Exploring the Role of Reactivity Gradients in Direct Dual Fuel Stratification.
Typical compression ignition engines operate with conventional diesel combustion (CDC), in which high-reactivity fuels such as diesel are burned in a high-temperature, mixing-controlled (i.e., diffusion-limited), process. In other words, the rate of combustion is determined by the speed at which the fuel can vaporize and mix with oxygen. When simultaneously mixing and burning fuel, there is a tendency for some areas in the reaction zone to have high equivalence ratio ([PHI]), resulting in high levels of particulate matter (PM) and mono-nitrogen oxides (N[O.sub.x]) . These are both criteria pollutants which contribute to smog formation and human respiratory illness, and are therefore subject to strict regulations. Exhaust after-treatment devices are required on modern diesel engines in order to meet emissions targets in most markets. However, this after-treatment typically comes at the expense of fuel economy, performance, and/or secondary reducing agents, adding complexity and upfront and recurring costs [2,3].
Much of the work on advanced combustion strategies has been focused on reducing the dependence on after-treatment while maintaining high thermal efficiency. One promising area of research is low-temperature combustion (LTC), which uses long ignition delays (achieved by premixing the fuel and oxygen) to mitigate the formation pathways for PM and N[O.sub.x], while simultaneously reducing heat transfer [4,5,6]. One of the biggest drawbacks to the use of long ignition delays is that ignition becomes kinetically controlled, and therefore subject to small stochastic variations in boundary conditions. As a consequence, combustion control becomes more difficult, confining operation to a smaller range than CDC. Improving the control and operating range of LTC strategies has been an active area of research [7,8].
Homogeneous charge compression ignition (HCCI), one of the first LTC strategies to be developed, is by definition a fully premixed mixture of fuel and air, with a kinetically-controlled ignition event which is path-dependent on temperature, pressure, equivalence ratio, and fuel properties [9,10]. Many researchers have explored the benefits of HCCI, which include high efficiency and orders-ofmagnitude reduction of PM and N[O.sub.x] when compared to CDC [7,8,9,10,11]. While this would seem to nicely address the shortcomings of CDC, the HCCI strategy has its own significant drawbacks. The volumetric nature of HCCI heat release results in a high peak heat release rate (PHRR) and peak pressure rise rate (PPRR), leading to excessive combustion noise even at moderate loads. This can be mitigated by retarding combustion phasing and extending combustion duration, but both come at the expense of reduced efficiency. This leads to a trade-off between efficiency and noise, and limits the peak practical load [8,12]. Because ignition is kinetically controlled, HCCI is also very sensitive to boundary conditions, a problem that is exacerbated by a lack of fast-response combustion phasing control variables typically available in more traditional combustion strategies, such as fuel injection and spark timing.
As it became evident that a fully homogenous charge was not suitable over a wide range of operating conditions, a variety of partially premixed compression ignition (PPCI) strategies were developed with the idea of maintaining the benefits of HCCI while adding control. In general, PPCI strategies seek to increase the operating range and stability by direct-injecting either a portion or all of the fuel to create stratification of the fuel charge, and therefore equivalence ratio and temperature, within the cylinder. This causes variations in local ignition delay throughout the cylinder, which has the overall effect of increasing combustion duration and retarding combustion phasing, resulting in decreased PHRR and noise. Stratification can be increased by increasing the amount of direct-injected fuel and also by retarding its injection, and the primary difference between the various strategies is the duration and timing of the fuel injections. These strategies range from mostly-premixed, such as partial fuel stratification (PFS) , to direct-injecting most or all of the fuel just before combustion starts, as with partially premixed combustion (PPC) [13,14,15]. Between these extremes exists a continuum of stratification levels and methods of achieving them, with each having their relative merits [16,17,18]. Successful operation has been demonstrated over a wide range of speeds and loads, requiring optimization of stratification and fuel selection, boost level, exhaust gas recirculation (EGR), intake temperature, and other variables for each combination of load and speed. While many of the problems with HCCI have been mitigated to some extent, there still exist several issues to address:
1. High levels of unburned hydrocarbons (HC) and carbon monoxide (CO) due to low combustion temperatures and premixed fuel entering crevice regions.
2. High boost is required at high loads in order to maintain low equivalence ratio. This is compounded by low exhaust enthalpy, resulting in the requirement of high combined turbocharger efficiency. If a supercharger is used instead, there will be a significant penalty to brake thermal efficiency in order to drive the required boost.
3. High EGR is required at high loads in order to maintain sufficient ignition delay to avoid excessive combustion noise.
4. Lack of sufficiently independent control over duration and phasing of heat release with a single fuel, resulting in a tradeoff between efficiency and noise as combustion phasing is retarded with increasing load.
Reactivity controlled compression ignition (RCCI) seeks to addresses the issue of combustion duration control by extending the concept of control by stratification to include fuel reactivity. By injecting two fuels with different levels of resistance to auto-ignition, both the reactivity and equivalence ratio are stratified throughout the cylinder. This adds an additional fast-response control parameter in the form of fuel chemistry, enabling the global fuel properties to change across the operating map. Simulations and experiments of RCCI are typically performed with port-injection of low-reactivity fuel, such as gasoline (often modeled as iso-octane), and direct-injection of high-reactivity fuel, such as diesel (often modeled as n-heptane). Examples include the work of Inagaki et al. , Kokjohn et al. [20,21,22], Hanson et al. [23,24], and Splitter et al. [25,26,27,28,29]. Experimental operation with RCCI has demonstrated gross indicated thermal efficiencies in excess of 56% and in-cylinder levels of soot and N[O.sub.x] below the EPA 2010 Heavy Duty emissions regulations . As with other LTC strategies, RCCI produces relatively high levels of HC and CO, which are primarily caused by low wall temperatures and the inability to completely oxidize fuel that has entered crevice regions .
