In-Cylinder Cold-Flow Analysis - ‘A Comparison of Crank-Slider Engine and Crank-Rocker Engine’

Modern spark-ignition internal combustion engines are being optimized using different geometrical and conditional variations, for better air-fuel mixing and efficient combustion [1-4]. For example, Port or Indirect injection was used instead of carburettor and eventually, direct injection instead of port injection was introduced as the most efficient method so far. Downsizing, Supercharging and Turbocharging are other methods to increase engine efficiency and to reduce the exhaust emissions.

It has been observed and proved that the flame propagation is rapid during the combustion stroke if the air-fuel mixing is efficient during intake and compression strokes [5]. Therefore, the effective air-fuel mixing is considered and promoted during the optimization of IC engines. It is of major importance to study the behaviour of engine fluid flows and to improve engine quality, it is extremely important to understand the flow properties in the cylinder. This is because engine efficiency is significantly related to the in-cylinder flow conditions [5, 6]. The greater the flow motion, greater the turbulence, more relative flow eddies are generated which results in efficient air-fuel mixing, and vice versa.

The combustion chamber geometry is of importance when fluid-flow is discussed. This geometry includes the cylinder liner, the piston head crown, and head roof [7-9]. Turbulence flow fields have a great effect on the efficient combustion process. The total process of thorough mixing during the intake phase gives rise to turbulent flow conditions which is then intensified by the piston motion and the geometry of combustion chamber [10, 11]. Also, there are many fluid flow characteristics that depends upon the geometry of the engine cylinder [12]. For example, the swirl or tumble flows, length scale, vorticity etc. These geometry parameters include intake port design, bore/stroke ratio, and cylinder head shape.

High-intensity turbulence, including great energy eddies, tumble, and swirls, is one of the most significant factors for smoothening the ignition process. In spark-ignition engines, tumble motion is most occurred [13, 14]. Different intake port geometries generate varying turbulent flows which can be very significant for improving the air-fuel mixing properties. These intake properties include the intake manifold geometry, intake port design, intake valve configurations, valve lift, piston geometry, piston position and combustion chamber geometries [2].

A single rotor Wankel engine was developed and tested at wide open throttle [15]. Researchers considered it a promising alternative to reciprocating engines. They concluded that the performance in terms of power, torque and efficiency were improved by increasing the energy fraction of hydrogen in the fuel mixture, which also reduced the HC and NO_X emissions. Hence, this engine can be used for hybrid applications.

New Concept Rocker Engine [16] was introduced which has two opposing pistons attached to rocker linkages that are coupled to a common crankshaft in the center. This was described as an improvement over crank-slider engines in terms of kinematics, compactness, and its ability to modify the piston stroke and compression ratio as per required. The authors claimed that this engine has 8 times lower piston-to-liner force than the conventional engine.

Another Opposed Piston Engine was designed, tested and patented with a slightly different configuration [17]. The researchers observed that the engine produced high thermal efficiency and reliability with lower fuel consumption and maintenance. As an improvement to the previous designs, a novel crank-rocker (C-R) engine, which is a type of toroidal engine, was invented in Universiti Teknologi PETRONAS (UTP), Malaysia with the intent to decrease the frictional losses present in the conventional crank-slider engine [18]. The engine geometric dimensions were benchmarked from the Modenas ACE115 crank-slider engine.

In the present paper, CFD analysis is carried out to investigate the airflow behaviour within the crank-rocker engine at cold-flow conditions and its comparison with the conventional crank-slider engine. Re-Normalization Group (RNG) k-ε turbulence model is used with operating conditions set at standard temperature and pressure i.e., 303 K and 101325 Pa. This study is useful in understanding the flow behaviour and velocity field properties within the C-R engine and how it differs from the conventional crank-slider engines.

1.1 Theoretical background

In order to understand the effect of air induction system and geometric variations on the combustion characteristics, visualization of in-cylinder flows can be practiced [19]. Numerical simulation is a reliable tool to demonstrate the effects of valve lifts and varying intake configurations. A recent increase in computational fluid dynamics (CFD) is observed due to its effectiveness for the advanced design in IC engines [20]. Complete air flow throughout the cycle can be visualized with the help of cold-flow simulations [6]. This includes the capturing of induced air, predicting the development of swirl, tumble, and squish flows, mixing of fuel droplets with induced air. In CFD, mass, momentum and energy are solved using balanced equations for discretised geometry. These set of equations are called Navier-Stokes equations which govern the fluid motion and are subjected to initial and boundary conditions [21].

