Thermodynamic Study Into Rotary Internal Engine

Thermodynamic Study Into Rotary Internal Engine

ABSTRACT In modern technology era, fuel consumption always has been main topic as the source of fuel is decreasing day after day. Many research and experiment has been made to improve the performance of the engine consequently will reduce the fuel consumption of the engine. A lot of development was made from the intake stroke until the exhaust stroke such fuel injection and compressor to increase air fuel ratio before injected to the engine. There are many type of engine has been produced nowadays which are reciprocating engine, diesel engine and rotary engine.

Ideally the performance of the rotary engine is higher than reciprocating engine but the rate of unburned fuel in rotary engine is higher than reciprocating engine. The project was done to investigate the flow field of the rotary engine, fuel concentration and fuel injection near the spark plug, and the pressure and temperature distribution in the rotary engine. The methodology to the project was simulate and calculate using computational fluid dynamic software which are GAMBIT and FLUENT.

The project was assumed as steady state problem since in order to do dynamic mesh; a lot of consideration must take to finish the task. In GAMBIT, the geometry of the rotary engine was built until meshing task. After that, mesh file was exported into FLUENT in order to declare boundary condition, operating condition, type of turbulence model, and material properties. After simulation was done, the result on temperature, pressure and density has been recorded into data and translated it into graph. The graph was compared to theoretical and journal that related to this project.

Page | x CHAPTER 1 INTRODUCTION 1. 1. Overview 1. 1. 1. Background Felix Wankel, German Engineer produces a great rotary engine on February 1, 1957 after 6-years working on this engine. Basic idea is almost same compare to 4-stroke engine. The Wankel rotary engine is a fascinating beast that features a very clever rearrangement of the four elements of the Otto cycle that is intake, compression, power and exhaust. One great advantage is its high operating speeds that allow the engine to produce twice as much power as a reciprocating engine like Otto cycle at the same weight.

Figure 1. 1: Complete cycle of Wankel rotary engine [1] Page | 1 The advantages of this rotary engine are its compactness, higher engine speed, inherent balance, and smoothness. The disadvantages are its sealing and leakage problem, lower efficiency and higher unburned hydrocarbon emissions resulting from the flattened combustion chamber shape. As mention before, the flow field in the combustion chamber and the fuel concentration around the spark plugs should be optimized in rotary engine in order to improve their fuel economy. 1. 1. 2.

Motivation The motivation for this research is due to ideology of the Wankel rotary engine which is capable of producing higher efficiency and performance compare to Otto engine. Nevertheless, the flaws on the Wankel rotary engine especially on the flow field in the combustion chamber and fuel concentration around the spark plug in this rotary engine has rised the issue on the optimize on the engine performance. Typically, each stroke in rotary engine are separated by the rotor within the chamber. This rotor is shaped like a triangle whose sides have been curved outward to allow for a better seal against the chamber wall.

The rotor is set up on an eccentric cam which allows for the rotor to have all three of its corners touching the chamber wall at all times. For an example on one of the Wankel engine, Aixro XR 50 that produced maximum power 48 horsepower (hp) when the engine is running at about 8500 rotations per minute (rpm) [11]. Compare to any piston engine is out on the market today, a single cylinder capable to produce 12hp. This Page | 2 means a single rotary engine can produce quadruples horsepower than a single piston when comparing to their weight rotary engine only around 30 pounds and piston engine weighs nearly 50 pounds.

Table 1. 1 illustrates the comparison between Wankel engine specifications to the piston-typed engine. Table 1. 1: Comparison between Wankel engine with piston engine Wankel Engine Horsepower Weight Torque Max engine speed 48 hp 30 pounds 34 lb/ft at 4500 rpm 10300 rpm Piston Engine 12 hp 50 pounds 32. 5lb/ft at 6000rpm 7000 rpm The most motivation of this project dealing with the fuel consumption in the Wankel rotary engine. If to compare to the Otto-cycle engine, Otto-cycle engine can lean more fuel rather than Wankel rotary engine.

In order to overcome this problem, study on flow field in the combustion chamber, fuel concentration, fuel injection and temperature distribution around the spark plug are necessary in order to know the thermodynamic properties of the Wankel rotary engine especially in combustion chamber. 1. 2. Problem Statement The problems with the oil and fuel consumption in the Wankel rotary engine that caused by the existence of the unburned area enhance its performance Page | 3 due to incomplete combustion in the combustion chamber.

Simulation and analysis using Gambit and Fluent software to the flow field on the rotary internal engine especially on temperature distribution, fuel concentration and injection near the spark plug as scope of the thermodynamic aspect. 1. 3. Objective of the Project The objective of this project is to investigate into the thermodynamic aspects of the rotary internal engine. Investigate into the flow field and increasing the performance of the rotary internal engine through the optimization of the fuel concentration around the spark plug in the combustion chamber using numerical simulation.

This includes: a) to study the flow field in the combustion chamber b) to study effect on the fuel concentration and injection c) to study the temperature distribution through the fuel injection using spark plug 1. 4. Scope of Work The project is focusing in thermodynamic aspect that is studying and simulating cycle-resolved measurement of the fuel concentration and the effect of the present air/fuel ration on the fuel concentration near the spark plug. Thus, the relationship between the fuel concentration and the combustion characteristics under lean-burn conditions will be studied.

