G AS DIRECT INJECTION COMPARED TO DIESEL COMBUSTION

pressures of up to 50 MPa and is described in Section 3.12. In the experiments, pure methane, the main component of natural gas, is used.

First, the gas injection combustion is compared to the diesel combustion. In this experiment, the gas pressure is set to 30 MPa and an injector nozzle with a diameter of 1.2 mm applied. Air conditions in the cylinder at the compression end are 10 MPa and 820 K, simulating a current GI engine. For the diesel pilot injection, 60 MPa injection pressure and a nozzle diameter of 0.16 mm are used. As in current gas injection engines, the pilot oil is injected before the gas injection.

The injected mass of pilot oil is equivalent to 3 – 4 % of the total heat release. The experimental conditions are summarised in Table 29, and the recorded injection timings from the gap sensor signals are shown in Figure 106.

Table 29: Experimental conditions to compare gas direct injection with diesel combustion

Experiment GI with diesel pilot injection Diesel

Fuel Pilot fuel (diesel) Gas (CH4) Diesel

Charge air conditions Compression pressure at injection

start 10 MPa

Compression temperature at

injection start 820 K

Oxygen concentration 21.0 %

Injection conditions

Nozzle hole diameter 0.16 mm 1.2 mm 0.5 mm

Injection pressure 60 MPa 30 MPa 100 MPa

Injection start setting [deg. ATDC] -7.2 -5.5 -5.0

Injection end setting [deg. ATDC] 11 23 23

Figure 106: Gap sensor signals from injections

A comparison of the rate of heat release and integrated heat release from the two fuels is shown in Figure 107. The rate of heat release of the gas injection rises more steeply just after the ignition, maybe because of the pilot fuel combustion. While the maximum rate is a little lower in the gas injection, the two fuels burn showing similar rates of heat release.

Figure 107: Rate of heat release and integrated heat release of high-pressure gas injection compared to diesel

This result of comparing the rates of heat release (Figure 107) looks similar to the data from MAN Diesel & Turbo presented in [112] and shown in Figure 108.

-0.3 0.35 1

-20 -10 0 10 20 30

Gap sensor [-]

deg. CA

Pilot fuel

Gas injection (GI) Diesel

-5 0 5 10 15 20 25

-0.3 0 0.3 0.6 0.9 1.2 1.5

-20 0 20 40 60

kJ/deg. kJ

deg. CA

Gas injection (GI) Diesel

Figure 108: Rate of heat release of gas injection compared to diesel, quoting from MAN Diesel & Turbo Turbo [112]

Examples of emission data are shown in Table 30: gas injection combustion emits much lower carbon monoxide but higher unburned hydrocarbons emissions than diesel. However, unburned methane from this experiment is estimated to be 0.5 – 0.6 % of the total injected gas. This value is obtained from comparison experiments in a non-reactive environment, where approximately 14,000 ppm unburned hydrocarbons are measured. This is much less than with lean burn gas combustion, where 2 – 4 % methane slip can be observed. The measured unburned hydrocarbons emissions are in accordance with the MAN Diesel & Turbo test results in [117], where 0.19 g/kWh unburned hydrocarbons with heavy fuel oil and 0.39 g/kWh unburned hydrocarbons with liquefied natural gas operation are reported. NO_x emission from gas injection combustion is about three quarters that of diesel. This could be due to the lower adiabatic flame temperature owing to the lower C/H ratio of methane. However, unlike the homogeneous combustion of lean burn gas engines (Section 5.1), a drastic reduction of NO_x emission, such as -80 %, is impossible.

