Experimental Investigation of Diesel-Ethanol Premixed Pilot-Assisted Combustion (PPAC) in a High Compression Ratio Engine.
Low temperature combustion (LTC) strategies have the potential to significantly reduce the oxides of nitrogen (NOx) and particulate matter (PM) from modern diesel engines [1,2,3,4]. Through precise control over the mixture formation process and air path system management, the clean combustion process can be optimized to minimize the energy efficiency penalty typically incurred as a result of the low combustion temperatures [5, 6]. Since the start of the new millennium, intensive research has been reported on a number of LTC strategies to achieve clean combustion; the most widely reported ones being the modulated kinetics (MK) combustion (smokeless rich combustion) , homogeneous charge compression ignition (HCCI) combustion [3, 8] and recently, the dual-fuel combustion of fuels with contrasting properties [1, 4, 9].
LTC strategies fundamentally target a long ignition delay (up to several milliseconds) by separating the fuel injection from the combustion event in order to prepare a well-mixed, lean/EGR diluted fuel-air charge . In the MK combustion, this is realized with a near-TDC single fuel injection along with an aggressive application of EGR (50 to 70%) to provide sufficient mixing time before the onset of ignition while still retaining control over the combustion phasing through injection timing adjustment [7, 11]. While NOx emissions are typically reduced at moderate EGR levels, achieving ultra-low soot emissions require very high EGR levels that typically results in a narrow engine operating regime, where minor changes in
EGR can result in large spikes in the soot emissions (EGR decrease) or significantly deteriorated combustion efficiency and unstable combustion (EGR increase). The ability to maintain a sufficiently long ignition delay greatly diminishes as the engine load level (fuelling amount) or the engine speed increases. Although, the implementation of MK combustion in production vehicles was done as early as 2004, the proposed MK II combustion that extended the strategy to higher loads and higher speeds was never commercially applied.
In contrast, HCCI combustion commonly utilizes multiple incylinder fuel injections at very early injection timings (up to -100[degrees]ATDC) to prepare a near-homogeneous cylinder charge. Whereas the long mixing time results in very low soot emissions, high EGR flowrates (50 to 60%) are still required to reduce NOx to ultra-low levels in high compression ratio engines. The decoupling of the auto-ignition timing from the injection timing typically results in a pre-TDC combustion phasing (0 to -12[degrees]ATDC) and therefore requires complex management and control of the intake charge temperature and EGR rate for combustion phasing adjustment [3, 12]. The advanced phasing also results in very high pressure rise rates and peak firing pressures that limit the attainable engine load levels. As such, no implementation of HCCI combustion has been successfully commercialized.
The modern dual-fuel combustion strategy employs two fuels of contrasting reactivity to manage the cylinder charge preparation and the ignition control under lean, EGR diluted conditions. A low-reactivity, volatile fuel (gasoline and ethanol being the most commonly reported ones) is used to prepare a highly-homogeneous, bulk cylinder charge. The high volatility permits the port injection of the fuel (simpler injection system) while the low-reactivity ensures that auto-ignition does not occur under the diesel-like high compression ratios. The reactivity of the cylinder charge is then modified by introducing a high-reactivity 'pilot' fuel, typically by means of a diesel direct injection (DI) to manage the auto-ignition timing.
Two variations of the dual-fuel combustion exist, depending on the injection timing of the high-reactivity fuel. In Reactivity Controlled Compression Ignition (RCCI) combustion, the cylinder charge reactivity is modified through two early diesel pilot injections (1st pilot: -80 to -50[degrees]ATDC; 2 (nd) pilot: -45 to -30[degrees]ATDC) to provide adequate control over the combustion initiation and phasing [4, 13]. The mixing time for the last pilot injection is long enough such that the soot emissions are ultra-low. Previous results indicate that moderate to high engine loads are achievable with RCCI combustion while retaining or exceeding diesel-like thermal efficiencies . However, a relatively complex combustion phasing control through pilot timing management 'on-the-fly' is required to enable transient operation.
The other implementation of dual-fuel combustion is the Premixed Pilot-Assisted Combustion (PPAC) that employs a near-TDC pilot injection timing to provide diesel-like direct control over the combustion initiation and phasing [9, 14]. While up to 2 diesel pilot injections may be used to achieve the necessary fuel blending for reactivity modification, the timing of the last pilot is always close to TDC. This greatly simplifies the combustion control requirements and provides a linear correlation of combustion phasing with the pilot injection timing that also enables easier handling of transient operation. Ultra low emissions of NOx and soot are obtained at low-to-mid loads; however, at higher engine loads and speeds, the relatively shorter mixing time for the near-TDC pilot may adversely affect the soot emissions. Therefore, the pilot injection quantity optimization is critical to minimize the impact on the soot emissions.
The implementation of both the MK and HCCI combustion strategies is further hindered by the requirements of flowing large amounts of EGR at elevated intake boost pressures. First, the simultaneous availability of high boost pressure and high EGR flow rates at low loads typical for these combustion modes is usually not possible due to the constraints of the turbocharging systems (energy availability in the exhaust is limited at such engine operations). Second, the effectiveness of EGR in terms of the intake charge dilution is a function of the engine load and intake pressure, and therefore requires a precise control on the intake charge dilution for maintaining stable operation close to the combustion stability limits . Last, the disproportionate response times of the air-path and combustion systems compound the control problem during transient events such as variations in the engine load demand. The current practice is to momentarily close the EGR valve during tip-in or tip-out events since the EGR valve opening for steady-state operation is mostly based on look-up tables calibrated during the development stage, with a few model-based corrections. Therefore, the implementation of these combustion modes over the speed/load range is difficult without the ability to track and enable the use of EGR during transients .
The pathway to a clean-burning, high efficiency engine should optimize both the combustion features and the fuel properties as shown in Figure 1. It can be observed that the PPAC strategy combines the benefits of both the HCCI and MK combustion and thereby, overcomes a number of challenges of MK and HCCI combustion. The long mixing time, together with the low reactivity of the bulk fuel usually minimizes or even eliminates the EGR requirements for postponing the onset of ignition or to control the combustion phasing. Since soot emissions are very low, only low-to-moderate amounts of EGR are required to achieve ultra-low NOx emissions. The load range of dual-fuel combustion is much wider than that for the MK and HCCI combustion, which minimizes the need to carryout mode-switching on-the-fly to enable clean combustion across the engine operating map.
