Printer Friendly

Boosted premixed-LTGC / HCCI combustion of EHN-doped gasoline for engine speeds up to 2400 rpm.

ABSTRACT

Previous work has shown that conventional diesel ignition improvers, 2-ethylhexyl nitrate (EHN) and di-tert-butyl peroxide (DTBP), can be used to enhance the autoignition of a regular-grade E10 gasoline in a well premixed low-temperature gasoline combustion (LTGC) engine, hereafter termed an HCCI engine, at naturally aspirated and moderately boosted conditions (up to 180 kPa absolute) with a constant engine speed of 1200 rpm and a 14:1 compression ratio. In the current work the effect of EHN on boosted HCCI combustion is further investigated with a higher compression ratio (16:1) piston and over a range of engine speeds (up to 2400 rpm). The results show that the higher compression ratio and engine speeds can make the combustion of a regular-grade E10 gasoline somewhat less stable. The addition of EHN improves the combustion stability by allowing combustion phasing to be more advanced for the same ringing intensity. The high-load limits of both the straight (unadditized) and additized fuels are determined, and the additized fuel is found to achieve a higher maximum load at all engine speeds and intake pressures tested, if it is not limited by lack of oxygen. The results reveal that the higher loads with EHN are the result of either reduced intake temperature requirements at naturally aspirated conditions or a reduction in heat release rate at higher intake pressures. Such effects are also found to increase the thermal efficiency, and a maximum indicated thermal efficiency of 50.1% is found for 0.15% EHN additized fuel at 1800 rpm and 180 kPa intake pressure. Similar to previous studies, the nitrogen in EHN increases NOx emissions, but they remain well below US-2010 standards. Higher engine speeds are found to have slightly lower NOx emissions for additized fuel at intake boosted conditions.

Keywords: HCCI, LTGC, Autoignition Reactivity, Combustion Stability, Ignition Improvers, 2-Ethylhexyl Nitrate

CITATION: Ji, C., Dec, J., Dernotte, J., and Cannella, W., "Boosted Premixed-LTGC / HCCI Combustion of EHN-doped Gasoline for Engine Speeds Up to 2400 rpm," SAE Int. J. Engines 9(4):2016.

INTRODUCTION

Well premixed low-temperature gasoline combustion (LTGC), also known as homogenous charge compression ignition (HCCI) [1], and partially stratified LTGC (HCCI-like) engines, are considered promising techniques to provide high thermal efficiencies (comparable to diesel engines), ultra-low [NO.sub.x] emissions, and near zero particulate matter (PM) emissions [2, 3, 4, 5, 6, 7, 8, 9, 10]. High-volatility fuels, like gasoline, are advantageous for LTGC/HCCI, as these fuels are relatively easy to form into a premixed or partially premixed charge. As a result, conventional gasoline with an Anti-Knock Index (AKI) = 87 or above has been widely studied in HCCI and HCCI-like engines [e.g., 4, 5, 6]. Despite its high-volatility, the low autoignition reactivity of conventional gasoline is considered to be a handicap for its application in these engines, especially at naturally aspirated and low-to-intermediate intake boost conditions. For example, at naturally aspirated conditions, a large amount of hot residuals or a high intake temperature ([T.sub.in]) is necessary for autoignition due to gasoline's low autoignition reactivity [4] under LTGC/HCCI conditions. As a result, charge densities are much lower than those of typical engines, which is an important factor limiting the loads of LTGC/HCCI engines.

Mixtures of primary reference fuels (PRF) with low octane number (ON) have been used in the studies of HCCI engines [e.g., 11, 12, 13] to overcome the low reactivity of conventional gasoline. Alternatively, low-octane gasoline has been shown to be more suitable for HCCI engines compared with conventional gasoline at naturally aspirated conditions [14]. However, these fuels are not readily available in the market. Recent studies show that some conventional diesel ignition improvers, such as 2-ethylhexyl nitrate (EHN), which is used to improve the Cetane Number (CN) of diesel fuels, can be used to enhance the autoignition of conventional gasoline in HCCI engines [e.g., 15, 16, 17, 18, 19]. Reitz and co-workers [15, 16, 17] found that EHN can enhance the autoignition reactivity of low reactivity fuels such as gasoline (hydrocarbon only, no ethanol), E10 (gasoline containing 10 vol.% ethanol), ethanol, and methanol. Hosseini et al. [18] indicated that a small amount of EHN additive advanced the low-temperature heat release (LTHR) and high exhaust gas recirculation (EGR) rates were required to retard the combustion phasing.

A previous study by Ji et al. [19] was conducted to understand the effect of adding EHN to regular-grade (87 AKI) E10 gasoline on combustion in an HCCI engine with compression ratio (CR) 14:1 and 1200 rpm engine speed. The study found that EHN addition was effective for enhancing the HCCI autoignition reactivity of the E10 gasoline. The required [T.sub.in] can be reduced significantly with a very small amount of this additive ([less than or equal to]0.4%) at naturally aspirated conditions. The reduction in required [T.sub.in] increased the charge density with a commensurate increase in gross indicated mean effective pressure ([IMEP.sub.g]). This effect, combined with the greater allowable CA50 retard with good stability for the additized fuel, allowed the high-load limit to be increased substantially at naturally aspirated and low-boost conditions, by 30% at intake pressure, [P.sub.in] = 100 kPa and 22% at [P.sub.in] = 130 kPa. Ji et al. [19] also found that the EHN produced a small increase in the NOx emissions for HCCI combustion approximately proportional to the additive concentration. About 30% of the nitrate groups in the EHN were converted to NOx. This observation is consistent with that reported by Dempsey [17] and Ickes [20].

The objective of the current work is to further investigate the effects of EHN on E10 gasoline combustion in an LTGC engine. Fully premixed fueling, i.e. HCCI combustion is used to eliminate any potential for changes in mixture formation from affecting the results. Compared to our previous work [19] a higher CR = 16:1 is used for most studies, and engine speeds are varied up to 2400 rpm for naturally aspirated and moderately boosted conditions (up to 180 kPa absolute). Engine speed has been shown to have a significant effect on LTHR and intermediate temperature heat release (ITHR), combustion stability, and operating range [21, 22, 23, 24, 25] of HCCI engines. The effects of EHN on the high-load limits at each intake pressure are studied and compared at engine speeds of 1200, 1800, and 2400 rpm. The combustion stability at high engine speed and high CR is investigated and discussed. The LTHR and ITHR, which can affect stability, of both EHN-additized fuel and straight (unadditized) fuel are also compared at various engine speeds. Finally, the effects of EHN on gross indicated thermal efficiency (TE) and NOx emissions are also investigated at different engine speeds.

The next section describes the experimental facility, data acquisition and analysis techniques, fuel/additive properties, and test conditions. Following this, the results are presented in two parts. The effects of CR = 16:1 compared to CR = 14:1 on the autoignition enhancement of additized and straight fuels in the HCCI engine are first presented and discussed. Then, the second part shows the effects of engine speed on combustion stability and high-load limits of both additized and straight fuels at various [P.sub.in]s. The peak TE and NOx emissions at various [P.sub.in]s. and engine speeds for both fuels are measured and discussed as well. Finally, the study is summarized, and conclusions are drawn in the last section.

EXPERIMENTAL SETUP

Engine Facility

As with our previous study [19], the HCCI research engine was derived from a Cummins B-series six-cylinder diesel engine, which is a typical medium-duty diesel engine with a displacement of 0.98 liters/cylinder. As shown in Fig. 1a, the engine has been converted for single-cylinder operation by deactivating cylinders 1-5. Two CR (14:1 and 16:1) custom pistons are used to study the effect of CR on the autoignition reactivity of the additized fuel. The combustion chamber geometry of two pistons at TDC is shown in Fig. 1b. The CR = 14:1 piston provides an open combustion chamber with a large squish clearance and a quasi-hemispherical bowl and the CR = 16:1 piston has a broad shallow bowl. Both pistons provide a small topland ring crevice with less than 0.9% of the TDC volume. The engine specifications and operating conditions are listed in Table 1.

Premixed fueling is used in this study. The premixed fueling system, shown at the top of the schematic in Fig. 1a, consists of a gasoline direct injector (GDI) mounted in an electrically heated fuelvaporizing chamber and appropriate plumbing to ensure thorough premixing of the vaporized fuel with the air and EGR upstream of the intake plenum. A positive displacement fuel flow meter (Max Machinery P002) is used to determine the amount of fuel supplied.

The intake air is supplied by an air compressor and precisely metered by a sonic nozzle. Real EGR is used in this work. The cooled real EGR is introduced well upstream of the intake plenum to ensure that the intake charge is well mixed. The exhaust pressure is controlled by throttling the exhaust flow so that it is higher than the intake pressure (typically 1-3 kPa in this work), which allows the EGR to flow back into the intake. Since the effect of EGR is determined by the amount of combustion product gases delivered back into the engine, the concentration of complete stoichiometric products (CSP) in the intake is calculated from the EGR level at each operating condition, and the CSP and/or the intake oxygen concentration are used for the analysis in this study. Also, an equivalence ratio based on the total charge-mass, rather than air alone, is used in this work. This equivalence ratio, referred to as charge-mass equivalence ratio ([[phi].sub.m]) is defined as [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] where F/C is the mass ratio of fuel and total inducted charge gas (i.e. fresh air and EGR), and [(F/A).sub.stoich] is the mass ratio of stoichiometric fuel/air mixture for complete combustion. This provides a convenient and consistent way to compare data with the same supplied energy content per unit charge mass (i.e., the same dilution level) for operating conditions with different fuels and different EGR levels. Note that [[psi].sub.m] is the same as conventional air-based [psi] when no EGR is used. It should also be noted that the air-based [psi] is [less than or equal to] 1 for all conditions presented, so combustion is never overall fuel rich.