By adjusting the ratio of high- and low-reactivity fuel, the RCCI strategy allows the shape and phasing of heat release to be adjusted with much greater fidelity than is possible in single-fuel strategies. For a given set of operating conditions, the combustion phasing can be controlled entirely by the fuel ratio, and to lesser degree, the timing of the high-reactivity fuel injections. However, the operating range is limited by the reactivity range of the fuels being used. As load increases, the fraction of high-reactivity fuel required to achieve a given combustion phasing decreases. If the load is increased to a high enough point, the fraction of high-reactivity trends to zero, and there is no longer any direct control over the combustion process, as the combustion strategy has essentially become HCCI. For operation with gasoline and diesel, this point is reached well before full load is achieved, although it can be extended by decreasing compression ratio (CR) at the expense of efficiency. Load can also be extended by decreasing the reactivity of the low-reactivity fuel, with examples such as gasoline-ethanol blends  and natural gas . However, if operation is desired with conventional fuels at compression ratios typical of diesel engines, it is evident that some additional method of control is required. One path is to explore the possibility of stratifying both fuels.
In our previous work , we demonstrated that control over the heat release event could be further improved by coupling RCCI with a diffusion-limited injection of gasoline in a strategy called direct dual fuel stratification (DDFS), which is illustrated alongside RCCI and PPC in Figure 1. Like RCCI, the DDFS strategy utilizes an early injection of gasoline that is almost completely premixed, along with a direct injection of diesel to establish a reactivity gradient along which the first stage of combustion will propagate. As this paper will show, the timing and quantity of the diesel injection affords a large degree of control over the early heat release. DDFS also uses an injection of gasoline just before top dead center (TDC), which was inspired by the highly-stratified injection schedule used in PPC. However, unlike PPC, in which the late gasoline injection is still afforded some mixing time before the start of combustion, the late gasoline injection in DDFS enters after the conclusion of low-temperature heat release (LTHR) into an already-burning charge, and its burn rate is limited primarily by diffusion rather than kinetics. The present work will focus on the role of the reactivity gradients created by the diesel injection and how they enable control over the heat release profile in the DDFS strategy.
The experimental setup was identical to that used in the previous work , and as such, only a brief description of the relevant components will be provided here. The experiments were performed on a 2.44 L Caterpillar 3401E Single Cylinder Oil Test Engine (SCOTE), with specifications listed in Table 1, laboratory setup as illustrated in Figure 2, and modified piston shown in Figure 3.
Dual Direct Injector Head
A stock engine head was modified to allow the mounting of two common rail injectors (CRI), shown in Figure 4. Full details of the design can be found in previous work . The injector specifications are listed in
Table 2. The naming convention for the injectors is that the first subscript refers to the physical location (1 = exhaust side, 2 = intake side), and the second subscript is the number of holes in the nozzle. The stock 7-hole nozzle was used for injecting diesel, but a 10-hole nozzle was used for gasoline injection, as this was seen to reduce soot levels in previous work .
Measurements and Methodology
Cylinder pressure was measured with a Kistler 6125C21U20 piezoelectric pressure transducer in conjunction with a Kistler 510 charge amplifier. The cylinder pressure traces were processed as described in the previous work , and quantities such as PPRR and sound pressure level (SPL) were calculated from individual cycles, thereby generating a statistical distribution for each variable.
Heat release calculations for experimental combustion work are often performed by using a single-zone heat-addition model that assumes the contents of the cylinder can be modeled as an ideal gas at uniform temperature and composition with fixed mass. The chemical heat addition Qch is then defined as the rate of conversion from chemical to sensible internal energy, and can be represented as
[[??].sub.ch] = [mc.sub.v] dT/dt (+) P dV/dt - [[??].sub.w] (1)
where m is the total mass in the cylinder and [[??].sub.w] is the rate of heat transfer into the cylinder from the walls. Equation (1) can be combined with the ideal gas law and simplified into the commonly-used expression shown in Equation (2), in which [gamma] represents the ratio of specific heats.
[[??].sub.ch] = (1/[gamma]-1) v dP/dt+ ([gamma]/[gamma]-1) p dV/dt - [[??].sub.w] (2)
When used with experimental data, this approach requires models for both [gamma], which is dependent on temperature and composition, and [[??].sub.w], which is typically dominated by turbulence. While models for both exist for the typical spark-ignited gasoline and compression-ignited diesel combustion regimes, no such models exist for the dual-fuel, multistage combustion presented in this work. As a substitute, we make an additional assumption that the process is locally polytropic, with polytropic index n. This allows for [[??].sub.w] to be represented directly as
[[??].sub.w] = [mc.sub.n] dT-dt (3)
where [c.sub.n] is the polytropic specific heat capacity, which is given by
[C.sub.n] = [C.sub.v] (_ R/n-1) (4)
By substituting Equations (3) and (4) into Equation (1) and proceeding the same as before, we arrive at an expression for the apparent heat release rate (AHRR) in terms of the polytropic index n:
AHRR = [[[??].sub.ch] = (1/n-1) V dP/dt + (n/n-1) p dV/dt (5)
The result in Equation (5) is quite similar to that from Equation (2), with the difference being that the substitution of n for [gamma] has the effect of including the heat transfer.