Page | 4 CHAPTER 2 LITERATURE REVIEW 2. 1. Introduction About half of century after the birth of Wankel rotary engine, a lot of experiments and researches have been performed to improve and develop the performance. Both mechanical and thermodynamic of Wankel rotary engine is different and has been improved since the first engine is produced. Recently, a new side exhaust port rotary engine has been developed [2] to reduce the unburned fuel and hydrocarbons through it. Upgraded of fuel injectors and spark plug also improve its fuel economy.

Basically, the most important thing need to be optimized is that the flow field in the combustion chamber and the fuel concentration around the spark plug because this is the significant to the thermodynamic process in the engine. Since the fuel concentration has been obtained, temperature distribution and pressure distribution also can be measure because these factors are related to fuel concentration. A lot of method and experiment [3] have been done in order to measure the fuel concentration in the engine especially in the combustion chamber.

One of the methods that often implemented is that using gas sampling method. Unfortunately this method cannot obtain the successive data. The time-lag problem exists seriously when fast-flame ionization detector (FID) system is applied due to the sampling tube [4]. Page | 5 2. 2. Fuel Concentration and Injection in Wankel Rotary Engine There are several methods such as FID method, LIF method, Raman Spectroscopy and In-situ Infrared Absorption Method in order to investigate fuel concentration and injection near a spark plug.

Previously, most of the research has been done at reciprocating engine because this engine is widely use compare to Wankel rotary engine. These methods also can be applying to rotary engine since the cycle-resolved process and the working principle both rotary engine and reciprocating engine are the same. 2. 2. 1. Fast-flame Ionization Detector (FID) Research related to the Fast-flame Ionization Detector (FID) that has been investigated by Cheng et. al. [4], found that this method found to be effective to measure the cycle-resolved Hydrocarbon (HC).

This method has been proven it is applicable and reliable in understanding the HC oxidation mechanisms when it comes to HC measurements in the exhaust port. The purpose of this method is to measure HC concentration of unburned mixture in cylinder and hence it is an extremely important measurement in order to have a complete combustion in the combustion chamber. Cheng et. al investigation has shown that the flow rates of FID are controlled by allowing each gas which is depending on the experimental sample of hydrocarbon to pass through capillary and dependent on the pressure drop across and the cross section of the capillary.

A short probe needed in order to satisfy bottle-neck to faster response. The flame Page | 6 chamber must be maintained at the pressure below that of the sample source because the sample needs to be drawn directly into the FID. The device as a Figure 2. 1 should be carefully set and calibrated in order to get the absolute HC concentration. They also have concluded that FID is good enough to apply when the method with terms of chamber pressure. Figure 2. 1: Fast FID sampling unit [4] However, Younggy [5] has investigated that the problem in poor repeatability f the measurements is existed due to the unknown intrinsic problem of the sampling system which stems from the alternating supply of unburned mixture and burned gas into the FID within one cycle. Then, this problem will be follow up with a serious time-lag problem. Hence, the in-cylinder sampling system still needs careful operation and calibration in order to get a better result in Fast-flame Ionization Detector. Page | 7 2. 2. 2. Raman Spectroscopy Egermann et. al. [6] have investigate that the concentration of oxygen gas, Nitrogen gas, vapor, and iso-octane which can be measured by Linear Raman scattering.

They have found that the influence of air entrainment on mixture formation, a separation between pure air and air/fuel mixture spectra was possible. The air/fuel ratios evaluated by a droplet passed spectra since pure gas-phase is the only thing will pass through it. They have indicated a suitable place in the chamber to be selected to measure the fuel concentration with high spatial and the temporal resolution. The experiment setup included Intensified Charge Coupled device (ICCD) cameras; Glan Prism and holographic notch filter.

The function of ICCD cameras is allow one dimensional spectroscopic imaging, Glan Prism for the separation of the polarization direction from each other and holographic notch filter for remove the elastically scattered light. The outlet valves and inlet valves have been line up properly in order to have an origin place in every axis. Hence, this origin place is corresponds to the position which directly underneath the spark plug. Schroeder et. al. [7] have investigated that the recorded spectra must be evaluated by automatically working software since the liquid phases of a species are different.

This method is suitable to reciprocating engine, and will produce good result for one-dimensional case and combine with the laser-induced fluorescence (LIF). This method also offers a potential Page | 8 quantitatively two-dimensional mapping of the mixture formation. However, when it comes to commercial rotary engine it is complicated method because of the need for optical windows is not really possible when apply to rotary engine. 2. 2. 3. Laser-induced Fluorescence (LIF) LIF measurement is almost same method with previous method used an optical solution to measure the fuel concentration in chamber.

The different is LIF can measure in two-dimensional mapping and Raman Spectroscopy only can measure for one-dimensional mapping. Research that has been investigated by Han et. al. [8] showed that the total vaporphase fluorescence is proportional to the instantaneous mass evaporation rate when the concentrations of the multiple LIF tracers are adjusted. Figure 2. 2: Schematic diagram of evaporation experiment [8] Page | 9 They used the saturated vapor like laser beam scan by the LIF where the air will be heated by fuel or tracer mixture.