The intake pressures varied from 100 kPa (simulating naturally aspirated conditions) to 180 kPa for the current study. All pressures given are absolute. The engine is fully preheated and kept at 100[degrees]C during all the tests by means of electrical heaters on the cooling water and lubricating-oil circulation systems. In addition, the intake tank and plumbing are preheated to 50 - 60[degrees]C to avoid condensation of the fuel or water from the EGR gases. An auxiliary heater mounted close to the engine provided precise control of the intake temperature to maintain the desired combustion phasing. Intake temperatures ranged from 60 - 145[degrees]C. The engine was operated with a speed range between 538 to 2400 rpm to study the effect of engine speed on required [T.sub.in], and LTHR and ITHR. The high-load limits are determined and studied at 1200, 1800, and 2400 rpm.

Data Acquisition

Cylinder pressure is measured with a transducer (AVL QC33C) mounted in the cylinder head approximately 42 mm off center. The pressure transducer signals are digitized and recorded at 0.25[degrees] Crank Angle (CA) increments for 100 consecutive cycles for most measurements. The combustion is found to be more unstable with a large coefficient of variation (COV) of [IMEP.sub.g] near high-load limits at high speeds, and thus more cycles are used near high-load limits. More details about the combustion stability and sampling cycles will be given in the next section. The cylinder-pressure transducer is pegged to the intake pressure near bottom dead center (BDC) where the cylinder pressure reading is virtually constant for several degrees. Intake temperatures are monitored using K-type thermocouples mounted in the two intake runners close to the cylinder head. Firedeck temperatures are monitored with a K-type thermocouple embedded in the cylinder head so that its junction is about 44 mm off the cylinder center and 2.5 mm beneath the surface. Surface temperatures are estimated by extrapolating the thermocouple reading to the surface, using the thickness of the firedeck and assuming that its back surface is at the 100[degrees]C cooling-water temperature. For all data presented, 0[degrees]CA is defined as TDC (top dead center) intake (so TDC compression is at 360[degrees]CA). This eliminates the need to use negative crank angles or combined bTDC, aTDC notation.

The crank angle of the 50% burn point (CA50) is used to monitor the combustion phasing, and the 10% burn point (CA10) is used as a representative marker for the hot-ignition point. CA10 and CA50 are determined from the cumulative apparent heat-release rate (AHRR), computed from the cylinder-pressure data (after applying a 2.5 kHz low-pass filter [10]). The start of heat release is set at the minimum point on the AHRR curve before the main heat release peak. For two-stage ignition cases, this minimum point is located between the LTHR peak and main heat release peak. This method provides a consistent measure to compare CA10 and CA50 of main combustion event although LTHR is excluded from the burn-duration calculation. Computations of CA10 and CA50 are performed for each individual cycle, disregarding heat transfer and assuming a constant ratio of specific heats (gamma, [gamma] = [c.sub.p]/[c.sub.v]) [26]. The average of 100 consecutive individual-cycle CA10 or CA50 values are then used to monitor CA10 or CA50 during operation and for the values reported. The reported pressure rise rates (PRRs) and ringing intensities (RIs, see Eq. 1) are computed from the same low-pass-filtered pressure data. For each cycle, the maximum (max.) PRR is analyzed separately with a linear fit over a moving [+ or -]0.5[degrees] CA window. Similar to CA50, these individual-cycle values are then averaged over the 100-cycle data set.

The acceptable knock limit for HCCI engines is often defined in terms of a max. allowable PRR (dP/d[theta], where [theta] is a variable representing [degrees]CA). However, this does not correctly reflect the potential for knock under boosted conditions where the cylinder pressure changes significantly. In this work, the correlation for RI developed by Eng [27] is used as a measure of the potential for engine knock and to avoid knocking operation:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

where [(dP/dt).sub.max], [P.sub.max], and [T.sub.peak] are the max. values of PRR (in real time), pressure, and peak temperature, respectively, [gamma] is the ratio of specific heats, and R is the gas constant. Based on the onset of an audible knocking sound and the appearance of obvious ripples on the pressure trace, a ringing criterion of 5 MW/[m.sup.2] is selected as the ringing limit for operation without knock. A previous study by the authors [28] shows that RI is a good criterion for avoiding knocking combustion and its detrimental effects.

A second method of computing the heat-release rate (HRR) is used for detailed HRR-curve analysis. Here, the heat release is computed in a more refined way from the ensemble-averaged pressure trace (with the 2.5 kHz low-pass filter applied), using the Woschni correlation for heat transfer [26]. The results of this detailed HRR analysis are used for comparisons of the early part of the heat release, leading up to hot ignition, and for comparisons of the shape of the main, high-temperature heat release rate with and without EHN addition.

Exhaust emissions data are also acquired, with the sample being drawn from the downstream of exhaust plenum using a heated sample line (See Fig. 1(a)). CO, C[O.sub.2], HC, NOx, and O2 levels are measured using standard exhaust-gas analysis equipment, as in our previous studies [e.g., 19]. In addition, a second CO2 meter monitored the intake gases just prior to induction into the engine, which allowed the EGR fraction of the intake gases to be computed from the ratio of the intake and exhaust CO2 concentrations.

Fuel/Additives Properties and Test Conditions

A Chevron commercial regular-grade gasoline containing 10 vol.% ethanol (supplied by Chevron) with nominal AKI = 87, hereafter referred as CCG-E10, is used in the current study. The fuel properties can be found in the previous study [19]. Ignition improver EHN with a purity of 97% is used in the current work (most of the impurities are isomers). The additive is mixed well with CCG-E10 to a 0.15 vol.%, and the physical properties of fuel/additive mixtures (e.g. density) are calculated from the known values for CCG-E10 and the additive, based on this ratio.

For each condition tested, the high-load limit is approached by first operating the engine at a relatively stable condition, and then gradually increasing the fueling rate to the knock/stability limit or oxygen/stoichiometric limit [19], whichever came first. During this process, the RI was held constant at 5MW/[m.sup.2] for conditions with boosted intake pressure. For naturally aspirated operation, a lower RI of 3 MW/[m.sup.2] was used because it produced more stable combustion. Incomplete burn cycles or partial burn cycles occurred intermittently at high engine speeds when approaching high-load limits, yet combustion was still under control. Such partial burn cycles resulted in more unburned hydrocarbons in the residual gas and would often cause a significantly higher heat release in next cycle, resulting in a high RI and knock, which might damage the engine. The onset of these abnormal combustion cycles increased COV of [IMEP.sub.g] above 2%. In order to avoid such a circumstance,

the high-load limit here is determined as COV of IMEPg [less than or equal to] 2%. This will be discussed in greater detail in the next section.

Throughout the current work, care was taken to ensure that the data were collected under conditions where the HCCI combustion was clean (NOx < 0.1 g/kWh, near-zero soot), complete (combustion efficiency > 95%), stable (COV of IMEP [less than or equal to] 2%), and not knocking (ringing intensity [less than or equal to] 5 MW/[m.sup.2]), unless otherwise noted.

RESULTS AND DISCUSSION

Effect of Compression Ratio

Increasing the CR from 14:1 to 16:1 causes both the temperature and pressure to increase more rapidly during the compression stroke, producing higher temperatures and pressures near TDC if the intake conditions are kept constant. Because these increases in both temperature and pressure will enhance the autoignition reactivity, adjustments must be made to prevent overly advanced combustion (CA50), which results in a high peak PRR and engine knock. In the current work, data were acquired at the same intake pressures for both CRs, and either the intake temperature or EGR level was adjusted to retard CA50 sufficiently to maintain a RI [less than or equal to] 5 MW/[m.sup.2] to prevent knock. Most intake-boosted data were acquired at RI = 5 MW/[m.sup.2], which gives the most advanced CA50, and hence highest thermal efficiency, without knock. However, for naturally aspirated operation, RI = 3 MW/[m.sup.2] was used because it gives better combustion stability, i.e. less likely to lead to runaway knock or misfire [31].

Straight Unadditized Fuel (no EHN added)

Figure 2 provides a comparison of the pressure and temperature traces prior to hot ignition for CRs of 14:1 and 16:1, for both naturally aspirated ([P.sub.in] = 100 kPa) and boosted ([P.sub.in] = 180 kPa) operation. For both intake pressures, the engine speed is 1200 rpm, and loads are similar for both CRs, with the [IMEP.sub.g] ~~ 483 and 830 kPa for [P.sub.in] = 100 and 180 kPa, respectively. As evident in Fig. 2a, with the higher CR piston, the cylinder pressure reaches a higher pressure before hot ignition, which enhances the autoignition reactivity. To compensate for this pressure-induced reactivity enhancement and the increased temperature rise during compression with the higher CR, [T.sub.in] is reduced by 33[degrees]C (from 135[degrees]C for CR = 14:1 to 102[degrees]C for CR = 16:1) for the [P.sub.in] = 100 kPa data in Fig. 2. For both CRs at medium-to-high intake-boost conditions, the pressure-induced autoignition enhancement becomes so large that [T.sub.in] adjustment alone is no longer sufficient to control CA50. Therefore, [T.sub.in] is reduced to 60[degrees]C (the minimum value used in this study so that fuel condensation does not occur in the intake system), and EGR is added to retard CA50 to maintain RI = 5 MW/[m.sup.2]. As indicated in the legend of Fig. 2, for [P.sub.in] = 180 kPa, the amount of EGR must be increased from 27.7% for CR = 14:1 to 46.0% for CR = 16:1 at the same load.

It is interesting to note the differences in charge temperature for the two CRs and the two boost levels, as shown in Fig. 2b. For [P.sub.in] = 100 kPa, [T.sub.in] is 33[degrees]C cooler for CR = 16:1, but the difference in charge temperature diminishes as the piston moves towards TDC due to the greater temperature rise with compression for CR = 16:1. However, just prior to the onset of combustion (~5[degrees] CA before CA10), the temperature is still lower with CR = 16:1, even though CA10 is the same for both cases. This temperature difference indicates the effect of the greater pressure enhancement of autoignition for CR = 16:1. With [P.sub.in] increased to 180 kPa, the pressure enhancement is so large that [T.sub.in] is reduced to 60[degrees]C for both CRs, and the charge temperature remains well below that of the [P.sub.in] = 100 kPa data up through the time of hot ignition in agreement with the much higher pressures throughout the cycle for [P.sub.in] = 180 kPa (Fig. 2a). In addition EGR must be added for [P.sub.in] = 180 kPa, and the EGR gases have a relatively low [gamma], which acts to reduce the temperature rise with compression. The higher EGR levels for CR = 16:1 mitigate the temperature rise with compression relative to CR = 14:1, and the temperature prior to hot autoignition is only slightly higher for CR =16:1 (Fig. 2b).

