**1. Introduction**

A diesel engine is one of the most indispensable power generation systems and is mainly used in industrial, public transportation, power generation, heavy-duty machinery, and agricultural applications due to their higher fuel-conversion efficiency, power output, torque and reliability compared to gasoline engines [1]. Furthermore, the emissions such as carbon monoxide (CO), hydrocarbon (HC) and carbon dioxide (CO2) from a diesel engine is much lesser compared to the gasoline engine emissions. However, diesel engine remains to be an important source of pollution as their usage leads to release of nitrogen oxides (NOx), black smoke, particulate matter (PM) and sulfur oxides (SOx) that are detrimental to both environment and human health [2]. In fact, the emissions

from diesel engines have been classified as carcinogenic by the International Agency for Research on Cancer (IARC). It is based on sufficient evidence that high exposure to diesel emissions can be a risk of lung cancer [3], soot emissions can cause cardiovascular diseases [4], while NOx emissions can cause ground level ozone [5], smog [6] and acid rain [7].

On the other hand, fluctuating petroleum prices, fossil fuel depletion, energy demand escalations and stringent emission regulations have intensified the look-out of the scientific community for alternative renewable fuels in place of the existing fossil fuel. Therefore, non-conventional types of fuel, made from biological resources such as biofuel and biodiesel, have been researched. These studied aimed to tackle the problems that arise due to the comparable properties with that of fossil fuels. However, this alternative fuel not only has high viscosity and boiling point, but also low volatility and calorific values. Biofuel, particularly refined palm oil (RPO) is readily made, safe to be handled and stored and is renewable [8]. However, direct use of this oil degrades the engine performance if operated for a prolonged period due to their high viscosity and low volatility, which causes filters, fuel lines and injectors to be clogged. In addition, engine problems such as piston ring sticking, carbon deposit build-up and lubricating oil thickening were also observed [8], therefore it would be necessary to overhaul, repair and replace some parts of the diesel engine.

The prominent factors that control and govern the combustion process depend on the air motion within the engine, charge temperature, compression ratio, spray structure, burning rate, piston bowl geometry, injection strategies, auto ignition fuels, and fuel molecular structure [9]. There is a significant influence of piston bowl geometry with respect to the combustion and the amount of emission as it strongly affects the mixing of air and fuel prior to the start of injection (SOI). Jing Li et al. [10] investigated the effect of piston bowl geometry on combustion and emissions using a high viscous fuel (biodiesel). The studies inferred that at low engine speed, the shallow depth re-entrance combustion chamber (SCC) piston exhibited better engine performance. Using computational fluid dynamics (CFD) analysis, Hamid et al. [11] discovered that the SCC piston had an ability to generate high swirl, tumble and cross tumble ratios. They also observed that the turbulence kinetic energy was increased, had a well-organized flow and a better air fuel mixture, especially for high viscous fuel applications. Consequently, the combustion efficiency had been improved and reduced incomplete combustion. In addition, the SCC piston is a preferable design since it can be run with high viscosity alternative fuel in the diesel engine.

The inherent long carbon chains become a limitation to biofuels as their nature results in high viscosity and density. The in-cylinder airflow rate was low and produced an adverse effect, as previously mentioned, because of the physiochemical nature of the alternative fuel, that produced an undesired injection profile which degraded cone angle and increased the length of penetration spraying due to high viscosity. Several measures have been taken to lower the viscosity of the alternative fuel for producing an injection profile closed to the diesel, e.g., blending alternative fuel with fossil fuel [12], preheating the alternative fuel [13], emulsification of alternative fuel [14], nano-fluid additives [15], mixing with low and high viscosity biofuel [16] and adjusting injection timing [17–19]. In general, these techniques partially minimized those aforementioned issues, but are still lacking certain requisites compared to the engines using petrol and diesel fuels.

Theoretically, when the in-cylinder airflow rate increased, the swirl, tumble, cross tumble ratios and kinetic energy of turbulence also increases and accelerates the in-cylinder evaporation and diffusion. These effects generally will enhance engine performance such as higher engine power and lower brake specific fuel consumption (BSFC). To enhance and improve the in-cylinder airflow rate and its characteristics, several strategies can be implemented, such as redesigning the airflow intake manifold, modifying the piston bowl and guide vanes to guide the inlet airflow [20].

Therefore, this research will investigate numerically the effect of the numbers of GVD incorporated with the SCC pistons to enhance and organize the in-cylinder airflow rate and its characteristics. Based on the previous literature, the geometry of the guide vanes consists of four main parameters: vane number, angle, height and length. Nevertheless, this research is limited to the numbers of GVD regardless of other parameters as including several parameters tend to make the analysis complex and increases the computational time consumption. Therefore, the remaining parameters were kept constant according to the previous researchers [21,22]. The design model of the SCC piston is modification suited on the YANMAR L70 engine specifications. The details of the designs is described in the following sections.