Table 30: Emissions from the gas injection and diesel experiments

Fuel Gas injection (GI) Diesel

CO [ppm] 8 – 10 44

HC [ppm] 69 – 87 21

NOx [ppm] 346 - 393 531

Direct photos of the gas injection flame set against the diesel flame are displayed in Figure 109;

both series are adjusted to the same total heat release. 100 MPa injection pressure and a nozzle diameter of 0.5 mm are applied for the diesel combustion. Seeing the gas injection combustion process in the left-hand series of Figure 109, a bright pilot flame successfully ignites the gas spray at -4 deg. ATDC. After that, the gas flame outruns the pilot flame and penetrates the right edge of the combustion chamber.

Compared to the diesel flame in the right-hand series of Figure 109, the gas flame looks less bright, as expected from the lower C/H ratio of methane and has a wider flame width. Although the gas injection pressure of 30 MPa (the pressure difference between the nozzle and the compressed air in the combustion chamber is only 20 MPa) is much lower than the diesel injection pressure of 100 MPa, the penetration of the gas flame is equivalent to the diesel flame, for example at 0 deg. and 6 deg. ATDC.

A special feature of the gas injection case is that the visible flame looks unable to come up to the nozzle side from the ignition point during and after the injection period. This may directly relate to the fuel: the diesel spray typically shows break up, atomisation and evaporation, as explained in Section 4.1. These processes support mixing with air. On the other hand, methane is already gaseous; hence, these processes are missing, and mixing with air requires more time.

The behaviour of the after-burning flame at 22 deg. ATDC also appears to be slightly different for the two fuels. However, such differences become clearer when a cylindrical combustion chamber with air swirl is applied at the next stage; this is explained in Section 5.8.

Figure 109: Direct photography of gas injection with pilot spray (left) and diesel spray combustion (right)

Gas direct injection (GI) Diesel

-4

0

6

16

22

deg. ATDC

To demonstrate the cleanliness of the gas injection combustion, soot formation in the flame is visualised using the back diffused laser technique. By applying a band pass filter where only laser wavelength passes, soot concentration formed in the flame can be visualised as black parts as explained in Section 3.4. For the gas injection combustion, a glow plug is used as the ignition source which eliminates any soot generated from the diesel pilot spray. The results are shown in Figure 110. While the diesel combustion has a thick black core in the flame, the gas injection combustion shows only small black parts near the front of the flame. It should be noted that for the gas injection combustion, pure methane is used. Hence, natural gas may consist of components that form slightly more soot during the combustion process.

Figure 110: Soot generated in gas injection (left) and in diesel combustion (right) detected using the back diffused laser technique

In the following experiments, shadowgraph and back diffused laser images are taken at the same time. This allows for the simultaneous visualisation of hot and burning gases, and soot formation. In order for this to happen two cameras, two lasers at different wavelengths and two half mirrors are required. The optical setting is illustrated in Figure 111.

Gas (GI) ignited by glow plug Diesel

5

7

10

12

deg. ATDC

Figure 111: Optical setting to gather back diffused laser and shadowgraph data simultaneously

Two experiments are carried out: one uses a glow plug as an ignition source and the other is carried out with diesel pilot fuel. The comparison, given in Figure 112, allows the pilot fuel to be tracked and a distinction to be made between the soot formed from the diesel pilot fuel and that formed from the methane. It should be noted that the position of the glow plug is suboptimal and hence, points of ignition vary in the experiments. However, the chosen result clearly reflects the tendency and helps gain an insight into the gas injection combustion process. At -1 deg. ATDC, the pilot fuel has already ignited the gas, and soot from the diesel pilot fuel can be observed. At this time, the glow plug has not ignited the gas. At 1 deg. ATDC, the pilot fuel injection is still in progress and soot seems to be concentrated at the flame front and at the upper part of the gas flame. At 7 deg. ATDC, no more pilot fuel is injected; however, diesel pilot fuel is still burning and the soot formation is visible. With the glow plug, the gas is ignited as well and burns very cleanly. Both experiments look very similar at 13 deg. ATDC: only gas is burning and small amounts of soot have formed near the front of the flame.

Figure 112: Shadowgraph and back diffused laser data from pilot injection compared to ignition by glow plug