Figure 2c shows the early part of the HRR profiles (corrected using Woschni equation for heat transfer [26]) for CR = 16:1 and 14:1 at [P.sub.in] = 100 and 180 kPa at the same conditions as Fig. 2a and 2b. The HRR is normalized by the total heat release (THR) to remove effects due only to differences in the amount of fuel supplied. At [P.sub.in] = 100 kPa, no LTHR is found for either CR because the charge temperature in this region (15~25CA before CA10) is high (870K~940K) due to the high [T.sub.in] required as shown in Fig. 2b. However, Fig. 2c does indicate a slightly stronger ITHR for CR = 16:1 compared to CR = 14:1 (at [P.sub.in] = 100 kPa) as a result of the lower charge temperature and higher pressure (see Fig. 2b and 2a). The ITHR period in Fig. 2c begins at about -15[degrees] CA relative to CA10 and continues to hot ignition. At [P.sub.in] = 180 kPa, LTHR is present for both CRs, but the amount for CR = 16:1 is about half that for CR = 14:1 even though temperatures are similar as shown in Fig. 2b. Due to the higher in-cylinder pressure, more EGR is required at CR = 16:1 (see values in the legend of Fig. 2), which reduces the intake [O.sub.2] and suppresses the LTHR at CR=16:1. The difference in HRR between the two CRs becomes smaller in the ITHR region and eventually vanishes before hot ignition (by -5[degrees] CA relative to CA10), as shown in Fig. 2c. As shown in previous studies [e.g., 4, 5, 6, 19], the combustion stability in an HCCI engine is strongly related to the amount of ITHR. Stronger ITHR allows more retarded CA50 with good combustion stability and therefore, an increased high-load limit with an acceptable RI. Due to the opposite effect of CR on the amount of ITHR at [P.sub.in] = 100 kPa compared to 180 kPa, it is expected that the effect of CR on stability near the high-load limit will vary with [P.sub.in]. Figure 3a shows the high-load limits of straight fuel with no EHN addition at various [P.sub.in]s. with CR = 16:1 for an engine speed of 1200 rpm. For comparison, the high-load limits with CR = 14:1 from our previous study [19] are plotted as well. Similar to CR = 14:1, at CR = 16:1, all high-loads are still limited by knock/stability at all [P.sub.in]s. At [P.sub.in] = 100 kPa, the required [T.sub.in] for the max. load is reduced from 130[degrees]C to 92[degrees]C (a difference of 42[degrees]C) by increasing the CR from 14:1 to 16:1. As [P.sub.in] is increased from 100 to 180 kPa, autoignition reactivity is enhanced for both CRs, but the reactivity is always greater for CR = 16, so a lower [T.sub.in] and/or more EGR is required than for CR = 14. For example, Fig. 3a shows that for CR = 14, increasing [P.sub.in] from 100 to 130 kPa, reduces the required [T.sub.in] to 93[degrees]C, but for CR = 16, [T.sub.in] must be reduced to 60[degrees]C (the minimum value allowed), and EGR is also required to reach the desired CA50 for RI = 5 MW/[m.sup.2], as shown in Fig. 3b. Note that EGR is indicated as percentage of complete stoichiometric products, CSP, in the intake, and it is also shown by the amount of intake [O.sub.2]. Similarly, at [P.sub.in] = 160 and 180 kPa, [T.sub.in] is reduced to 60[degrees]C for both CRs, but more EGR (or CSP) is required and less intake [O.sub.2] is available for CR = 16:1 compared with CR = 14:1 as shown in Fig. 3b. With the high EGR levels, the intake [O.sub.2] is reduced to 11.3% at [P.sub.in] = 180 kPa for CR = 16.

The max. [IMEP.sub.g] changes a little more moderately as a function of [P.sub.in] at CR = 16:1 compared with CR = 14:1, as shown in Fig. 3a. At [P.sub.in] = 100 and 130 kPa, the max. [IMEP.sub.g]s for CR = 16:1 are respectively, 9.7% and 7.6% higher than those for CR = 14:1, but at [P.sub.in] = 160, the max. [IMEP.sub.g]s are nearly the same, and at 180 kPa, it is 3.6% lower for CR = 16:1. As discussed earlier, the LTHR and ITHR are weaker for CR = 16:1 compared with CR = 14:1 at high [P.sub.in]s., which reduces combustion stability, decreasing the high-load limit. In addition, the faster expansion rate with CR = 16:1 also contributes to the decreased combustion stability at high-loads. Due to the reduced combustion stability, the CA50s at max. loads for [P.sub.in] = 160 and 180 kPa are expected to be less retarded for CR = 16:1 compared with CR = 14:1, as shown in Fig. 3c. At naturally aspirated and low intake boost conditions ([P.sub.in] = 130 kPa), however, the required [T.sub.in] is lower for the higher CR. This increases the charge density, which increases in-cylinder charge mass, and therefore, results in a higher maximum load, as discussed in [19] and shown in Fig. 3a. The lower [T.sub.in] also favors LTHR and ITHR as shown in Fig. 2c. Such slightly higher ITHR for CR = 16:1 acts to improve stability, which allows a more retarded CA50 now as shown in Fig. 3c.

Figure 4 compares the COV of [IMEP.sub.g] at CR = 16:1 and 14:1 for fueling ([[psi].sub.m]) sweeps up to the high-load limit points shown in Fig. 3 for [P.sub.in] = 100 and 180 kPa. The COV is found to be below 1% for all the stable points shown in Fig. 4. The last points with CR = 16:1 at [P.sub.in] = 100 and 180 kPa have the COV > 6%, and it is hard to control combustion or keep it stable. Such loads are considered to have unstable combustion and are not counted for the max. loads shown in Fig. 3. It is noted that for CR = 16:1, at the high-load limit, the COV increases dramatically, from 1% to 6%, for less than a 1% increase in the [IMEP.sub.g], as shown in Fig. 4. Similar behavior was also found at CR = 14:1, but the high COV points at slightly high loader loads were not recorded, so are not shown here. This suggests that when approaching the high-load limit, the combustion stability is very sensitive to fueling rate at this engine speed (1200 rpm). Under such circumstances, the combustion is hard to control and easily changes to knocking or misfre runaway. Note that the differences in the lowest load of the sweeps in Fig. 4 are not related to the stable operating range but are simply artifacts of where the data sweeps for the high-load limits were started.

For both [P.sub.in] = 100 and 180 kPa, the COV is a little higher for CR = 16:1 compared to CR =1 4:1. This is thought to be mainly caused by the faster expansion rate with CR = 16:1. For [P.sub.in] = 100 kPa, the slightly higher ITHR at CR = 16:1 (Fig. 2c) tends to mitigate this effect, and the difference in COV between the two CRs is very small. In contrast, at [P.sub.in] = 180 kPa, Fig. 2c shows that ITHRs are quite similar for the two CRs and the COV is noticeably greater for CR = 16:1 over the whole load range. However, it's important to note that the COVs are under 1% for both CRs, so the small increase with CR = 16:1 would not have a noticeable impact on engine performance.

Figure 5a shows a comparison of the TE between CR = 16:1 and 14:1 at [P.sub.in] = 100 and 180 kPa. The TE at CR = 16:1 is consistently 1.2% higher than CR = 14:1 over the entire range of loads for [P.sub.in] = 100 kPa. For [P.sub.in] = 180 kPa, it is about 1.2% higher at low-to-intermediate loads but only ~0.3-0.5% greater at high loads. The higher expansion ratio with CR = 16:1 is the main reason for these increased TEs, but at [P.sub.in] = 100 kPa, TE is also improved by the increase in [gamma] that results from the lower required [T.sub.in] (as also evident from the lower peak temperatures presented below). This TE increase for [P.sub.in] = 100 kPa is mitigated (by about 0.2%-units) by a lower combustion efficiency (CE) for CR = 16:1 as shown in Fig. 5b, which is caused by the lower charge temperature. At [P.sub.in] = 180 kPa, TE is 1.2% higher for low-to-intermediate loads at CR = 16:1 compared with CR = 14:1 due not only to the higher CR, but also to a higher CE (see Fig. 5b) because more EGR is used (see Fig. 3b) so more fuel gets a second chance to burn [4]. As EGR is also required at higher loads with CR = 14:1, the difference in TE gets smaller as the difference in CE decreases with increased load as shown in Fig. 5b. Additionally, EGR decreases [gamma] which is detrimental to TE, and this effect is greater at CR = 16:1 since EGR levels are higher. Although the effect of increased CR still dominates, causing an increase in TE, the net effect of the other factors is that the increase in TE is smaller at high-loads compared with low-to-intermediate loads for [P.sub.in] = 180 kPa.

Additized Fuel (EHN added)

High-load limits of 0.15% EHN additized fuel with CR = 16:1 at various [P.sub.in]s. are shown in Fig. 6. The high-load limits of 0.35% DTBP with CR = 14:1 from our previous study are shown here for comparison as the autoignition reactivity enhancement with 0.35% DTBP has been shown to be very similar to that of 0.15% EHN [19]. Similar to the cases with CR = 14:1, adding EHN enhances the autoignition reactivity with CR = 16:1, which reduces the required [T.sub.in] from 92[degrees]C to 60[degrees]C at [P.sub.in] = 100 kPa. This decrease in [T.sub.in] increases the charge density, allowing the additized fuel to reach a higher max. [IMEP.sub.g] than straight fuel at CR = 16:1. Similar results were also found for CR = 14:1, as shown in Fig. 6 [19]. Also similar to CR = 14:1, the additized fuel at CR = 16:1 is knock/stability limited at [P.sub.in] = 100 kPa and shows a similar high-load limit, which is mostly due to the similar required [T.sub.in] ~~ 60[degrees]C. At [P.sub.in] = 130 kPa, the additized fuel with CR = 16:1 is found to be both knock/stability and oxygen/stoichiometric limited, and it shows a 10.5% higher max. [IMEP.sub.g] than straight fuel even though the required [T.sub.in] = 60[degrees]C for both fuels. This is because adding EHN with CR = 16:1 enhances the ITHR which allows a more retarded CA50 with good stability, which in turn allows higher loads (i.e. higher fueling rates for higher [[psi].sub.m]) with no knock compared to the straight fuel. Similar behavior was previously observed for CR = 14:1 [19]. Fig. 6 also shows that the additized fuel gives a slightly higher max. [IMEP.sub.g] for CR = 16:1 at [P.sub.in] = 130 kPa, compared with additized fuel for CR = 14:1.