Recent LTC research has focused on renewable, oxygenated fuels such as ethanol and butanol for the low-reactivity fuels [16,17,18]. Ethanol fuel has gathered particular interest because of its high volatility that allows it to be injected into the intake port to form a highly premixed cylinder charge and results in significant lowering of the compression-end temperature and pressure (charge cooling effect due to its heat of vaporization). Unlike gasoline (octane number: 87-94) which tends to auto-ignite as the load increases, the low reactivity of ethanol (research octane number: 110-115) allows for its effective use in high compression ratio diesel engines.
Previous work including that by the authors has demonstrated the potential of the PPAC strategy to meet the challenging task of achieving high energy efficiency together with ultra-low NOx and soot emissions [9, 19]. However, a number of issues have been identified that need to be addressed to make PPAC a viable strategy for implementation in production engines. These issues include a methodology to select the optimum distribution of fuel quantity between the bulk and pilot fuels, the use of a single-pilot vs twin-pilots based on the engine load and smoke limit, the improvement of the combustion efficiency at light loads as well as load range extension and management to cover the full operating range of the engine.
The first part of this research has been carried out on a single-cylinder research engine and is reported here. In this work, detailed empirical investigations of the diesel-ethanol PPAC are carried out at 1500 rpm (consistent with speeds in the heavy-duty vehicle drive cycle) to determine the optimum boundary conditions of PPAC, including the minimum ethanol fraction for ultra-low NOx & soot emissions, effect of single pilot vs twin pilot strategies on emissions and ignition controllability, reducing the EGR requirements and achieving high efficiency full-load operation. To quantify the minimum ethanol-to-diesel quantity ratio, a 'Heat Release Profile Distribution'' parameter is proposed and validated for identification of the transition from the diesel pilot-dominated heterogeneous combustion to predominantly premixed, homogenous ethanol combustion and for optimizing the fuel quantity distribution between the two fuel quantities for low emissions and high efficiency. Test results for the single-pilot PPAC with NOx-soot emission compliance are presented for the full load range of the test engine.
The experimental work was conducted on a single-cylinder diesel engine based on a modified 4 cylinder diesel engine and connected to an eddy current dynamometer. The 4-cylinder engine was configured to operate in a 3 cylinders-to-1 cylinder configuration (single-cylinder mode) where 3 cylinders operate in the conventional diesel mode at low load for stable engine operation while the single-cylinder (cylinder # 1) is used for conducting systematic research with independent control of EGR, intake boost pressure, exhaust backpressure and fuel injection timing and quantity . A schematic of the test setup is shown in Figure 2 and the specifications of the test engine are given in Table 1.
Ethanol fuel was injected at the back of the intake valve in the intake manifold using a gasoline port injection (PFI) system, suitably calibrated for ethanol fuel delivery. A constant injection timing of 10[degrees]ATDC (during intake stroke) was used for the ethanol injection throughout the tests. The diesel fuel was directly injected into the combustion chamber using the engine's common rail solenoid injector-based fuel system. The specifications for the port-fuel injector and the diesel injector are given in Table 2. No modification was made to the stock piston.
The intake manifold pressure and exhaust backpressure were controlled to [+ or -]0.5kPa with PID controllers while the EGR valve opening was precisely controlled using the controller area network (CAN), implemented through LabVIEW-based systems. The fuel injection timing and quantity as well the common-rail injection pressure controls were implemented with a FPGA-based real-time deterministic system. The engine coolant temperature was precisely controlled to 90[degrees]C with an external conditioning system to minimize the discrepancies in the testing results.
The engine indicating system consisted of an AVL GU13P cylinder pressure transducer mounted through the glow-plug access-hole and connected to the combustion analysis system through a Kistler 510B charge amplifier. The pressure data was resolved at 0.1[degrees]CA and 200 consecutive combustion cycles were recorded for each data point. The data was processed both online and offline to compute combustion metrics such as the indicated mean effective pressure (IMEP), start of combustion (SOI), net heat release rate, cumulative heat release and the crank angle of 50% heat release (CA50) [21, 22]. In this work, the pumping work is included in all the calculations and the reported values of IMEP and indicated efficiency are on a net basis.
The measured exhaust emissions included NOx, HC, CO, C[O.sub.2] and [O.sub.2] concentrations while the intake concentrations of C[O.sub.2] and [O.sub.2] were measured after EGR. The gaseous emissions were averaged for 10 seconds after attaining steady state operation. The particulate matter (PM) in the exhaust was measured using an AVL 415S smoke meter. Hydrocarbon-free, dry combustion air was supplied from a combustion-air conditioning system at a fixed temperature of 24[degrees]C and 30% RH. EGR was implemented using a combination of the exhaust backpressure and EGR valve opening. The EGR rate was estimated as the ratio of intake C[O.sub.2] concentration to the exhaust C[O.sub.2] concentration.
The spectral analysis of the PPAC exhaust was carried out with an MKS MultiGas 2030 Fourier Transform Infrared Spectroscopy (FTIR) continuous gas analyser. The system has a spectral resolution of 0.5-128 [cm.sup.-1] and performs 1 scan/second @0.5 [cm.sup.-1]. The exhaust sample gas for the FTIR was also supplied from the standard sample conditioning unit (heated sampling line & filter, chiller); therefore, only the light hydrocarbon species have been analysed and reported in this work due to the above mentioned limitation of the exhaust sampling setup. The hydrogen concentration in the exhaust was measured using a V & F H-Sense [H.sub.2] gas analyser, based on the electron pulse ionization mass spectrometry principle.
The fuel properties are given in Table 3 [23,24]. Commercially available laboratory-grade anhydrous ethyl alcohol (ethanol) with a research octane number of 110-115 and research-grade ultra-low sulphur diesel (ULSD) fuel with a cetane number of 46.5 were used foe all the tests.
The quantity of ethanol at any operating point was quantified by the 'Ethanol Fraction' (EF) on an energy input basis and defined as the ratio of the energy contributed by the ethanol quantity to the total energy input from both the fuels.
