Figure 1 shows a schematic illustration of the configuration of the test device. A single-cylinder CRDI diesel engine equipped with a sac-type injector was used to perform the test; the specifications of the test engine are summarized in Table 2. Various devices were constructed to perform the test. Table 3 lists the main functions of the devices constructed, as shown in Figure 1. The single-cylinder CRDI diesel engine maintained a constant speed via a 22 kW electric motor. The energy generated by the engine was dissipated by a braking resistor connected to an electric motor.

Figure 1: 
Schematic illustration of experimental system

In this study, a single fuel-injection timing strategy without pilot injection was applied. The test was conducted while adjusting the fuel-injection timing from before the top dead center (BTDC) 30 crank angle (CA) to TDC at 5 CA units. The experiment was performed with the speed of the test engine fixed at 1100 rpm, a fuel injection amount of 10 mg, and a fuel-injection pressure of 40 MPa. The in-cylinder pressure was averaged by measuring 300 cycles to minimize cycle-to-cycle fluctuations. Table 4 lists the experimental conditions for reference. The physicochemical properties of the test fuel, compared with those of the diesel fuel used in naval vessels, are listed in Table 5.

Based on the same operating conditions of the engine, adjusting the fuel injection timing can be an effective strategy for improving engine vibration and noise. To confirm this hypothesis, a method for quantifying engine vibration and noise is required. To date, the most effective method for analyzing engine vibration and noise is by using a measurement device [16]-[18]. However, in this study, the RI and CNL were quantified by applying a theoretical methodology that has been verified by researchers, without using a direct measuring device. The RI representing knocking combustion can be expressed as follows [19]:

Where γ represents the ratio of specific heat, dPdtpeak the peak ROPR, P_peak the peak in-cylinder pressure, T_peak the peak in-cylinder temperature, R the ideal gas constant, and β = 0.05 ms a constant value. The RI was simulated by calculating the ROPR and in-cylinder temperature using the measured in-cylinder pressure. The ROHR (Equation (2)) was calculated based on the first law of thermodynamics and Equation (3). Meanwhile, Equation (4) was applied to the in-cylinder temperature [20][21].

In Equation (2), P is the measured in-cylinder pressure, V the cylinder volume to corresponding to the CA of the test engine, and γ the specific heat ratio. In Equations (3) and (4), P, V, and T represent the in-cylinder pressure, volume, and in-cylinder temperature, respectively; m_intake is the intake air mass at intake-valve closing; ρ_air is the air density; V_ivc is the cylinder volume at intake-valve closing; v_l is the function of valve lift; ϴ_vod,_intake is the intake-valve opening duration; ϴ_vod,_exhaust is the exhaust-valve opening duration; ϴ_IVO and ϴ_EVO are the intake-valve opening and exhaust-valve opening angles, respectively. Because the ROPR was applied as a time ratio in the RI equation, the ROPR that had been calculated using the CA ratio was changed to the time ratio to quantify the RI. To quantify the CNL characteristics, the statistical analysis proposed previously and the equation that established the experimental correlation between the variables of the peak ROHR and combustion period were applied in this study. The equation to quantify the CNL can be expressed as follows [16][21]:

Where CD represents the combustion duration between 10% and 90% MFB, and ROHR_peak is the peak rate of the pressure rise.

Figure 2 shows the effects of the fuel-injection timing on the in-cylinder pressure, ROHR, and ROPR.

Figure 2: 
In-cylinder pressure, ROHR, and ROPR characteristics with various injection timings corresponding to crank angle

As the fuel-injection timing advanced, the peak in-cylinder pressure, ROHR, and ROPR values increased. When the fuel injection was advanced and the in-cylinder pressure and temperature were low, the mixture formation improved significantly owing to a longer ignition delay, thereby resulting in a higher premixed combustion intensity, and hence higher peak ROHR and ROPR values. In addition, because the advanced combustion directly affected the compression stroke, it exerted an adverse effect. A higher mixture-formation intensity and ignition delay reduced the overall combustion period, which consequently decreased the indicated mean effective pressure (IMEP).