#### **2. Methodology**

### *2.1. Computer Simulation*

There are four main steps in order to investigate the airflow rate and its characteristics, namely to draw the SCC piston and GVD, mesh the parts, define their boundary condition and analyze the cold flow IC engine. SolidWorks 2017 and ANSYS-FLUENT v15 software were utilized to prepare the model and analyze the in-cylinder airflow in the transient engine cycle without combustion. The complete details of the computer simulation setup is described in the following section.

#### *2.2. Guide Vane Design (GVD)*

GVD is designed specifically to enhance the air velocity due to swirling flow generated. The geometry of GVD consists of vanes number (N), height (Hv), length (*l*) and angle (θ) which are illustrated in Figure 1 and the specification of GVD is shown in Table 1. GVD dimensions play an important parameter in generating optimized in-cylinder air flow characteristics. They will guide the intake airflow into the combustion chamber, generate turbulence phenomena and sustain swirl momentum until the end of the expansion stroke. The increasing swirl flow will produce high convective heat transfer coefficient inside the combustion chamber [23]. Nevertheless, if the number of vanes increases, it tends to obstruct the airflow and affect volumetric efficiency [24]. By considering this, our research has a limit to eight guide vanes (V8) starting with a base (without vane). The vane twist angle was fixed at 35◦ angle.

**Figure 1.** GVD design. (**a**) Front view; (**b**) Side view\*\*.

**Table 1.** Specification of guide vane design (GVD).


#### *2.3. Shallow Depth Re-Entrance Combustion Chamber (SCC)*

To improve the mixing of air and fuel, most of the important modifications were performed on the engine design. The nature of formation of mixture in the engine cylinder is predominantly dependent on the shape of the combustion chamber and the piston bowl design. Running the emulsified biofuel with high viscosity will deteriorate the injection profile. Hwang et al. [25] carried out studies on the injection profile using waste cooking oil biodiesel. They observed that the penetration length was longer, and the cone angle was shorter compared to petrol and diesel fuels. To mitigate these issues, many researchers [26] suggested that the piston bowl design needs to be modified for smooth running with high viscosity fuel. They discovered that the SCC piston bowl design (as shown in Figure 2) is recommended since it can organize the airflow well; swirl ratio (*Rs*), tumble ratio (*RT*), cross tumble ratio (*RCT*) and break the penetration length of injection to facilitate enhanced mixing with the surrounding air. Therefore, the effect of the GVD and SCC piston combination will be focused on this research.

**Figure 2.** Schematic diagram for shallow depth re-entrance combustion chamber (SCC) piston bowl geometry design (all in mm) [11]. Reproduced with permission from Hamid et al., (Renewable Energy); published by (Elsevier), (2018).

#### *2.4. Engine Model*

The geometry of the engine model was adapted from the experimental Yanmar model type L70AE-DTM CI generator with a four-stroke direct injection, vertical cylinder, one cylinder, one intake valve and one exhaust valve. The technical specification of the engine is given in Table 2.


**Table 2.** Technical specifications of Yanmar L70AE CI engine generator.

#### **3. Simulation Setting**

The engine geometry of Yanmar L70AE-DTM was modelled using SolidWorks 2017. The GVD, intake runner, exhaust runner, cylinder, intake valve and the exhaust valve have been modelled separately and assembled together as illustrated in Figure 3. The assembled model of the engine was exported to the CFD software, namely ANSYS-FLUENT v15. The software was used to construct a solver, comprising of mathematical computations that simulates and analyses cold flow. To compute the parameters representing fluid flow, both valves were set as solid domains as in reality while the others were set as a fluid flow domain. The moving boundaries were the main challenges in order to simulate a 3D IC engine such as piston bowl, valve and cylinder. The moving grid and remapping mesh were the common strategies used by many researchers [27,28]. The mesh generation on this research was based on assembly level meshing technique. The CFD simulation setting was based according to the cold flow IC engine published by ANSYS Inc. [29].


**Figure 3.** Schematic diagram of modelling engine Yanmar L70AE-DTM configuration.

The equations governing the fluid flow which formed the basis for simulation are the conservation of mass, momentum and energy (energy equation) [30,31]. The conservation of mass is derived based on the control volume and the corresponding differential Equation [32]. It is written as:

$$\frac{\partial p}{\partial t} + \nabla(\rho \mathcal{U}) = 0 \tag{1}$$

ρ is the fluid density and *U* is three-dimensional flow velocities in the *x*, *y* and *z* directions.