Furthermore, air has been forced to produce a beam like shown in Figure 2. 2. The evaporation data from the top chamber is used to characterize droplet evaporation in an engine since LIF intensity proportional to the instantaneous mass evaporation rate. The LIF also image can be converted into fuel mass distributions condition when it the image meets the desired condition which in turn allows calculation of local fuel equivalence ratios. Their investigation has proven that PLIF images represent the mass distribution of fuel and the equivalence mass distribution can be calculated.

However, the problem will occur when this method apply to rotary engine. 2. 2. 4. In-situ Infrared (IR) Absorption Method Recently, an in-situ laser infrared (IR) absorption method is commonly applied to perform cycle-resolved measurements of the fuel concentration in engine. Nishiyama et. al [9] has performed investigation using this method applying the optical spark plug sensor with a doublepass measurement length as shown in Figure 2. 3. The new sensor has been introduced to provide quantitative cycle-resolve fuel concentration measurements around the spark plug with high temporal resolution.

Page | 10 In-situ laser infrared absorption method is newly involved in measurement of fuel concentration. This method is a replacement from previous method due to their disadvantages and it is found unsuitable for rotary engine. Tomita et. al. [10] has found that the molar absorption conditions which are dependent on both pressure and temperature must be investigated in order to measure the fuel concentration accurately. They have concluded that the drastic changes in the pressure and temperature inside the combustion chamber with piston movement. Figure 2. : Schematic diagram and photograph of an IR spark plug sensor [9] In order to obtain fuel concentration for combustion diagnostics, 3. 392 µm He-Ne lasers particularly has been used. By measuring the concentration of a homogeneous methane-air mixture in a compressionexpansion engine, the result of measurement accuracy was confirmed. Page | 11 Kawahara et. al. [11] used an in situ laser infrared absorption method by using a spark plug sensor and a 3. 392µm He-Ne laser as the light source in order to measure the fuel concentration near a spark plug in the combustion chamber of the rotary engine.

The wavelength that has been produced will coincide with the absorption line of hydrocarbons. Furthermore, the new IR spark plug sensor is been used so that it can provided quantitative cycle-resolved fuel concentration measurements around the spark plug with a high temporal resolution. Inhomogeneity mixture was strong during on the early compression stroke, and this will be decreased when mixture throughout the compression stroke. Figure 2. 4: Experimental setup on the rotary engine with infrared spark plug sensor system [11] Page | 12

A commercial rotary engine which is Mazda 13B engine has two plug holes in one rotor housing. The experimental setup for this experiment is shown in Figure 2. 4. Pressure transducer is been setup on the trailing side (T) of the spark plug to measure the pressure in the combustion chamber. For leading side (L) of the spark plug was replaced with spark plug sensor. This sensor used to send the measurement detail to vibration isolator to measure the fuel concentration in the combustion chamber. In the vibrator isolator, there was a 3. 392µm He-Ne laser and an infrared detector.

Light from the laser was guided to the sensor through and passed through the in-cylinder gas. After that, the transmitted light is guided to the IR detector through the second fiber and a band-pass filter. At every degree of rotation of the eccentric shaft, in-cylinder pressure, infrared signal intensity and top dead center signal will be recorded. Kawahara et. al. have concluded that there was a strong relationship between the fuel concentration measured with the spark plug sensor and the combustion characteristics during the initial combustion period due to the increasing of molar density.

The transmissivity was slightly increased at the time spark plug will spark and gradually retuned to unity after the flame passed through the spark plug sensor. This research is currently in progress and the full result on fuel concentration and cyclic variability will be known in future since rotary engine a more complicated compare to reciprocating engine. Page | 13 As for this project, simulation and analysis using Gambit and Fluent program originally want to compare the result with previous experiment that has been proposed by Kawahara even it is just prerequisite result.

In other word, simulation result is been compare with experimental result. The condition and properties of the experimental setup must be exactly same especially the specifications of the test engine. Fuel concentration will be measure and investigate by using simulation after having an analysis on rotary engine working and the combustion characteristic of the rotary engine. 2. 3. Temperature Distribution in Wankel Rotary Engine The understanding on temperature distribution in combustion chamber is important because it has a strong correlation with the combustion characteristic of the engine.

For internal combustion engine, before enter combustion chamber temperature is directly proportional with pressure in the combustion chamber due to compression process. During the compression process, as the rotor continues its motion, the volume of the chamber gets smaller and the air/fuel mixture gets compressed. Based on the previous experiment done by Masaki et. al. [2] and Kawahara et. al. [11], two spark plugs are been used because the combustion chamber is long. Therefore, applying with only one spark plug can cause the flame spread at slow pace.

