Experimental study of glow plug assisted compression ignition

Fuel 197 (2017) 111–120

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Experimental study of glow plug assisted compression ignition Changsheng Yao, Tianyuan Zhou, Fuyuan Yang ⇑, Yaodong Hu, Jinli Wang, Minggao Ouyang Department of Automotive Engineering, State Key Laboratory of Automotive Safety and Energy, Tsinghua University, Beijing 100084, PR China

a r t i c l e

i n f o

Article history: Received 5 October 2016 Received in revised form 8 January 2017 Accepted 1 February 2017

Keywords: Dieseline Multi-mode combustion GA-CI Glow plug Pressure sensor glow plug

a b s t r a c t The purpose of this study is to investigate the effects of glow plug assist on a four-cylinder compression ignition (CI) engine fueled with dieseline. Pressure sensor glow plugs (PSG) are utilized to assist low temperature combustion (LTC) at low to medium load range, and the glow plugs are required for fast response and accurate output control. Therefore, a glow plug control unit (GPCU) is developed, and a closed-loop power feedback control algorithm is used. Equipped with PSGs and GPCU, the engine is tested at three speeds under varying loads. As a part of multi-mode combustion for CI engines, glow plug assisted combustion (GA-CI) appears earlier combustion phases and higher peak in-cylinder pressure. GA-CI can effectively reduce cycle-to-cycle variations and misfire is avoided, especially at low load conditions. The nature of the glow plug assisting process is to enhance the in-cylinder temperature and fuel reactivity. Plus, glow plugs also play a role in triggering the auto-ignition of the pre-mixture. With glow plug assist, NOx emissions rise slightly but are still below 0.2 g/kW h, whereas particulate matter (PM) emissions drop sharply and decrease to under 0.02 g/kW h. Since CO emissions and HC emissions decrease as well, the combustion efficiency is significantly enhanced, with the maximum being over 98%. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction As a combination of gasoline spark ignition and diesel compression ignition, Homogeneous Charge Compression Ignition (HCCI) was proposed and widely investigated due to its high thermal efficiency and low particulate matter (PM) and NOx emissions. However, difficulties in controlling combustion phases and chemical reaction processes drove engine researchers to shift their focus to other low temperature combustion (LTC) modes in order to extend the load limit and improve the combustion performance [1]. In such studies, many assisting methods have been utilized to help the process of compression ignition be more stable and efficient. Spark Assisted Compression Ignition (SACI) and Partially Premixed Compression Ignition (PPCI) were developed as two of the most promising alternative combustion modes for HCCI [2–5]. SACI is described as a bridge across the gap between HCCI and spark ignition (SI) [5]. In the SACI concept a spark plug is used as an additional means of combustion control. It starts propagating a flame and then the rising temperature triggers auto-ignition of the rest of the fuel-air mixture. A negative valve opening (NVO) is usually utilized to trap hot residuals that form at higher precombustion temperature and induce the mixture to tend toward

⇑ Corresponding author. E-mail address: [email protected] (F. Yang). http://dx.doi.org/10.1016/j.fuel.2017.02.008 0016-2361/Ó 2017 Elsevier Ltd. All rights reserved.

auto-ignition [6–8]. However, SACI is still limited in medium to high load conditions due to high cycle to cycle variation [9]. PPCI combines conventional diesel combustion (CDC) and HCCI. In the PPCI concept, gasoline-like fuels are directly injected into the cylinder during the compression stroke, and the end of injection (EOI) is ahead of the start of combustion (SOC) [10,11]. Precise injection control is needed for the formation of the partial premixture. A large amount of exhaust gas recirculation (EGR) is used to prolong the ignition delay in order to form the pre-mixture of fuel and air. Researchers have found that fuels with a low octane number and a high volatility are ideal for PPCI, and studies have been carried out to find the proper fuels compression ignition [12–23]. For instance, Wang et al. [17,18] and Yao et al. [19] tested naphtha under different loads. Wang et al. [20] studied the emission characteristics of biodiesel. Manente et al. [20,21] and Splitter [23] conducted research on a blend of ethanol and gasoline. Liu [24,25] and Tong [26] investigated using a blend of gasoline and polyoxymethylene dimethyl ethers (PODE). Among the above test fuels, the blend of gasoline and diesel, which is also called ‘dieseline’ [27,28], could be the most feasible solution because it is the easiest to obtain in most commercial markets. Therefore, dieseline seems to have the highest potential to be popularized. According to the research of Xu et al. [27,28], Weall and Collings [29] and Han et al. [30,31], dieseline can maintain the thermal efficiency of diesel while reducing the total concentration and mean diameter of PM emissions significantly.