[mathematical expression not reproducible] (1)
RESULTS AND DISCUSSION
A comprehensive test matrix was setup to investigate and understand the impact of different combustion and fuelling parameters on the emissions and performance of ethanol-diesel PPAC. The parametric analysis was carried out by conducting sweep tests for the ethanol fraction, diesel pilot quantity, EGR and boost pressure. The targeted emission levels were 0.2 g/kWh for NOx and 0.01 g/kWh for soot across the load range, based on the US EPA 2010 emission standards for heavy-duty on-road compression-ignition engines.
All the testing was carried out at 1500 rpm, consistent with speeds observed in the US EPA 'Not-To-Exceed" and 'European Stationary Cycle' for heavy-duty vehicle drive cycles. The results presented in this work are representative of the trends observed during the experimental work over the complete load range of the engine.
Combustion Characteristics with Ethanol Fraction Variation
The first objective was to determine the optimum boundary conditions of PPAC, including the minimum ethanol fraction for ultra-low NOx & soot emissions. To this end, an examination of the heat release data revealed that the location of the crank angle of 50% heat release (CA50) within the combustion process correlates well with the changing quantity of ethanol. Based on these analyses, the authors have defined a 'Heat Release Profile Distribution' (HRPD) parameter as follows:
HRPD = [[[theta].sub.CA50] - [[theta].sub.SOC]/[[theta].sub.EOC]-[[theta].sub.SOC]] (2)
The term ([[theta].sub.CA50] - [[theta].sub.SOC]) represents the time required for accumulation of heat release up to CA50 and normalizing this term by the total combustion duration ([[theta].sub.EOC]-[[theta].sub.SOC]) provides a measure of the rate at which heat is released during the early phase of combustion. The values of the heat release parameters are calculated from the online cycle-by-cycle heat release analysis. The start of combustion (SOC) is taken as the first zero-crossing after the pilot start of injection (SOI) timing and the end of combustion (EOC) is taken as the crank angle at which 90% of the total heat has been released.
The results of an EF sweep test are presented in Figure 3. The sweep was started from EF: 0.3 and EGR was increased to 43% to reduce NOx to almost 0.2 g/kWh. The ethanol fraction was then progressively increased (while reducing the pilot quantity) to maintain the IMEP at 9.8 bar. EGR was held constant and the combustion phasing was maintained at 7[degrees]ATDC throughout by adjusting the pilot injection timing to account for the changing ignition delay with increased EF.
The combustion at low ethanol fractions (0.3 to 0.46) is characterised by an initial high rate of heat release that can be attributed to the diesel pilot-dominated combustion (large diesel pilot quantity). The maximum pressure rise rate is >10 bar/[degrees]CA and the HRPD ~0.32 stays almost constant till EF: 0.4. As the ethanol quantity is further increased (>0.50), the first peak of heat release reduces/disappears and the heat release profile indicates a transition to a premixed, homogenous ethanol-dominated combustion. This is characterized by a drop in the maximum pressure rise rate due to the reduced initial burn rate (low reactivity of ethanol) as well as an increase in the ignition delay ([[tau].sub.ID]) to > 1 ms. The HRPD increased rapidly with the ethanol fraction and then stabilized around 0.68.
The emission results for this EF sweep test are shown in Figure 4. The initial EGR-diluted, diesel pilot-dominated combustion (EF: 0.3) exhibits the traditional NOx-soot trade-off with high soot emissions (0.06 g/kWh) but as the EF increases, the soot emissions show a steady drop, consistent with the increasing homogeneity of the cylinder charge. With HRPD > 0.5, the soot emissions drop below the target value of 0.01 g/kWh. The higher ethanol fraction also results in lower NOx emissions because of two predominant factors: first, increased cylinder charge cooling occurs due to the high enthalpy of vaporization ([[DELTA]H.sup.vap]) and the relatively low boiling point of ethanol that help to lower the peak compression-end pressures (~ 10 bar in this case) and temperatures (mean cylinder temperature estimation reduces by >100K) as the port-injected ethanol undergoes a phase change during the compression process. Second, the reduced initial burn rate implies that the maximum combustion temperatures are also lower, thereby inhibiting NOx production.
The HC emissions initially increase with the ethanol fraction, consistent with the reducing reactivity of the cylinder charge but then approach a constant value as the homogenous combustion prevails. Previous studies have linked the HC emissions to the size of the crevice volume and the leaner air- fuel ratio with ethanol [25, 26]. An interesting trend is observed with the CO emissions which increase at first but then decrease as the homogeneous combustion of ethanol becomes predominant. The partial oxidation of unburned fuel is one of the CO production mechanisms and the increasing HC emissions at the start may contribute to the CO production. Based on hydrogen gas measurement in the exhaust gas, it is observed that the hydrogen production during the diesel-ethanol combustion follows a profile similar to that of CO as shown in Figure 5. Similar trends are also observed with chemical kinetics simulations and measurement of CO-H2 concentration for MK combustion (LTC with high EGR levels) .
Since hydrogen is known to accelerate the dissociation of C[O.sub.2] into CO, it is believed that the presence of H, OH radicals initially promote the chemical dissociation of C[O.sub.2] but the further lowering of the combustion temperatures with increased ethanol fraction suppresses the hydrogen production as well as the C[O.sub.2] dissociation, resulting in lower CO emissions. It is important to note that the start of reduction in the CO emissions coincides with the HRPD approaching a value of 0.50.
The above results indicate that the HRPD is able to capture the transition to homogeneous combustion or in other words, the de-coupling of the NOx and soot trade-off. Similar tests were carried out at different engine load levels (4 bar to 16 bar IMEP) to determine the reliability of the HRPD in identifying the minimum required ethanol fraction and one test result is presented in the next section.
Determining Minimum Required Ethanol Fraction
The performance of the HRPD parameter in identifying the minimum required ethanol fraction was evaluated at 11 bar IMEP and 2 bar abs boost pressure. The engine-out NOx emissions of 0.23 g/kWh were obtained with the diesel only (EF: 0) test and were set as the baseline target for the diesel ethanol PPAC.
The results for the combustion analysis and emissions are shown in Figure 6. The reduction in NOx beyond 0.23 g/kWh with diesel (EF: 0) was restricted by the high soot emissions of 0.87 g/kWh (smoke: 4.3 FSN). However, the HC and CO emissions were low, a characteristics of the conventional high-temperature diesel combustion. As the ethanol quantity was increased, the NOx-soot trade-off transformed into a trade-off between the increasing HC (lower reactivity, lean charge) and the decreasing soot emissions (high homogeneity). The HRPD was almost constant at low EF and then started to increase as the EF approached a value of 0.5. Based on the results presented earlier, the ethanol quantity was increased until the HRPD value crossed the threshold of 0.50; the soot emissions were observed to reduce to 0.005 g/kWh, meeting the target value of < 0.01 g/kWh. The corresponding ethanol fraction was 0.57.