The assumption of a polytropic process is quite good during the compression and expansion strokes, but is obviously incorrect during the combustion process. However, for the purposes of calculating combustion parameters such as PHRR and crank angle at 50% of total heat release (CA50), the use of an effective polytropic index produces reasonable and repeatable results even during combustion mode changes. For the results shown here, n was calculated for both the early compression stroke and late expansion stroke. The value for the compression stroke was used until the start of combustion, and the value for the expansion stroke was used after combustion. The transition between the two values was modeled by a logistic function centered at the location of PHRR and with a width proportional to the duration of heat release in an iterative procedure. Note that by using the measured polytropic index for the start and end points, flat tails are guaranteed on the resulting accumulated heat release traces.
PM measurements of filter smoke number (FSN) and mass per volume (mg/[m.sup.3]) were performed with an AVL 415s smoke meter and averaged among 5 samples of 2 L volume each, taken with paper saving mode off. The mass density reported by the smoke meter, which is representative only of the black carbon in the exhaust, was used to calculate the PM level in g/kW-hr. Investigations of the PM generated by RCCI have shown that unlike CDC, it often has a very high fraction of organic carbon (>95%) and a significantly different size distribution, which is bimodal under certain conditions [35,36,37]. Thus, we do not expect that the mass density reported by the smoke meter will be an accurate measure of total PM when using RCCI or similar combustion strategies such as DDFS, and these measurements are therefore referred to as soot for the remainder of the paper. All gaseous emissions measurements were performed with a 200-series Horiba emissions bench. Hydrocarbon (HC) emissions were measured on a wet basis using a heated sampling system. All other emissions were sampled on a dry basis. EGR rate was determined by the ratio of intake to exhaust C[O.sub.2] concentrations. Gaseous emissions were averaged for 120 seconds after attaining steady-state for several minutes. The emissions levels prescribed for N[O.sub.x], non-methane hydrocarbons, CO and PM within the EPA emissions standards for heavy-duty diesel engines that were phased in from 2007-2010  will be referred to collectively as EPA 2010 HD. These values are provided for reference only, and the methods by which they are measured for compliance with federal law may vary considerably from those used here.
The engine was operated at 1300 rev/min and a nominal load of 0.9 MPa gross indicated mean effective pressure (IMEPg) for all tests. The fuels used were EPA Tier II EEE Gasoline and 2007 Certification Diesel, both of which were supplied by Haltermann Solutions. Selected properties of each fuel are listed in Table 3. As the operator's console was located inside the engine test cell adjacent to the engine, noise and safety considerations dictated a conservative peak cylinder pressure of 15 MPa and a PPRR of 1.5 MPa per crank angle degree (CAD) to be imposed as upper limits for the duration of the experiment. The location of TDC was fixed at the beginning of each experiment by setting peak motored cylinder pressure to 0.4[degrees] before top dead center (BTDC). This loss angle is based on hardware determinations performed by Caterpillar for this engine platform, and agrees well with hot and cold motored traces taken on this particular engine setup. Data was acquired at approximately steady-state conditions, as determined by monitoring the pressure and temperature of the surge tanks as well as the power output and emissions levels over a span of several minutes.
All parameter sweeps are presented as variations on the baseline operating conditions, shown in Table 4. The nominal target load was maintained by fixing the total delivered fuel energy across all tests. The intake surge tank and EGR temperatures were chosen to represent practical operating conditions for this load, while avoiding condensation in the EGR cooler and intake systems .
RESULTS AND DISCUSSION
Fuel and Injection Pressure
The baseline condition for the second injection (diesel fuel) is at a timing that would make it a so-called "squish conditioning" injection in RCCI . The idea is that the diesel fuel injected in this window creates a region of higher reactivity and equivalence ratio in the squish region, causing combustion to start there rather than everywhere in the charge nearly simultaneously, as occurs in HCCI. In other words, the reactivity and equivalence ratio stratification created by this injection extends the heat release and gives control over when it begins. With the ability to inject both fuels directly, one of the first questions we want to address is the relative impact of reactivity and equivalence ratio on this injection, i.e., what would be the impact of using gasoline instead of diesel for this injection?
The fuel test matrix is shown in Table 5, and all other conditions were kept at the baseline values shown in Table 4. Because gasoline was injected at 100 MPa and diesel at 50 MPa in the baseline case, an intermediate point with diesel at 100 MPa was also performed so that the fuel effect could be compared directly. The duration was adjusted to maintain constant total fuel energy, as indicated in Table 5.
The pressure and heat release traces for the second injection fuel test are shown in Figure 5. It is apparent that injection pressure had little effect when diesel was used. The phasing was slightly advanced due to the shorter duration at higher pressure, which increased the local reactivity. Note that the cases with diesel are exceptionally stable, with the shaded regions of one standard deviation being essentially indiscernible from the ensemble average values for pressure and heat release. When the second injection fuel was switched from diesel to gasoline, the pressure and duration were the same, and therefore, the retarded combustion phasing and high instability of the gasoline case can be attributed to a fuel effect. The main fuel differences here are volatility, which influences mixing, and reactivity, which influences ignition delay. This figure shows the limited control that is available with equivalence ratio stratification alone, and the flexibility that is gained by direct-injecting a highly reactive fuel.
The in-cylinder pressure metrics for the second injection fuel test are shown in Figure 6. The CA50 for the gasoline case was retarded by [approximately equal to]10[degrees] and the PPRR and SPL increased significantly due to the shorter combustion duration and higher PHRR. The gross indicated mean effective pressure (IME[P.sub.g]) was also observed to decrease by roughly 10% when operating with gasoline alone.