This phenomenon will eventually causing the power of the engine drops due to the time lagging in the combustion chamber. Page | 14 Consequently, more unburned fuel and hydrocarbon will be sent through the exhaust port. The understanding and instigation on the fuel concentration in the combustion chamber will aid on the study of the temperature distribution in the rotary engine. These two characteristics are dependent on each other since it is related to the pressure distribution in the combustion chamber. Up to date has, the research and experimental investigations on the temperature distribution in the rotary engine are lacking. . 4. Summary on literature review For this project, the comparison between experimental and analytical will make a greater result on the study on the effect of the preset air/fuel ration on the fuel concentration and injection, and temperature distribution near the spark plug based on the previous experiment and researched that has been done by others. Hence, the relationship between these factors with the combustion characteristics will be benefit and useable in the future modification of the Wankel rotary engine. Page | 15

CHAPTER 3 METHODOLOGY 3. 1. Introduction As mention before this project includes the analysis and simulation for comparison to the investigation performed by Kawahara et. al. [11]. GAMBIT 2. 3. 1. 6 and FLUENT 6. 3. 26 were used in this project. GAMBIT was used to generate the mesh and pre-process the model before exported to FLUENT. FLUENT was used to process to further simulate the model and post-processing the result obtained. 3. 2. Method on Gambit software This investigation consists of a two dimensional computational fluid dynamic problem.

As basic information, two dimensional projects are related to face geometry instead volume geometry that suitable for three dimensional projects. Figure 3. 1 is illustrates the procedures to generate the mesh in gambit. Page | 16 Plotting the point vertex using specific equation Creating the edge based on vertex that has been plotted Creating the face from wireframe Meshing the face Declaring the boundary condition Save the project and ready to read in FLUENT Figure 3. 1: Steps on GAMBIT software Page | 17 3. 2. 1. Geometry Creation The Wankel rotary engine has been modeled using the equation for both housing and rotor.

The housing of the Wankel rotary engine consists of two side parts and an epitrochoid as shown in Figure 3. 2. Research has that has been investigated by Husni [12], the inner curve of the epitrochoid expressed by the coordinates of point, p (x, y) like shown in Equation 3. 1: x = e cos 3? + (R + a) cos ? y = e sin 3? + (R + a) sin ? Where: e is eccentricity, R is the generating radius a is clearance distance or amount parallel transfer. Equation 3. 1 Figure 3. 2: Peritrochoid curve of housing [12] Page | 18 The engine rotor is a triangular shape, has three convex faces, each of which acts like a piston.

The inner envelope of the Peritrochoid is a basic curve forming the contour of the rotor. From Modelaford website [13], coordinates of point, p (x, y) can generate the rotor surface curve by two equations as follows: x = R cos 2? + (3e2/ 2R) (cos 8? – cos 4? ) + e (1 – 9e2/ R2 (sin2 3? )) (cos 5? + cos ? ) y = R sin 2? + (3e2/ 2R) (sin 8? – sin 4? ) + e (1 – 9e2/ R2 (sin2 3? )) 1/2 (sin 5? – sin ? ) …………………. Equation 3. 2 1/2 x = R cos 2? – (3e2/ 2) (sin 6? ) (sin 2? ) + 2e (1 – 9e2/ R (sin2 3? )) 1/2 (cos 3a) (cos ? ) y = R sin 2? – (3e2/ 2) (sin 6? ) (cos 2? ) + 2e (1 – 9e2/ R (sin2 3? )) 1/2 (cos 3a) (cos ? …………………. Equation 3. 3 The rotor’s contour Equation 3. 3, ideally the apexes of triangle rotor are fully fay with the inner surface of cylinder as show in Figure 3. 3, which is very suitable for research and analysis on academic theory. For equation 3. 2, there are gaps in between apexes of triangle rotor and the inner surface of cylinder, this clever design with space is just nicely fill in by ape seal; it is valuable for practical application of engine operation. For this project, equation 3. 2 has been chosen since it is valuable for practical application besides when plotting equation 3. in gambit there is some error occur which the rotor exceed the geometry of the housing. Page | 19 3. 2 3. 3 Figure 3. 3: Contour of the rotor [13] In order to complete the model of Wankel rotary engine, the equation must be solving by angle of rotation from 0° to 360°. For contour of the rotor, the equation is solve with varies angle of rotation that is from 30° to 90° for rotor face 1, 150° to 210° for rotor face 2 and 270° to 330° for rotor face 3. After finishing plotting the point vertex, the edge will be creating and thus face will be creating from wireframe.

A complete modeling Wankel rotary engine is shown like Figure 3. 4. Detail and specification on Wankel rotary engine is shown like Table 3. 1, the value of e, R and a are refer to this table. Page | 20 Figure 3. 4: Complete modeling of Wankel rotary engine Table 3. 1: Engine geometrical data [11] Variable R [mm] e [mm] a [mm] B [mm] R/e [-] R+a [mm] V, [cm3] Value 105 15 3 80 10. 0 108 654 x 2 motor Description Generating radius Eccentricity Clearance distance Housing width Ratio Engine capacity volume Page | 21 3. 2. 2. Grid Generation For two dimensional projects, only two element options for meshing that are quad and tri.

The element of quad is better than tri since it will give more detail mesh. For this project, Quad pave is more suitable than Quad map since map will give uniform mesh and pave will give an unstructured mesh. Thus, this project using quad pave with spacing value is 0. 7 in order to get very smooth result when iterate on FLUENT. The project that completed with meshing task will be as shown in Figure 3. 5. After finishing mesh task, boundary condition will be declared before the project will be export in FLUENT. For rotor of the engine, mesh will be modified in FLUENT so that it will be dynamic mesh since it is moving face.