It is interesting to note that with HRPD at 0.51 (EF: 0.57), the trends in the maximum rate of pressure rise and the CO emissions were found to be consistent to those presented in Figure 3 and Figure 4. Both these parameters exhibit a reduction, highlighting the changeover to the PPAC.
An important aspect is the trend observed in the EGR rate. As the ethanol fraction was increased, the EGR rate had to be reduced to maintain NOx at 0.23 g/kWh. This can be correlated to the reduced combustion temperatures that result in lower NOx emissions due to the high activation energy of NO formation reactions . This implies that the homogeneous diesel-ethanol PPAC may need less exhaust gas to be recirculated for achieving the same NOx reduction when compared to the diesel only LTC strategies. This is critical because the necessity of providing a high boost pressure and high EGR flow rates is one of the impediments to the successful implementation of LTC strategies in production engines. As such, a reduction in the requirement of high EGR rates to lower NOx not only reduces the sensitivity of the combustion process to small changes in EGR and but also lessens the requirements for implementing fast, precise EGR control.
The cylinder pressure and heat release traces for this test are shown in Figure 7. The cylinder charge cooling is clearly observed from the cylinder pressure traces and is an important benefit of using ethanol as the bulk fuel for power production. The heat release traces indicate that a complete transition to the homogeneous ethanol-dominated combustion (absence of first heat release rate peak) is not an absolute necessity. Rather, the pilot diesel quantity should be reduced such that the pilot injection should mainly serve as the ignition source, ensure stable combustion and should contribute minimally to the emissions.
At EF: 0.57, the EGR was increased to the baseline level (EGR: 36.8%; EF: 0) and the results are shown in Figure 8. The higher intake dilution reduces the NOx to ultra-low values (0.138 g/kWh) without any soot or HC penalty. The heat release profile indicates a significantly reduced burn rate (-110 J/[degrees]CA) compared to the lower EGR test point (-180 J/[degrees]CA at EGR: 32.5%) and highlights the effectiveness of EGR at higher engine loads.
The results, comparing the diesel (EF: 0) and PPAC (EF: 0.57) points are summarized in Table 4. It can be seen that the clean PPAC achieves a higher thermal efficiency despite the lower combustion efficiency. Based on the authors' previous work on diesel-ethanol zero-dimensional engine cycle simulations coupled with empirical observations , the fuel energy distribution indicates that this efficiency improvement can be attributed to the lowered compression-end pressures and temperatures that result in reduced compression work, a reduction in heat transfer over the engine cycle and lower exhaust temperatures The results also confirm that the decrease in the combustion efficiency is more than offset by the reduced heat transfer and lower exhaust temperatures that highlight an improved energy utilization with respect to the total input energy. The maximum rate of pressure rise is also similar between the two combustion strategies. These test results also highlight the potential benefit of the HRPD parameter to optimize the fuel quantity distribution between the two fuels for low emissions and high efficiency.
A limited spectral analysis of the exhaust stream for the PPAC results in the previous section (Figure 6 to Figure 8) is presented to gain insights into the effect of the ethanol fraction on the principal components (diesel and ethanol) of the HC emission as well as its impact on the aromatics and aldehydes emissions. The speciation results are shown in Figure 9. The predominant component of the HC emissions is observed to be the unburnt ethanol fuel. At higher ethanol fractions, the relative percentage of the diesel component decreases, with a corresponding increase in the ethanol component within the HC emissions. These findings can be attributed to the combination of three factors: low relativity of the ethanol, a long mixing time and the reduced in-cylinder temperatures at higher ethanol fractions .
The aldehydes (formaldehyde and acetaldehyde) show a consistent increase with the ethanol fraction while the aromatics represented by toluene increased moderately. The hydrogen concentrations followed the trend of the CO emissions (Figure 6) with peak concentration of ~ 120 ppm. Previous speciation studies by the authors on the diesel MK combustion at similar load levels and boundary conditions identified a strong correlation of hydrogen with the CO production . While the PPAC CO-hydrogen trends agree with those earlier findings, the PPAC hydrogen concentrations are found to be 2-3 times lower than those for the diesel MK combustion at the same excess-air ratios.
PPAC Phasing Control
One of the principal benefits of PPAC is the close-to-TDC pilot injection timing that enables diesel-like control over the combustion phasing while providing a sufficiently long ignition delay to ensure separation between the fuelling and the main combustion event. The results from the previous sections indicate that the pilot quantity should also just be sufficient to provide a stable source of ignition without the chance of misfire. If the quantity is higher, the soot emissions may increase as the initial combustion corresponds to a diesel-like heat release while a lower than optimal quantity may result in partial burning or misfire. To evaluate the pilot injection characteristics, the results for 3 SOI sweeps at different EF were performed and the results from the heat release analysis are shown in Figure 10. The boost pressure and EGR was kept the same during these tests.
The CA50 exhibits an almost linear relationship with the SOI timing, independent of the ethanol fraction. This is significant because the pilot SOI can provide a direct control over the combustion phasing, similar to the conventional diesel operation where the timing of the diesel injection regulates the combustion phasing. It is pertinent to mention here that as the SOI timing is advanced beyond -20[degrees]ATDC (not shown here), the combustion phasing retards which signifies the phasing control being moved over to the reactivity of the cylinder charge.
The ignition delay profiles are observed to be similar except for the EF: 0.65 case and close-to-TDC SOI timings when the ignition delay shows a greater increase. This can be attributed to the lower reactivity of the cylinder charge at higher ethanol fraction and the reducing temperatures at the start of the expansion stroke that delay the onset of ignition.