The emissions and efficiency results for the second injection fuel test are shown in Figure 7. Increasing the pressure of the diesel injection resulted in a slight increase in N[O.sub.x] and a slight decrease in CO, HC, and soot as a result of the shorter injection duration, which caused areas of higher reactivity and temperature. This led to a small increase in both combustion efficiency ([[eta].sub.comb]) and gross indicated thermal efficiency ([[eta].sub.gross]). For the gasoline case, the N[O.sub.x] was nearly identical to the 100 MPa diesel case, suggesting similar peak temperatures. However, the retarded combustion phasing led to incomplete combustion of HC and CO, which resulted in a significant decrease in [[eta].sub.comb]. The retarded phasing also decreased the effective expansion ratio, which when combined with the decreased [[eta].sub.comb], led to a [[eta].sub.gross] reduction of almost 10% relative to the 100 MPa diesel case. The values for turbocharger efficiecy ([[eta].sub.turbo]) and volumetric efficiency ([[eta].sub.vol]) were essentially unchanged between cases, indicating that the boundary conditions were consistent throughout.
The only benefit of using gasoline for the second injection was a decrease in soot by over an order of magnitude. It is unknown whether this was due primarily to soot being generated directly by diesel fuel components when it was used for the second injection, or because using diesel fuel advanced combustion, causing the third injection to be diffusion-limited upon entering a hotter and more reactive charge. When gasoline was used as the second injection, the third injection was partially premixed because of the extra time before combustion begins, and none of the benefits of the DDFS strategy were realized. Therefore, we observe that equivalence ratio stratification alone is insufficient to achieve this strategy, and that reactivity stratification is a key enabler for high-fidelity heat release control.
Timing and Duration
Having established that it is advantageous to use diesel for the second injection, the effects of its timing and duration were investigated. When changing duration, fuel had to be removed from one of the other injections in order to maintain constant total fuel energy. If fuel were removed from the first injection, the equivalence ratio at [SOI.sub.3] would remain the same, but the total charge reactivity would be higher. Changes due to this shift in fueling would be relatively easy to understand within the context of previous RCCI work. If fuel were instead removed from the third injection, the equivalence ratio and reactivity at [SOI.sub.3] would increase, and the quantity of fuel in the diffusion-limited regime would be decreased. It would be difficult to separate the compounded effect of these changes, as each would have an impact on combustion phasing and emissions. Therefore, the total fueling was moderated by changing the duration of the first injection.
The test matrix for the timing and duration sweeps of the second injection is shown in Table 6, and all other conditions were kept at the baseline values shown in Table 4. Because there was only a single diesel injection, it was possible to measure the energy content of that injection directly from fuel flow, and therefore the duration sweep was equivalent to a sweep of diesel energy fraction. The [SOI.sub.2] sweep around the baseline value of [Dur.sub.2] was the most inclusive, ranging from -200 to -10[degrees]ATDC. From this sweep, a region of interest between -80 to -30[degrees]ATDC was identified, and the [SOI.sub.2] sweeps at the other levels were constrained to this range.
Pressure and heat release traces from the longest SOI2 sweep are presented in two figures for clarity. The results from -200 to -30[degrees]ATDC are shown in Figure 8. The points in the sweep between -200 and -100[degrees]ATDC have been excluded for clarity, as there was little change in combustion phasing within that range, indicating that the diesel was mostly premixed for [SOI.sub.2] before -100[degrees]ATDC. The effect of diesel stratification is evident for [SOI.sub.2] between -100 and -40[degrees]ATDC, in which the combustion phasing advanced as the diesel stratification increased. At -30[degrees]ATDC, the large bump at the beginning of the heat release suggests that the diesel may be transitioning to a diffusion-limited burn. This notion is reinforced by Figure 9, which shows the results from -30 to -10[degrees]ATDC. Here we observe that combustion phasing retards with SOI2, representing a reversal of the trend from -200 to -30[degrees]ATDC.
As with the previous plots, the LTHR traces are presented in two figures corresponding to the same range of [SOI.sub.2]. In Figure 10, the LTHR appears to have advanced and increased slightly in magnitude for [SOI.sub.2] between -200 and -50[degrees]ATDC. There was a slight shift in the beginning of the LTHR at -40[degrees]ATDC, and then a major change at -30[degrees]ATDC, which can be interpreted as the onset of diffusion-limited combustion of the diesel injection. For [SOI.sub.2] after -30[degrees]ATDC, shown in Figure 11, the LTHR was significantly diminished and was presumably a product only of the premixed gasoline from the first injection and any residual fuel from the previous cycle.
The in-cylinder pressure metrics for the sweeps of [SOI.sub.2] and [Dur.sub.2] are shown in Figure 12. There appears to be little effect of stratification for [SOI.sub.2] earlier than -120[degrees]ATDC, suggesting that the diesel fuel in this range is effectively premixed. CA50 advanced as [SOI.sub.2] was retarded from -120 to -40[degrees]ATDC, which was observed in Figure 8 to simultaneously increase the combustion duration and reduce the PHRR, thereby reducing noise in this range. This effect was increased at higher diesel energy fraction. IME[P.sub.g] was also observed to increase in this range, which will be addressed with the following figures. For [SOI.sub.2] from -30 to -10[degrees]ATDC, CA50 was retarded with [SOI.sub.2], with a corresponding increase in noise and decrease in IME[P.sub.g]. We can conceptually separate these results into three ranges: premixed for [SOI.sub.2] before -120[degrees]ATDC, RCCI between -120 and -40[degrees]ATDC, and diffusion-limited after -40[degrees]ATDC.