Experimental results consistently indicate that the combustion is more stable for the same [[psi].sub.m] when the in-cylinder mixture is near the stoichiometric limit, as occurs for CR = 16:1 due to the higher level of required EGR, rather than fuel lean cases, as for CR = 14:1 with a low or intermediate level of EGR. The reason for this phenomenon remains unclear, but the increased stability allows the additized fuel to reach a higher [[psi].sub.m] at CR = 16:1 than at CR = 14:1, and thus, a higher max. [IMEP.sub.g]. The higher TE with CR = 16:1 also contributes to the higher [IMEP.sub.g].

As noted above, for CR = 16:1, the max. [IMEP.sub.g] point for the additized fuel had reached the oxygen limit at [P.sub.in] = 130 kPa due to the amount of EGR required. Therefore, as [P.sub.in] is further increased (further increasing the fuel reactivity), [[psi].sub.m] must be reduced to accommodate the additional EGR fraction required to prevent overly rapid autoignition and keep RI = 5MW/[m.sup.2]. As a result, the rate of increase in max. [IMEP.sub.g] with [P.sub.in] is less for

[P.sub.in] > 130 kPa, and at [P.sub.in] = 160 kPa, the max. [IMEP.sub.g] for CR = 16:1 for the additized fuel falls below that of CR = 14:1. With CR = 14:1, the additized fuel becomes both knock/stability and oxygen/stoichiometric limited at [P.sub.in] = 160 kPa. For [P.sub.in] > 160 kPa, the CR = 14:1 additized fuel is oxygen limited, and its curve shows a similar change in slope to that observed for CR = 16:1 at [P.sub.in] > 130 kPa, due to the need to accommodate the additional required EGR fraction. Because of this oxygen-limit effect and the greater amount of EGR required for CR = 16:1, the max. [IMEP.sub.g] of additized fuel for CR = 16:1 is 10.5% and 8.2% lower than for CR = 14:1 at [P.sub.in] = 160 and 180 kPa, respectively. Similarly, this oxygen-limit effect causes the max. [IMEP.sub.g] of the additized fuel at CR = 16:1 to be 5.2% and 6.7% lower than that of the straight fuel for [P.sub.in] = 160 and 180 kPa, respectively.

Effect of Engine Speed

Speed Effects on [T.sub.in] and PRR

Figure 7 shows the [T.sub.in] and temperature at BDC ([T.sub.bdc]) required to maintain CA50 = 372[degrees]CA for straight fuel with [psi] = [[psi].sub.m] = 0.38 and [P.sub.in] = 100 kPa as a function of engine speed for both CR = 16:1 and 14:1. [T.sub.bdc] is estimated using the method described in Ref. 29 (which accounts for heat transfer during the intake, residual gases, and dynamic heating). The results indicate that the required [T.sub.in] and [T.sub.bdc] at CR = 16:1 are 40 and 30[degrees]C lower, respectively, than those for CR = 14:1 at all engine speeds tested. This reduction refects the greater increase in pressure and temperature during the compression stroke with CR = 16:1. Similar to CR = 14:1, at CR = 16:1, the required [T.sub.in] increases rapidly at low speeds but is almost constant at engine speeds [greater than or equal to] 1200 rpm. The rapid increase in required [T.sub.in] and [T.sub.bdc] at lower speeds occurs because LTHR reactions are present at low speeds, but their magnitude decreases rapidly as speed increases since there is insufficient time for these relatively slow reactions before the compression process raises the charge temperature above the temperature where LTHR reactions can occur. As a result, [T.sub.in] must be increased to keep CA50 at 372[degrees]CA, which further reduces the LTHR, as discussed in greater detail with respect to Fig. 9 and 10 below. For both CRs, the [T.sub.bdc] curves parallel the [T.sub.in] curves up to about 1200 rpm, but above 1200 rpm, [T.sub.bdc] keeps rising despite the almost-constant [T.sub.in], due to the increased dynamic heating at higher speeds [29]. Thus, [T.sub.bdc] increases monotonically over the speed sweep, which compensates for the reduced time for autoignition reactions with increased speed.

Figure 8a shows the max. PRR in both bar/deg and MPa/ms, and RI corresponding to the CR = 16:1 data in Fig. 7. The [T.sub.in] curve from Fig. 7 is also reproduced for reference. As can be seen, PRR in bar/deg decreases monotonically with increased speed; however, PRR in MPa/ms decreases at low speeds as the effect of LTHR disappears (see previous paragraph), but it increases over the rest of the speed sweep, albeit with an indication of fattening for speeds [greater than or equal to] 2100 rpm. As might be expected, since PRR in MPa/ms is the relevant term for acoustic oscillations [27], the RI curve generally tracks the PRR (MPa/ms) curve. This indicates that if RI were to be held constant at 3 MW/[m.sup.2] CA50 would have to be more advanced at speeds lower than about 1900 rpm, and more retarded at higher speeds, compared to the constant CA50 = 372[degrees]CA used here.

A similar speed sweep at [P.sub.in] = 180 kPa, but with CA50 = 374[degrees]CA, is presented in Fig. 8b. As discussed previously, at these higher boost conditions the fuel is more reactive, so [T.sub.in] is maintained at 60[degrees]C and EGR is used to control CA50 at all engine speeds tested. The EGR requirement (presented as CSP) is found to monotonically decrease with increased engine speed (analogous to the monotonic increase in [T.sub.bdc] for [P.sub.in] = 100 kPa). As speed increases from 900 to 1800 rpm, the max. PPR (MPa/ms) and RI increase, as observed for speed increases above 900 rpm for [P.sub.in] = 100 kPa in Fig. 8a. However, the max. PRR (MPa/ms) and RI become nearly constant for speeds [greater than or equal to] 1800 rpm (CA50 held constant at 374[degrees]CA). The reason for this change in the trend of RI with engine speed at 1800 rpm is not currently understood, but it is not related to a decrease in CE, which remains above 97.5% over the entire speed sweep.

Speed Effects on LTHR and ITHR, CR = 16:1

Normalized HRRs for straight fuel with [[psi].sub.m] = 0.38 and CR = 16:1 at various engine speeds are shown in Fig. 9. CA10 is kept constant at 368.3~368.9[degrees]CA for [P.sub.in] = 100 kPa and 370.4~370.7[degrees]CA for [P.sub.in] = 180 kPa. At 717 rpm with [T.sub.in] = 60[degrees]C and [P.sub.in] = 100 kPa, straight fuel shows significant LTHR and ITHR, as evident in Fig. 9a. Increasing the engine speed decreases the time available for the early low-tointermediate temperature reactions, so the temperature must be increased to compensate, as shown in Fig. 10 and discussed above with respect to Fig. 7. These higher temperatures further suppress the early reactions, and the combined effects of reduced time and increased temperature produce substantial reductions in the LTHR and ITHR with increased speed. As can be seen, LTHR is no longer noticeable for engine speeds [greater than or equal to] 900 rpm at [P.sub.in] = 100 kPa. However, these early reactions are enhanced with increased boost pressure, and Fig. 8b shows significant LTHR at 900 rpm for [P.sub.in] = 180 kPa. Similar to [P.sub.in] = 100 kPa, increasing engine speed reduces the LTHR and ITHR at [P.sub.in] = 180 kPa as the reduced time available for these relatively slow reactions decreases. Also, the temperature increases despite [T.sub.in] being kept constant at 60[degrees]C, as discussed above with respect to Fig. 8. This temperature increase, shown in Fig. 10b, results from the combination of increasing dynamic heating (discussed earlier with Fig. 7) and reduced EGR levels (Fig. 8) with increased engine speed. The latter results in a higher [gamma] which increases the temperature rise with compression. Thus, LTHR and ITHR are suppressed at higher engine speed even for [P.sub.in] = 180 kPa, as shown in Fig. 9b.

Speed Effects on Stability, CR = 16:1

The weaker LTHR and ITHR at higher speeds is expected to reduce combustion stability, and therefore, allow less CA50 retard for controlling the RI to be [less than or equal to] 5 MW/[m.sup.2] to avoid knock as the load is increased at a given [P.sub.in] by increasing the fueling rate ([[psi].sub.m]), as discussed earlier in this work. The COV of IMEP (for 100-cycle samples) is shown in Fig. 11 for load sweeps using the straight fuel with [P.sub.in] = 100 and 180 kPa at 1200, 1800, and 2400 rpm. (Note that the 1200 rpm data is the same as presented above in Fig. 4.) At 1200 rpm, the COV is less than 1% for the stable points, and then it quickly increases to 6% beyond a certain load. At this highest load point shown for both [P.sub.in] = 100 and 180 kPa, the load is still knock/stability limited, and controlling it to prevent runaway knock or misfre is difficult. Beyond the last points shown, it could not be controlled. At higher engine speeds, the COV is larger for a given load, and it increases more rapidly with increased load than that at 1200 rpm. This indicates that the combustion is less stable at higher engine speeds, in agreement with the reduced ITHR shown in Fig. 9. As a result, the max. load with an acceptable COV of IMEP [less than or equal to] 2% decreases with increased engine speed for both intake pressures.

It is interesting to note that despite the higher COV at 1800 and 2400 rpm, the combustion timing near the maximum allowable load is more easily controlled to prevent runaway knock or misfre at these higher speeds compared to the high-load points at 1200 rpm. However, the maximum load is less at higher speeds, and at this same [IMEP.sub.g], the COV is low and combustion timing stable for 1200 rpm. Also at these higher speeds, as the COV increases to ~2% with increased load, the magnitude of the COV begins to vary widely for the 100-cycle sample size used here.(1) This is particularly evident for the 1800 rpm, [P.sub.in] = 100 kPa data in Fig. 11. Notice that COV data from two 100-cycle samples taken in close succession are presented for same [IMEP.sub.g] = 506 kPa, and that the COV is 1.1% for one point and 2.9% for the other. This behavior was only noticed for this higher-speed operation (i.e. not at 1200 rpm) with CR = 16:1, but an extensive investigation of conditions where it might occur was not conducted for CR = 14:1.