Figure 3. 5: Complete task in GAMBIT Page | 22 3. 2. 3. Angle Identification The project is done in steady state condition instead unsteady state condition even in real condition for rotary engine is unsteady state condition. In order to run the project with unsteady state condition, the project must be consider dynamic mesh and user define function since the rotor moving in different center of gravity and different grid motion. User define function is a coding C programmed that must be compiled into FLUENT illustrate the motion of the rotor by some coding of C programmed.

Besides, it take a long time in order to avoid negative volume in FLUENT even the C programmed is right. Alteration on spacing value in mesh and smoothing in mesh is needed to avoid this negative volume exist. Thus, the complication of dynamic mesh, the project is defined as a steady state condition. For this project, five different angle is been chosen in order to observe temperature distribution, pressure distribution, air fuel ratio, and thermal boundary layer in combustion chamber. Table 3. 2 illustrates the angle identification that has been done in GAMBIT. Page | 23 Table 3. : Angle Identification No Angle View 1 -270° 2 -90° 3 0° 4 120° 5 200° Page | 24 3. 3. Method on Fluent software Fluent software contains the broad physical modeling capabilities needed to model flow, turbulence, heat transfer, and reactions for any fluid computational dynamic problem. It has advanced solver technology provides fast, accurate CFP results, flexible moving and deforming meshes. Figure 3. 6 shows the basic steps in fluent. Read . msh file from GAMBIT Checking and scaling the grid Defining suitable solver with appropriate equation Defining materials properties

Defining operating conditions and boundary conditions Run the project Figure 3. 6: Steps on FLUENT software Page | 25 3. 3. 1. Material properties Based on the Kawahara et. al. [11], the best hydrocarbon fuel is optimum gasoil fuel. Thus, from FLUENT database, gasoil-vapor (c8h18) has been choose so that the computational fluid dynamic project is similar with real life condition which used gasoline vapor to injected into rotary engine. The air fuel ratio injected into inlet vent is been calculated based on different type of density since the area and velocity in inlet vent is set to be constant.

The density of air and fuel is referred to Table A1 and Table A2. 3. 3. 2. Boundary Condition Boundary condition is one of the important properties in CFD simulation to introduce the physical processes. Improper sets of boundary conditions may introduce nonphysical influences on the simulation system hence it will be ruin the result off the simulation. For housing face, it has to declare fixed wall because there is no any movement on the housing. These boundary conditions are necessary to be set: ? At inlet vent, velocity and pressure of the air and fuel is need to set based on Fluent database ?

The outlet vent has a function like exhaust so that any combustible thing will be carry out by this vent ? At the inlet temperature (spark plug), the temperature is being apply by periodic under specific spark condition (400kPa, 600K) ? The wall of the rotary engine is set to constant temperature, 300K for comparison Page | 26 Inlet Vent T e m p I n l e t Outlet Vent Figure 3. 7: Boundary condition Based on the Figure 3. 7, inlet is set to be velocity inlet with a different speed which is 50 m/s, 100 m/s and 150 m/s.

For the temperature inlet, the condition set is to be 600K and 400kPa. The outlet vent is set to be 300K and 101. 325kPa in order to ensure the combustion is process in normal atmospheric condition. Page | 27 3. 3. 3. k-? model The k-? model is the most widely used turbulence model in practical engineering application. This is because it is very stable and most accurate result in order to find the solution to turbulence model. The k-? model employs additional partial differential equations for the turbulence kinetic energy k. The final form of the ? -equation is: …………. Equation 3. 4 Equation 3. 4 illustrate the equation of the k-? model that has been used to solve the turbulence model in FLUENT. There are terms and condition that describe to this equation: ? Term 1 – The transport of ? by the fluctuating velocities and is modeled in the same way as the similar term in the k equation ? Term 2 – Express the augmentation of the dissipation rate by the mean motion ? Term 3 – The decay of the dissipation rate and is assumed to be proportional to the dissipation rate itself divided by the decay time scale of the turbulence, k/? Term 4 – correlation of gradients of fluctuating velocity; neglected high Reynolds number flows Page | 28 3. 3. 4. Conservation of thermal energy Thermal energy can be defined as a specialized term that refers to the part of the internal energy of a system which is the total present kinetic energy resulting from the random movements of atoms and molecules. By ignoring fluctuation of density, laminar viscosity, thermal conductivity and specific heat, the ensemble-averaged stagnation enthalpy equation give the final from like Equation 3. 5. ……………. Equation 3. 3. 3. 5. Coherent Flame Model Coherent Flame Model (CFM) is one of combustion energy motion that always used in Computational Fluid Dynamic problem. The CFM is applied for both premixed and non-premixed conditions on the basis of a laminar flamelet concept, whose velocity SL and thickness ? L are mean values integrated along the flame front, only dependent on the pressure, the temperature and the richness in fresh gases. This model assumes the reaction relatively within thin layers that separate the fresh unburned gas from the fully burnt gas. Equation 3. is used for lean combustion case: Page | 29 ……………. Equation 3. 6 Where: ? ?L – the mean laminar fuel consumption rate per unit surface along the flame front ? ? ? ?fu,fr – the partial fuel density of fresh gas ? fr – the density of the fresh gas yfu – is the fuel mass fraction in the fresh gas Page | 30 CHAPTER 4 RESULT AND DISCUSSION 4. 1. Introduction All the simulation works were done to the flow field on the rotary internal engine and discussed in this chapter. The results obtained are presented in form of table, graph or pictures. This chapter covers topics as follow ? ? ?