The increase in combustion duration with advancing CA50 initially seems to be counterintuitive from the perspective of a premixed combustion strategy, and can be better understood by examining the behavior at the end of the main heat release event, as shown in Figure 13. The heat release from the third injection can be thought of as a discrete event that is partially coincident with the premixed combustion. Depending on when the third injection occurs and what the in-cylinder conditions are at that time, there may be some ignition delay in which there is time for the third injection to premix before burning, leading to a premixed spike before a diffusion-limited burn, as is typical of CDC. It is evident from Figure 13 that the change in the in-cylinder conditions caused by the retarding of the second injection has the effect of advancing the diffusion-limited burn at the end of the main heat release event while also decreasing its magnitude. It can also be seen that the end of the heat release occured at a nearly identical point regardless of the second injection timing, as evidenced by the merging of the heat release profiles near 15[degrees]ATDC. This indicates that under these conditions, the duration of the third injection is long enough to ensure that the last fuel to enter the cylinder burns in a diffusion-limited manner that is essentially independent of whatever occurred beforehand. In other words, the third injection allows us to effectively pin the end of the heat release in place, such that advancing the start of heat release has the effect of increasing combustion duration, which reduces PHRR and noise.
The emissions results for the sweeps of [SOI.sub.2] and [Dur.sub.2] are shown in Figure 14. As with the pressure metrics, there appears to be little stratification for [SOI.sub.2] earlier than -100[degrees]ATDC, with the exception of an increase in soot at -120[degrees]ATDC. The N[O.sub.x] emissions were unaffected by the diesel fuel stratification until the diffusion-limited regime was encountered, but for HC, CO, and soot, there was a trough in the RCCI range, with a minimum between -60 and -50[degrees]ATDC. The reduction of HC and CO was improved with higher diesel fraction, but soot had the opposite trend, indicating that the soot was generated either by the diesel fuel itself, or as a consequence of the advanced combustion phasing that occurred with increased diesel fraction.
The efficiency results for the sweeps of [SOI.sub.2] and [Dur.sub.2] are shown in Figure 15. As a result of the reduced HC and CO in the RCCI range, we observe an increase in [[eta].sub.comb], which combined with the advanced CA50, resulted in increased [[eta].sub.gross] and [[eta].sub.net], with a peak at -60[degrees]ATDC. For the case with 7% diesel, the peak value of [[eta].sub.gross] (-60[degrees]ATDC) was 4% higher on a relative basis than the value at the most premixed condition (-200[degrees]ATDC). The values for [[eta].sub.turbo] and [[eta].sub.vol] were essentially constant, indicating that the conditions in the exhaust and intake tanks did not change significantly during these sweeps.
Extended Duration Sweep
The results from the timing and energy fraction sweeps of the second injection suggest that the quantity of diesel fuel injected has a strong effect on the combustion phasing, noise, emissions, and efficiency when operating in the RCCI range of injection timings. In order to better understand these effects, an extended sweep of diesel energy fraction was performed at the baseline timing, as described in Table 7. All other conditions were kept at the baseline values shown in Table 4.
The pressure and heat release traces for the extended sweep of diesel energy fraction are shown in Figure 16. For 0% diesel, the combustion was essentially PPC with an early pilot and a late stratified injection, resulting in retarded combustion phasing and rapid heat release. As the diesel fraction was increased, the early HRR advanced and became more gradual. As the start of main heat release advanced up to and beyond the start of the third injection, this caused the third injection to transition from partially-premixed to almost entirely diffusion-limited. This is further evidenced by the behavior at the end of the main heat release event, as shown in Figure 17. As the diesel fraction was increased, the end of heat release transitioned from a short bump and tail with relatively high cyclic variability to an elongated bump and tail with low variability, which looks quite similar to the diffusion-limited regime encountered in CDC. It can also be seen that the tails became coincident near 15[degrees]ATDC (in a similar way to those shown in Figure 13) once the diesel fraction was increased beyond 4%.
The LTHR traces for the extended sweep of diesel energy fraction are shown in Figure 18. As the diesel fraction increased, the peak of the LTHR advanced and increased in magnitude, leading to an earlier start of main heat release.
The in-cylinder pressure metrics for the extended diesel energy fraction sweep are shown in Figure 19. As the diesel energy fraction increased, CA50 advanced in an asymptotic manner, which was mirrored by an asymptotic decrease in noise and increase in IME[P.sub.g]. The reduction in noise was a direct result of the reduced PHRR caused by the advanced CA50.
The emissions results for the extended sweep of diesel energy fraction are shown in Figure 20. As with the sweeps of [SOI.sub.2], N[O.sub.x] emissions were unaffected by varying the diesel fraction. HC and CO decreased asymptotically as the diesel fraction increased, indicating that the high reactivity of diesel promoted oxidation. Soot increased from essentially zero with no diesel, to near the EPA 2010 HD limit with 18% diesel. The source of this soot could be the diesel fuel itself, or it could from the late gasoline injection, which becomes increasingly diffusion-limited as the combustion phasing advances due to the increasing diesel fraction. One way to gain more insight on the source of the soot would be to explore the use of other high-reactivity fuels with a lower sooting potential, such as cetaneimproved gasoline, oxygenated diesel fuels, or dimethyl ether.
The efficiency results for the extended diesel energy fraction sweep are shown in Figure 21. The essentially constant values for [[eta].sub.turbo] and [[eta].sub.vol] indicate consistent boundary conditions throughout the sweep. Because of the decrease in HC and CO, [[eta].sub.comb] increased asymptotically with diesel fraction. Both [[eta].sub.gross] and [[eta].sub.net] increased quickly at first due to improving [[eta].sub.comb] and advancing CA50, but as the diesel fraction increased beyond 11%, a significant fraction of the heat release occurred before TDC, and the duration also increased, limiting the work extraction. Based on the diminishing returns for [[eta].sub.comb] and [[eta].sub.gross] and the steady increase in soot with increasing diesel fraction, an optimum value would appear to be in the range of 6 to 12% diesel, which justifies the continued use of 7% diesel as the baseline point at this load.