The HRPD profiles demonstrate that the parameter can provide useful feedback, independent of the SOI timing. The HRPD profiles for the low EF cases (0.3, 0.49) remain almost constant (~0.28) from SOI timings of 0 to -12[degrees]ATDC while the HRPD profile for the EF: 0.65 sweep displays an increasing trend with advancing SOI timings. The corresponding emission results are shown in Figure 11. Although the soot emissions are overall low, the soot emissions for the low EF cases show an initial increase with the advancing SOI. In addition to the higher diesel pilot quantity, the HRPD value of 0.28 points to the fact that the initial rate of heat release is quite high and that a significant part of the combustion occurs within a few crank angle degrees. Hence, the probability of the occurrence of soot-producing fuel-rich zones is high as the mixing time for the pilot is less. On the other hand, the HRPD for the EF: 0.65 displays an increasing trend and the soot emissions are low (less pilot quantity). The NOx emissions show an increasing trend with the earlier pilot SOI timing. The NOx emissions at EF: 0.65 are slightly higher than the other SOI sweep tests since the load (9.5 bar IMEP) was higher (EGR and intake boost pressure being the same).
The effect of the pilot timing at a fixed ethanol fraction (EF: 0.65) on the cylinder pressure and heat release rate is shown in Figure 12. The pilot timing provides flexibility in terms of attaining the desired combustion phasing, the maximum pressure rise rate and the maximum cylinder pressure. The combustion can be duly phased to suit the emissions, efficiency or stability demands with varying EGR, intake pressure and engine load.
The cylinder pressure characteristics for the SOI sweep test (EF: 0.65) are summarized in Table 5. It should be noted that these SOI sweeps were not optimized for NOx emissions (25% EGR, 18% intake oxygen), maximum cylinder pressure rise rate or efficiency but rather for investigating the trends of the emissions, combustion metrics and efficiency with the pilot injection timing/quantity.
Single-vs-Twin Pilot Investigation
The near-TDC diesel pilot timing provides a diesel-like direct control over the combustion initiation and phasing, and eliminates the need for a complex phasing control system as required for HCCI or RCCI types of combustion modes. However, the PPAC combustion is more susceptible to higher soot emissions if the pilot quantity is not optimized. To minimize this soot penalty, it is worthwhile to investigate the splitting of the near-TDC pilot into two pilot injections to achieve the necessary fuel blending for reactivity modification while maintaining the timing of the last pilot close to TDC.
The results for such an investigation are shown from Figure 13 to Figure 16. EGR sweeps at two ethanol ratios (0.52 and 0.59) with a single pilot were carried out to achieve the stated emissions targets for both NOx and soot emissions. The EGR cooling was reduced to increase the intake charge temperature and a lower diesel fuel injection pressure of 90MPa was selected. It can be observed in Figure 13 that while the NOx emissions of 0.2 g/kWh can be achieved at both ethanol fractions, the soot emissions are higher by 1 to 2 orders of magnitude. Therefore, a further increase in the ethanol fraction (along with higher fuel injection pressure or cooled EGR) is the only way to reduce the soot emissions with the single pilot PPAC. Noted, the HRPD value for both these EFs was -0.42.
Since the diesel pilot quantity accounted for 41-48% of the total fuel energy input for these EGR sweeps, the possibility to divide the pilot quantity over 2 injections was explored so that the last pilot injection serves only as the ignition source and provides phasing control. This should help to minimize the trade-off between the NOx reduction and the increase in soot emissions. Therefore, the EGR sweep with lower ethanol quantity (EF: 0.52) and a single pilot injection was taken as the baseline and the single diesel pilot was replaced with 2 pilot injections; the first pilot delivering 45% of the total pilot quantity at -60[degrees]ATDC and the second pilot delivering 55% of the total pilot quantity at 0[degrees]ATDC. The results in Figure 13 show a drastic reduction not only in the soot emissions but a significant reduction in the NOx emissions as well with the twin pilot strategy.
The NOx-soot trade-off is shown in Figure 14. The red box indicates the points at the same EGR rates (39-41%). The soot emissions decrease by an order of magnitude once compared on the same NOx basis and move the NOx-soot trade-off closer to the origin - the ultimate target of any emission reduction strategy. It needs to be highlighted here that the PPAC with the twin pilots can further reduce the EGR requirements as the higher homogeneity of the mixing charge and the reduced diesel quantity for ignition result in NOx reduction at lower intake dilution levels.
The emission comparison in Figure 15 shows that the CO and HC emissions also decrease with the twin-pilot strategy. The first pilot in the twin pilot PPAC conditions the cylinder charge so that the charge within the squish has a high propensity to burn (similar to the first pilot for the RCCI strategy). It is pertinent to mention here that the twin-pilot strategy does not place any additional demands on the fuel injection and control system as the modern common-rail diesel injections systems are capable of running multiple injections per cycle. However, at low engine loads, it may not be possible to split the pilot injections into two since the minimum pilot injection quantity dictated by the fuel injection system characteristics may already contribute a significant proportion of the total fuel energy input.
Figure 15. Emissions for the Single-vs-Twin Pilot Tests (EF: 0.52) Indicated Emissions [g/kWh] 1 Pilot 2 Pilots NOx 0.37 0.23 Soot 0.044 0.007 CO 15.7 10.8 HC 2.84 2.25 Note: Table made from bar graph.
The heat release rate profiles for the single-pilot and the twin-pilot PPAC at similar EGR rates are shown in Figure 16. The reduced fuelling quantity of the 2nd pilot eliminates the first peak of the heat release that is typically associated with the burning of the pilot injection. The early pilot prepares the ethanol charge for combustion by increasing its reactivity (charge conditioning) which is evident by the gradual increase in the heat release rate close to TDC (low temperature reactions). Although the ignition delay is not affected, the mixing time is increased due to the short injection energizing time (reduced pilot quantity). The HRPD for the twin-pilot PPAC was 0.66 compared to the HRPD of 0.42 for the single-injection PPAC. The maximum pressure rise rate increased from 4.4 bar/[degrees]CA to 9.6 bar/[degrees]CA but the stability of the combustion improved as the coefficient of variance (COV) of IMEP decreased from 2.97% (EF: 0.52; 1 pilot) to 1.83% (EF: 0.52, twin pilot).
PPAC Load Range
The observations and lessons learnt during this work were applied to optimize the PPAC combustion across the load range as follows:
1. To meet the NOx target, an appropriate EGR level has to be applied.
2. The pilot injection quantity needs to be minimized by modulating the ethanol fraction using HRPD feedback to ensure that stable combustion can be achieved with minimal soot emission penalty.