Result and observation on the flow field in the combustion chamber Result on the effect of the fuel concentration and injection Result on the temperature distribution and the pressure distribution through the fuel injection using spark plug 4. 2. Result and observation on the flow field in the combustion chamber Test on thermal boundary layer on the combustion chamber thickness will assist to understand the flow field in the combustion chamber. The concept of the thermal boundary layer is to define the properties of the streamlines whether is it laminar, transitional or turbulent flow. Figure 4. show the thermal boundary layer in the combustion chamber of the Rotary Engine that has been observed in FLUENT. Page | 31 Figure 4. 1: Thermal boundary layer in combustion chamber From Figure 4. 1, it can be simplify and construct a streamlines passing through boundary layer shown in Figure 4. 2 Based on that Figure 4. 2, the right hand side is the beginning of the streamlines and the left hand side means the final state of the streamlines. We can see that, there is transitional streamlines flow exist at the beginning of the streamlines. Figure 4. 2: Streamlines flow in combustion chamber Page | 32

The parameter when on rotor at the ignition process, the spark plug will inject the temperature at the turbulent state. When it comes to combustion chamber, the turbulent boundary layer has been mixed up with surrounding air to become transitional boundary layer and lasted with laminar boundary layer. This phenomenon occurs because of the reducing value of density of the fluid, velocity of the fluid and thus the Reynolds number of certain region. Based on the theoretical of the boundary layer, these two boundary layers, turbulent and transitional must be reduced so that the engine will be less vibrating.

This is because there will be unsteady flow exist in turbulent and transitional boundary layers. Until now, it is impossible to get the perfect and constant steady flow because the effect of the spark plugs that produced turbulent flow. In real life engineering flow, we cannot neglect transition and turbulent flow, thus we can reduce the length of time of their occurrence by minimizing any possibilities factor that has been catalyst to unsteady flow live longer such that roughness along the surface, free-streams disturbance, vibrations, and curvature of the wall contribute to an earlier transition location.

In order to reduce unsteady flow exist in the combustion chamber, the surface in the combustion chamber must be very smooth so that there is no disturbance that can affect the stream flow. Page | 33 4. 3. Result on the effect of the fuel concentration and injection The study on fuel concentration and injection around the spark plug is important in order to increase the performance of the rotary engine. Figure 4. 3, Figure 4. 5 and Figure 4. 7 illustrate the experimental graph. (A/F)? = 13. 1 17 16 15 A/F 14 13 12 11 -60 -50 -40 -30 -20 -10 0 Ecccentric shaft angle (°)

Figure 4. 3: Computational result on A/F versus angle – 13. 1 Figure 4. 4: A/F versus eccentric shaft angle – 13. 1 [11] Page | 34 (A/F)? = 14. 7 17 16 15 A/F 14 13 12 11 -60 -50 -40 -30 -20 -10 0 Ecccentric shaft angle (°) Figure 4. 5: Computational result on A/F versus angle – 14. 7 Figure 4. 6: A/F versus eccentric shaft angle – 14. 7 [11] Page | 35 (A/F)? = 16. 1 25 23 21 A/F 19 17 15 13 11 -60 -50 -40 -30 -20 -10 0 Ecccentric shaft angle (°) Figure 4. 7: Computational result on A/F versus angle – 16. 1 Figure 4. 8: A/F versus eccentric shaft angle – 16. [11] Page | 36 25 23 21 19 A/F Kawahara 13. 1 Kawahara 14. 7 Kawahara 16. 1 17 Comp 13. 1 Comp 14. 7 15 Comp 16. 1 13 11 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 Ecccentric shaft angle (°) Figure 4. 9: Comparison between computational with Kawahara’s result on A/F versus eccentric shaft angle Page | 37 The graph is been plotted after the data get from certain angle since the density of the mixture is depending on pressure and temperature. Since the velocity and area was assumed constant, the air fuel ratio is only depending on the density of the air and fuel.

From experimental graph, the air fuel ratio near spark plug maintained a constant value at the preset air fuel ratio when the intake mixture is 13. 1 and 14. 7. For a leaner case, (A/F)? = 16. 1, the air fuel ratio decrease gradually before reaching the preset air fuel ratio. The delay between preset value at 16. 1 with 14. 7 and 13. 1 in order approaching the preset air fuel ratio is about 20°. This phenomenon will affect the initial stage of combustion and the stability of lean mixture combustion. The experimental graph was compared to the research that has investigated by Kawahara et. l [11] which has been indicated at Figure 4. 4, Figure 4. 6, and Figure 4. 8. Both experimental graph and journal graph found to have an agreement. Based on these graph, the richer case = 13. 1 is the most suitable properties for intake mixture because there was only small delay in order to reach strong homogeneity mixture before the spark timing. This homogeneity will reduce unburned fuel in combustion chamber and hence the fuel consumption will decrease subsequently the performance of the rotary engine will increase because the efficiency of the engine.