The importance of reactivity stratification was highlighted by replacing diesel with gasoline, which proved to be much less effective at controlling the start of heat release. Consequently, all of the other tests were performed with diesel as the second injection. Increasing injection pressure was observed to have only minor benefit, indicating that high-pressure injection hardware may not be necessary for the high-reactivity fuel. Sweeps of [SOI.sub.2] identified three possible regimes for the second injection: premixed, RCCI, and diffusion-limited. The RCCI regime was superior in several ways, including improved control over CA50 and noise, substantial reduction of HC, CO, and soot, improved indicated efficiency, and N[O.sub.x] comparable to the premixed regime. Increasing the energy fraction of the diesel injection was shown to reduce noise and improve indicated efficiency, albeit with diminishing returns and increased soot. The relative importance of these factors would determine the optimum value for the diesel energy fraction, but the baseline value of 7% was seen to offer a reasonable compromise.
DDFS has the unique ability to independently control combustion duration and combustion phasing by combining reactivity stratification with a diffusion-limited gasoline injection. There are multiple ways to take advantage of this ability: Efficiency can be improved across the engine map by always operating at optimum combustion phasing and by using increased compression ratios, which are enabled by the increased combustion duration and reduced noise; combustion duration control without the need to retard phasing enables high load operation without sacrificing efficiency; and the combination of the diesel and late gasoline injections offer robust fast-response control mechanisms with a wide range of authority.
[1.] Heywood, J.B., Internal Combustion Engine Fundamentals, McGraw-Hill, 1988.
[2.] Theis, J., Ura, J., Li, J., Surnilla, G. et al., "N[O.sub.x] Release Characteristics of Lean N[O.sub.x] Traps During Rich Purges," SAE Technical Paper 2003-01-1159, 2003, doi:10.4271/2003-01-1159.
[3.] Johnson, T., "Diesel Emissions in Review," SAE Int. J. Engines 4(1):143-157, 2011, doi: 10.4271/2011-01-0304.
[4.] Kalghatgi, G., Risberg, P., and Angstrom, H., "Partially Pre-Mixed Auto-Ignition of Gasoline to Attain Low Smoke and Low N[O.sub.x] at High Load in a Compression Ignition Engine and Comparison with a Diesel Fuel," SAE Technical Paper 2007-01-0006, 2007, doi:10.4271/2007-01-0006.
[5.] Kimura, S., Aoki, O., Ogawa, H., Muranaka, S. et al., "New Combustion Concept for Ultra-Clean and High-Efficiency Small DI Diesel Engines," SAE Technical Paper 1999-01-3681, 1999, doi:10.4271/1999-01-3681.
[6.] Yanagihara, H., "A study of DI diesel combustion under uniform higher-dispersed mixture formation", JSAE Review, 1997, 18, pg. 247-254.
[7.] Sjoberg, M. and Dec, J., "EGR and Intake Boost for Managing HCCI Low-Temperature Heat Release over Wide Ranges of Engine Speed," SAE Technical Paper 2007-01-0051, 2007, doi:10.4271/2007-01-0051.
[8.] Dec, J., Yang, Y., and Dronniou, N., "Boosted HCCI - Controlling Pressure-Rise Rates for Performance Improvements using Partial Fuel Stratification with Conventional Gasoline," SAE Int. J. Engines 4(1):1169-1189, 2011, doi: 10.4271/2011-01-0897.
[9.] Shibata, G. and Urushihara, T., "Auto-Ignition Characteristics of Hydrocarbons and Development of HCCI Fuel Index," SAE Technical Paper 2007-01-0220, 2007, doi: 10.4271/2007-01-0220.
[10.] Shibata, G. and Urushihara, T., "Realization of Dual Phase High Temperature Heat Release Combustion of Base Gasoline Blends from Oil Refineries and a Study of HCCI Combustion Processes," SAE Int. J. Engines 2(1):145-163, 2009, doi:10.4271/2009-01-0298.
[11.] Dec, J. and Yang, Y., "Boosted HCCI for High Power without Engine Knock and with Ultra-Low N[O.sub.x] Emissions - using Conventional Gasoline," SAE Int. J. Engines 3(1):750-767, 2010, doi: 10.4271/2010-01-1086.
[12.] Yun, H., Wermuth, N., and Najt, P., "High Load HCCI Operation Using Different Valving Strategies in a Naturally-Aspirated Gasoline HCCI Engine," SAE Int. J. Engines 4(1):1190-1201, 2011, doi:10.4271/2011-01-0899.
[13.] Manente, V., Tunestal, P., Johansson, B., and Cannella, W., "Effects of Ethanol and Different Type of Gasoline Fuels on Partially Premixed Combustion from Low to High Load," SAE Technical Paper 2010-01-0871, 2010, doi: 10.4271/2010-01-0871.
[14.] Manente, V., Johansson, B., Tunestal, P., and Cannella, W., "Influence of Inlet Pressure, EGR, Combustion Phasing, Speed and Pilot Ratio on High Load Gasoline Partially Premixed Combustion," SAE Technical Paper 2010-01-1471, 2010, doi:10.4271/2010-01-1471.
[15.] Shen, M., Tuner, M., Johansson, B., and Cannella, W., "Effects of EGR and Intake Pressure on PPC of Conventional Diesel, Gasoline and Ethanol in a Heavy Duty Diesel Engine," SAE Technical Paper 2013-01-2702, 2013, doi: 10.4271/2013-01-2702.