The conservation of linear momentum is derived based on the Newton's second law where the surface forces are the control volumes and forces are the body of the control volume. It can be written as:

$$\frac{\partial(\rho \, \mathrm{d}I)}{\partial t} + \nabla(\rho \, \mathrm{d}I \times \mathrm{d}I) = \nabla p - \nabla \pi + \mathrm{S}\_{\mathrm{M}} \tag{2}$$

*p* is the fluid pressure, τ is the strain rate and *SM* is a momentum source. This equation is also known as the Navier–Stoke equation [33].

The rate of energy change inside the fluid element is also known as the energy equation and it is given by:

$$\frac{\partial(\rho h\_{\text{tot}})}{\partial t} - \frac{\partial \rho}{\partial t} - \nabla(\rho \, \mathcal{U} h\_{\text{tot}}) = \nabla(\lambda \nabla T) + \nabla(\mathcal{U} \cdot \tau) + \mathcal{U} \cdot \mathcal{S}\_M \tag{3}$$

*htot* and λ are the total enthalpy and thermal conductivity, respectively.

Shear Stress Transport (SST) is a two-equation eddy-viscosity model that was used in this numerical study. This model is a combination of the *k-*ω and *k-*ε turbulence models. It is a low Reynolds number model. It resolution has similar requirements to the *k-*ω model and the low Reynolds number *k-*ε turbulence model, but its formulation abolishes some weakness displayed by pure *k-*ω and *k-*ε turbulence models. While the *k-*ω model pertains to the inner boundary layer, the *k-*ε model plays a role in the outer region. The combinational model overcame the limitation of shear stress until 5% turbulence intensity in the adverse gradient region, wherein it is sufficient to consider fully developed flow turbulence. The conditions of temperature and pressure were set at 300 K and 1 atm respectively. The detailed information on the setting used for the models and their limitations can be referred to ANSYS FLUENT v15–Solver Theory Guide [26].

On the basis of the physical boundary conditions of the engine, the simulation was carried out in two distinct phases of analysis; intake analysis and intake port analysis. Intake analysis is related to the intake runner, while intake port analysis corresponds to the clearance volume prior to the downward movement of the piston and intake runner in the *y*-direction and drawing of air into the cylinder. The components that are not applicable in this analysis were suppressed since there was no contribution to the results and to reduce the computation time during calculations.

The results of the intake analysis were transferred to the compression and expansion analysis for further simulations. During the compression and expansion analysis, the intake and exhaust valves remained closed while air was compressed during the upward movement of the piston towards the top dead center (TDC) and expanded during the downward movement towards the bottom dead center (BDC). The only domain applicable during this analysis was the cylinder volume, where the volume changed due to the motion of the piston being progressed up and down via the moving mesh. To retain the stability of the simulation progress, the time step must be small enough to simulate the moving mesh.

#### **4. Results and Discussion**

The in-cylinder airflow rate and its characteristics are well-recognized such that they can enhance the evaporation, diffusion and combustion process. As a result, it can be utilized for emulsified biofuel to improve engine performance and reduce engine emissions. Results and discussion will focus on the events taking place within the fuel injection period or ignition delay, the time difference between start of injection (SOI) and start of combustion (SOC). Due to the default setting of Yanmar L70AE-DTM (manufacturer setting), the fuel injection is at 14◦ before TDC, therefore the results will be covered at a crank angle (CA) from 346◦ SOI until 352◦ SOC.

#### *4.1. Numerical Validations*

The purpose of the experimental setup is to validate the numerical simulation, which was carried out in the test rig of single a cylinder of Yanmar engine as shown in the Figure 4. During the experiment setup, a high sensitivity water-cooled precision type sensor Kistler 7061B, magnetic pickup shaft encoder and TDC position optical sensor were used to measure the in-cylinder pressure and crank angle data. A sensor of type 7061B was screwed directly into the standard M14 hole. The position of the sensor was mounted near the valve for better accuracy of the values. K-type thermocouple was used to measure the boundary temperatures, at the intake and exhaust boundaries, and the thermocouple was positioned as close as possible to the cylinder head. The SCC piston was used to validate the experimental data.

**Figure 4.** The engine Yanmar L70 setup test rig.

Figure 5 shows the in-cylinder pressure against the crank angle (θ) diagram from the simulation and experiment results without combustion at a rotational speed of 2000 rpm. The data measurement for the intake temperature and pressure are 302 K and 1.02 bar, respectively. The graph shows the variation of pressure between 0◦ to 540◦. Based on the figure, a reasonable agreement between experimental and numerical results with a slight difference of about 7% peak pressure is witnessed. This minor deviation might be due to the gas seepage from the cylinder into the crankcase.