To better understand the reason for this behavior, Fig. 12 provides a detailed comparison for two 100-cycle samples at the same operating condition as the duplicate points mentioned in the preceding paragraph: [IMEP.sub.g] = 506 kPa, [[psi].sub.m] = 0.42, 1800 rpm, and [P.sub.in] = 100 kPa. One sample in Fig. 12 corresponds to the 1.1%-COV point in Fig. 11, while the other is another 100-cycle sample, also taken in close succession, but having a COV = 6.4% (not shown in Fig. 11). As can be seen for the 1.1% COV data in Fig. 12 (top plot), the [IMEP.sub.g] of almost every cycle is near the average value; however, for the 6.4% COV data, several cycles show a significantly lower [IMEP.sub.g] than the average, particularly cycle 15. These low-[IMEP.sub.g] cycles are thought to be caused by incomplete combustion, and are often followed by a cycle with a high RI (middle plot) and somewhat higher [IMEP.sub.g], as occurs for cycle 16. The data indicate that the explanation for this behavior is that for some reason, cycle 15 is randomly more retarded as evident in the bottom plot of Fig. 12, so its combustion is less complete.(2) This results in a low [IMEP.sub.g] and a relatively high level of unburned hydrocarbons (UHC) in the residual gas, and it is likely that some of these UHC compounds are more reactive than the original fuel. Thus, the next cycle has a greater autoignition reactivity, and a somewhat higher fuel + UHC energy content, so its CA50 is more advanced and its [IMEP.sub.g] little higher, the combination of which causes the higher RI.

Examination of a 1000-cycle sample shows that these poor-combustion cycles appear to occur randomly. The COV of these 1000 cycles is 3.1%, even though individual consecutive 100-cycle samples within this larger sample have COVs varying from 1.3 to 6.6%. Despite the [IMEP.sub.g] of these poor-combustion cycles dropping as low as about 250 kPa, the engine still fires on the following cycle and combustion phasing remains under control. It does not decay to misfire or runaway to high knock as typically occurs at 1200 rpm and at other conditions. Because uncertainty of the COV is large, 500-cycle or 1000-cycle samples were used when approaching the high-load limits in this study, i.e. for loads giving a COV of [IMEP.sub.g] approaching or higher than 2%. Thus, for the high-load limits shown later in this section, care was taken to ensure that COV is [less than or equal to] 2% and that there was no incomplete combustion cycles. At many conditions reported in this article, the load is limited by these criteria (termed a "high-COV limit") rather than increasing the load to the knock/stability limit, which would include many poor-combusting cycles for these operating conditions. The reason that the engine recovers from poor combustion for these conditions, requiring the use of a COV limit, whereas this did not occur at other conditions, such as 1200 rpm, is still uncertain.

Speed Effects on the High-Load Limit, CR = 16:1

High-load limits of both straight fuel and 0.15% EHN additized fuel at various engine speeds for a range of [P.sub.in]s. are plotted in Fig. 13a. The max. [IMEP.sub.g] of the additized fuel is oxygen limited at [P.sub.in] = 160 and 180 kPa for 1200 rpm, and [P.sub.in] = 180 kPa for 1800 rpm, and all the other points are limited by combustion stability or COV < 2%. The max. [IMEP.sub.g] is found to decrease with increased engine speed at a given [P.sub.in] if it is not [O.sub.2] limited because combustion becomes less stable at higher speeds, as discussed previously with respect to Fig. 11. Adding 0.15% EHN to the straight fuel increases the max. [IMEP.sub.g] at all engine speeds and [P.sub.in]s. tested, except when the maximum load is [O.sub.2] limited (see Fig. 13a).

Fig. 13b quantifies the increase in max. [IMEP.sub.g] when 0.15% EHN is added to the fuel. As can be seen, the additive increases the max. [IMEP.sub.g] about 7-10% at [P.sub.in] = 100 and 130 kPa for all engine speeds, but at higher [P.sub.in]s. the effect of the additive varies with engine speed. For 1200 rpm, the 10.4% increase in max. load at [P.sub.in] = 130 kPa is the largest increase, and for [P.sub.in] = 160 and 180 kPa, the max. [IMEP.sub.g] with the additive drops significantly (becomes negative) due to the high levels of EGR required, which limits the available [O.sub.2]. For 1800 rpm, the benefit of the additive increases with increased [P.sub.in] up to [P.sub.in] = 160 kPa, where it gives a 12.3% increase in max. [IMEP.sub.g]. However, further increasing [P.sub.in] to 180 kPa causes a decline in the max.-load increase as the max. load point with the additive becomes [O.sub.2] limited.

For 2400 rpm, the benefit peaks with the 11.3% increase at [P.sub.in] = 130 kPa, and for higher [P.sub.in]s., the benefit declines even though the max. load remains stability limited (i.e. it is not [O.sub.2] limited). The reason for this is not well understood, but it appears to be related to the finding that the allowable CA50 advance (for RI = 5 MW/[m.sup.2]) with the additive compared to the straight fuel is less for 2400 rpm than for 1800 rpm, as discussed below with respect to Fig. 15.

To better understand the reasons for the increase in max. [IMEP.sub.g] with the additive, the COV of [IMEP.sub.g] for straight fuel and additized fuel is shown in Fig. 14 for the three engine speeds with [P.sub.in] = 100 and 180 kPa. The results show that the COV of [IMEP.sub.g] increases with increased engine speed at a given [[psi].sub.m] for both the base fuel and the additized fuel at both [P.sub.in]s. (Note that the straight fuel data in Fig. 14 are the same as those in Fig. 11.) The higher COV at high engine speeds indicates that combustion is less stable for the same [[psi].sub.m], and as a result, the max. [[psi].sub.m] that can be attained is reduced with increased speed, commensurate with the [IMEP.sub.g] data in Fig. 13a.

At [P.sub.in] = 100 kPa, the 0.15% EHN additized fuel has a similar COV to the straight fuel at a given [[psi].sub.m] for all speeds, and the max. [[psi].sub.m] is also similar for the straight and additized fuels except for 1200 rpm where the [[psi].sub.m] for the additized fuel is just slightly higher. Since the [[psi].sub.m]s are similar, the increase in max. [IMEP.sub.g] at [P.sub.in] = 100 kPa results mainly from the increased charge density (greater charge mass) with 0.15% EHN due to the lower required [T.sub.in]s, which are given in Fig. 13a.

At [P.sub.in] = 180 kPa, however, 0.15% EHN additized fuel shows a lower COV of [IMEP.sub.g] than the straight fuel for the same [[psi].sub.m], particularly at the higher [[psi].sub.m]s acquired for each speed. This improved combustion stability with the 0.15% EHN fuel allows fueling to be increased to give a higher max. [[psi].sub.m], except for 1200 rpm where the max. [[psi].sub.m] with the additized fuel is [O.sub.2] limited as discussed above. Thus, the improved stability with the 0.15% EHN additized fuel, which allows a higher [[psi].sub.m], is the reason that higher loads ([IMEP.sub.g]) were obtained with the additized fuel (see Fig. 13a) at the higher [P.sub.in]s. where [T.sub.in] = 60[degrees]C for both the straight and additized fuels. Interestingly, Fig. 13b shows that the increase in max. [IMEP.sub.g] with the additive is similar at [P.sub.in] = 100 kPa and at the higher [P.sub.in]s., except for the [O.sub.2] limited conditions, even though the cause of the increase shifts from a density increase at the essentially the same [[psi].sub.m] ([P.sub.in] = 100 kPa) to a [[psi].sub.m] increase at essentially the same density ([P.sub.in] = 180 kPa).

The results in Fig. 15 show that increasing the boost pressure allows CA50 for the additized fuel to be progressively more advanced compared to the straight fuel, for the same [[psi].sub.m] and RI. At [P.sub.in] = 100 kPa, Fig. 15a shows that CA50 is similar for the additized and straight fuels. As [P.sub.in] is increased to 130 kPa, CA50 is 2~3[degrees]CA more advanced for the additized fuel at 1200 rpm only, as shown in Fig. 15b. As [P.sub.in] is further increased to 160 and 180 kPa, a more advanced CA50 is found for the additized fuel at all three engine speeds, as shown in Fig. 15c and d. This more advanced CA50 is considered to be the main reason for the better combustion stability (i.e. lower COV) of the additized fuel compared to the straight fuel at [P.sub.in] = 180 kPa as shown in Fig. 14b. The relationship between combustion stability and CA50 is more clearly evident in Fig. 16, which shows the COV of [IMEP.sub.g] plotted against CA50 for [P.sub.in] = 180 kPa. These results show that at a given CA50, the COV of [IMEP.sub.g] is similar for both the additized and straight fuels for each engine speed. Similar phenomena are also found for other [P.sub.in]s. It should be noted that higher engine speeds still result in a higher COV for a given CA50, especially at high loads. This is thought to be caused by the faster expansion rate at higher speeds.

The data in Fig. 15 also suggest that the observed CA50 advancement may related to the intake [O.sub.2] concentration, which varies due to differences in EGR addition and is also plotted in Fig. 15 for the various [P.sub.in]s. At [P.sub.in] = 100 kPa (Fig. 15a), no EGR is added (except for a relatively small amount for the higher [[psi].sub.m]s with the 0.15% EHN additized fuel at 1200 rpm) and the 0.15% EHN fuel does not allow a CA50 advancement at any speed. With [P.sub.in] increased to 130 kPa (Fig. 15b), much more EGR is required to control the autoignition reactivity of the additized fuel. For 1200 rpm, which has the lowest intake [O.sub.2], CA50 for the additized fuel is significantly advanced compared to the straight fuel. At [P.sub.in] = 160 and 180 kPa (Fig. 15c and 15d), the intake [O.sub.2] of the 0.15% EHN additized fuel is further reduced, falling below 15% at the higher [[psi].sub.m]s for all three speeds. Along with this reduction in [O.sub.2], notable CA50 advancement can be observed for all engine speeds.