From this research, the flaw of the rotary engine was decrease on the understanding on fuel concentration around spark plug. Page | 38 4. 4. Result on the temperature distribution and the pressure distribution The understanding on temperature distribution and the pressure distribution is important to inform about the fluid properties at certain angle based on the temperature and pressure. Besides, for the manufacturing of this engine, temperature and pressure around the engine must be simulate to ensure the material that has been used can withstand these two mechanical properties.

The temperature distribution has been made on three different inlet velocities and the data and graph is shown in Table 4. 1 and Figure 4. 10 respectively. Angle of the rotor was set at constant value so that it can show how the effect of velocity enhanced temperature. Table 4. 1: Temperature – Speed data Temperature (K) Speed (m/s) (-270°) 50 100 150 285. 38 285. 37 285. 25 (-90°) 320. 22 325. 28 322. 86 (-0°) 570. 32 568. 48 569. 56 (120°) 494. 99 490. 15 493. 57 (200°) 465. 62 468. 66 465. 76 Page | 39 Temperature vs Angle 600 Temperature (K) 500 400 300 200 100 0 -300 -200 -100 0 100 200 50 m/s 100 m/s 150 m/s Angle (°)

Figure 4. 10: Computational result on temperature versus angle Based on the Figure 4. 10 the highest temperature value is at 0° due to the spark of the spark plug that has been set at 600K. The values of temperature almost constant from -270° until -100° because of small compression occur at intake process. After that, at compression stage the graph shown an increasing value of temperature until it reach nearly 0°. At the compression stage, the fluid in the chamber will be compress and there will be increasing value of temperature. After that, there is a negative slope means that the value of temperature is decreasing.

When applying different speed of intake velocity, the graph shown these three values of velocity did not affect value temperature around chamber of rotary engine. Based on theoretical about compressible fluid, when there are increasing of speed in compress area, the result on temperature and pressure will be different. Page | 40 Since this project is real life engineering problem, the ideal gas law is inconsiderable because real gas has many different residual property. The ideal gas law is most accurate monatomic gases which molecular size and density of the fluid is neglected.

For real gas problem, the equation of state that can be used is van der Walls model. The van der Walls model shown that when pressure of the fluid increase will make the increasing value of temperature. Based on Figure 4. 10, the profile of the graph was identical on each other even though different speed of the inlet velocity is applied. There is an error occur because the graph of pressure versus angle shown that pressure can be affected when using a different speed of the inlet velocity like shown if Figure 4. 11. The graph of the velocity versus angle seems disagreed to the van der Walls model of real gases state.

The slope of the graph is supposed to be different with the speed of inlet velocity. The probability on why there is an error occurred on the graph temperature versus angle because computational work was done in steady state condition. This condition will approach unexpected result since the real rotary engine operates in unsteady state condition. Another probability is the boundary condition which was been set in FLUENT maybe not accurate value because the data is taken from the real operating rotary engine which is unsteady state condition.

Page | 41 For the pressure distribution in chamber of the rotary engine, the methodology in order to get a result is made exactly like temperature distribution and the only different is the selection of the contour properties. The result has been record in table shown in Table 4. 2 and translates the data with the graph form like shown in Figure 4. 11. Table 4. 2: Pressure – Speed data Pressure (Pa) Speed (m/s) (-270°) 50 100 150 0. 08124 0. 41347 0. 844051 (-90°) 8000. 8477 34023. 375 77458. 836 (0°) 48309. 363 199338. 09 527379. 5 (120°) 0. 9566 0. 115796 0. 561995 Pressure Vs Angle 600000 500000 Pressure (Pa) 400000 300000 200000 100000 0 -270 -180 -90 0 90 180 50 m/s 100 m/s 150 m/s Angle (°) Figure 4. 11: Computational result on pressure versus angle Page | 42 From Figure 4. 11, the slope of the graph at the early angle (intake stroke) is slightly increasing because the pressure at the intake stroke not much different until the rotor reach compression stroke. Around -90°, the slope of the graph is rise suddenly due to compression process occur at the compression stroke.

It is rising until the angle of the rotor reach 30°. After that, the slope the graph is negative means that the rotor has reach combustion stroke. At combustion stroke, the pressure will be drop after has been compress and ignite. This graph is compared to the pressure history of the rotary engine that has written by Kawahara et. al. [11]. The graph is shown in Figure 4. 12. Figure 4. 12: Pressure history of the rotary engine with rotor position in the housing [11] Page | 43 Kawahara et. al. investigated that an apex seal on the rotor edge through trailing plug was -225°.

At that stage, the intake port was left open until the angle reach -205°. The pressure rise suddenly at -17° cause by the mixture has just been ignited. After that, at angle of 105° the next apex seal passed through the trailing plug. From Figure 4. 11, the graph of pressure versus angle seems to agree to the pressure history of the rotary engine since the contour of the both graph is slightly same. Compare to the pressure history, eccentric shaft angle from 17° until 0°, the slope of the graph is different before and after these angle.

