paper 2014-01-1596, 2014, doi:10.4271/2014-01-1596.
[10.] Curran, S., Hanson, R., and Wagner, R., "Effect of E85 on RCCI Performance and Emissions on a Multi-Cylinder Light-Duty Diesel Engine," SAE Technical Paper 2012-01-0376, 2012," doi: 10.4271/2012-01-0376.
[11.] Dempsey, A., Curran, S., and Reitz, R., "Characterization of Reactivity Controlled Compression Ignition (RCCI) Using Premixed Gasoline and Direct-Injected Gasoline with a Cetane Improver on a Multi-Cylinder Engine," SAE Int. J. Engines 8(2):859-877, 2015, doi:10.4271/2015-01-0855.
[12.] Belaid-Saleh, H., Jay, S., Kashdan, J., Ternel, C. et al., "Numerical and Experimental Investigation of Combustion Regimes in a Dual Fuel Engine," SAE Technical Paper 2013-24-0015, 2013, doi: 10.4271/2013-24-0015.
[13.] Zhang, Y., Sagalovich, I., De Ojeda, W., Ickes, A. et al., "Development of Dual-Fuel Low Temperature Combustion Strategy in a Multi-Cylinder Heavy-Duty Compression Ignition Engine Using Conventional and Alternative Fuels," SAE Int. J. Engines 6(3):1481-1489, 2013, doi:10.4271/2013-01-2422.
[14.] Helmantel, A. and Denbratt, I., "HCCI Operation of a Passenger Car DI Diesel Engine with an Adjustable Valve Train," SAE Technical Paper 2006-01-0029, 2006, doi:10.4271/2006-01-0029.
[15.] Nevin, R., Sun, Y., Gonzalez D. M., and Reitz, R., "PCCI Investigation Using Variable Intake Valve Closing in a Heavy Duty Diesel Engine," SAE Technical Paper 2007-01-0903, 2007, doi:10.4271/2007-01-0903.
[16.] He, X., Durrett, R., and Sun, Z., "Late Intake Valve Closing as an Emissions Control Strategy at Tier 2 Bin 5 Engine-Out N[O.sub.x] Level," SAE Int. J. Engines 1(1):427-443, 2009, doi:10.4271/2008-01-0637.
[17.] M odiyani, R., Kocher, L., Van Alstine, D. G., Koeberlein, E., Stricker, K., Meckl, P., & Shaver, G. (2011). Effect of intake valve closure modulation on effective compression ratio and gas exchange in turbocharged multi-cylinder engines utilizing egr. International Journal of Engine Research, 12(6), 617-631.
[18.] Stricker, K., Kocher, L., Koeberlein, E., Van Alstine, D., & Shaver, G. M. (2012). Estimation of effective compression ratio for engines utilizing flexible intake valve actuation. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 226(8), 1001-1015.
[19.] "Recovery Act-Systems Level Technology Development, Integration, and Demonstration for Efficient Class 8 Trucks (SuperTruck) and Advanced Technology Powertrains for Light Duty Vehicles (ATP-LD)," DE-FOA-0000079, Department of Energy: Washington, D.C., 2009.
[20.] Ickes, A., Wallner, T., Zhang, Y., and De Ojeda, W., "Impact of Cetane Number on Combustion of a Gasoline-Diesel Dual-Fuel Heavy-Duty Multi-Cylinder Engine," SAE Int. J. Engines 7(2):860-872, 2014, doi:10.4271/2014-01-1309.
[21.] Kassa, M., Hall, C., Ickes, A., and Wallner, T., "In-Cylinder Oxygen Mass Fraction Estimation Method for Minimizing Cylinder-to-Cylinder Variations," SAE Technical Paper 2015-01-0874, 2015, doi:10.4271/2015-01-0874.
[22.] Schwoerer, J., Kumar, K., Ruggiero, B., and Swanbon, B., "Lost-Motion VVA Systems for Enabling Next Generation Diesel Engine Efficiency and After-Treatment Optimization," SAE Technical Paper 2010-01-1189, 2010, doi:10.4271/2010-01-1189.
[23.] GT-Power User's Manual, "Multiple Cylinder Turbocharged DI Diesel Engine." GT-Suite Engine Performance Tutorials Version 7.4, Gamma Technologies, 2014. 69-92.
Illinois Institute of Technology, MMAE Department
10 West 32nd Street
Chicago, IL 60616, USA
Experimental data utilized in this study was generated by a project supported by the U.S. Department of Energy (DOE), the National Energy Technology (NETL) office, under cooperative agreement "SuperTruck - Development and Demonstration of a Fuel-Efficient Class 8 Tractor & Trailer" DOE Contract: DE-EE0003303
The submitted manuscript has been partially created by UChicago Argonne, LLC, Operator of Argonne National Laboratory ("Argonne"). Argonne, a U.S. Department of Energy Office of Science laboratory, is operated under Contract No. DE-AC02-06CH11357. The U.S. Government retains for itself, and others acting on its behalf, a paid-up nonexclusive, irrevocable worldwide license in said article to reproduce, prepare derivative works, distribute copies to the public, and perform publicly and display publicly, by or on behalf of the Government.
Mateos Kassa and Carrie Hall
Illinois Institute of Technology
Andrew Ickes and Thomas Wallner
Argonne National Laboratory
Table 1. Engine Specification Displacement Volume 12.4 L Number of Cylinders 6 Stroke 166 mm Bore 126 mm Connecting Rod 251 mm Compression Ratio 17 Diesel Fuel System 2200 bar common rail Air System 2-stage turbocharger (variable vanes on the high pressure turbine)
|Printer friendly Cite/link Email Feedback|
|Author:||Kassa, Mateos; Hall, Carrie; Ickes, Andrew; Wallner, Thomas|
|Publication:||SAE International Journal of Engines|
|Article Type:||Technical report|
|Date:||Jun 1, 2016|
|Previous Article:||Exploring the Role of Reactivity Gradients in Direct Dual Fuel Stratification.|
|Next Article:||Experimental Investigation of Diesel-Ethanol Premixed Pilot-Assisted Combustion (PPAC) in a High Compression Ratio Engine.|