3. The timing of the diesel pilot is preferred close to the TDC to retain direct control on the combustion phasing.
4. The not-to-exceed hardware limits in terms of the maximum cylinder pressure (160 bar) and the maximum pressure rise rate (less than 15 bar/[degrees]CA) should be met.
The load optimization for the single pilot injection PPAC is reported in this work. The test targeted a simultaneous reduction of the NOx and soot emissions while minimizing the combustion stability and efficiency to achieve clean and efficient PPAC combustion. The cylinder pressures and the heat release rates across the load range (4 bar to 18 bar IMEP) are shown in Figure 17. The intake boost pressure was increased with load to maintain a lean, EGR-diluted homogeneous cylinder charge. At higher engine loads, the combustion phasing had to be retarded to stay within the prescribed hardware limits.
The low reactivity of the ethanol fuel and the minimum fuel injection quantity of the pilot diesel injection restrict the maximum ethanol fraction that can be practically utilized at low engine loads to achieve clean and stable combustion. The CO and HC emissions are very high at the 4 bar IMEP point as shown in Figure 18. While NOx can be reduced below 0.2 g/kWh with EGR, the combustion efficiency will tend to deteriorate further and may increase the combustion cyclic variability. These factors limit the EF to about 0.51 in this case and any further reduction in the engine load is prohibited by the high CO and HC emissions.
The low load performance may be improved by switching off the EGR cooling as a higher intake temperature results in an earlier initiation of the ethanol evaporation process and therefore, a further lowering of the compression-end temperatures . Therefore, NOx emissions are lowered, reducing the intake dilution requirements that help to keep the soot emissions in check.
As the ethanol fraction is increased with load (EF: 0.84 at 16 bar IMEP), the first peak of the heat release reduces and the soot emissions generally meet the targeted level of 0.01 g/kWh up to 16 bar IMEP as shown in Figure 18. At higher loads, a further increase in the boost pressure is limited by the peak firing pressure hardware limitation of 160 bar for the test engine, which results in soot emissions being slightly higher than the targeted 0.01 g/kWh. It should be noted that the legislated emission limits are based on the cycle-averaged emission measurements taken during the course of the complete test cycle. These cycles (such as the FTP-75) rarely require the engine to operate at the full load point. As such, momentary excursions of the soot emissions above the target levels may be compensated by the ultra-low soot emissions at part-load operating points.
The load range optimization results are summarized in Table 6. The optimization achieved a maximum pressure rise rate less than 15 bar/[degrees]CA which is generally considered as the threshold for the combustion noise levels in large heavy-duty compression-ignition engines . The peak cylinder pressure constraint was met at all operating points.
Stable engine operation (quantified by the CO[V.sub.IMEP] < 3%) was established with low NOx and soot emissions across the load range with the PPAC strategy. The thermal efficiency (on a net indicated basis) was comparable to the diesel-only operation and at higher loads, the efficiency was better than that obtained with conventional high temperature diesel combustion on the test engine.
From this work, a minimum pilot quantity that corresponds to about 15% of the total energy input was deemed sufficient to meet the requirements of ensuring stable combustion with minimal soot penalty at high loads (17-19 bar IMEP). It is pertinent to add here that the application of the PPAC with a twin-pilot configuration may provide a substantial relief in the CO/HC emissions as the lower NOx and soot emissions reduce the necessity of higher intake dilution through EGR application. Noted that the twin pilot PPAC at low loads may reduce the ethanol fraction as the smallest pilot injection quantity per shot is governed by the injector characteristics (standard flow rate reference: 1.2 mg/stroke @30 MPa for the tested solenoid injector; as low as 0.7 mg/stroke for a piezo electric type). However, the early pilot (# 1) prepares the ethanol charge for auto-ignition by increasing its reactivity so that the last pilot (# 2) results in a faster burn rate and improved stability. Thus, a reduction in the HC and CO emissions may be possible. One potential issue at higher loads can be the increased rate of pressure rise (due to the faster burn rates) and therefore, the fuel distribution between the two injections needs a careful evaluation (this strategy is currently being tested and the optimization results will be reported in a subsequent paper).
The premixed pilot-assisted combustion (PPAC) of diesel-ethanol fuels, enabled with a single-pilot injection strategy has been successfully demonstrated over a wide load range (4 bar IMEP to 18.4 bar IMEP) using production-level hardware with no modifications. This work can be summarized as follows:
(1.) The proposed 'heat release profile distribution' (HRPD) parameter enables the identification of the minimum ethanol fraction for decoupling the NOx-soot trade-off.
(2.) A HRPD value of > 0.50 indicates the transition to ethanol-dominated homogeneous combustion that reduces the soot emissions to < 0.01 g/kWh and allows the NOx emission target of 0.2 g/kWh to be met with moderate EGR rates.
(3.) The low load limit of the PPAC strategy is governed by the combustion efficiency penalty and the combustion stability concerns. It may be feasible to switch to EGR-diluted diesel-only operation at idle to low-engine loads to meet the emission and stability targets.
(4.) The near-TDC pilot injection timing provides a direct, linear control over the combustion initiation and phasing.
(5.) PPAC provides a diesel-like thermal efficiency with ultra-low NOx and soot emissions. The CO and HC emissions are higher but can potentially be managed with an oxidation catalyst-based exhaust aftertreatment system.
(6.) The optimization of the pilot injection quantity is essential to minimize the associated soot emission penalty. The twin-pilot configuration has the potential to mitigate these issues and a detailed investigation is required to confirm these preliminary findings.
[1.] Asad, U., Divekar, P., Zheng, M., and Tjong, J., "Low Temperature Combustion Strategies for Compression Ignition Engines: Operability limits and Challenges," SAE Technical Paper 2013-01-0283, 2013, " doi:10.4271/2013-01-0283.
[2.] Kimura, S., Aoki, O., Kitahara, Y., and Aiyoshizawa, E., "Ultra-Clean Combustion Technology Combining a Low-Temperature and Premixed Combustion Concept for Meeting Future Emission Standards," SAE Technical Paper 2001-01-0200, 2001, doi:10.4271/2001-01-0200.
[3.] Asad, U., Zheng, M., Ting, D. S. K., Tjong, J., "Implementation Challenges and Solutions for Homogeneous Charge Compression Ignition Combustion in Diesel Engines," Journal of Engineering for Gas Turbines and Power. 137 (10):101505-1-2,2015, 10.1115/1.4030091.
[4.] Reitz, R. D., Duraisamy, G., "Review of high efficiency and clean reactivity controlled compression ignition (RCCI) combustion in internal combustion engines," Progress in Energy and Combustion Science. 46:12-71,2014, http://dx.doi.org/10.1016/j.pecs.2014.05.003.
[5.] Asad, U. and Tjong, J., "A Zero-Dimensional Intake Dilution Tracking Algorithm for Real-Time Feedback on Exhaust Gas Recirculation," SAE Int. J. Engines 8(4):1856-1865, 2015, doi:10.4271/2015-01-1714.
[6.] Asad, U. and Zheng, M., "Efficiency & Stability Improvements of Diesel Low Temperature Combustion through Tightened Intake Oxygen Control," SAE Int. J. Engines 3(1):788-800, 2010, doi:10.4271/2010-01-1118.
[7.] Akihama, K., Takatori, Y., Inagaki, K., Sasaki, S. et al., "Mechanism of the Smokeless Rich Diesel Combustion by Reducing Temperature," SAE Technical Paper 2001-01-0655, 2001, doi:10.4271/2001-01-0655.
[8.] Gan, S., Ng, H. K., Pang, K. M., "Homogeneous Charge Compression Ignition (HCCI) combustion: Implementation and effects on pollutants in direct injection diesel engines," Applied Energy. 88 (3):559-67,2011, http://dx.doi.org/10.1016/j.apenergy.2010.09.005.
[9.] Asad, U., Kumar, R., Zheng, M., Tjong, J., "Ethanol-fueled low temperature combustion: A pathway to clean and efficient diesel engine cycles," Applied Energy.2015, http://dx.doi.org/10.1016/j.apenergy.2015.01.057.
[10.] Asad, U., Zheng, M., "Evaluation of Diesel Low Temperature Combustion Fuel-Injection Strategies at Different Engine Loads." ASME Paper ICEF2010-35172; 2010.
[11.] Asad, U., Zheng, M., Han, X., Reader, G. et al., "Fuel Injection Strategies to Improve Emissions and Efficiency of High Compression Ratio Diesel Engines," SAE Int. J. Engines 1(1):1220-1233, 2009, doi:10.4271/2008-01-2472.
[12.] Reader, G. T., Asad, U., Zheng, M., "Energy efficiency trade-off with phasing of HCCI combustion," International Journal of Energy Research. 37 (3):200-10,2013, 10.1002/er.1900. "
[13.] 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.
[14.] Divekar, P., Asad, U., Han, X., Chen, X., et al., "Study of Cylinder Charge Control for enabling Low Temperature Combustion in Diesel Engines," Journal of Engineering for Gas Turbines and Power. 136 (9):091503-1-7,2014, 10.1115/1.4026969.
[15.] Asad, U., Tjong, J., Zheng, M., "Exhaust gas recirculation - Zero dimensional modelling and characterization for transient diesel combustion control," Energy Conversion and Management. 86:309-24,2014, http://dx.doi.org/10.1016/j.enconman.2014.05.035.
[16.] De Ojeda, W., Bulicz, T., Han, X., Zheng, M. et al., "Impact of Fuel Properties on Diesel Low Temperature Combustion," SAE Int. J. Engines 4(1):188-201, 2011, doi:10.4271/2011-01-0329.
[17.] Senda, J., Kawano, D., Kawakami, K., Shimada, A., et al. Fuel Design Concept Research for Low Exhaust Emissions by Use of Mixing Fuels. The 5th International Symposium on Diagnostics and Modelling of Combustion in Internal Combustion Engines (COMODIA 2001). Kyoto, Japan2001. p. 343-50.
[18.] Lu, X., Han, D., Huang, Z., "Fuel design and management for the control of advanced compression-ignition combustion modes," Progress in Energy and Combustion Science. 37 (6):741-83,2011, 10.1016/j.pecs.2011.03.003.
[19.] Divekar, P. S., Asad, U., Tjong, J., Chen, X., et al., "An engine cycle analysis of diesel-ignited ethanol low-temperature combustion," Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering.2015, 10.1177/0954407015598244.
[20.] Asad, U., Kumar, R., Han, X., Zheng, M., "Precise Instrumentation of a Diesel Single-Cylinder Research Engine," Measurement. 44 (7):1261-78,2011, 10.1016/j.measurement.2011.03.028.
[21.] Asad, U., Zheng, M., "Diesel Pressure Departure Ratio algorithm for combustion feedback and control," International Journal of Engine Research. 15 (1):11,2014, 10.1177/1468087412461268.
[22.] Asad, U., Zheng, M., "Fast Heat Release Characterization of a Diesel Engine," International Journal of Thermal Sciences. 47:1688-700,2008, 10.1016/j.ijthermalsci.2008.01.009.
[23.] Dillon, H. E., Penoncello, S. G., "A Fundamental Equation for Calculation of the Thermodynamic Properties of Ethanol," International Journal of Thermophysics. 25 (2):321-35,2004, 10.1023/B:IJOT.0000028470.49774.14.
[24.] Commercial Alcohols, Material Safety Data Sheet - Ethyl Alcohol (Anhydrous), 2014, www.comalc.com/wp-content/uploads/2014/08/MSDS-English-Ethyl-Alcohol-Anhydrous-aug-14.pdf, Accessed on January 01, 2016.
[25.] Jung, G. S., Sung, Y. H., Choi, B. C., Lee, C. W., et al., "Major sources of hydrocarbon emissions in a premixed charge compression ignition engine," International Journal of Automotive Technology. 13 (3):347-53,2012, 10.1007/s12239-012-0032-5.
[26.] Tutak, W., "Bioethanol E85 as a fuel for dual fuel diesel engine," Energy Conversion and Management. 86 (0):39-48,2014, http://dx.doi.org/10.1016/j.enconman.2014.05.016.
[27.] Asad, U., Mendoza, A., Xie, K., Jeftic, M., et al., "Speciation Analysis of Light Hydrocarbons and Hydrogen Production During Diesel Low Temperature Combustion." ASME Paper ICEF2011-60130; 2011.
[28.] Turns, S. R. An introduction to combustion: concepts and applications: McGraw-Hill; 1996.
[29.] Zhang, Y., Sagalovich, I., De Ojeda, W., Ickes, A. et al., "Development of Dual-Fuel Low Temperature Combustion Strategy in a Multi-Cylinder Heavy-Duty Compression Ignition Engine Using Conventional and Alternative Fuels," SAE Int. J. Engines 6(3):1481-1489, 2013, doi:10.4271/2013-01-2422.
Usman Asad, Ph.D., P.Eng.
Mechanical, Automotive & Materials Engineering
University of Windsor, Ontario, Canada N9B 3P4
The authors are grateful for support from the University of Windsor, the Canada Research Chair Programme, the Natural Sciences and Engineering Research Council of Canada, the Canadian Foundation for Innovation, Ontario Research Fund, AUTO 21[TM] (a member of the Network of Centres of Excellence of Canada programme) and the Ford Motor Company of Canada. This research received no specific grant from any funding agency in the public, commercial, or not-forprofit sectors.
[[tau].sub.id] - ignition delay
[degrees]CA [theta] - degree crank angle
AI - auto-ignition
ATDC - after top-dead-centre
CA50 - crank angle of 50% heat release
CAN - controller area network
CO - carbon monoxide
COV - coefficient of variance
DI - direct injection
(dp/d[theta]) - rate of pressure rise
EF - ethanol fraction
EGR - exhaust gas recirculation
EOC - end of injection
FSN - filter smoke number
HC - unburnt hydrocarbons
HCCI - homogeneous charge compression ignition
HRPD - heat release profile distribution
IMEP - indicated mean effective pressure
Ind - indicated
LTC - low temperature combustion
[??] - mass flow rate
MK - modulated kinetics
ms - milli-second
NOx - oxides of nitrogen
p - pressure
PFI - port fuel injection
PM - particulate matter
PPAC - premixed pilot-assisted combustion
ppm - parts per million
RCCI - reactivity controlled compression ignition
RH - relative humidity
rpm - revolution per minute
SOC - start of combustion
SOI - start of injection
T - temperature
TDC - top dead centre
f - fuel
int - intake
inj - injection
max - maximum
Usman Asad, Ming Zheng, and Jimi Tjong
University of Windsor
Table 1. Test Engine Specifications Type Single Cylinder, 4 stroke Dl Diesel Displacement 499.5 [cm.sup.3] Bore x stroke 0.086 m x 0.086 m Compression ratio 18.2:1 Injection System Common-rail Maximum injection pressure 180 MPa Peak firing pressure 16 MPa (160 bar) Table 2. Details of the Fuel Injection Systems Dl PFI Type Solenoid Solenoid Number of Holes 6 4 Hole Diameter [ [micro]m] 130 - Included Spray Angle 155[degrees] 50[degrees] Injection Pressure [MPa] 60-150 0.4-0.8 Table 3. Fuel Properties Fuel ULSD Ethanol Density [15[degrees]C, kg/[m.sup.3]] 846 788 Viscosity [40[degrees]C, cSt] 2.5 1.52 Cetane Number [-] 46.5 8-11 Research Octane Number [-] 25 110-115 Lower Heating Value [MJ/kg] 43.1 27 Oxygen Content [% mass] 0 34.78 Final Boiling Point [1 bar, [degrees]C] 340.7 78.3 C-H-O Ratio [-] [C.sub.1][H.sub1.87] [C.sub.1][H.sub.3][O.sub.0.5] Fuel ULSD Ethanol Table 4. Comparison between Diesel (EF:0) & PPAC (EF:0.57) Tests EF[-] 0.0 0.57 EGR [%] 36.8 37.2 Ind. NOx [g/kWh] 0.27 0.138 Smoke [FSN] 4.33 0.18 Ind. Soot [g/kWh] 0.873 0.008 Ind. HC [g/kWh] 0.1 3.5 Ind. CO [g/kWh] 4.3 10.8 Combustion Efficiency [%] 99.4 97.8 Ind. Thermal Efficiency [%] 42.8 43.5 [(dp/d[theta]).sub.max] [bar/[degrees]CA] 9.4 8.9 CA50 [[degrees]ATDC] 6.5 7.1 HRPD 0.35 0.57 Table 5. Summary of Cylinder Pressure Metrics for SOI Sweep Test (EF: 0.65) EF: 0.65; EGR: 25% Diesel SOI [[degrees]ATDC] 0 -6 -14 CA50 [[degrees]ATDC] 13.1 5.1 -2.7 IMEP [bar] 7.9 9.5 9.2 [P.sub.max] [bar] 69.2 137.4 156.2 [(dp/d[theta]).sub.max] [bar/[degrees]CA] 2.8 18.6 26.1 Ind. NOx [g/kWh] 1.69 4.25 8.41 Ind. Soot [g/kWh] 0.0015 0.0009 0.0011 COVlMEP [%] 3.86 1.88 3.06 Ind. Thermal Efficiency [%] 36.9 43.2 41.8 Table 6. Summary of the PPAC Load Range Optimization Tests IMEP [bar] 4 8 12 16 18.4 EF[-] 0.51 0.68 0.76 0.84 0.85 CA50 [[degrees]ATDC] 4.6 7.5 10.2 13.1 15.5 [P.sub.max] [bar] 98 120 146 156 159 [(dp/d[theta]).sub.max] 4.1 8.4 10.3 11 12.4 [bar/[degrees]CA] Ind. NOx [g/kWh] 0.21 0.13 0.12 0.2 0.19 Ind. Soot [g/kWh] 0.002 0.011 0.01 0.01 0.015 Ind. CO [g/kWh] 29.9 18.8 7.2 8.3 5.2 Ind. HC [g/kWh] 4.4 2.6 1.9 2.6 3 [COV.sub.lMEP] [%] 3.1 2 1.9 2.8 3 Ind. Thermal 40.9 43.5 44.1 45.1 46.3 Efficiency [%]
|Printer friendly Cite/link Email Feedback|
|Author:||Asad, Usman; Zheng, Ming; Tjong, Jimi|
|Publication:||SAE International Journal of Engines|
|Article Type:||Technical report|
|Date:||Jun 1, 2016|
|Previous Article:||Cylinder-to-Cylinder Variations in Power Production in a Dual Fuel Internal Combustion Engine Leveraging Late Intake Valve Closings.|
|Next Article:||Experimental Optimization of a Small Bore Natural Gas-Diesel Dual Fuel Engine with Direct Fuel Injection.|