[16.] Hardy, W. and Reitz, R., "A Study of the Effects of High EGR, High Equivalence Ratio, and Mixing Time on Emissions Levels in a Heavy-Duty Diesel Engine for PCCI Combustion," SAE Technical Paper 2006-01-0026, 2006, doi:10.4271/2006-01-0026.
[17.] Hanson, R., Splitter, D., and Reitz, R., "Operating a Heavy-Duty Direct-Injection Compression-Ignition Engine with Gasoline for Low Emissions," SAE Technical Paper 2009-01-1442, 2009, doi:10.4271/2009-01-1442.
[18.] Sellnau, M., Moore, W., Sinnamon, J., Hoyer, K. et al., "GDCI Multi-Cylinder Engine for High Fuel Efficiency and Low Emissions," SAE Int. J. Engines 8(2):2015, doi:10.4271/2015-01-0834.
[19.] Inagaki, K., Fuyuto, T., Nishikawa, K., Nakakita, K. et al., "Dual-Fuel PCI Combustion Controlled by In-Cylinder Stratification of Ignitability," SAE Technical Paper 2006-01-0028, 2006, doi:10.4271/2006-01-0028.
[20.] 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.
[21.] Kokjohn, S., Hanson, R., Splitter, D., Kaddatz, J. et al., "Fuel Reactivity Controlled Compression Ignition (RCCI) Combustion in Light- and Heavy-Duty Engines," SAE Int. J. Engines 4(1):360-374, 2011, doi:10.4271/2011-01-0357.
[22.] Kokjohn, S., Reitz, R., Splitter, D., and Musculus, M., "Investigation of Fuel Reactivity Stratification for Controlling PCI Heat-Release Rates Using High-Speed Chemiluminescence Imaging and Fuel Tracer Fluorescence," SAE Technical Paper 2012-01-0375, 2012.
[23.] 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.
[24.] Hanson, R., Kokjohn, S., Splitter, D., and Reitz, R., "Fuel Effects on Reactivity Controlled Compression Ignition (RCCI) Combustion at Low Load," SAE Int. J. Engines 4(1):394-411, 2011, doi: 10.4271/2011-01-0361.
[25.] Splitter, D., Reitz, R., and Hanson, R., "High Efficiency, Low Emissions RCCI Combustion by Use of a Fuel Additive," SAE Int. J. Fuels Lubr. 3(2):742-756, 2010, doi: 10.4271/2010-01-2167.
[26.] Splitter, D., Kokjohn, S., Rein, K., Hanson, R. et al., "An Optical Investigation of Ignition Processes in Fuel Reactivity Controlled PCCI Combustion," SAE Int. J. Engines 3(1):142-162, 2010, doi:10.4271/2010-01-0345.
[27.] Splitter, D., Hanson, R., Kokjohn, S., and Reitz, R., "Reactivity Controlled Compression Ignition (RCCI) Heavy-Duty Engine Operation at Mid-and High-Loads with Conventional and Alternative Fuels," SAE Technical Paper 2011-01-0363, 2011, doi:10.4271/2011-01-0363.
[28.] Splitter, D., Hanson, R., Kokjohn, S., Wissink, M. et al., "Injection Effects in Low Load RCCI Dual-Fuel Combustion," SAE Technical Paper 2011-24-0047, 2011, doi: 10.4271/2011-24-0047.
[29.] Splitter, D., Wissink, M., Kokjohn, S., and Reitz, R., "Effect of Compression Ratio and Piston Geometry on RCCI Load Limits and Efficiency," SAE Technical Paper 2012-01-0383, 2012, doi:10.4271/2012-01-0383.
[30.] 40 C.F.R. [section] 86.007-11
[31.] Nieman, D., Dempsey, A., and Reitz, R., "Heavy-Duty RCCI Operation Using Natural Gas and Diesel," SAE Int. J. Engines 5(2):270-285, 2012, doi:10.4271/2012-01-0379.
[32.] Wissink, M. and Reitz, R., "Direct Dual Fuel Stratification, a Path to Combine the Benefits of RCCI and PPC," SAE Int. J. Engines 8(2):878-889, 2015, doi:10.4271/2015-01-0856.
[33.] Wissink, M., "Direct Injection for Dual Fuel Stratification (DDFS): Improving the Control of Heat Release in Advanced IC Engine Combustion Strategies", Doctoral Dissertation, University of Wisconsin - Madison, 2015.
[34.] Shahlari, A.J., Hocking, C., Kurtz, E., Ghandhi, J., "Comparison of Compression Ignition Engine Noise Metrics in Low-Temperature Combustion Regimes", SAE Int. J. Engines 6(1):541-552, 2013, doi:10.4271/2013-01-1659.
[35.] Prikhodko, V. Y., Curran, S. J., Barone, T. L., Lewis, S.A., Storey, J. M., Cho, K., Wagner, R. M., Parks, J. E. II, "Diesel Oxidation Catalyst Control of Hydrocarbon Aerosols from Reactivity Controlled Compression Ignition Combustion," ASME IMECE2011-64147, 2011.
[36.] Kolodziej, C., Wissink, M., Splitter, D., Hanson, R. et al., "Particle Size and Number Emissions from RCCI with Direct Injections of Two Fuels," SAE Technical Paper 2013-01-1661, 2013, doi:10.4271/2013-01-1661.
[37.] 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.
The authors would like to acknowledge support from Caterpillar.