The difference exist because the project is applied on the steady state flow not in dynamic flow since it hard to iterate the solution when it comes to the user define function in FLUENT. So, only certain angle has been measured and the spark timing is set to be steady state after the angle of eccentricity shaft reach 17°. When applying different speed of inlet velocity, the pressure graph will be different with an increasing value of slope. From the graph in Figure 4. 4, speed of inlet velocity enhanced the pressure the whole stroke in the rotary engine. The propose of applying different speed of inlet velocity is to ontribute to the speed of rotor since the project is steady state problem and no moving part involved. Page | 44 CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS 5. 1. Conclusion In combustion chamber, there are three types of flow involved after spark timing which are turbulence flow, transitional flow and laminar flow. Based on the theoretical on thermal boundary layer, the best flow in order to get optimum level on the performance of the engine is the laminar flow since transitional flow and turbulent flow will produce the vibration to the engine which cause the performance engine will be reduces.

Investigation on fuel concentration and injection around spark plug, the best air fuel ration for intake mixture is 13. 1 because the time in order to mixture reach homogeneity is less than preset air fuel ratio value for 14. 7 and 16. 1. Hence, the performance of the engine is at optimum level when using preset air fuel ration at 13. 1 since unburned area in combustion chamber will be small compare to preset air fuel ratio at 14. 7 and 16. 1 The highest pressure and temperature distributions when eccentric shaft angle around 0°.

This is due to spark timing on -17° which catalyst the mixture to burn and hence pressure and temperature will increase instantly. The study pressure and temperature is important to ensure that the material to build the rotary engine can withstand the maximum pressure and temperature when rotary engine is operating. Page | 45 5. 2. Recommendation In order to improve the performance of the rotary engine and simulation on computational fluid dynamic of the rotary engine, some recommendations are given below: 1. The performance of the rotary engine a.

The surface of the overall chamber in engine must be avoided to any disturbance that catalysts the period of turbulent flow exist in combustion chamber. b. The best preset air fuel ratio to the rotary engine is 13. 1 because of the less period of time to make mixture become homogeneity. c. Two spark plugs must be applied in order to reduce the unburned area that exists in rotary engine. 2. Simulation on computational fluid dynamic of the rotary engine a. The simulation of the project must be in unsteady state condition because the otary engine operating in that condition. b. User defined function must be apply in the project to programmed the simulation, to declare boundary condition and the properties of material. Page | 46 References [1] Merriam Webster website [online]. 2011. [Accessed 8 December 2010]. Available form: http://visual. merriam-webster. com/transportmachinery/road-transport/types-engines/rotary-engine-cycle. php [2] O. MASAKI, T. TASHIMA, S. RITSUHARU, F. SUGURU, E. HIROSHI. Developed Technologies of the New Rotary Engine (RENESIS).

Society of Mazda Motor Corporation [online], 2004. [3] H. ZHAO and N. LADOMMATOS. Engine Combustion Instrumentation and Diagnostics, Society of Automotive Engineers, 2001. [4] W. K. CHENG, T. SUMMERS, and N. COLLINGS. The fast-response flame ionization detector. Journal of Progress in Energy and Combustion Science [online], 1998, 24(2), p. 89-124. [5] S. YOUNGGY. Measurement of In-Cylinder Equivalence Ratio during Starting Using a Fast FID. Journal of Mechanical Science and Technology [online], 1997, 11(6), p. 726-736. [6] J. EGERMANN, W. KOEBCKE, W. IPP and A. LEIPERTZ.

Investigation of the mixture formation inside a gasoline direct injection engine by means of linear Raman spectroscopy. Journal of Proceedings of the Combustion Institute [online], 2000, 28(1), p. 1145– 1152. [7] H. W. SCHROEDER, and H. W. KLOCKNER. Raman Spectroscopy of Gases and Liquids. Journal of IEEE [online], 1979, 15(11) p. 161– 163. Page | 47 [8] D. HAN, and R. R. STEEPER. An LIF equivalence ratio imaging technique for multicomponent fuels in an IC engine. Journal of Proceedings of the Combustion Institute [online], 2002, 29(1), p. 727734. [9] N. KAWAHARA, E. TOMITA, T. KADOWAKI, T. HONDA and H.

KATASHIBA. In situ fuel concentration measurement near a spark plug in a spray-guided direct-injection spark-ignition engine using infrared absorption method. Journal of Experiment in Fluids [online], 2010, 49(4), p. 925-936. [10] E. TOMITA, N. KAWAHARA, A. NISHIYAMA, and M. SHIGENAGA. In situ measurement of hydrocarbon fuel concentration near a spark plug in an engine cylinder using the 3. 392 µm infrared laser absorption method: application to an actual engine. Journal of Measurement Science and Technology [online], 2003, 14(8), p. 1357. [11] N. KAWAHARA, E. TOMITA, K. HAYASHI, M. TABATA, K.

IWAI, and R, KAGAWA. Cycle-resolved measurements of the fuel concentration near spark plug in a rotary engine using an in situ laser absorption method. Journal of Proceedings of the Combustion Institute [online], 2007, 32(2), p. 3033-3040. [12] T. I. HUSNI. CFD investigations of mixture formation, flow and combustion for multi-fuel rotary engine. PHD thesis. University Cottbus, Denmark [13] Modelaford website [online]. 2007. [Accessed 18 February 2011]. Available form: http://modelaford. ca/Henry/rotary_math. htm Page | 48 APPENDICES Table A1: Properties of air Page | 49 Table A2: Properties of fuels Page | 50