Cylinder-Pressure and HRR Analysis

To better understand the reason that CA50 can be more advanced for RI = 5 MW/[m.sup.2] with the 0.15% EHN-additized fuel (see Fig. 15), changes in the cylinder pressure and HRR traces with 0.15% EHN addition were investigated in greater depth. These HRR traces have been corrected for heat transfer using Woschni correlation [26] and normalized by the THR. Figure 17 shows the cylinder pressure and normalized HRR profiles of both 0.15%-EHN additized and straight fuels for [[psi].sub.m] = 0.32 at 1800 rpm. CA50 is kept constant at 371.5[degrees]CA for [P.sub.in] = 180 kPa, and at 369[degrees]CA for [P.sub.in] = 100 kPa for both fuels. For [P.sub.in] = 180 kPa, the additized fuel has a lower PRR and [P.sub.max] compared with straight fuel as shown in Fig. 17a. This results from the lower HRR as shown in Fig. 17b. This lower PRR at the same CA50 allows CA50 of the additized fuel to be significantly advanced compared to the straight fuel for the same RI = 5 MW/[m.sup.2], as shown in Fig. 15d. However, at [P.sub.in] = 100 kPa, this effect is not observed (see Fig. 17a and b), in agreement with the similar CA50s with and without the additive shown in Fig. 15a.

Figure 18a shows additional comparisons between the pressure profiles of the additized and straight fuels at [P.sub.in] = 180 kPa, 1800 rpm, and constant RI = 5MW/[m.sup.2] for [[psi].sub.m] = 0.3 and 0.36. For both [[psi].sub.m]s, the additized fuel has a significantly more advanced CA50 for the same RI = 5MW/[m.sup.2]. This more advanced CA50 results in a higher [P.sub.max] (see Fig. 18a), which allows a slightly larger PRR (+3.7%) for the same RI, as evident from Eq. 1. Thus, the beneft of the reduced HRR/PRR for the additized fuel is compounded. First, CA50 can be more advanced for the same HRR/PRR compared to the straight fuel, but this advancement increases [P.sub.max] which reduces the RI for a given PRR. This allows CA50 to be even further advanced to reach a higher PRR that gives the same RI as the straight fuel.

Conversely, if CA50 of the additized fuel is retarded to be the same as that of the straight fuel, it shows a much lower [P.sub.max] and PRR, as shown in Fig. 18b (and also observed for [[psi].sub.m] = 0.32 in Fig. 17a). As a result, the RI for the additized fuel is significantly reduced to 2.0MW/[m.sup.2] for [[psi].sub.m] = 0.3 and to 1.2MW/[m.sup.2] for [[psi].sub.m] = 0.36, as shown in Fig. 18b. It should be noted that despite these much lower RIs for the additized fuel at these [[psi].sub.m]s, the COV of [IMEP.sub.g] is about the same for both fuels since they have the same CA50, as discussed above with respect to Fig. 16. It should be noted that this effect of the additive would be beneficial for reducing the noise levels without sacrificing stability [28].

Figure 19 presents the HRR curves (corrected for heat transfer and normalized by the THR) corresponding to the pressure traces in Fig. 18. As can be seen in Fig. 19a, the additized and straight fuels have similar HRR profiles when CA50 is adjusted to give the same RI. However, if CA50 of the additized fuel is retarded to match the CA50 of the straight fuel, the additized fuel has a broader HR and lower peak HRR than the straight fuel as shown in Fig. 19b, which is in agreement with changes in the pressure trace in Fig. 18. At [[psi].sub.m] = 0.3, the additized fuel shows the onset of a second peak in HRR with RI = 5MW/[m.sup.2] (Fig. 19a), and this double-peak becomes more prominent when CA50 is retarded to match that of the straight fuel (Fig. 19b).

This double peak in the HRR is thought to result from a combined effect of low peak temperature ([T.sub.peak]), high EGR, and low [[psi].sub.m].(3) As discussed by Sjoberg and Dec [30], the CO-to-C[O.sub.2] conversion rate first slows as [T.sub.peak] drops below ~1500 K, and then the conversion becomes progressively less complete if [T.sub.peak] continues to fall. Fig. 20b shows that [T.sub.peak] of the 0.15% EHN additized fuel with [[psi].sub.m] = 0.3, CA50 = 369.7[degrees]CA and RI = 2.0 MW/[m.sup.2] is slightly below 1500 K (1492 K), which will slow the CO-to-C[O.sub.2] conversion, and contribute to the double peak in the HRR.

In addition to [T.sub.peak], EGR and [[psi].sub.m] could also help produce this double peak in the HRR. As shown in Fig. 19a, the onset of a double peak in the HRR can be observed for 0.15% EHN additized fuel with [[psi].sub.m] = 0.3, CA50 = 366.6[degrees]CA and RI = 5 MW/[m.sup.2], although Fig. 20a shows that the [T.sub.peak] is 1598K at this condition. It should also be noted that the straight fuel for [[psi].sub.m] = 0.3 in Fig. 19a shows no indication of a double peak in the HRR even though it has a very similar [T.sub.peak] = 1587 K (see Fig. 20a). The much higher EGR concentration for the additized fuel (which results in a greater intake [O.sub.2] reduction) appears to be the most likely cause for the difference in HRR. Note that the additized fuel requires more EGR, in order to suppress its greater autoignition reactivity to maintain the desired combustion phasing. This reduced [O.sub.2] is thought to also reduce the CO-to-C[O.sub.2] conversion rate, slowing the high-temperature reactions responsible for the HRR from the main combustion.(4) Similarly, even higher EGR levels are required for the [[psi].sub.m] = 0.3 additized fuel when CA50 is retarded to 369.7[degrees]CA and RI = 2.0 MW/[m.sup.2] (Fig. 19b and 20b). As a result, the strong double peak in the HRR at this condition is thought to result from a combination of reduced intake [O.sub.2] and the low [T.sub.peak] = 1492 K discussed above.

Meanwhile, no double peak is found for 0.15% EHN additized fuel with [[psi].sub.m] = 0.36, CA50 = 374.9[degrees]CA, and RI = 1.2 MW/[m.sup.2] even though EGR is high and [T.sub.peak] is 1513 K, just 21 K hotter than the [T.sub.peak] = 1492 K additized-fuel condition that showed a strong double peak for [[psi].sub.m] = 0.3, CA50 = 369.7[degrees]C A (Fig. 20b). This suggests that [[psi].sub.m] may also be a key parameter related to a double peak in the HRR. However, It should be noted that the mass-average charge temperature for [[psi].sub.m] = 0.36 remains above 1500 K for several crank-angle degrees, which would also contribute to the more rapid combustion in addition to the slightly higher [T.sub.peak].

Thermal Efficiency

Adding EHN increased the TE at [P.sub.in] = 180 kPa over a wide range of engine loads (i.e. [[psi].sub.m]) for all engine speeds tested, as shown in Fig. 21a. Similar behavior was observed for other [P.sub.in]s. The increased efficiency with EHN addition is caused by a combined effect of changes in required [T.sub.in], EGR, and CA50, with the relative importance of these parameters changing with operation conditions. For example, at [P.sub.in] = 100 kPa, EHN addition reduces the [T.sub.in] required to reach a given RI (e.g. from 117 to 90[degrees]C for the data in Fig. 17b), which increases [gamma] (for greater work extraction) and reduces heat-transfer losses. However, CA50 is virtually unchanged by additive addition at [P.sub.in] = 100 kPa (see Fig. 15a) so it does not contribute to the changes in TE. At intake-boosted conditions, [T.sub.in] is typically the same (60[degrees]C) with or without the additive, but the additized fuel requires more EGR which increases CE, and its more advanced CA50 increases the effective expansion ratio, both of which increase the TE. This gain is slightly mitigated by the reduction in [gamma] due to the increased EGR. The max. TE is also found to increase with increased [P.sub.in] for both fuels as shown in Fig. 21b. Also for both fuels, 1800 rpm shows a higher max. TE for a given [P.sub.in] compared with 1200 and 2400 rpm, although the difference in max. TE between 1800 and 2400 rpm is quite small for the additized fuel. The highest TE found was 50.1% for 0.15% EHN additized fuel at 1800 rpm with [P.sub.in] = 180 kPa.

NOx Emissions

NO.x emissions for all fueling rates ([[psi].sub.m]) at both [P.sub.in] = 100 and 180 kPa are found to be well below the US-2010 standard of 0.27g/kWh for both the straight and 0.15% EHN additized fuels as shown in Fig. 22. At [P.sub.in] = 100 kPa, for [[psi].sub.m] below about 0.4, the NOx emissions are ultra-low and near the detection limit (~1 ppm). The slight differences between the various datasets in Fig. 22a are not considered significant, since they amount to only a few tenths of a ppm in the NOx measurement, which is in the "noise" of the NOx analyzer.(5) However, as [[psi].sub.m] increases above 0.4, [T.sub.peak] increases above 1800K and thermal NOx starts become measureable causing an upward trend to the data (Fig. 22a). The 1200 rpm straight-fuel data show a more rapid increase in NOx emissions compared with 0.15%-EHN additized fuel at this speed. This more rapid increase is thought to result from the higher intake temperature required to obtain autoignition without the additive. The 1200 rpm straight-fuel data also show a more rapid increase with [[psi].sub.m] than the 1800 rpm straight-fuel data. This is thought to be the result of the CA50 being more retarded for the 1800 rpm data to maintain RI = 5 MW/[m.sup.2] (as shown in Fig. 15a), which acts to reduce [T.sub.peak]. Also, the greater stability at 1200 rpm allowed higher [[psi].sub.m]s, and therefore higher combustion temperatures, than at 1800 or 2400 rpm, so the maximum NOx values are highest at this speed.