AHRR - Apparent heat release rate
ATDC - After top dead center
BTDC - Before top dead center
CA50 - Crank angle at 50% of total heat release
CAD - Crank angle degrees
CDC - Conventional diesel combustion
CI - Confidence interval
CO - Carbon monoxide
C[O.sub.2] - Carbon dioxide
CR - Compression ratio
[CRI.sub.n,m] - Common rail injector in position n with m holes
DDFS - Direct dual fuel stratification
[Dur.sub.i] - Duration of ith injection
EGR - Exhaust gas recirculation
EPA - Environmental Protection Agency
EVC - Exhaust valve closure
EVO - Exhaust valve opening
HC - Hydrocarbons
HCCI - Homogeneous charge compression ignition
HRR - Heat release rate
IME[P.sub.g] - Gross indicated mean effective pressure
[Inj.sub.i] - ith injection
IVC - Intake valve closure
IVO - Intake valve opening
LTC - Low-temperature combustion
LTHR - Low-temperature heat release
N[O.sub.x] - Oxides of nitrogen
PPCI - Partially premixed compression ignition
PFI - Port fuel injector
PM - Particulate matter
PHRR - Peak heat release rate
PPC - Partially premixed combustion
PPRR - Peak pressure rise rate
PSD - Power spectral density
RCCI - Reactivity controlled compression ignition
SCOTE - Single cylinder oil test engine
SOI; - Start of injection i
SPL - Sound pressure level
TDC - Top dead center
Martin Wissink and Rolf Reitz
University of Wisconsin
Table 1. 3401E SCOTE specifications Displacement [L] 2.44 Bore [mm] 137.2 Stroke [mm] 165.1 Connecting Rod Length [mm] 261.6 Squish Height [mm] 1.57 Number of Valves 4 LVO ([degrees]ATDC) 335 LVC ([degrees]ATDC) -143 EVO ([degrees]ATDC) 130 EVC ([degrees]ATDC) -355 Swirl Ratio (stock) 0.7 Piston Lype Articulated Compression Ratio 14.88:1 Table 2. Common rail fuel injector specifications Injector Name [CRI.sub.1,7] [CRI.sub.2,10] Body Style Bosch CRI 2 Series Body Part Number 0986435088 Nozzle Part Number DLLA148P1347 DLLA143PV3201426 Nozzle Angle [[degrees]] 148 143 Hole Diameter [[micro]m] 141 117 Number of Holes 7 10 Hydraulic Flow (cc per 30 440 440 sec @ 10 MPa) Table 3. Selected properties of test fuels Fuel Haltermann Product Code Test Method EPA Tier II 2007 Cert. EEE Diesel HF0437 HF0582 Density [kg/L] ASTM D4052 0.742 0.858 H/C Ratio [mol/mol] ASTM E191 1.845 ASTM D5291 1.771 Oxygen [weight %] ASTM D4815 None Detected Sulfur [ppm] ASTM D5453 5 9 Net Heating Value [kJ/g] ASTM D240 42.715 42.145 Research Octane Number ASTM D2669 96.8 Motor Octane Number ASTM D2700 88.6 Cetane Number ASTM D613 45.0 Cetane Index ASTM D4737 44.8 Distillation Curve [[degrees]C] Initial Boiling Point 31 189 10% 52 220 50% ASTM D86 103 271 90% 156 322 End Boiling Point 208 345 Table 4. Target operating conditions for DDFS baseline case Fuel Diesel Gasoline Injector [CRI.sub.1,7] [CRI.sub.2,10] Pressure [MPa] 50 100 Energy Ratio [%] 7 93 Parameter Value Parameter Value Speed [RPM] 1300 [Inj.sub.1] Fuel Gasoline EGR [%] 40 [SOI.sub.1] [[degrees]ATDC] -340 [T.sub.EGR] 90 [Dur.sub.1] [ms] 1.4 [[degrees]C] [T.sub.in] 50 [Inj.sub.2] Fuel Diesel [[degrees]C] [P.sub.in] [kPa] 186.1 [SOI.sub.2] [[degrees]ATDC] -60 [P.sub.exh] [kPa] 206.8 [Dur.sub.2] [ms] 0.6 [PHI] [-] 0.58 [Inj.sub.3] Fuel Gasoline [PHI]' [-] 0.35 [SOI.sub.3] [[degrees]ATDC] -5 IME[P.sub.g] 0.9 [Dur.sub.3] [ms] 0.8 [MPa] [E.sub.fuel] 4.7 [kJ/cyc] Table 5. Test matrix for second injection fuel and injection pressure Inj2 Fuel Diesel (baseline) Diesel Gasoline Pressure [MPa] 50 100 100 Duration [ms] 0.6 0.45 0.45 Table 6. Test matrix for timing and energy fraction of second injection Diesel Energy Fraction [%] 4 7 11 [Dur.sub.1] [ms] 1.48 1.40 1.30 [Dur.sub.2] [ms] 0.49 0.60 0.71 [SOI.sub.2] [[degrees]ATDC] -80 to -30 -200 to -10 -80 to -30 Table 7. Test matrix for extended sweep of diesel energy fraction Diesel Energy Fraction [%] 0 1 4 7 11 18 Duri [ms] 1.67 1.60 1.48 1.40 1.30 1.20 [Dur.sub.2] [ms] 0 0.35 0.49 0.60 0.71 0.87
|Printer friendly Cite/link Email Feedback|
|Author:||Wissink, Martin; Reitz, Rolf|
|Publication:||SAE International Journal of Engines|
|Article Type:||Technical report|
|Date:||Jun 1, 2016|
|Previous Article:||Conventional and Low Temperature Combustion Using Naphtha Fuels in a Multi-Cylinder Heavy-Duty Diesel Engine.|
|Next Article:||Cylinder-to-Cylinder Variations in Power Production in a Dual Fuel Internal Combustion Engine Leveraging Late Intake Valve Closings.|