112

C. Yao et al. / Fuel 197 (2017) 111–120

Nomenclature CA10 CA50 CDC CI COV ECU EGR EOI GA-CI GPCU HCCI IMEPm

crank angle when 10% heat is released crank angle when 50% heat is released conventional diesel combustion compression ignition coefficient of variation engine control unit exhaust gas recirculation end of injection glow plug assist compression ignition glow plug control unit homogenous charge compression ignition measured indicated mean effective pressure

Based on LTC research with dieseline, Yang et al. [32] proposed a multi-mode combustion concept for compression ignition engines. In this study, low temperature combustion (LTC) modes with direct injection were divided into two detailed branches [1,32]. One is early direct-injection LTC, known as PPCI, and it is used for the medium load range due to its higher thermal efficiency and low emission levels. Near-top dead center (TDC) direct-injection LTC is used for medium to high load range because emissions of PPCI deteriorate at such load conditions. However, challenges remain in low load combustion control. Due to differences in fuel reactivity, stable combustion is more difficult to achieve for gasoline-type fuel and in particular for high octane fuels. Some previous studies used spark plug assist [2–5], NVO [33,34], rebreathing [14], and intake heating [35], requiring the diesel engine to have special modifications. In the multi-mode combustion concept, Yang proposed Glow Plug Assisted Compression Ignition (GA-CI) for low load range, which utilized a glow plug as an additional means for increasing the fuel reactivity and helping trigger the pre-mixture to auto-ignite. Of note, glow plugs are commonly equipped on diesel engines and the glow plug used in this study is also a mass-produced product so extra modifications are not necessary. Generally, glow plugs are designed for cold start assist in diesel engines, and their use for cold start has been widely investigated [36]. Other than cold start assistance, Cheng et al. studied the performance of natural gas ignition assisted with glow plugs. Mueller et al. applied glow plugs for methanol ignition [37]. Ambekar et al. investigated glow plug assist for spray combustion of liquid nitromethane [38]. At the same time, using glow plugs for a gasoline premixed combustion is a relatively new area. Manente et al. applied glow plugs on a model engine [39]. Borgqvist et al. studied the effect of glow plugs on gasoline PPCI combustion, and combustion stability and efficiency were compared with glow plugs turned on or turned off [34]. However, further research on glow plug assist is missing. In this paper, as a part of the multi-mode combustion concept, GA-CI is proposed and studied via engine experiments. First, the characteristics of the glow plugs are presented and a selfdeveloped control unit is introduced. Second, engine experiments are undertaken at different loads and speeds, and a comparable study of engine performance with or without glow plug assist is presented in this paper. 2. Glow plugs and the glow plug control unit for GA-CI 2.1. Pressure sensor glow plugs (PSG) The Pressure Sensor Glow Plug (PSG) is a mass-produced resistance-type pressure sensor integrated with a fast response

LTC MPRR NVO PM PPCI PSG PWM SACI SI SOC TDC

low temperature combustion maximum pressure rise rate negative valve opening particulate matter partially premixed compression ignition pressure sensor glow-plug pulse-width modulation spark assisted compression ignition spark ignition start of combustion top dead center

glow plug, which means PSG is a sensor as well as an actuator. In previous studies, the pressure sensor function of PSG played an important role in developing a closed-loop combustion control system based on in-cylinder pressure [40–42]. In this paper, we focus on the glowing function of PSG and using it as one of the actuators of the control system to assist PPCI at low loads. Zhou et al. undertook experiments to study the glowing performance of three different glow plugs [43]. Their results showed that PSG could achieve much higher temperatures than another two glow plugs at both high and low voltages. At the same time, PSG had the widest high temperature region and the fastest temperature response among the test glow plugs. The excellent performance makes PSG suitable for GA-CI. The structure of PSG is shown in Fig. 1 [44] consisting of such elements as a glow plug heating rod, a glow plug body, the glow current terminal, a pressure measurement diaphragm, a printed circuit board, and sensor contacts. The dimensions and electric characteristics of PSG are listed in Table 1.

Fig. 1. Components of PSG: 1 – Plug, 2 – Circuit board with electronics, 3 – Glow plug body, 4 – Glow plug heating rod, 5 – High voltage connection, 6 – Measuring diaphragm, 7 – Gasket [44].

Table 1 PSG specifications. Dimensions Rod diameter (mm) Rod length (mm) Rod thickness (mm) Thread Total length (mm)

3.3 (front), 4 (rest) 24 0.75 M10 148.5

Electrical characteristics Maximum voltage (V) Time to reach 1000 °C (s)

11 <2

113

C. Yao et al. / Fuel 197 (2017) 111–120

2.2. Glow plug control unit (GPCU) Based on the experimental research on PSG’s glowing characteristics [43], a microcontroller based GPCU is developed to drive the PSGs and regulate output power. The basic structure of the GPCU is shown in Fig. 2. Fig. 3 is a picture of the GPCU. PSGs are driven by smart power switches and the effective output voltage is regulated through changing the duty cycle of pulse-width modulation (PWM). The GPCU also supports CAN communication with the engine control unit (ECU), and current detection for closed-loop control. The algorithm used on the GPCU is an output power based closed-loop control, and each PSG can be driven and independently controlled. In this study, the output power of glow plugs is 80 W, which is regulated by the GPCU through a PWM duty cycle of 60%. 3. Engine and experiment setup 3.1. Apparatus Equipped with PSGs and the GPCU, a four-cylinder 1.9 L common-rail diesel engine was used in this study. The detailed engine parameters are given in Table 2. The schematic of the engine system is shown in Fig. 4. Engine speed was controlled by an electrical dynamometer Horiba HT250. A low-pressure EGR loop was installed on the engine. The original Bosch-developed ECU was replaced with a Tsinghua University self-developed ECU, which was utilized to control flexible injections as well as change the boost pressure and the EGR rate. In-cylinder pressure was measured using four PSGs. The pressure data were recorded with a resolution of 0.375°CA using a Dewetron 5000 data acquisition platform. Combustion analysis data were calculated based on the average cylinder pressure of 100 consecutive cycles. CA10, and CA50 were defined as the crank angle when 10% and 50% heat was released, respectively. The fuel consumption measurement was taken by a To-Ceil digital fuel meter. The PM emissions were measured by a Cambustion DMS 500 M106 fast engine particulate analyzer, whereas total HC emissions were measured by a Cambustion HFR 500 fast HC analyzer. A Horiba Mexa-584L was used to measure the NOx and CO emissions. 3.2. Fuels In this paper, a blend fuel of diesel and gasoline was tested. The blend is G70D30 indicating that the blending ratio of gasoline and

Glow Plug1 Glow Plug2 Glow Plug3 Glow Plug4

Smart Power Switch

ADC1

Smart Power Switch Smart Power Switch Smart Power Switch

ADC3

Driver

Output

Fig. 3. GPCU.

Table 2 Engine parameters. Parameter

Description

Engine model Engine type Compression ratio Cylinder Turbocharger Displacement (L) Stroke (mm) Bore (mm) Rated power (kW)) Max. torque (N m) Low idle speed (rpm) Maximum torque speed (rpm) Maximum power speed (rpm) Fuel injection system

SQR481A 4-Valve compression ignition 17.5 Inline 4 Variable nozzle turbine 1.905 92.4 81 93 271 900 2800 3900 Common rail

diesel is 70:30 by volume. In our previous research [32,42], G70D30 was tested and had been confirmed as one proper fuel for PPCI, so we continued to use this fuel to study the impact of glow-plug assist in this paper.

BDM Interface

ADC2

CAN (MSCAN)

ADC4 Diagnose

CAN Controller Interface

Communication

RAM

PWM MCU

ROM

I/O EEPROM

V_Battery

Digital Core VCC

Fig. 2. Structure of GPCU.

BDM COM CAN BUS

114

C. Yao et al. / Fuel 197 (2017) 111–120

Fig. 4. Engine control system structure.

Table 3 Specifications of gasoline, diesel and the test fuel. Parameter

Gasoline

Diesel

G70D30

RON CN Sulphur (m/m) Aromatics (V/V) 10% evaporation temperature (°C) 50% evaporation temperature (°C) 90% evaporation temperature (°C) Density@20 °C (kg/m3) LHV (lower heat value) (MJ/kg)

93 – 0.001% 35.5% 59.7 107.3 160.2 755.4 43.5

– 52.6 0.001% 29.6% 214.8 266.1 333.6 839.3 42.9

(73.2) (25.4) – – – – – 780.5 43.3

IMEPm ¼ IMEP indicated work þ IMEPphysical heat

The baseline fuels were commercial RON93 gasoline and 0# diesel sold on the Beijing market. The specifications of these baseline fuels and the test fuel G70D30 are listed in Table 3. The numbers in the brackets mean that they are estimated values according to Eq. (1) [45].

CN ¼ 68:54  0:59  RON

The main injection timing was scanned from 25°ATDC to 10°ATDC. EGR rates of 10%, 15%, 20%, 25%, 30% and 35% were scanned. The load conditions of operating points were presented as values of measured Indicated Mean Effective Pressure, which was calculated from the measured in-cylinder pressure and marked as IMEPm. In this study, glow plugs added extra heat into the cylinders with glow plug assist, so IMEPm consisted both indicated work and physical heat from glow plug as shown in Eq. (2). The effects on IMEPm of glow plug assist are discussed in following sections.

ð1Þ

ð2Þ

The engine was operated at different loads from 0.2 MPa to 0.6 MPa IMEPm, which is a low to medium load range. Three engine speeds 900 rpm, 1200 rpm, 1500 rpm were tested. All the experiments were taken after the engine was warmed up. The coolant temperature, intake temperature and fuel temperature was kept at 353 K, 313 K and 303 K, respectively. 4. Results and discussion 4.1. Characteristics of GA-CI at a constant speed

3.3. Methods This study mainly focuses on the influence of glow-plug assist, so comparisons between engine performance with or without glow plug assist were performed. Except for glow plug assist or not, the other control parameters were fixed at each operating point in order to isolate the effect of glow plug assist. The operating points presented in this paper were chosen according to sweeps of control parameters. A double injection strategy with a 2 mg/stroke pilot injection was used. In order to enhance engine efficiency and to keep the emissions as low as possible, sweeps of the main injection timing, pilot injection timing, and EGR rate were performed. The pilot injection timing is between 10° and 20° before the main injection.

In this section, experiments were conducted under stable conditions at 1500 rpm. The operating points and operating strategies are listed in Table 4. As shown in Table 4, the engine load of each test point increases with glow plug assist. This is for two reasons: better combustion processes due to glow plug assist, which is the main reason, and the physical heating effect of glow plugs. And as load increases, a higher EGR rate is needed to keep NOx emissions under a sufficiently low level. 4.1.1. Operating characteristics Fig. 5 shows comparisons of in-cylinder pressure and heat release rate with or without glow plug assist at the load of 0.41 MPa IMEPm. The pressure and heat release rate curves in (a)

115

C. Yao et al. / Fuel 197 (2017) 111–120 Table 4 Operating points and operating strategies at 1500 rpm. Engine loads IMEPm without glow plug assist (MPa) IMEPm with glow plug assist (MPa)

0.25 0.32

0.33 0.37

0.41 0.47

0.52 0.54

0.60 0.60

Operating strategies Pilot injection timing (°ATDC) Main injection timing (°ATDC) EGR (%) Intake air pressure (kPa)

33 20 10 103.00

32 20 15 103.00

33 24 25 103.00

33 21 30 103.22

33 21 30 107.28

(c) Pmax Variation - Heating 7 6.5 6

0

Heating Pmax (MPa)

In-Cylinder Pressure (MPa)

No Heating 6

4

-20

-10

0

10 20 Crank Angle (°)

20 40 60 80 100 Cycle Number

(d) Pmax Variation - No Heating 5.5 5 4.5

2

0 -30

Heat Release Rate (J/°)

Pmax (MPa)

(a)

8

0

20 40 60 80 100 Cycle Number

30

40

50

60

30

40

50

60

(b)

50 30 10 -10 -30

-20

-10

0

10 20 Crank Angle (°)

Fig. 5. In-cylinder pressure and heat release rate with/without glow plug assist.

52 50 Indicated Efficiency [%]

and (b) are the average value computed from 100 consecutive cycles. From Fig. 5(a), the two pressure curves are almost coincident before SOC. After SOC, the maximum in-cylinder pressure (Pmax) with glow plug assist is much higher than without assist. Fig. 5(c) and (d) compare the real Pmax of all 100 cycles with and without glow plug assist. The variation of Pmax is obviously reduced with glow plug assist. From Fig. 5(b) the influence of glow plugs on the combustion process is clearer. The SOC is elevated, whereas the heat release process is faster, and the peak value is higher. The four sub-figures of Fig. 5 indicate that the glow plugs help stabilize by shifting the combustion phase earlier. From Figs. 6–13, the operating characteristics, emissions characteristics, and combustion characteristics of different operating points are analyzed. In the following figures, circle dots indicate engine performance without glow plug assist, while square dots show the impact of glow plug assist. It needs to be stressed that the trends along the lines cannot be compared because the operating strategies were chosen from sweeps with different control parameters applied at various loads. Therefore, the main focus of these figures is the difference between the values of circle dots and square dots and the trends of those difference values.

48 46 44 42 40 38

w/o Glow plug assist w/ Glow plug assist

36 34 0.2

0.3

0.4 0.5 0.6 IMEPm [MPa]

0.7

Fig. 6. Indicated efficiency with/without glow plug assist.

116

C. Yao et al. / Fuel 197 (2017) 111–120

0.4

52

w/o Glow plug assist w/ Glow plug assist

48

MPRR[MPa/°CA]

Revised Indicated Efficiency [%]

50

46 44 42 40

0.2

0.1

38

w/o Glow plug assist w/ Glow plug assist

36 34 0.2

0.3

0.4 0.5 0.6 IMEPm [MPa]

0.7

Fig. 7. Revised indicated efficiency with/without glow plug assist.

30 25

Crank Angle [°CA]

0.3

20

CA50 w/o Glow plug assist CA50 w/ Glow plug assist CA10 w/o Glow plug assist CA10 w/ Glow plug assist

5 0

0.4 0.5 0.6 IMEPm [MPa]

0.4 0.5 0.6 IMEPm [MPa]

0.7

Fig. 10. MPRR with/without glow plug assist.

Indicated efficiency ð%Þ ¼

10

0.3

0.3

In Fig. 6, the trend of indicated efficiency versus IMEPm is described. The calculation of indicated efficiency is based on Eq. (3).

15

-5 0.2

0 0.2

0.7

IMEPm  V d  engine speed=2 _f LHV f  m

ð3Þ

In Eq. (3), Vd is displacement volume, LHVf is the lower heating value of test fuel G70D30 and m_ f is fuel mass flow rate. From Fig. 6, the indicated efficiency with glow plug assist is higher than without. As load increases, GA-CI shows a high economy potential. The highest indicated efficiency in Fig. 6 is above 51%. However, the physical heat from glow plugs is considered as indicated work in Eq. (3). Since glow plugs consume electric power, the improvement shown in Fig. 6 cannot directly indicate that glow plugs have improved the combustion process. In order to further analyze this issue, revised indicated efficiency is proposed and shown in Eq. (4). The electric power used to drive glow plugs (Pglow-plug) are subtracting from indicated power calculated from IMEPm and Pglow-plug is controlled using GPCU and fixed as 80 W in this study.

Fig. 8. Combustion phases with/without glow plug assist.

Rev ised indicated efficiency ð%Þ ¼

IMEPm  V d  engine speed=2  P glow plug _f LHV f  m ð4Þ

15

COV of IMEP [%]

w/o Glow plug assist w/ Glow plug assist

10

5 3

0 0.2

0.3

0.4 0.5 0.6 IMEPm [MPa]

0.7

Fig. 9. COV of IMEP with/without glow plug assist.

Revised indicated efficiency is plotted in Fig. 7. The first four test points appear higher indicated efficiency with glow plug assist, which means glow plug assist is not merely providing extra heat physically. The last point shows a different feature and this is because glow plug assist increased heat transfer and exhaust loss at this point. As shown in Fig. 7, glow plug assist improves economy performance of most test points. This improvement is attributed to two parts: adjustment of combustion phases and better combustion process. Fig. 8 illustrates that glow plug assist advances combustion phases at each operating load condition and improvement of combustion process is presented in Section 4.1.3. From Fig. 8, CA50 and CA10 both shift approximately 5°CA, which means that glow plug assist is an effective method to adjust the combustion timing. Since the CA10 is moved up, the ignition delay, which can be defined as the gap between EOI and CA10, is shortened by using glow plug assist. This phenomenon is interesting because, normally, assist methods for PPCI are used to prolong the ignition delay in order to promote the premixing process. However, the timing of ignition delay with glow plug assist is

117

C. Yao et al. / Fuel 197 (2017) 111–120

0.4

10

ISPM [g/kW·h]

8

ISNOx [g/kW·h]

w/o Glow plug assist w/ Glow plug assist

w/o Glow plug assist w/ Glow plug assist

6 4

NOx Threshold 2

0.3

0.2

0.1

PM Threshold 0 0.2

0.3

0.4

0.5

0.6

0.02 0 0.2

0.7

0.3

IMEPm [MPa]

0.4

(a)

0.7

100 w/o Glow plug assist w/ Glow plug assist

15

10

5

w/o Glow plug assist w/ Glow plug assist

80

ISCO [g/kW·h]

ISHC [g/kW·h]

0.6

(b)

20

0 0.2

0.5

IMEPm [MPa]

60 40 20

0.3

0.4

0.5

0.6

0.7

0 0.2

0.3

IMEPm [MPa]

0.4

0.5

0.6

0.7

IMEPm [MPa]

(c)

(d)

Fig.11. Emission characteristics with/without glow plug assist’.

100

Energy Ratio ( Pcombustion /Pglow plug )

Combustion Efficiency [%]

98

30

w/o Glow plug assist w/ Glow plug assist

96 94 92 90 88 86 0.2

0.3

0.4

0.5

0.6

0.7

IMEPm [MPa]

25 20 15 10 5 1 0 0.2

0.3

0.4

0.5

0.6

0.7

IMEPm [MPa] Fig. 13. Energy ratio at 1500 rpm.

Fig. 12. Combustion efficiency with/without glow plug assist.

approximately 20–25°CA, which is sufficiently long for premixing. Plus, with glow plug assist, more stable combustion is realized, meaning that better combustion effects are achieved, even though the ignition delay is shortened, as can be seen from the next figures. Fig. 9 illustrates the Coefficient of Variation (COV) of IMEPm with and without glow plug assist. As the load becomes lower,

the possibility of misfire rises; therefore, the COV of IMEPm increases at lower load conditions. With glow plug assist, the values for COV of IMEPm at all operating points decrease significantly. At 0.32 MPa IMEPm, the COV value reduces to 6%, and at other loads, the COV values are approximately 3%. J. Heywood defines 10% as the comfort limit for combustion variation [46], and researchers usually consider 3% or 5% as the standard for stable combustion to avoid misfire [47–50]. Therefore, the results of COV of IMEPm confirm that glow plug assist helps stabilize the

118

C. Yao et al. / Fuel 197 (2017) 111–120

combustion process, and it can reduce the variation to a relatively low range in addition to the results shown in Fig. 5(c) and (d). The reason for the effect is that glow plugs improve the in-cylinder conditions by increasing the air temperature and forming a local hot region that helps ignite the premix charge. The heating characteristics of glow plug assist are stable [43], and the hot region created by glow plug assist is relatively stable as well. Previously, the ignition of PPCI was relatively unpredictable because it was determined by charge conditions, mixing processes and compression processes. However, with a hot region assisting ignition, the premix charge tends to achieve an earlier but more stable combustion in each cycle. The Maximum Pressure Rise Rate (MPRR) that represents the noise of combustion is plotted in Fig. 10. Glow plug assist enhances MPRR, but the enhancement is relatively low. The maximum value of MPRR with glow plug assist is at 0.60 MPa IMEPm, which is 0.21 MPa/°CA and it is still considered a low value in PPCI research [18,48,49]. 4.1.2. Emission characteristics Fig. 11(a)–(d) describe the NOx, PM, HC, and CO emissions trends versus the engine load, respectively. All the emission results are calculated based on IMEPm. In this study, sweeps of control parameters were taken in order to enhance engine efficiency and to reduce the NOx and PM emissions at the same time. The emissions criteria of NOx and PM are set as references and drawn in Fig. 11(a) and (b) with a black solid line. The threshold values are 2.0 g/kW h and 0.02 g/kW h, which are the limiting values of the Stage V National Emissions Standard of China for diesel engines. Fig. 9 shows that glow plug assist can significantly decrease the PM emissions to the threshold value and slightly enhance the NOx emissions while keeping them within the NOx threshold. This decrease in PM emissions is observed because glow plug assist improves the in-cylinder thermal environment and results in higher in-cylinder pressure and temperature leading to a better mixing process before ignition. Plus, though glow plug assist raises the NOx emissions, the EGR rates are still moderate so there is still space for reducing NOx emissions. Therefore, the glow plug assist helps reduce NOx and PM emissions at the same time. Fig. 11(c) and (d) shows the HC and CO emissions as a function of the engine load. HC and CO emissions are serious problems for PPCI, especially with high octane fuel [18]. When the glow plug works, HC emissions drop rapidly and the CO emissions decrease as well. The significant decreases are mostly attributed to improved oxidation processes because of increased in-cylinder pressure and temperature. 4.1.3. Combustion characteristics As mentioned above, the better economy performance with glow plug assist is achieved by adjustment of combustion phases and improvement of combustion process. In order to better understand the combustion process with glow plug assist, combustion efficiency is computed by Eq. (5) [46].

P

Combustion efficiency ð%Þ ¼ 1 

i xi LHV i  100 _ f =ðm _ aþm _ f Þ  LHV f  ½m

ð5Þ

In Eq. (5), xi are the mass fractions of CO and HC, and the LHVi are the lower heating values of the species. The m_ f and m_ a denote mean mass flow rate of fuel and fresh air. As shown in Fig. 12, GA-CI enhances combustion efficiency at every operating point. The combustion efficiency values are all above 92%, and the highest value is approximately 98%. Higher in-cylinder temperature and improved oxidation processes are the main reasons for the test results. Therefore, glow plug assist essentially changes the incylinder thermal environment and improves both premixing and combustion processes.

Although GA-CI improves indicated efficiency through enhancing combustion efficiency and advancing combustion phases, glow plugs also consume electric power and heat the charge. And the heat produced by glow plugs has been calculated as part of IMEPm. So it is necessary to compare electric energy consumption and recovered energy from the improved combustion efficiency. In order to present the comparison, a variable named Energy Ratio is defined as recovered energy to electric energy consumption, which is calculated by Eq. (4).

P Energy Ratio ¼

_ aþm _ fÞ  ðm PGlowplug

i Dxi LHV i

ð6Þ

In Eq. (6), 4xi are the reduced mass fractions of CO and HC via glow plug assist. If the Energy Ratio is greater than one, it means recovered energy is larger than electric energy consumption. The Energy Ratios of each test point are shown in Fig. 13. From Fig. 13, values of Energy Ratio are all above one, even greater than ten, which means recovered energy from improved combustion efficiency is much more than the electric energy consumed by the glow plug. It indicates that the glow plug assist improves the combustion process effectively in addition to providing extra heat physically. 4.2. Characteristics of GA-CI at different speeds In this section, experiments performed in Section 4.1 are also conducted at 900 rpm and 1200 rpm. Since most characteristics at these speeds appear as similar trends to the results at 1500 rpm, the details are not repeated in this paper. In order to show the effects of GA-CI over the operating range directly, two of the most important impacts of GA-CI on engine performance are illustrated as follows. First, the combustion stability is shown in Fig. 14. The color of the MAPs indicates the COV of IMEPm. The cooler the color is, the more stable the combustion process is. Each dot of the two MAP dots is using the same control parameters, except for applying glow plug assist or not. Comparing the two MAP dots, it can be observed that glow plug assist helps reduce the COV of IMEPm at most points. In the left MAP, the COVs of IMEPm at loads below 0.25 MPa IMEPm are above 10%. With GA-CI the COV value of most points at low loads is 7–10%. In fact, the COV values of both MAPs could be further reduced by sacrificing emissions or efficiency. However, this comparison is enough to explain the effect of glow plug assist on combustion stability. Therefore, GA-CI is an effective way to obtain stable combustion and to avoid misfire. Second, the impact on combustion efficiency of GA-CI is plotted in Fig. 15. The warmer color of the regions means higher combustion efficiency. With glow plug assist, the combustion process of most points is improved and the combustion efficiency across the medium load region exceed 98%. This result means that less unburned materials are generated, and both emissions and efficiency progress with GA-CI. As a part of multi-mode combustion, GA-CI can effectively reduce PM emissions at low to medium loads. The combustion efficiency and stability are increased. Plus, the pressure sensor function and the glow plug function are integrated together with the closed-loop combustion control system, which includes a dualcore ECU and GPCU. The dual core ECU has been introduced in previous research [42]. It consists of one normal engine control unit and a combustion analysis tool. GPCU can communicate with the ECU via the CAN network. As a result of the work performed in this paper, the closed-loop combustion control system can detect online misfire with the pressure sensor function of PSGs and implement fast feedback control using the glow plug function. Moreover, glow plug assist could be widely used as a valid means

119

C. Yao et al. / Fuel 197 (2017) 111–120

COV [%] w/o Glow plug assist

COV [%] w/ Glow plug assist 20

4

0.55

0.35

8

10

0.25

12

1000

1200

14

4

12 0.4 7 0.35

8

5

10

6

8

0.3 7

0.25

14 16

0.2

1400

1000

Speed [rpm]

6 4

9

0.2

0.45

16

3

10

0.3

IMEPm [MPa]

IMEPm [MPa]

6

0.5 6

0.4

5

8

0.45

4

6

0.5

18

0.55

1200

8

1400

2

Speed [rpm]

Fig. 14. MAP of combustion stability.

Combustion Efficiency [%] w/o Glow plug assist

Combustion Efficiency [%] w/ Glow plug assist 98

0.55

0.55

94

0.4 94

0.2

96 96

88

94 90

86

88

92

0.25

88 8 84 6

90

92 90

0.35

92

0.25

98

94

0.4

0.3 92

0.3

0.45

94

IMEPm [MPa]

92

96

0.45

0.35

96

0.5

90

IMEPm [MPa]

0.5

98

1000 1200 1400 Speed [rpm]

0.2

1000 1200 1400 Speed [rpm]

84

Fig. 15. MAP of combustion efficiency.

for combustion optimization and emission reduction in future engine research. 5. Conclusions This study focuses on GA-CI as part of a multi-mode combustion concept for compression ignition engines. The glow plug is integrated with a resistance pressure sensor named PSG, and a control unit for GA-CI is developed. Plus, the characteristics of GA-CI are investigated. With glow plug assist, one dieseline fuel (G70D30) is tested at low to medium load range. The conclusions drawn from this study are as follows: 1. PSG is a mass-produced pressure sensor glow plug. A GPCU is developed for control of the PSG, and it can achieve fast and precise control for glow plugs. The control algorithm is a closed-loop control method based on output power. A surface temperature closed-loop control algorithm is designed and will be verified in future work.

2. With the same operating strategy, peak in-cylinder pressure and the heat release rate of GA-CI are higher. With glow plug assist the load increases slightly and the combustion phases move forward because higher in-cylinder temperature and fuel reactivity boost the pre-mixture to auto-ignition. The cycle-tocycle variations are effectively reduced and the combustion stability is enhanced. For most operating points presented, the values of COV of IMEPm have been decreased to below 5% under GA-CI. 3. With glow plug assist, NOx emissions rise but are still under the threshold value, and 0.2 g/W h. PM emissions decrease dramatically to below 0.02 g/kW h at most operating points. CO emissions and HC emissions drop simultaneously due to the higher combustion temperature and the more complete oxidation process. 4. Under GA-CI, combustion efficiency is improved. As the engine load increases, combustion efficiency with glow plug assist appears higher than 98%. A variable named Energy Ratio is proposed as recovered energy to electric energy consumption

120

C. Yao et al. / Fuel 197 (2017) 111–120

and values of the ratio are all above ten, which means the glow plug assist improves the combustion process effectively in addition to providing extra heat physically. Acknowledgments This work is supported by the State Key Research Program of China under contract No. 2016YFB0101400, the research fund of State Key Laboratory of Automotive Safety and Energy under contract No. ZZ2016-072 and International Science & Technology Cooperation Program of China under contract No. 2016YFE0102200. References [1] Dec JE. Advanced compression-ignition engines-understanding the in cylinder processes. Proc Combust Inst 2009;32(2):2727–42. [2] Cairns A, Blaxill H. The effects of combined internal and external exhaust gas recirculation on gasoline controlled auto-ignition. SAE technical paper; 2005. [3] Szybist J, Nafziger E, Weall A. Load expansion of stoichiometric HCCI using spark assist and hydraulic valve actuation. SAE Int J Engines 2010;3(2010-012172):244–58. [4] Olesky LKM, Middleton RJ, Lavoie GA, et al. On the sensitivity of low temperature combustion to spark assist near flame limit conditions. Fuel 2015;158:11–22. [5] Manofsky L, Vavra J, Assanis DN, et al. Bridging the gap between HCCI and SI: spark-assisted compression ignition. SAE technical paper; 2011. [6] Manente V, Johansson B, Tunestal P, Sonder M, Serra S. Gasoline partially premixed combustion: high efficiency, low NOx and low soot by using an advanced combustion strategy and a compression ignition engine. Int J Vehic Des 2012;59(2/3):108. [7] Nüesch S, Hellström E, Jiang L, et al. Mode switches among SI, SACI, and HCCI combustion and their influence on drive cycle fuel economy. In: 2014 American control conference; 2014. p. 849–54. [8] Guardiola C, Triantopoulos V, Bares P, et al. Simultaneous estimation of intake and residual mass using in-cylinder pressure in an engine with negative valve overlap. IFAC-Pap OnLine 2016;49(11):461–8. [9] Kalghatgi GT, Risberg P, Ångström HE. Advantages of fuels with high resistance to auto-ignition in late-injection, low-temperature, compression ignition combustion. SAE technical paper; 2006. [10] Kalghatgi GT, Risberg P, Angstrom HE. Partially pre-mixed auto-ignition of gasoline to attain low smoke and low NOx at high load in a compression ignition engine and comparison with a diesel fuel. SAE technical paper; 2007. [11] Hanson R, Splitter D, Reitz RD. Operating a heavy-duty direct-injection compression-ignition engine with gasoline for low emissions. SAE technical paper; 2009. [12] Ra Y, Loeper P, Reitz RD, et al. Study of high speed gasoline direct injection compression ignition (GDICI) engine operation in the LTC regime. SAE technical paper; 2011. [13] Sellnau MC, Sinnamon J, Hoyer K, et al. Part-load operation of gasoline directinjection compression ignition (GDCI) engine. SAE technical paper; 2013. [14] Sellnau M, Foster M, Hoyer K, et al. Development of a gasoline direct injection compression ignition (GDCI) engine. SAE Int J Engines 2014;7(2014-011300):835–51. [15] Sellnau M, Moore W, Sinnamon J, et al. GDCI multi-cylinder engine for high fuel efficiency and low emissions. SAE Int J Engines 2015;8(2015-01-0834): 775–90. [16] Ciatti S, Subramanian SN. An experimental investigation of low-octane gasoline in diesel engines. J Eng Gas Turb Power 2011;133(9):092802. [17] Wang B, Wang Z, Shuai S, et al. Combustion and emission characteristics of multiple premixed compression ignition (MPCI) fuelled with naphtha and gasoline in wide load range. Energy Convers Manage 2014;88:79–87. [18] Kim K, Wang Z, Wang B, et al. Load expansion of naphtha multiple premixed compression ignition (MPCI) and comparison with partially premixed compression ignition (PPCI) and conventional diesel combustion (CDC). Fuel 2014;136:1–9. [19] Yao Changsheng, Yang Fuyuan, Wang Jinli, Ouyang Minggao, Huang Haiyan. Injection strategy study of compression ignition engine fueled with naphtha. SAE technical paper; 2015. [20] Chen P, Ibrahim U, Wang J. Experimental investigation of diesel and biodiesel post injections during active diesel particulate filter regenerations. Fuel 2014;130:286–95. [21] Manente V, Johansson B, Tunestal P. Partially premixed combustion at high load using gasoline and ethanol, a comparison with diesel. SAE technical paper; 2009.

[22] Manente V, Tunestal P, Johansson B, et al. Effects of ethanol and different type of gasoline fuels on partially premixed combustion from low to high load. SAE technical paper; 2010. [23] Splitter DA, Reitz RD. Fuel reactivity effects on the efficiency and operational window of dual-fuel compression ignition engines. Fuel 2014;118:163–75. [24] Liu H, Wang Z, Wang J, et al. Performance, combustion and emission characteristics of a diesel engine fueled with polyoxymethylene dimethyl ethers (PODE 3–4)/diesel blends. Energy 2015;88:793–800. [25] Wang Z, Liu H, Ma X, et al. Homogeneous charge compression ignition (HCCI) combustion of polyoxymethylene dimethyl ethers (PODE). Fuel 2016;183: 206–13. [26] Tong L, Wang H, Zheng Z, et al. Experimental study of RCCI combustion and load extension in a compression ignition engine fueled with gasoline and PODE. Fuel 2016;181:878–86. [27] Zhang F, Xu H, Zhang J, et al. Investigation into light duty dieseline fuelled partially-premixed compression ignition engine. SAE Int J Engines 2011;4 (2011-01-1411):2124–34. [28] Rezaei SZ, Zhang F, Xu H, et al. Investigation of two-stage split-injection strategies for a dieseline fuelled PPCI engine. Fuel 2013;107(9):299–308. [29] Weall A, Collings N. Investigation into partially premixed combustion in a light-duty multi-cylinder diesel engine fuelled gasoline and diesel with a mixture of gasoline and diesel. SAE technical papers; 2007. [30] Han D, Ickes AM, Bohac SV, et al. Premixed low-temperature combustion of blends of diesel and gasoline in a high speed compression ignition engine. Proc Combust Inst 2011;33(2):3039–46. [31] Han D, Ickes AM, Assanis DN, et al. Attainment and load extension of highefficiency premixed low-temperature combustion with dieseline in a compression ignition engine. Energy Fuels 2010;24(6):3517–25. [32] Yang F, Yao C, Wang J, et al. Load expansion of a dieseline compression ignition engine with multi-mode combustion. Fuel 2016;171:5–17. [33] Shi L, Deng K, Cui Y, et al. Study on knocking combustion in a diesel HCCI engine with fuel injection in negative valve overlap. Fuel 2013;106:478–83. [34] Borgqvist P, Tunestal P, Johansson B. Comparison of negative valve overlap (NVO) and rebreathing valve strategies on a gasoline PPC engine at low load and idle operating conditions. SAE Int J Engines 2013;6(2013-01-0902): 366–78. [35] Weall A, Collings N. Gasoline fuelled partially premixed compression ignition in a light duty multi cylinder engine: a study of low load and low speed operation. SAE technical paper; 2009. [36] Hasegawa M, Nishizawa T, Imaoka Y, et al. Improvement of combustion stability under cold ambient condition by mixture control. SAE Int J Engines 2013;6(2013-01-1303):1021–34. [37] Mueller C, Musculus M. Glow plug assisted ignition and combustion of methanol in an optical DI diesel engine. SAE technical paper; 2001. [38] Ambekar A, Bhangale R, Chatterjee R, et al. Glow-plug-assisted combustion of nitromethane sprays in a constant volume chamber. Appl Therm Eng 2015;76:462–74. [39] Manente V, Tunestal P, Johansson B. A study of a glow plug ignition engine by chemiluminescence images. SAE technical paper; 2007. [40] Yang F, Gao G, Ouyang M, et al. Research on a diesel HCCI engine assisted by an ISG motor. Appl Energy 2013;101:718–29. [41] Fang C, Yang F, Ouyang M, et al. Combustion mode switching control in a HCCI diesel engine. Appl Energy 2013;110:190–200. [42] Wang J, Yang F, Ouyang M. Dieseline fueled flexible fuel compression ignition engine control based on in-cylinder pressure sensor. Appl Energy 2015;159: 87–96. [43] Zhou T, Yao C, Hu Y, et al. Research on performance and temperature control of glow plugs for PPCI low load assist. IFAC-Pap OnLine 2016;49(11):223–30. [44] Pressure Sensor Glow Plug (PSG) for DIESEL engines. [EB/OL]. ; 2014. [45] Morris W. Method relates diesel cetane, octane ratings. Oil Gas J 2007;105 (45):58–60. [46] Heywood JB. Internal combustion engine fundamentals. New York: McgrawHill; 1988. [47] Lee K, Cho S, Kim N, et al. A study on combustion control and operating range expansion of gasoline HCCI. Energy 2015;91:1038–48. [48] Yang B, Yao M, Cheng WK, et al. Experimental and numerical study on different dual-fuel combustion modes fuelled with gasoline and diesel. Appl Energy 2014;113:722–33. [49] Chen G, Shen Y, Zhang Q, et al. Experimental study on combustion and emission characteristics of a diesel engine fueled with 2,5-dimethylfuran– diesel, n-butanol–diesel and gasoline–diesel blends. Energy 2013;54:333–42. [50] Liu H, Wang Z, Wang J, et al. Effects of gasoline research octane number on premixed low-temperature combustion of wide distillation fuel by gasoline/ diesel blend. Fuel 2014;134:381–8.