$\begin{aligned} \frac{\partial \mathbf{v}}{\partial t}+(\mathbf{v} . \nabla \mathbf{v}) &=-\frac{1}{\rho} \nabla p v_{o} \nabla^{2} \mathbf{v} \\ \nabla \cdot \mathbf{v} &=0 \end{aligned}$             (1)

where, v(x,t) is the fluid velocity, p is the pressure, ρ is density and v_o is the molecular kinematic viscosity. The air-fuel mixture upon entering the combustion chamber through the intake valve flows in all possible directions. Large-scale flow motions are energized by boundary conditions which eventually cascade to smaller scales and thus the boundary conditions are of importance. Renormalizing the above-mentioned Navier-Stokes equations derives the Re-Normalization Group (RNG) k-ε variation of Reynold’s Averaged Navier-Stokes (RANS) turbulence model, where k is the transported variable turbulent kinetic energy and ε is the transported variable rate of dissipation of turbulent kinetic energy.

$\frac{\partial}{\partial t}(\rho k)+\frac{\partial}{\partial x_{i}}\left(\rho k u_{i}\right)=\frac{\partial}{\partial x_{j}}\left[\left(\mu+\frac{\mu_{t}}{\sigma_{k}}\right) \frac{\partial k}{\partial x_{j}}\right]+P_{k}-\rho \varepsilon$               (2)

$\frac{\partial}{\partial t}(\rho \epsilon)+\frac{\partial}{\partial x_{i}}\left(\rho \epsilon u_{i}\right)=\frac{\partial}{\partial x_{j}}\left[\left(\mu+\frac{\mu_{t}}{\sigma_{\epsilon}}\right) \frac{\partial \epsilon}{\partial x_{j}}\right]+C_{1 \epsilon} \frac{\epsilon}{k} P_{k}-C_{2 \epsilon}^{*} \rho \frac{\epsilon^{2}}{k}$             (3)

where,

$C_{2 \epsilon}^{*}=C_{2 \epsilon}+\frac{C_{\mu} \eta^{3}\left[1-\frac{\left(2 S_{i j} S_{i j}\right)^{\frac{1}{2}} k}{\epsilon \eta_{o}}\right]}{1+\beta \eta^{3}}$          (4)

This version of RANS is suitable especially for moderately complex behaviours like separating and swirling flows for in-door small scales eddies. The turbulent viscosity and the numerical constants in the RNG equations can be derived from the standard k-ε model. Due to the restrictions of cylinder liner and piston wall, the flow performs rotational motion. The component of flow around the axis of cylinder is called swirl flow and the component having its axis perpendicular to the cylinder axis is called tumble flow. The swirl ratio and tumble ratio are defined mathematically as:

$R_{s}=\frac{\omega_{s}}{2 \pi N} ; R_{t}=\frac{\omega_{t}}{2 \pi N}$         (5)

where, R_s, R_t, ω_s, ω_t and N are the swirl ratio, tumble ratio, angular velocity of flow around swirl axis, angular velocity of flow around tumble axis and engine angular speed, respectively.

The crank-rocker (C-R) engine was successfully designed, fabricated, and tested and its schematic as well as the final assembly are shown in Figure 1. The curved piston does not slap against the cylinder surface as in the case of conventional crank-slider engine, this reduces the frictional losses in the engine [22].

1a.png

(a)

1b.png

(b)

Figure 1. (a) Design of Crank-Rocker Engine and (b) Fabricated and Assembled C-R Engine [18]

Another important factor that can make the crank-rocker engine more powerful than the conventional crank-slider engine is the possibility of using biofuel for combustion. This is because biofuels are future source and possible alternative of the engine fuels. One of the main advantages of having biofuel is the reduction of greenhouse emissions, as many researchers have asserted the importance of biofuels [23]. Biofuel in the form of LPG and natural gas can be used in crank-rocker engine.

The analytical procedure was carried out to evaluate the Indicated Torque (IT) and Power (IP) comparison of crank-rocker and crank-slider engines whereas, the experimental dynamometer analysis was performed on the crank-rocker engine to evaluate the HC and NOx emissions. These results are shown in Figure 2. In the light of these experimental and analytical results, the advantages of crank-rocker engine over the conventional crank-slider engine are summarized and tabulated (Table 1).

2a.png

(a)

2b.png

(b)

Figure 2. (a) Indicated Torque (IT) and (b) Indicated Power (IP) verses engine speed comparison of crank-rocker engine and crank-slider engine [18]

Table 1. Comparison between crank-rocker engine and crank-slider engine [18, 22, 24]

The improvement of engine performance parameters is a result of the curved cylinder and the turbulence intensity inside the combustion chamber because the geometry has significant effect on airflow which results in better mixing, better combustion, and reduced exhaust emissions. To validate this hypothesis, a comparison of various CFD simulations is presented on the crank-rocker and crank-slider engines during the intake stroke. The numerical cold-flow analysis is performed using the renowned commercial CFD software CONVERGE. Combustion and exhaust strokes analyses are beyond the scope of this research as mostly the intake flow structures describe the overall thorough combustion and engine efficiency.