At [P.sub.in] = 180 kPa, the maximum [[psi].sub.m] is only 0.44, and [T.sub.in] = 60[degrees]C for both fuels, except for the three lowest [[psi].sub.m] straight-fuel points at 2400 rpm for which [T.sub.in] = 64, 69, and 72[degrees]C. As a result, the highest [T.sub.peak] is well below 1800 K, and there is virtually no thermal NOx produced. For the straight fuel, the small differences in NOx with engine speed amounts to only a few tenths of a ppm (see Fig. 22b), and they are not thought to be the result of increased HC and OHC emissions with increased speed, as discussed in the preceding paragraph for [P.sub.in] = 100 kPa. For the additized fuel, NOx emissions are significantly higher than for the straight fuel despite [T.sub.peak] being similar (well below 1800 K for the highest [[psi].sub.m]s), as shown in Fig. 22b. This increase in NOx is attributed to the nitrate group on the EHN. At 1200 rpm about one-third of the nitrogen in the EHN is converted to NOx emissions in agreement with our previous work [19]. NOx emissions for the additized fuel are found to be less at 1800 and 2400 rpm, with only about one-sixth of the nitrogen in the nitrate group being converted to NOx at the highest loads examined for 1800 rpm. This suggests that the formation of NOx from the nitrate group in EHN is sensitive to engine speed, with lower engine speeds allowing more time for the nitrogen in the EHN to be converted to NOx. Also, a comparison between Fig. 22a and 22b shows a much higher conversion of nitrogen from the EHN to NOx at [P.sub.in] = 180 kPa compared to [P.sub.in] = 100 kPa. It is interesting to note that these dependencies of the conversion of the nitrogen in the EHN to NOx on speed and pressure are the same as those for low- and intermediate-temperature chemistry, which can lead to greatly enhanced autoignition reactivity at low speeds (see Fig. 7) and higher [P.sub.in] (see Fig. 2 and 3).

CONCLUSIONS

The effect of enhancing the autoignition reactivity of a regular-grade E10 gasoline by adding 0.15% 2-ethylhexyl nitrate (EHN) on the performance of an HCCI engine has been investigated over a wide range of conditions. The investigation included the effects of increasing the compression ratio (CR) from 14:1 to 16:1, engine speed variations up to 2400 rpm, intake pressures ([P.sub.in]) from 100 to 180 kPa, and a wide range of fueling rates up to the high-load limit. Over this range of conditions, comparisons were made of the autoignition reactivity, combustion stability, high-load limits, heat release rates (HRRs), thermal efficiency, and NOx emissions between the 0.15%-EHN additized fuel and the base gasoline without additive (straight fuel). The following conclusions can be drawn.

1. Increasing the CR from 14:1 to 16:1 increases the autoignition reactivity, reducing the required [T.sub.in] at naturally aspirated and low intake-boost conditions, and requiring more EGR at moderate and high intake-boost conditions.

2. The higher CR increases the thermal efficiency at both low- and high-boost conditions, but the increase is greater at low-boost due to the decreased [T.sub.in], which also increases the high-load limit at low boost.

3. Combustion stability decreases slightly with increased CR probably due to the higher expansion rate and greater EGR requirements, but the COV of [IMEP.sub.g] remains below 1% up to the high-load stability limit.

4. The LTHR and ITHR are found to decrease with increased engine speed at both low and boosted [P.sub.in]s. LTHR is not noticeable for engine speed > 900 rpm. High charge temperatures due to high [T.sub.bdc] at high engine speeds also suppresses the LTHR and ITHR.

5. In agreement with the lower ITHR with increased speed, the autoignition reactivity and combustion stability decrease with increased engine speeds. Additionally, combustion must be more retarded with increased engine speeds to maintain the same ringing intensity, which also contributes to the reduced combustion stability (i.e., a higher COV of [IMEP.sub.g] at higher speeds).

6. A lower maximum load ([IMEP.sub.g]) is achieved at higher engine speeds due to the reduced combustion stability for both the straight and additized fuels, except when the load is oxygen limited at lower speeds, as occurs for the additized fuel at high boost pressures.

7. At lower [P.sub.in], the higher reactivity of the additized fuel reduces the required [T.sub.in] increasing charge density. This increases the load, for a given [[psi].sub.m] (charge-mass equivalence ratio).

8. For intake boosted conditions, adding 0.15% EHN is very effective for increasing the combustion stability. It reduces the HRR, and therefore the PRR, allowing a more advanced CA50 for the same ringing intensity, which increases combustion stability for a given load.

9. The addition of 0.15%-EHN increases the high-load limits for all engine speeds and boost pressures that are not oxygen limited. For low [P.sub.in], the load increase is mainly the result of the reduced [T.sub.in], which increases the charge density. For boosted conditions, the reduced HRR/PRR with the additive allows a higher fueling rate ([[psi].sub.m]) at the stability limit, increasing the maximum load.

10. The reduced HRR for the additized fuel at boosted conditions appears to result from reduced high-temperature reaction rates caused by a combination of lower combustion temperatures and the increased EGR levels required to control combustion timing. The increased EGR reduces the oxygen available for combustion, slowing the high-temperature reactions that control the HRR in addition to slowing the autoignition reactions that control combustion timing.

11. Adding EHN increases the thermal efficiency at all intake pressures tested. At lower [P.sub.in], the improvement results from the lower required [T.sub.in] which increases the specific-heat ratio ([gamma]) and reduces heat-transfer losses. At higher [P.sub.in], the improvement results from the more advanced CA50 due to the reduced HRR, and from a higher combustion efficiency because of the increased EGR. A maximum indicated TE of 50.1% was achieved for 0.15%-EHN additized fuel at 1800 rpm and [P.sub.in] = 180 kPa.

12. At low [P.sub.in] (100 kPa) NOx emissions are very low (near the detection limit ~1 ppm) for both the straight and 0.15%-EHN additized fuel for [[psi].sub.m] below about 0.4. This is attributed to the low [T.sub.peak] (< 1800 K), resulting in almost no thermal NOx, and the apparently slow conversion of the nitrogen in the nitrate group in EHN to NOx at low pressures. As [[psi].sub.m] increases above 0.4 at [P.sub.in] = 100 kPa, thermal NOx begins to form, but is only significant for 1200 rpm for which higher maximum loads ([[psi].sub.m]) could be achieved due to the greater stability at this speed. The highest NOx levels are still more than a factor of 2 below US2010 standards.

13. For boosted operation ([P.sub.in] = 180 kPa), there is no significant thermal NOx. NOx from the nitrogen in the EHN, appears to dominate the NOx emissions, but total NOx remains extremely low for all conditions studied, more than a factor of 6 below US2010 standards. NOx from the EHN decreases with increased engine speed due to the reduced time for NOx-forming reactions.

REFERENCES

[1.] Dec, J., Yang, Y., Dernotte, J., and Ji, C., "Effects of Gasoline Reactivity and Ethanol Content on Boosted, Premixed and Partially Stratified Low-Temperature Gasoline Combustion (LTGC)," SAE Int. J. Engines 8(3):935-955, 2015, doi:10.4271/2015-01-0813.

[2.] Christensen, M., Johansson, B., Amneus, P., and Mauss, F., "Supercharged Homogeneous Charge Compression Ignition," SAE Technical Paper 980787, 1998, doi:10.4271/980787.

[3.] Christensen, M. and Johansson, B., "Supercharged Homogeneous Charge Compression Ignition (HCCI) with Exhaust Gas Recirculation and Pilot Fuel," SAE Technical Paper 2000-01-1835, 2000, doi:10.4271/2000-01-1835.

[4.] Dec, J., and Yang, Y., "Boosted HCCI for high power without engine knock and with ultra-low NOx emissions - using conventional gasoline," SAE Int. J. Engines 3(1): 750-767, 2010, doi: 10.4271/2010-01-1086.

[5.] Dec, J, Yang, Y., and Dronniou, N., "Boosted HCCI - controlling pressure-rise rate 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.

[6.] Dec, J., Yang, Y., and Dronniou, N., "Improving Efficiency and Using E10 for Higher Loads in Boosted HCCI Engines," SAE Int. J. Engines 5(3):1009-1032, 2012, doi:10.4271/2012-01-1107.

[7.] Sjoberg, M. and Dec, J., "Smoothing HCCI Heat-Release Rates Using Partial Fuel Stratification with Two-Stage Ignition Fuels," SAE Technical Paper 2006-01-0629, 2006, doi:10.4271/2006-01-0629.

[8.] Yang, Y., Dec, J., Dronniou, N., Sjoberg, M., "Tailoring HCCI heat release rates with partial fuel stratification: Comparison of two-stage and single-stage ignition fuels," Proc. Combust. Inst. 33 (2): 3047-3055, 2011, doi:10.1016/j.proci.2010.06.114.

[9.] Yang, Y., Dec, J., Dronniou, N., Sjoberg, M. et al., "Partial Fuel Stratification to Control HCCI Heat Release Rates: Fuel Composition and Other Factors Affecting Pre-Ignition Reactions of Two-Stage Ignition Fuels," SAE Int. J. Engines 4(1):1903-1920, 2011, doi:10.4271/2011-01-1359.

[10.] Dec, J. and Sjoberg, M., "A Parametric Study of HCCI Combustion - the Sources of Emissions at Low Loads and the Effects of GDI Fuel Injection," SAE Technical Paper 2003-01-0752, 2003, doi:10.4271/2003-01-0752.

[11.] Dec, J., "Advanced compression-ignition engines-understanding the in-cylinder processes," Proc. Combust. Inst. 32(2): 2727-2742, 2009, doi: 10.1016/j.proci.2008.08.008.

[12.] Machrafi, H., Cavadias, S., Gilbert, P., "An experimental and numerical analysis of the HCCI auto-ignition process of primary reference fuels, toluene reference fuels and diesel fuel in an engine, varying the engine parameters," Fuel Processing Technology 89(11): 1007-1016, 2008, doi: 10.1016/j.fuproc.2008.03.007.

[13.] Lu, X., Shen, Y., Zhang, Y., Zhou, X., Ji, L., Yang, Z., Huang, Z., "Controlled three-stage heat release of stratified charge compression ignition (SCCI) combustion with a two-stage primary reference fuel supply," Fuel 90(5): 2026-2038, 2011, doi: 10.1016/j.fuel.2011.01.026.