The effect of the fuel-injection timing on the engine performance was confirmed. Figure 3 shows a comparison of the results of the IMEP, indicated torque, and thermal efficiency based on the fuel-injection timing. The formulas for calculating the IMEP, indicated torque, and thermal efficiency are as follows:

Figure 3: 
IMEP, Indicated torque, and Normalized indicated thermal efficiency characteristics with various injection timings

In Equations (6), (7), and (8), V_d represents the swept volume, mf the injection fuel mass, and Q_LEV the lower heating value of the biodiesel. As shown in Figure 3, the IMEP, indicated torque, and thermal efficiency increased as the fuel-injection timing approached the TDC. The in-cylinder pressure and temperature were high under retarded conditions. This reduced the ignition delay period, resulting in an improved combustion efficiency. As the fuel-injection timing was delayed, the adverse effects diminished, whereas the positive effects became more prominent; hence, the engine performance improved. In addition, retarding the fuel-injection timing while maintaining a constant heat release (m_fQ_LEV) improved the fuel conversion efficiency and combustion efficiency, which consequently improved the thermal efficiency.

To calculate the RI, combustion parameters, such as the peak ROPR, peak in-cylinder pressure, and peak in-cylinder pressure were considered. According to various studies, an acceptable RI value is 5 MW/m² [19][22]. Figure 4 presents the results of RI, peak ROPR, and peak in-cylinder pressure vs. fuel-injection timing. As mentioned above, the advanced fuel-injection timing increased the intensity of the premixed combustion, which consequently increased the peak in-cylinder pressure and ROPR. In addition, the peak in-cylinder temperature increased, as per Equations (3) and (4). RI is an index used to quantitatively evaluate the strength of knocking by combustion and is associated closely with engine noise and vibration. The highest RI was indicated at BTDC 25 CA; in fact, the RI decreased as the fuel-injection timing was retarded toward the TDC. The RI decreased because the retarded fuel injection suppressed the premixed combustion and improved the combustion stability. In particular, it was evaluated to be below the RI limit of BTDC 15 CA.

Figure 4: 
Ringing intensity, peak ROPR, peak in-cylinder pressure, and Peak gas mean temperature vs. injection timing

Figure 5 provides the analysis results of the effects of the fuel-injection timing on the CNL, the peak ROHR, and the overall combustion duration. In this study, the variables for evaluating the CNL based on the fuel-injection timing were the peak ROHR and overall combustion duration that affected the combustion process [16][21]. As the fuel-injection timing advanced, the combustion duration decreased and the peak ROHR increased. Under more advanced conditions, the additional time for mixture formation and ignition delay increased the premixed combustion intensity and reduced the combustion period. A strong premixed combustion and a short combustion period resulted in a high CNL. As the fuel-injection timing was delayed, the CLN decreased gradually as the combustion phase (CA50) was generated in the expansion stroke.

Figure 5: 
Combustion noise level, peak ROHR, and combustion duration characteristics vs. injection timing

Figure 6 shows the effect of fuel-injection timing on CO, NOx, and peak gas mean temperature (GMT).

Figure 6: 
CO, NOx, and peak gas mean temperature vs. injection timing

Our analysis showed that CO, NOx, and the peak GMT increased as the fuel-injection timing advanced from the TDC. At the early injection point, the spray of the injected fuel developed rapidly because of the lower in-cylinder pressure and temperature. Therefore, local piston wall-wetting was generated, and incomplete combustion occurred owing to the weaker oxidation. This was considered a key contributor to the increase in CO. The NOx was considered thermal NOx, which is associated closely with the combustion gas temperature during the combustion reaction stage. Advancing the injection timing increased the intensity of the premixed combustion via the formation of a superior mixture with a longer ignition delay, as well as via an increase in the peak ROHR and combustion gas temperature. Significant NOx formation was observed as the injection timing became more advanced.