The flow velocity vectors are generated in the postprocessing of solver as shown in Figure 6. The static pressure at the intake port is at atmospheric pressure i.e., 101325 Pa. As the intake valve opens, from 30° CA to 210° CA, the pressure drops to around 100950 Pa, due to which the air-fuel mixture enters the cylinder. The port pressure immediately starts increasing as the intake valve closes. The average intake velocity at the intake port reaches a maximum of 25 ms^-1 at around maximum intake valve position. The plots of intake port pressure and average velocity are shown in Figure 7.

6a.png

(a)

6b.png

(b)

Figure 6. In-cylinder flow vectors visualization of (a) Crank-Rocker engine and (b) Crank-Slider engine during Intake at 500° CA

7a.png

(a)

7b.png

(b)

Figure 7. (a) Intake port static pressure and (b) Average intake flow velocity for both engines at cold-flow conditions

8a.png

(a)

8b.png

(b)

Figure 8. (a) Swirl ratio and (b) Tumble ratio trend within the cylinders of crank-rocker and crank-slider engines at cold-flow conditions

Swirl and tumble ratio profiles of the airflow within the cylinder region are analyzed. It can be clearly seen that during the intake from 30° CA, the swirl ratio increases reaching its maximum and slowly diminishes as piston reaches BDC. Again, the swirl flow increases during the exhaust phase as piston moves upwards, but this fluctuation is less than that during intake. Tumble flow, on the other hand, instantly jumps to maximum point at around 520° CA to 530° CA at the end of expansion at power stroke as seen in Figure 8.

During exhaust stroke, the tumble structure breaks down. When compared to crank-slider engine, the swirl trend of crank-rocker engine is almost the same except during the exhaust phase and just before the spark ignition at around 360° CA, the increase of 50% swirl in C-R engine is observed. This can be attributed to the geometrical profile of the curved cylinder along with the upward piston motion. This results in a higher swirl air movement and a more thorough air-fuel mixing as compared to the crank-slider engine just before spark ignition.

As for the tumble flow, the C-R engine has undoubtedly higher tumble ratio than the crank-slider engine throughout the engine cycle but at 360° CA, an increase of 33% in tumble is observed for crank-rocker engine. It can be attributed to the curved cylinder itself, as it facilitates the inherent shape of the tumble motion to be generated during the intake and compression strokes by the motion of piston. The swirl ratio and tumble ratio comparison for crank-rocker and crank-slider engines are presented in Figure 9.

The internal cylinder pressure is set at the atmospheric conditions i.e., 101325 Pa. During the intake, the cylinder pressure remains unchanged, however it starts increasing during the compression stroke till 390° CA. The pressure then decreases as the piston moves down during the expansion stroke. The maximum pressure within the crank-rocker engine is slightly higher than the crank-slider engine, which is desired. This is because the higher the peak pressure at the end of compression, the greater the temperature of air-fuel mixture. This helps in somewhat vaporization of fuel, which leads to thorough combustion.

The dynamic viscosity increases during intake and reaches a maximum value at the end of intake stroke and a second peak is found at the end of power stroke. After this, the dynamic viscosity deceases again. This is because the dynamic viscosity is an inverse function of fluid temperature. As the piston moves downwards during the intake, the temperature decreases, and thus the dynamic viscosity increases till it reaches the maximum value at around TDC and then the viscosity drops during the compression stroke because of increase in temperature. Same goes for the expansion stroke and we observe the second peak of viscosity which then drops again during the exhaust stroke. The cylinder pressure and dynamic viscosity trends of the crank-rocker engine and crank-slider engine are shown in Figure 9.

Vorticity is higher in case of crank-rocker engine just before the power stroke begins. The spark is supposed to ignite at around 360° CA just before the piston reaches TDC. At this point, the vorticity and TKE of crank-rocker engine is seen significantly higher than in case of crank-slider engine. Almost 100% increase is recorded for C-R engine at 360° CA, just before the spark, in terms of vorticity and turbulent kinetic energy trends as shown in Figure 10.

According to previous researchers [9, 26, 27], the higher the vorticity and TKE in the cylinder, the greater the thorough mixing, efficient combustion, and thus minimum emissions. This analytical result can be validated with the experimental findings of C-R engine for combustion emissions (Figure 11).

9a.png

(a)

9b.png

(b)

Figure 9. (a) Cylinder pressure and (b) Dynamic viscosity trend of crank-rocker and crank-slider engine at cold-flow conditions

10a.png

(a)

10b.png

(b)

Figure 10. (a) Vorticity and (b) Turbulent kinetic energy trend within the cylinders of crank-rocker and crank-slider engine at cold-flow conditions

11a.png

(a)

11b.png

(b)

Figure 11. Experimental emissions analysis for (a) HC and (b) NO_X of crank-rocker engine [24]

The authors hereby acknowledge financial support from Yayasan Universiti Teknologi PETRONAS (YUTP) Malaysia in lieu of Graduate Research Assistantship (GRA) Scheme under the grant cost centre: 015LC0-127.