[14.] Yang, Y., Dec, J., Dronniou, N., and Cannella, W., "Boosted HCCI Combustion Using Low-Octane Gasoline with Fully Premixed and Partially Stratified Charges," SAE Int. J. Engines 5(3):1075-1088, 2012, doi:10.4271/2012-01-1120.

[15.] 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.

[16.] Kaddatz, J., Andrie, M., Reitz, R., and Kokjohn, S., "Light-Duty Reactivity Controlled Compression Ignition Combustion Using a Cetane Improver," SAE Technical Paper 2012-01-1110, 2012, doi:10.4271/2012-01-1110.

[17.] Dempsey, A., Walker, N., and Reitz, R., "Effect of Cetane Improvers on Gasoline, Ethanol, and Methanol Reactivity and the Implications for RCCI Combustion," SAE Int. J. Fuels Lubr. 2013; 6(1):170-87. doi:10.4271/2013-01-1678.

[18.] Hosseini, V., Neill, W.S., Guo, H., Chippior, W.L., Fairbridge, C., and Mitchell, K., "Effects of different cetane number enhancement strategies on HCCI combustion and emissions," Int. J. Engine Res. 12: 89-108, 2011, doi: 10.1177/1468087410395873.

[19.] Ji, C., Dec, J., Dernotte, J., and Cannella, W., "Effect of Ignition Improvers on the Combustion Performance of Regular-Grade E10 Gasoline in an HCCI Engine," SAE Int. J. Engines 2014; 7(2). doi:10.4271/2014-01-1282.

[20.] Ickes, A., Bohac, S., and Assanis, D., "Effect of 2-ethylhexyl nitrate cetane improver on NOx emissions from premixed low-temperature diesel combustion. Energy and Fuels," 2009;23(10):4943-48. doi: 10.1021/ef900408e.

[21.] Sjoberg, M. and Dec, J., "Combined Effects of Fuel-Type and Engine Speed on Intake Temperature Requirements and Completeness of Bulk-Gas Reactions for HCCI Combustion," SAE Technical Paper 2003-01-3173, 2003, doi:10.4271/2003-01-3173.

[22.] Sjoberg, M. and Dec, J., "Effects of Engine Speed, Fueling Rate, and Combustion Phasing on the Thermal Stratification Required to Limit HCCI Knocking Intensity," SAE Technical Paper 2005-01-2125, 2005, doi:10.4271/2005-01-2125.

[23.] 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.

[24.] Sjoberg, M. and Dec, J., "Ethanol Autoignition Characteristics and HCCI Performance for Wide Ranges of Engine Speed, Load and Boost," SAE Int. J. Engines 3(1):84-106, 2010, doi:10.4271/2010-01-0338.

[25.] Aroonsrisopon, T., Foster, D., Morikawa, T., and Iida, M., "Comparison of HCCI Operating Ranges for Combinations of Intake Temperature, Engine Speed and Fuel Composition," SAE Technical Paper 2002-01-1924, 2002, doi:10.4271/2002-01-1924.

[26.] Heywood, J.B., Internal Combustion Engine Fundamentals. New York: McGraw-Hill; 1988.

[27.] Eng, J., "Characterization of Pressure Waves in HCCI Combustion," SAE Technical Paper 2002-01-2859, 2002, doi:10.4271/2002-01-2859.

[28.] Dernotte, J., Dec, J., and Ji, C., "Investigation of the Sources of Combustion Noise in HCCI Engines," SAE Int. J. Engines 7(2):2014, doi:10.4271/2014-01-1272.

[29.] Sjoberg, M. and Dec, J., "An Investigation of the Relationship Between Measured Intake Temperature, BDC Temperature, and Combustion Phasing for Premixed and DI HCCI Engines," SAE Technical Paper 2004-01-1900, 2004, doi:10.4271/2004-01-1900.

[30.] Sjoberg, M. and Dec, J., "An investigation into lowest acceptable combustion temperatures for hydrocarbon fuels in HCCI engines," Proc. Combust. Inst. 30(2):2179-2726, 2005, doi: 10.1016/j.proci.2004.08.132.

[31.] Sjoberg, M., Dec, J. E., "Influence of Fuel Autoignition Reactivity on the High-Load Limits of HCCI Engines", SAE Int. J. Engines 1(1): 39-58, SAE paper 2008-01-0054, 2008.

[32.] Sjoberg, M., Dec, J. E., and Cernansky, N. P., "The Potential of Thermal Stratification and Combustion Retard for Reducing Pressure-Rise Rates in HCCI Engines, based on Multi-Zone Modeling and Experiments," SAE Transactions, Vol. 114, Sec. 3, pp. 236-251, SAE paper 2005-01-0113, 2005.

ACKNOWLEDGEMENTS

The authors would like to thank the following people for help with this study: Kenneth St. Hilaire, Christopher Carlen, David Cicone, Alberto Garcia, and Gary Hubbard of Sandia National Laboratories for their dedicated support of the LTGC engine laboratory. Primary support for this investigation was provided by Chevron under WFO contract FI083070907-Z, managed by William Cannella. The work was performed at the Combustion Research Facility, Sandia National Laboratories, Livermore, CA. Support for establishing the HCCI lab facility was provided by the US Department of Energy, Office of Vehicle Technologies, managed by Gurpreet Singh and Leo Breton. Sandia is a multiprogram laboratory operated by the Sandia Corporation, a Lockheed Martin Company, for the US Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.

DEFINITIONS/ABBREVIATIONS

AHRR - apparent heat release rate

AKI - antiknock index

aTDC - after TDC

BDC - bottom dead center

bTDC - before TDC

CA - crank angle

CA10 - 10% energy released point

CA50 - 50% energy released point

CCG-E10 - Chevron commercial regular-grade gasoline containing 1- vol.% ethanol

CE - combustion efficiency

CN - cetane number

COV - coefficient of variation

[c.sub.p] - constant pressure heat capacity

[c.sub.v] - constant volume heat capacity

CR - compression ratio

CSP - complete stoichiometric products

DTBP - di-tert-butyl peroxide

E10 - gasoline containing 10% ethanol

EGR - exhaust gas recirculation

EHN - 2-ethylhexyl nitrate

EVC - exhaust valve close

EVO - exhaust valve open

[psi] - equivalence ratio (without EGR)

[[psi].sub.m] - charge-mass equivalence ratio

[gamma] - ratio of specific heats ([c.sub.p]/[c.sub.v])

GDI - gasoline direct injector

HCCI - homogenous charge compression ignition

HRR - heat release rate

[IMEP.sub.g] - gross indicated mean effective pressure

ITHR - intermediate temperature heat release

IVC - intake valve close

IVO - intake valve open

LTGC - low-temperature gasoline combustion

LTHR - low-temperature heat release

ON - octane number

[P.sub.in] - intake pressure

PM - particulate mate

[P.sub.max] - max pressure

PRF - primary reference fuels

PRR - pressure rise rate

R - gas constant

RI - ringing intensity

[T.sub.bdc] - temperature at BDC

TDC - top dead center

TE - gross indicated thermal efficiency

THR - total heat release

[T.sub.in] - intake temperature

[T.sub.peak] - peak temperature

UHC - unburned hydrocarbons

Chunsheng Ji, John Dec, and Jeremie Dernotte Sandia National Laboratories

William Cannella Chevron Energy Technology Company

(1.) For 2400 rpm, the load was not pushed clear to the instability limit since several data points were already acquired with COV [greater than or equal to] ~2%, which is considered to be the highest acceptable value.

(2.) For example, this might be caused by variation in heat transfer during the intake or compression strokes due to random variations in the turbulence. Since these data are near the knock/stability limit, a small change in CA50 can cause a relatively large change in [IMEP.sub.g].

(3.) Although the plot in Fig. 19b is for ensemble-averaged data, this double peak in the HRR is not an artifact of the ensemble averaging. All 100 of the individual cycles used for this ensemble-average plot also show a similar double peak in the HRR. Additionally, the standard deviation of CA50 for these 100 cycles is only about 1.5[degrees] CA, which is well less than the approximately 4[degrees] CA difference between the two peaks in the ensemble-averaged HRR curve.

(4.) The HRR profile is affected by both the chemical-kinetic rates and the amount of thermal stratification (for a well- mixed charge). Under naturally aspirated conditions, which require relatively high temperature for autoignition, kinetic rates are typically fast, and the heat-release duration is dominated by the effects of thermal stratification [30, 32]. However, under boosted conditions, particularly with the additized fuel, autoignition temperatures are much lower, resulting in relatively slow kinetic rates which can significantly affect the duration of the heat-release. Reduced [O.sub.2] levels also contribute to reduced kinetic rates.

(5.) The very small but consistent trends in the NOx for [[psi].sub.m] < 0.4, are thought to be caused by changes in HC or oxygenated HC (OHC) (e.g. formaldehyde) emissions. Trends in the HC emissions match the slight trends in NOx lower [[psi].sub.m], and previous work has indicated that the NOx analyzer can sometimes respond weakly to these emissions.

Table 1. Engine specifications and operating conditions.

Displaced (Single cylinder)    0.981 liters
Bore                         102 mm
Stroke                       120 mm
Connecting Rod Length        192 mm
Geometric Compression Ratio  14:1 and 16:1
No. of Valves                  4
IVO                            0[degrees]CA (*)
IVC                          202[degrees] CA (*)
EVO                          482[degrees] CA (*)
EVC                            8[degrees]CA (*)
Swirl Ratio                    0.9
Fueling system               Fully Premixed
Engine Speed                 538-2400 rpm
Intake Pressure (abs.)       100~180kPa
Intake Temperature           60~145[degrees]C
Coolant/Oil Temperature      100[degrees]C

0[degrees] CA is taken to be TDC intake. The valve-event timings
correspond to 0.1 mm lift.
COPYRIGHT 2016 SAE International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Ji, Chunsheng; Dec, John; Dernotte, Jeremie; Cannella, William
Publication:SAE International Journal of Engines
Article Type:Report
Date:Dec 1, 2016
Words:14453
Previous Article:Evaluation of knock behavior for natural gas - gasoline blends in a light duty spark ignited engine.
Next Article:Study on the double injection strategy of gasoline partially premixed combustion under a light-duty optical engine.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters