diesel RCCI combustion characteristics in a light duty diesel engine

diesel RCCI combustion characteristics in a light duty diesel engine

Applied Energy 199 (2017) 430–446 Contents lists available at ScienceDirect Applied Energy journal homepage: www.elsevier.com/locate/apenergy Effec...

7MB Sizes 0 Downloads 2 Views

Applied Energy 199 (2017) 430–446

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Effect of diesel injection strategies on natural gas/diesel RCCI combustion characteristics in a light duty diesel engine Kamran Poorghasemi a, Rahim Khoshbakhti Saray a,⇑, Ehsan Ansari b, Behrouz Khoshbakht Irdmousa b, Mehdi Shahbakhti b, Jeffery D. Naber b a b

Mechanical Engineering Department, Sahand University of Technology, Sahand New Town, Tabriz 51335-1996, Iran Department of Mechanical Engineering, Michigan Technological University, Houghton, MI 49931, USA

h i g h l i g h t s  NG/diesel RCCI engine was simulated by Converge CFD model.  By increasing the PR, the lower reactivity of NG causes lower combustion rate.  Increasing first injected fuel quantity results in higher HC and CO emissions.  Narrower spray angles have higher values of HC and CO emissions.

a r t i c l e

i n f o

Article history: Received 3 February 2017 Received in revised form 18 March 2017 Accepted 2 May 2017

Keywords: RCCI Combustion CFD simulation Spray angle Diesel injection pressure Injection timing

a b s t r a c t Reactivity controlled compression ignition (RCCI) combustion mode is an attractive combustion strategy due to its potential in satisfying the strict emission standards. In this study, the effects of direct injection (DI) strategies on the combustion and emission characteristics of a modified light duty RCCI engine, fueled with natural gas (NG) and diesel were numerically investigated. In this way, Converge CFD code employing a detail chemical kinetics mechanism was used for 3D simulation of combustion process and emissions prediction. NG with higher octane number (ON) is mixed with air through intake port, while diesel fuel with lower ON is directly injected into the combustion chamber during compression stroke by means of split injection strategy. The effects of several parameters, including the premixed ratio (PR) of NG, diesel fuel fraction in first and second injection pulses, first and second start of injection timing (SOI1 and 2), injection pressure and the spray angle on the engine performance and emission characteristics are investigated. The results indicate that these parameters have significant effects on the light duty RCCI engine performance and engine out emissions. Also, it was demonstrated that by decreasing the first injection pressure from 450 to 300 bar, the gross indicated efficiency increases by 5% and CA50 is retarded by 4 CAD. Moreover, by reducing the spray angle from 144° to 100°, the gross indicated efficiency decreases by 4% and CA50 is advanced by 6 CAD. The results showed that reduction in NOx emission is achievable, while controlling HC and CO emissions, by means of increasing the NG fraction, advancing the SOI1, increasing the fuel fraction in first DI injection with lower injection pressure and employing a wider injector spray angle. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Reactivity controlled compression ignition (RCCI) is a favorable dual-fuel low temperature combustion (LTC) with significant potential for improving thermal efficiency while reducing NOx and particular matter (PM) emissions compared to the conventional internal combustion (IC) engines [1]. In RCCI engines, two ⇑ Corresponding author. E-mail address: [email protected] (R.K. Saray). http://dx.doi.org/10.1016/j.apenergy.2017.05.011 0306-2619/Ó 2017 Elsevier Ltd. All rights reserved.

fuels with different reactivities are needed. The fuel with lower reactivity (such as NG or gasoline) provides the majority of the energy during the combustion process and the high reactive fuel (such as diesel or biodiesel) is directly injected into the combustion chamber during compression stroke [2–8]. Plenty of parametric studies have been done by researchers in recent years on dual fuel RCCI engines. Recently, engine researchers are interested on investigation of RCCI engines with NG/diesel Fuels [9–14]. NG is one of the most preferred fuels for using in RCCI engines due to its lower carbon to hydrogen ratio (C/H) which results in lower emissions

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

431

Nomenclature P T

pressure (kPa) temperature (K)

Abbreviations ALE arbitrary Lagrangian Eulerian aTDC after top dead center bTDC before top dead center BD burning duration CFD computational fluid dynamic CA50 crank angle of 50% heat released CO carbon monoxide CO2 carbon dioxide DI diesel injection EGR exhaust gas recirculation EVO exhaust valve open GIE gross indicated efficiency HRR heat release rate HC unburned hydrocarbons

like unburned hydrocarbons and CO compared to gasoline/diesel RCCI engines [9]. Also, the large reactivity gradient between NG and diesel, compared to gasoline and diesel, makes it attractive for RCCI combustion mode. The larger reactivity difference benefits to extend the combustion duration and reduce the peak cylinder temperature that leads to reduction in NOx formation [9,10]. Moreover, using NG instead of gasoline leads to a reduction in the flame temperature, which is another reason for reducing the NOx formation [10]. While the heating value of methane is higher than that of gasoline, the adiabatic flame temperature is not necessarily higher. This is because the adiabatic flame temperature also depends on the fuel air ratio [9,10]. In RCCI combustion mode, increasing the premixed burn rate by controlling the pilot and main DI timing and duration enable us to achieve lower local equivalence ratio and temperature [9]. As literature review in NG/Diesel RCCI combustion, some of the studies are summarized as below. Neiman et al. [10] used multidimensional CFD code, KIVA 3V, coupled with CHEMKIN chemistry and genetic algorithm (NSGA II) to optimize NG/diesel RCCI engine parameters including fuel fraction, DI SOI, DI fuel quantity split, injection pressure of diesel and amount of EGR. The results showed that split injection strategy was promising for an RCCI engine to achieve higher efficiency and lower emissions, while engine loads sweeping from low levels to high levels. Zoldak et al. [11] used KIVA-CHEMKIN code to study combustion, performance and emission characteristics of a NG/diesel RCCI engine. They illustrated that the in-cylinder peak pressure and pressure rise rate (PRR) are increased in NG/diesel RCCI combustion compared to the conventional diesel mode. Also, they numerically investigated the effect of direct injection of NG in RCCI mode on engine performance and emissions. They found out that the direct injection of NG leads to more stratification of the NG fuel in the combustion chamber and avoids over premixing of the incylinder mixture. So, this will control the PRR at higher engine loads [12]. Doosje et al. [13] focused on NG/Diesel RCCI combustion in a six-cylinder heavy duty engine. The engine could be operated in RCCI mode between 2 and 9 bar IMEP without EGR while satisfying Euro IV soot and NOx emissions standard. However, they couldn’t control HC emissions to obtain Euro IV standard level. Min et al. [14] experimentally investigated the effects of preinjection parameters on combustion and emissions in a pilot diesel/CNG engine. Their results indicate that early pre-injection in

IVC LTC NG NOx PFI PR PRR SA SOC SOI RCCI RI TDI

inlet valve closure low temperature combustion natural gas nitrogen oxides port fuel injection premixed ratio pressure rise rate spray angle start of combustion start of injection reactivity controlled compression ignition ringing intensity turbo charged direct injection

Subscript amb ambient exh exhaust int intake

compression stroke is useful solution to reduce in-cylinder pressure and HRR and lower the NOx emission. However, later preinjection timing resulted in higher HRR which increases NOx emission. Dahodwala et al. [15] experimentally studied the Diesel/CNG RCCI combustion mode. It was illustrated that RCCI combustion is achievable at low engine loads with higher NG replacement along with lower engine emissions. Bekdemir et al. [16] developed a multi-zone model to simulate NG/Diesel RCCI combustion in a heavy-duty diesel engine. Results showed that their map-based real-time RCCI model can be used to develop model-based NG/Diesel RCCI controls. Denbratt et al. [17] studied the influence of injection timing and duration of diesel injection on combustion phasing and pollutants of a heavy duty single cylinder engine. Also, they studied the effect of various compression ratios at different engine loads and speeds. The results confirmed that both low NOx and significantly low soot emissions can be achieved but HC emissions increases dramatically. Also, CA50 can be retarded by lowering the compression ratio from 17 to 14 while keeping unburned hydrocarbon emissions in acceptable range. Peykani et al. [18] investigated the effect of different injection strategies of diesel fuel on the combustion and emission characteristics of a heavy-duty NG/Diesel RCCI engine. They calibrated their model by means of Nieman et al. [10] simulations. Results showed that in a heavy duty RCCI engine, SOI1 and SOI2 and the diesel fuel quantity split are three main factors which affect the RCCI engine performance and emissions. Also, in order to reduce soot and NOx emissions, advancing the SOI1 and SOI2 timings and increasing the fuel split fraction in SOI1 are desired. According to the literature, the injection strategy of diesel fuel has significant effect in controlling the reactivity of the mixture in combustion chamber which affects RCCI engine emission and performance characteristics, but less comprehensive researches have been dedicated to computationally investigating the effects of injection strategy of diesel fuel in NG/diesel RCCI engines. To our knowledge, there is no systematical study on the emission and combustion characteristics of a light duty NG/diesel RCCI engines. Therefore, in the present study, the effects of diesel split injection strategy on the combustion and emission characteristics of a light duty NG/diesel RCCI engine are investigated. NG PR, SA, injection pressure of the first pulse, diesel SOI1 and SOI2 timings, and diesel fuel mass split between two injection pulses were selected as important parameters which are studied in details.

432

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

2. Model description The CONVERGE CFD code [19] was used for the RCCI engine simulations. Regarding combustion modeling, detailed chemistry method was used, using the SAGE solver [20]. In the SAGE model, CVODES solver is used to solve initial value problems of ODE systems. NG was considered to be homogeneous mixture with air at IVC (start of simulation). Also, the standard Droplet Discrete Model (DDM) was used to simulate injection process of the diesel fuel [20]. Spray atomization and break-up were modeled by using the hybrid KH-RT model [21,22]. Also, the physical properties of diesel fuel are represented by tetradecane (C14H30) for the spray and mixing processes [19]. The RNG k-e model was used as turbulence model [23]. Also, wall function model of Han and Reitz [23] was considered to wall heat transfer process. However, in order to decrease running time, the multi-zone chemistry solving method was employed [24]. This solver significantly decreases the calculations of combustion, considering computational cells with similar properties into the same zones in the chemistry calculations. In order to accurate prediction of chemical chemistry, 1 ls time step was selected for combustion chemistry calculations. In this research, a reduced chemical kinetics mechanism of CH4 oxidation developed by Bahlouli et al. [25,26] is used for CH4 burning. Also, the optimized Valeri mechanism for n-heptane (as a diesel surrogate), is applied for diesel fuel oxidation which was introduced by Rahimi et al. with 23 species and 139 reactions [27]. The reduced NOx mechanism proposed by Sun [28], which consists of four additional species and 12 reactions, is used for NOx emission prediction. Finally, the mechanism used for RCCI combustion modeling contains 57 species and 190 reactions.

3. Experimental setup and engine specifications All experiments were conducted in Advanced power system (APS) research facility at Michigan Technological University [29].

The schematic view of engine test bed is shown in Fig. 1. A 1.9L, inline 4 cylinder VW TDI engine is used in this study. As shown in Fig. 1, both high and low pressure EGR loops are set between intake and exhaust ports of the engine. The EGR flow is being controlled by a throttle valve, which is installed on low pressure loop EGR. Few modifications have been done on the intake manifold to install port fuel injectors (PFI). The modification allows the PFIs to spray right into the intake port. Variable geometry turbocharger (VGT) and common rail direct diesel injection system were left in the production form. An oxygen sensor is installed at the intake and exhaust ports of the engine to measure the oxygen concentration at both sides for EGR ratio measurement [29]. EGR ratio was calculated by Eq. (1):

Table 1 Fuel specifications [9]. Fuel type Total sulfur (ppm) Initial boiling point (°C) Final boiling point (°C) Cetane index Water content (ppm) Higher heating value (MJ/kg) Lower heating value (MJ/kg)

ULSD 7 184 363 48.7 34 45.68 42.89

Table 2 Engine specifications [9]. Cylinder arrangement Cylinder bore/stroke (mm) Geometric compression ratio IVC (aTDC) EVO (aTDC) Piston bowl Max. power (kW) Max. torque (N m)

Fig. 1. Engine test setup layout [9].

Inline 4 79.5/95.5 17:1 169 162 Mexican hat 66 @ 3750 RPM 210 @ 1900 RPM

433

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

Head

Liner

Piston

Crevice

Fig. 2. Geometry of the combustion chamber used for the computational study.

Table 3 Operation conditions for the light duty RCCI engine. Parameters

Case a

Case b

Case c

Case d

Speed (RPM) BMEP (Bar) Diesel flow rate (g/s) NG flow rate (g/s) Air flow rate (kg/h) SOI1/SOI2 (bTDC) Split type (%/%) T_IVC (K) Common Rail Pressure (bar) BR% EGR%

1300 4 0.071 0.50 60.736 Single - 20 Single 348 400 89 0%

1500 5 0.107 0.56 55.95 55/20 70/30 378 400 85 20%

2000 6 0.099 0.91 87.51 55/20 68/32 420 400 90 15%

1500 5 0.206 0.50 70.2 55/25 50/50 345 400 74 0%

EGR % ¼

ðO2 Þamb  ðO2 Þint  100 ðO2 Þamb  ðO2 ÞExh

ð1Þ

Two control valves are installed for cooling water to adjust the water flow to keep the oil and engine coolant temperature at 100 °C and 90 °C, respectively. The VW TDI engine is connected to 150 kW Dynamitic 8100 eddy-current absorption dynamometer with maximum 6000 RPM [30]. The DI diesel and PFI NG flow rates were measured by Micro Motion Coriolis flow meter (CMF010). High speed combustion data is being recorded by PCB model

Fig. 3. The validation of in-cylinder pressure and HRR for four cases.

434

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

175A01 installed in the glow plug hole. The high speed data including in-cylinder pressure, MAP and fuel flow is being recorded by ACAP (Advance Combustion Analysis Program) with 0.1 degree crank angle resolution. The DI fuel used for this work was ultra low sulfur diesel (ULSD). Fuel property is shown in Table 1. The PFI fuel was NG which is the main (>95%) composition of NG. Engine specifications are provided in Table 2.

A MotoTron MPC-555 Black oak ECU, which is a part of a target based rapid prototyping engine controller used to control and monitor the test engine during operation. [30]. The Bosch high pressure fuel pump, which can pressurize the fuel rail up to 2000 bar, are driven by timing belt. Also, Bosch CRDI injector with six nozzles (nozzle hole diameter = 0.165 mm) and 144° degree SA was used for direct injection of diesel fuel [9,30]. The exhaust gas

Fig. 4. The validation of Emissions for four cases of Table 3.

Fig. 5. Effects of PR of in-cylinder mixture on in-cylinder Pressure (left) & HRR (right) for the engine operating condition, case b.

435

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

measurement system is HORIBA MEXA-1600D/DEGR. Three analyzers are utilized in this system, which are CO/CO2 analyzer (AIA-260), O2/CO2 analyzer (INA-260), and total HC/NOx analyzer (FCA-266) [30]. 4. Model validation

Fig. 6. Effects of PR on GIE, CA50 and RI for the engine operating condition, case b.

In this study, the simulation started from the intake valve closing timing (169° bTDC) to the exhaust valve opening timing (162° aTDC). The initial in-cylinder mixture at IVC was assumed to be fully homogenous and uniform with the swirl ratio of 2.0 at the beginning of computation according to combustion chamber shape [9]. To have accurate in-cylinder geometry, 3D scan of real engine was used. Since the Mexican hat bowl of piston is not located in the center of piston, so the 360 degree of combustion chamber was considered in the simulation, as shown in Fig. 2. A structured Cartesian grid was used in the CFD code along with base cell size

Fig. 7. Effects of PR on the emissions for the engine operating condition, case b.

55% NG

75% NG

85% NG

-20 aTDC

-5 aTDC

20 aTDC

Fig. 8. Effects of PR on in-cylinder temperature for the engine operating condition, case b.

436

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

Fig. 9. Effects of SOI1 timing on in-cylinder Pressure (left) & HRR (right) for the engine operating condition, case b.

the spray region and boundaries such as Liner, Piston and crevice volumes was implied in start of simulation. The engine operating conditions for validation of simulation results are given in Table 3. As it is seen, four different engine speeds, BMEP, injection strategies and EGR ratios were used to achieve the model with acceptable accuracy. Fig. 3 shows comparison of the predicted in-cylinder pressure and heat release histories with the corresponding experimental data for all four cases. Experimental gross HRR was calculated by following equation (not included heat transfer and crevices effect):

dQ c dV 1 dP ¼ þ P V dh c  1 dh c  1 dh

Fig. 10. Effects of SOI1 timing on GIE, CA50 and RI for the engine operating condition, case b.

of 3 mm. Also, the CFD code uses an adaptive mesh refinement (AMR) which controls the cells size according to critical factors like temperature, velocity and species concentration. In this study, maximum embedding level was 3, [19]. A fixed refinement within

ð2Þ

c is the average slope of compression and expansion strokes in log P-V diagram. dV and dP over dh are the volume and pressure changes at the corresponding crank angle positions [9]. According to the mentioned comparison, the maximum difference between the experimental and predicted in-cylinder peak pressure is 3.5%. Also, the mean errors for SOC and CA50 prediction in four engine operating cases are 2.0 and 1.2 CAD, respectively. So, the CFD model can predict the engine combustion phasing and performance with acceptable accuracy. Fig. 4 shows the comparison between the simulated and measured engine emissions. It can be seen that NOx emission is under-predicted by the model. Also, the model predicts HC emission for four cases with errors around 4.6% for cases a, b and

Fig. 11. Effects of SOI1 timing on the emissions for the engine operating condition, case b.

437

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

SOI1= -85 aTDC

SOI1= -55 aTDC

SOI1= -45 aTDC

-15 aTDC

0 aTDC

20 aTDC

Fig. 12. Effects SOI1 on in-cylinder temperature for the engine operating condition, case b.

Fig. 13. Effects of SOI2 timing on in-cylinder Pressure (left) & HRR (right) for the engine operating condition, case b.

d, and 18% for case c. CO emission is predicted with error of 10% for case a, 6% for case b, 18% for case c and 8% for case d. Due to better prediction of in-cylinder pressure, combustion phasing and emissions for case b by the model, and also stable performance of the engine in middle speeds, case b was selected for parametric study in the following sections.

5. Results and discussions 5.1. Effect of energy based premixed ratio (PR) PR is defined as the energy provided by methane upon the total energy provided by both fuels:

438

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

quent removal of diesel fuel and elimination of this ignition points provided by removed diesel. This will hinder ignition and will cause longer ignition delay. Ringing intensity (RI) is used as an indicator of knocking in RCCI engine which can be computed by the below equation [31]:

 dP 2   0:5 dt max pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi MW 1 ¼ RI  cRT max m2 2c Pmax

Fig. 14. Effects of SOI2 timing on GIE, CA50 and RI for the engine operating condition, case b.

PR ¼

_ NG  LHV NG m  100 _ diesel  LHV diesel þ m _ NG  LHV NG m

ð3Þ

_ NG and m _ diesel represent mass flow rates of NG and diesel, where m respectively. Lower heating values of NG and diesel are represented by LHV NG (50 MJ/kg) and LHV diesel (42.89 MJ/kg), respectively [9]. Amount of total introduced energy is kept constant for all parametric sweeps. Effect of PR on HRR and in-cylinder pressure is presented in Fig. 5. In this figure, all premixed ratios provide the same amount of the total energy for combustion. The results in this graph demonstrate that sweeping the PR from 55% to 85%, increases the ignition delay, reduces the in-cylinder PRR and HRR, and higher peak pressure. Moreover, GIE is observed to increase from 35% to 42% and CA50 to be retarded from 10 to 0 CAD, as seen in Fig. 6. An important conclusion in this section is significant relation between GIE increase and combustion phasing delay which deserves to be emphasized. However, further increase in PR to over 85% results in significant drop in GIE and rise in CA50. The reason which can be set forth for this phenomenon is that increasing the PR to values beyond 85%, reduces the reactivity of mixture by removing reactive fuel out. This lowering the reactivity is very likely for misfiring. Similarly, increased ignition delay can be justified by increasing the NG fraction and consequent decrease in mixture reactivity. Increasing PR, requires reduction in diesel fuel quantity to keep the total introduced energy unchanged and conse-

ð4Þ

  where dP is the maximum rise rate of the in-cylinder pressure dt max (kPa/ms), c is the ratio of specific heats, Tmax is the maximum incylinder temperature (K), Pmax is the maximum in-cylinder pressure, and R (J/kgK), is the gas constant. The limit for RI is <5 MW/ m2 [31]. Fig. 6 demonstrates the effect of PR ratio on RI. It can be observed that RI decreases by increasing PR ratio. This phenomenon can be justified by considering HRR and pressure rise rate trends in Fig. 5. Based on the presented outcomes at Fig. 5, increasing the PR ratio results in decreasing the HRR and PRR. Therefore, RI decreases by power 2 according to Eq. (4). Fig. 7 depicts variations of NOx, unburned HC and CO emissions versus PR. Increasing the PR results in lower NOx emission. The reduced NOx emission is in result of retarded combustion due to lower reactivity. It reduces combustion duration and available time for NOx production and results in lower the NOx emission. Also, Fig. 7 shows CO emission increasing trend with increasing the PR. Carbon monoxide emission is mainly generated in rich and lower temperature regions. It could be concluded that the more NG in charge, the lower reactive zones results in lower temperature, which increases CO emission. As it is illustrated in right side of Fig. 7, HC emission increases with increasing the PR. The HC emission depends on the geometry such as crevices and squish volumes and the in-cylinder combustion temperature near cylinder walls. According to Fig. 8, when the PR increases, the in-cylinder temperature reduces which results in increase of HC emission. Moreover, it can be observed that the crevice volume and regions near to the walls are colder in higher PRs. This will reduce combustion efficiency and higher unburned HC emission. 5.2. Effect of injection timing 5.2.1. First injection timing (SOI 1) Investigation on the effects of the first injection timing (SOI1) was completed by sweeping SOI1 from 85° to 45° bTDC while

Fig. 15. Effects of SOI2 timing on the emissions for the engine operating condition, case b.

439

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

SOI2= -40 aTDC

SOI2= -20 aTDC

SOI2= 0 aTDC

-10 aTDC

0 aTDC

20 aTDC

Fig. 16. Effects SOI2 on in-cylinder temperature for the engine operating condition, case b.

Fig. 17. Effects of first injection pulse fuel quantity on in-cylinder Pressure (left) & HRR (right) for the engine operating condition, case b.

keeping SOI2 timing constant. SOI2 timing was at 20° bTDC for case b. For this case, 70% of the total diesel fuel was injected at the first pulse and 30% at the second pulse. Fig. 9 shows the in-cylinder pressure and HRR curves for several SOI1 timings. Advancing SOI1 timing from 45° to 85° bTDC results in delayed combustion and reduced peak in-cylinder pressure and HRR. Advancing the first diesel injection, increases the available mixing time, which results in more premixed mixture of methane and diesel, and lower local equivalence ratio gradients and higher homogeneity

of the mixture. Therefore, less reactive mixture is prepared because of the longer mixing time [32,33]. Fig. 10 shows the effect of advanced SOI1 timing on GIE, CA50 and RI. By advancing SOI1 timing, GIE increases from 41.5% to 45% and CA50 retards from 1° bTDC to 4° aTDC. RI showed a slight reduction trend. By advancing the SOI1, it seems that combustion is going to be more premixed as most of fuel is injected in the SOI1. Therefore, less local reactive mixture is produced and CA50 is retarded to near and after TDC with shorter burning duration which results in higher GIE. Also,

440

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

of in-cylinder temperature in Fig. 12 is the reason of the increasing trend of NOx emission. On the other hand, it can be concluded from Fig. 12 that spray is impinged on the wall and enters to the crevice volume which is the main source of HC and CO due to their lower temperature and lower volume to surface ratio. In later SOI1, like SOI1 at 45° bTDC, most of the fuel is injected into the piston bowl and squish area, where the temperature and reactivity gradient is high, so lower amounts of HC and CO are produced.

Fig. 18. Effects of the first injection pulse fuel quantity on GIE, CA50 and RI for the engine operating condition, case b.

advanced injection timing faces with lower in-cylinder temperature and pressure which adversely affects the atomization and evaporation of the fuel droplets. Fig. 11 shows the effect of SOI1 sweep on CO, HC and NOx emissions. As the SOI1 timing is retarded, NOx emission is increased while the levels of HC and CO are decreased. The increasing trend

5.2.2. Second injection timing (SOI 2) The second injection timing (SOI2) was swept from 40° to 0° bTDC whereas the SOI1 timing was kept constant at 55° bTDC. Fig. 13 shows the in-cylinder pressure and HRR for different SOI2 timings. As illustrated in this figure, by retarding the SOI2 from 40° to 0° bTDC, combustion phasing is retarded and the peak of in-cylinder pressure decreases. There is not enough time for the second diesel injection pulse to be mixed properly with incylinder mixture, hence ignition delay is increased. In addition, by increasing the delay in SOI2 timing, the injected diesel fuel in the first pulse is more mixed; therefore, local fuel reactivity of mixture reduces which will sweep the CA50 to a position after TDC as shown in Fig. 14. When SOI2 timing shifts to TDC, a bigger amount of released heat happens after TDC which results in GIE reduction. Moreover, there is no significant change in RI by retarding the SOI2 and it is below 5 MW/m2 for all the SOI2.

Fig. 19. Effects of the first injection fuel quantity on the emissions for the engine operating condition, case b.

Fig. 20. Effects of the first injection pressure on in-cylinder Pressure (left) & HRR (right) for the engine operating condition, case b.

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

Fig. 21. Effects of the first injection pressure on GIE, CA50 and RI for the engine operating condition, case b.

Fig. 15 shows the effect of SOI2 sweep on the emission trends. According to Figs. 14 and 16, due to the high local in-cylinder temperature near TDC, the amount of NOx is increased at earlier SOI2 timings. Also, as can be seen in Fig. 15, the lowest CO and HC emissions were achieved when SOI2 is advanced. As it can be observed in Fig. 16, in SOI2 at 40° bTDC the flame could reach whole combustion chamber while in SOI2 at 0° bTDC, there are more regions with lower temperatures. Therefore, the amounts of HC and CO emissions are high at later SOI2 timings due to lower combustion efficiency [18].

5.3. Effect of first injection diesel fuel quantity The next parameter which was considered in the parametric study, was the amount of fuel split between the two diesel injection pulses at constant PR. The fraction of diesel fuel injected at SOI1 was changed from 40% to 90% for case b at the engine operating condition which was presented in Table 3. Fig. 17 illustrates the simulated in-cylinder pressure history and HRRs for fuel fraction sweep at the first injection pulse. As can be seen, the maximum in-cylinder pressure and amount of HRR are increased, and the

441

CA50 is advanced when most of diesel fuel is injected in the first pulse. The influence of SOI1 mass fraction of diesel fuel on GIE, RI and CA50 are also described in Fig. 18. As it is shown in the figure, GIE and RI are decreased slightly when increasing first pulse diesel fuel fraction. The increased in-cylinder total reactivity with more SOI1 fuel fractions leads in a more advanced CA50. Maximum amount of GIE is achieved at 50% mass fraction of the first injection pulse for case b. Fig. 19 shows the simulated emission trends for different first injection pulse fuel mass percentage. CO and HC are slightly increased as the first injection pulse fuel is increased. With increasing the first injection fuel quantity, more diesel fuel impinges on the wall and is accumulated in the crevice volume which results in more HC and CO emissions. However, injecting of more diesel fuel in the second injection pulse creates further stratification of the in-cylinder mixture which rises the mixture reactivity. Hence, these richer zones with higher reactivity ignite earlier, rising combustion chamber temperature, and consequently producing more NOx emission.

5.4. Effect of diesel first pulse injection pressure Injection pressure is an important factor which affects mixing quality of the diesel fuel and also stratification of in-cylinder mixture [34]. In this research, split injection strategy was used to operate the engine at case b. Injection pressure of first pulse has two effects: 1 by increasing the injection pressure, droplets momentum increases which results in wall impingement of sprays [34], 2 - by increasing the injection pressure, due to the smaller size of droplets, evaporation rate increases and less liquid parcels will be remained in squish region [35]. These two factors will affect the stratification and local reactivity of mixture in combustion chamber. As shown in Fig. 20, increasing the first pulse injection pressure from 300 bar to 450 bar, while keeping the second pulse injection pressure at 400 bar, causes peak in-cylinder pressure to increase by about 15 bar. High injection pressure of the first pulse causes the droplets with smaller diameter to be evaporated near the bulk region, and less parts of droplets penetrate deeply near the cylinder walls. Consequently, local reactivity increases, while peak HRR decreases and combustion phasing is advanced as seen in Fig. 20.

Fig. 22. Effects of the first injection pressure on the emissions for the engine operating condition, case b.

442

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

Injection pressure= 300 bar

Injection pressure= 400 bar

Injection pressure= 450 bar

-15 aTDC

0 aTDC

20 aTDC

Fig. 23. Effects of the first injection pressure on in-cylinder temperature for the engine operating condition, case b.

Fig. 24. Effects of the injector included angle on in-cylinder Pressure (left) & HRR (right) for the engine operating condition, case b.

Following the above discussion, Fig. 21 shows GIE, CA50 and RI trends for several first injection pulse pressures. It can be seen that GIE reduces for higher injection pressures as CA50 occurs before TDC. Also, there is a slight reduction in RI by increasing the injection pressure of the first pulse and it can be explained by maximum HRR which is reduced at higher injection pressures.

Figs. 22 and 23 support above-mentioned descriptions and show that for injection pressure of 300 bar, more diesel fuel parcels are accumulated around the bowl and piston surface at 0° aTDC in comparison to the injection pressure of 450 bar. As a result, by increasing the first injection pressure, due to higher in-cylinder temperature in piston bowl and bulk region, NOx emissions

443

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

SA= 100

SA = 120

SA = 144

-10 aTDC

0 aTDC

20 aTDC

Fig. 25. Effects of the SA on in-cylinder temperature for the engine operating condition, case b.

Fig. 26. Effects of the SA on GIE, CA50 and RI for the engine operating condition, case b.

increase and because of slower flame propagation through all combustion chamber, less methane fuel is burned completely which causes rising in HC and CO emissions. 5.5. Effect of spray angle (SA) This research studies the effects of SA on engine performance and emission characteristics. Injector SA affects the distribution of diesel fuel parcels (reactivity gradient of in-cylinder mixture) in the combustion chamber. Hence, it may influence the combus-

tion and emissions characteristics of an RCCI engine. The base SA for this injector is 144°. Fig. 24 shows the variations of the in-cylinder pressure and HRR for different SAs. Simulation results show that in-cylinder peak pressure increases in about 5 bar by decreasing the SA from 150° to 100°. Lower SAs result in diesel fuel to be targeted toward the piston bowl region with higher temperature compared to other regions in the combustion chamber, which increase the incylinder peak pressure. The highest HRR is observed at SA equals to 100°. Fig. 25 shows that diesel fuel droplets are gathered inside the bowl for narrower SAs, whereas for wider SA, diesel fuel droplets target the walls and the squish regions. Due to the higher temperature of piston bowl, diesel spray impingement on the piston bowl is more preferred than the piston top surface impingement or the cylinder liner impingement [36]. Hence, combustion phasing (CA50) is advanced for narrower SA which results in lower GIE as shown in Fig. 26. The RI level reduces as SA increases. The RI values are less than the standard value of 5 MW/m2 for all SAs. Fig. 27 shows the variations of engine-out HC, CO and NOx emissions for various SAs. It is shown that higher levels of HC and CO emissions were produced within the narrower SAs (100° and 120°). The reason is that for wider SAs (140° and 150°), the diesel fuel droplets go toward the squish zone; thus, the ignition points ignite the mixture and flame spreads in the squish zone and the cylinder liner. Hence, complete burning of the mixture in the squish zone happens, while narrower SAs (100° and 120°) cause the mixture to burn near centerline of the cylinder. However,

444

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

Fig. 27. Effects of the SA on the emissions for the engine operating condition, case b.

SA= 100

SA = 120

SA = 144

-10 aTDC

-5 aTDC

0 aTDC

20 aTDC

Fig. 28. Effects of the SA on methane concentration for the engine operating condition, case b.

at narrower SAs, hydrocarbons, including homogeneously distributed NG and diesel mixture cannot be burned properly near the liner and crevice volumes. Because higher reactive mixture is formed in piston bowl and the lower reactive mixture is formed in the squish region as seen in Fig. 28. Also at crank angle position

of 20° aTDC, it can be observed that with wider SA, flame propagation in squish region is higher than the narrower SA. This results in lower HC and CO emissions as seen in Fig. 27. However, as it can be seen in Fig. 28, due to higher temperature of piston bowl, the burning rate of methane in the narrower SA is higher. Fig. 27 also shows

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

that the narrower SAs generally yield greater NOx emission due to the higher in-cylinder gas temperatures. It is seen from Fig. 25 that high temperature regions at SA of 100° are wider than the case with SA of 144° at 20° aTDC. 6. Conclusion A CONVERGE 3D/CFD combustion model with a detailed reaction mechanism was developed to simulate a light duty dieselNG RCCI engine. The validation results of the model against four different operating conditions showed that the model is able to predict the combustion phasing and performance with reasonable accuracy. A parametric study was done to study the effects of diesel injection strategies, first diesel injection pressure and injector SA on the RCCI engine performance and emissions characteristics. The conclusion of this study can be summarized as follows: 1. By increasing the PR, the lower reactivity of NG causes lower combustion rate and longer ignition delay. Therefore, the peak HRR, RI and combustion temperature decrease. Low temperature of combustion, due to shorter reaction time in higher PR, increases HC and CO emissions considerably and decreases NOx formation. 2. Diesel injection strategy, controls the local reactivity of the incylinder charge. In split injection strategy, as the SOI1 is retarded toward TDC, local fuel reactivity of mixture increases combustion temperature by raising local equivalence ratio in combustion chamber. This increases GIE and CA50 shifts over TDC. Also, by retarding the SOI1, NOx production increases while HC and CO decrease. 3. It was observed that retarding the SOI2 timing toward TDC, there is a slight increase in GIE while CA50 is retarded and RI increases. Also, by retarding the SOI2 timing, HC and CO increase while NOx decreases. The reason is that as SOI timing is delayed, the SOC is more retarded which reduces combustion duration and decreases the in-cylinder temperature. 4. Higher amount of diesel fuel in the first injection pulse increases the in-cylinder pressure and HRR, and advances the combustion phasing. By increasing the first injected fuel quantity, more diesel is accumulated in the crevice volume and on the cylinder wall, which results in higher HC and CO emissions. The increased local reactivity of the mixture from the second injection pulse raises combustion temperature, and consequently increase the NOx emission formation. 5. The results showed that higher first injection pressure increases the penetration of diesel droplets which causes wall impingement. On the other hand, high injection pressure increases the evaporation rate of droplets. Hence, stratification of diesel fuel increases and combustion phasing is advanced. Due to higher in-cylinder temperature in the piston bowl and the bulk region, NOx emission increases. Because of the lower flame propagation through combustion chamber, less premixed methane fuel is burned near the cylinder wall and crevice volume which causes higher HC and CO emissions. 6. It was founded that for wider SAs, burning of the fuel occurs in both the centerline and the squish regions, while the narrower SAs cause the mixture to be burned near the cylinder bulk. Therefore, the narrower SAs cause higher values of HC and CO emissions near the cylinder wall and crevice volumes. Also, NOx emission is greater with lower SA due to richer and locally higher reactive zone with higher combustion temperature. Future work In the most parametric studies, one parameter is swept at the same time and the others are kept constant. Due to complexity

445

of combustion process in the RCCI engines, it will be useful to investigate the interaction of the effective parameters to find the way to optimize the engine performance and emissions by more than one parameter at the same time. Hence, in next step, the authors will investigate the effects of the interaction of the studied parameters on the engine performance and emissions by means of statistical methods. Acknowledgments The authors would like to thank Paul Dice form Michigan Technological University for sharing his experiences with us for preparing the engine setup, and emission analyzer calibration. References [1] Reitz RD, Duraisamy G. Review of high efficiency and clean reactivity controlled compression ignition (RCCI) combustion in internal combustion engines. Prog Energy Combust Sci 2015;46:12–71. [2] Dempsey B, Walker N, Reitz RD. Effect of piston bowl geometry on dual fuel reactivity controlled compression ignition (RCCI) in a light-duty engine operated with gasoline/diesel and methanol/diesel. No. 2013-01-0264. SAE technical paper; 2013. [3] Walker N, Dempsey A, Andrie M, Reitz RD. Use of low-pressure direct-injection for reactivity controlled compression ignition (RCCI) light-duty engine operation. SAE Int J Engines 2013;6(2):1222–37. [4] Benajes J, Molina S, García A, Belarte E, Vanvolsem M. An investigation on RCCI combustion in a heavy duty diesel engine using in-cylinder blending of diesel and gasoline fuels. Appl Therm Eng 2014;63(1):66–76. [5] Li J, Yang WM, An H, Zhou DZ, Yu WB, Wang JX, et al. Numerical investigation on the effect of reactivity gradient in an RCCI engine fueled with gasoline and diesel. Energy Convers Manage 2015;92:342–52. [6] Curran S, Hanson R, Wagner R, Reitz RD. Efficiency and emissions mapping of RCCI in a light-duty diesel engine. No. 2013-01-0289. SAE technical paper; 2013. [7] Yang B, Yao M, Cheng WK, Li Y, Zheng Z, Li S. Experimental and numerical study on different dual-fuel combustion modes fuelled with gasoline and diesel. Appl Energy 2014;113:722–33. [8] Nazemi M. Modeling and analysis of reactivity controlled compression ignition (RCCI) combustion. Master Thesis in Mechanical Engineering Department. Michigan Technological University; 2015. [9] Ansari E, Poorghasemi K, Irdmousa BK, Shahbakhti M, Naber J. Efficiency and emissions mapping of a light duty diesel-natural gas engine operating in conventional diesel and RCCI modes. In: SAE international powertrain, fuel and lubricants conference, 2016-01-2309; 2016. [10] Nieman D, Dempsey A, Reitz RD. Heavy-duty RCCI operation using natural gas and diesel. SAE Int J Engines 2012;5(2):270–85. http://dx.doi.org/10.4271/ 2012-01-0379. [11] Zoldak P, Sobiesiak A, Bergin M, Wickman D. Computational study of reactivity controlled compression ignition (RCCI) combustion in a heavy-duty diesel engine using natural gas. SAE technical paper 2014-01-1321; 2014. http:// dx.doi.org/10.4271/2014-01-1321. [12] Zoldak P, Sobiesiak A, Wickman D, Bergin M. Combustion simulation of dual fuel CNG engine using direct injection of natural gas and diesel. SAE Int J Engines 2015;8(2):846 e58. http://dx.doi.org/10.4271/2015-01-0851. [13] Doosje E, Willems F, Baert R. Experimental demonstration of RCCI in heavyduty engines using diesel and natural gas. SAE technical paper 2014-01-1318; 2014. http://dx.doi.org/10.4271/2014-01-1318. [14] Xu M, Cheng W, Li Z, Zhang H, An T, Meng Z. Pre-injection strategy for pilot diesel compression ignition natural gas engine. Appl Energy 2016;179:1185–93. [15] Dahodwala M, Joshi S, Koehler E, Franke M. Investigation of diesel and CNG combustion in a dual fuel regime and as an enabler to achieve RCCI combustion. SAE technical paper 2014-01-1308; 2014. http://dx.doi.org/10. 4271/2014-01-1308. [16] Bekdemir C, Baert R, Willems F, Somers B. Towards control-oriented modeling of natural gas-diesel RCCI combustion. SAE technical paper, 2015-01-1745; 2015. http://dx.doi.org/10.4271/2015-01-1745. [17] Jia Z, Denbratt I. Experimental investigation of natural gas-diesel dual-fuel RCCI in a heavy-duty engine. SAE Int J Engines 2015;8(2):797–807. http://dx. doi.org/10.4271/2015-01-0838. [18] Paykani A, Kakaee AH, Rahnama P, Reitz RD. Effects of diesel injection strategy on natural gas/diesel reactivity controlled compression ignition combustion. Energy 2012. http://dx.doi.org/10.1016/j.energy.2015.07.112. [19] Richards KJ, Senecal PK, Pomraning E. CONVERGE (v2.2.0). Middleton (WI): Convergent Science, Inc.; 2014. [20] Dukowicz JK. A particle-fluid numerical model for liquid sprays. J Comput Phys 1980;35:229–53. [21] Beale JC, Reitz RD. Modeling spray atomization with the Kelvin-Helmholtz/ Rayleigh-Taylor hybrid model. Atomization Sprays 1999;9:623–50.

446

K. Poorghasemi et al. / Applied Energy 199 (2017) 430–446

[22] Schmidt DP, Rutland CJ. A new droplet collision algorithm. J Comput Phys 2000;164:62–80. [23] Han Z, Reitz RD. Turbulence modeling of internal combustion engines using RNG k-e models. Combust Sci Technol 1995;106:267–95. [24] Raju M, Wang M, Dai M, Piggott W, Flowers D. Acceleration of detailed chemical kinetics using multi-zone modeling for CFD in internal combustion engine simulations. SAE technical paper 2012-01-0135; 2012. http://dx.doi. org/10.4271/2012-01-0135. [25] Bahlouli K, Atikol U, Saray RK, Mohammadi V. A reduced mechanism for predicting the ignition timing of a fuel blend of natural-gas and n-heptane in HCCI engine. Energy Convers Manage 2014;79:85–96. [26] Poorghasemi K, Saray RK, Bahlouli K, Zehni A. 3D CFD simulation of a natural gas fuelled HCCI engine with employing a reduced mechanism. Fuel 2016;182:816–30. [27] Rahimi A, Fatehifar E, Saray RK. Development of an optimized chemical kinetic mechanism for homogeneous charge compression ignition combustion of a fuel blend of n-heptane and natural gas using a genetic algorithm. Proc Inst Mech Eng, Part D: J Autom Eng 2010;224(9):1141–59. [28] Kokjohn SL, Reitz RD. Reactivity controlled compression ignition and conventional diesel combustion: a comparison of methods to meet lightduty NOx and fuel economy targets. Int J Engine Res 2013:1–17.

[29] Kokjohn SL, Hanson R, Splitter D, Reitz RD. Experiments and modeling of dualfuel HCCI and PCCI combustion using in-cylinder fuel blending. SAE Int J Engines 2010;2(2):24–39. http://dx.doi.org/10.4271/2009-01-2647. [30] Polonowski CJ. Accelerometer based measurements of combustion in an automotive turbocharged diesel engine. Master thesis, Michigan Technological University, Mechanical Engineering-Engineering Mechanics; 2009. [31] Eng JA. Characterization of pressure waves in HCCI combustion. SAE technical paper 2002-01-2859; 2002. [32] Curran S, Hanson R, Wagner R. Reactivity controlled compression ignition combustion on a multi-cylinder light-duty diesel engine. Int J Engine Res 2012;13(3):216–25. [33] Ma SH, Zheng Z, Liu H, Zhang Q, Yao M. Experimental investigation of the effects of diesel injection strategy on gasoline/diesel dual-fuel combustion. Appl Energy 2013;109:202–12. [34] Poorghasemi K, Ommi F, Yaghmaei H, Namaki A. An investigation on effect of high pressure post injection on soot and NO emissions in a DI diesel engine. J Mech Sci Technol 2012;26(1):269–81. [35] Hariprasad T. Effect of injection pressure on performance of dual fuel diesel engine. SAE technical paper 2013-01-2887; 2013. [36] Lim J, Min K. The effects of spray angle and piston bowl shape on diesel engine soot emissions using 3-D CFD simulation. SAE technical paper 2005-01-2117; 2005.