Download as pdf or txt
Download as pdf or txt
You are on page 1of 22

Accepted Manuscript

Numerical analysis of flow dynamics for two piston bowl designs at different spray
angles

Dinesh Kumar Soni, Rajesh Gupta

PII: S0959-6526(17)30368-2

DOI: 10.1016/j.jclepro.2017.02.142

Reference: JCLP 9068

To appear in: Journal of Cleaner Production

Received Date: 13 October 2015

Revised Date: 13 February 2017

Accepted Date: 20 February 2017

Please cite this article as: Dinesh Kumar Soni, Rajesh Gupta, Numerical analysis of flow dynamics
for two piston bowl designs at different spray angles, Journal of Cleaner Production (2017), doi:
10.1016/j.jclepro.2017.02.142

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT

Highlights
 Two geometries of diesel engine can be simulated and compared by using numerical tool.

 Performance wise hemispherical bowl piston geometry is better than re-entrant bowl piston geometry at
mentioned operating conditions.

 Re-entrant bowl piston geometry produces lower emission than hemispherical bowl piston geometry at
different spray angle.
ACCEPTED MANUSCRIPT
1

1 Numerical analysis of flow dynamics for two piston bowl designs at different spray
2 angles
3 Dinesh Kumar Soni1*, Rajesh Gupta2
4
5 1Researchscholar, Department of Mechanical Engineering, MANIT, Bhopal (INDIA).
6 2AssociateProfessor, Department of Mechanical Engineering, MANIT, Bhopal (INDIA).
7 *Corresponding Author. E-mail: dinsoni2419@gmail.com; Tel: (+919424412177)

8
9 Abstract:-
10 The main objective of the present study is to find the need of technology transformation from
11 hemispherical bowl piston geometry to re-entrant piston bowl geometry. The two different piston bowl
12 geometries were examined by the application of different spray angles. The NO and Soot emission
13 characteristics for both geometries were measured and performance parameters were also evaluated to
14 justify the use of geometry as a part of combustion study. A commercial CFD simulation software
15 AVL FIRE was used to evaluate re-entrant and hemispherical bowl piston geometry. Easily available
16 hemispherical bowl piston geometry of Kirloskar single cylinder diesel engine was used for
17 comparison and validation purpose. The effect of spray angles such as 120°, 140° and 160° tested on
18 both geometry, while comparison has drawn for emissions and performance parameters. Results
19 indicated that, NO and Soot mass fraction of 120° and 160° spray angle is lowest respectively.
20 Whereas, NO emissions are reduced to 66% at 120° spray angle in case of a Re-entrant piston bowl
21 than hemispherical bowl piston geometry. However, the soot mass fraction shows same response for
22 both geometry at 160° spray angle.
23 Keywords: - Diesel engine, Emissions, Hemispherical geometry, AVL Fire, Re-entrant geometry, Spray angle.
24
Nomenclature Symbols
BDC Bottom Dead Center Da , Db , and Dr, Bowl Diameter (m)
TDC Top Dead Center Di Inner Bowl Diameter (m)
ATDC After Top Dead Center Dm Bowl Center Diameter (m)
CA Crank Angle (degree) Tm Bowl Center Depth (m)
NO Nitrogen Oxides R1 to R5 Radius (m)
BSFC Brake Specific Fuel Consumption S1 to S3 Distance (m)
BMEP Brake Mean Effective Pressure T Bowl Depth (m)
CFD Computational Fluid Dynamics Φ (Greek Symbol) Angle (degree)
25
26 1. Introduction:-
27 Emissions from diesel engine are not only harmful for the present, but it also unfavorable for the future
28 of the world. It must be control today by considering its hazardous effect on tomorrow. The severity of the
ACCEPTED MANUSCRIPT
2

29 emission problem has judged by a life cycle assessment of new diesel engine, and the results clearly show that,
30 the global warming and photochemical ozone formation have major impact of 19.47% and 17.54%
31 respectively, of total impacts of the diesel engine (Li et al., 2013). On this stage, numerous researches are
32 proceeding to control emissions from diesel engines by means of different facets of applied science. The
33 technology has new dimension to replace the diesel fuel completely with the combination of hydro refining
34 catalysts and natural triacylglycerols (TAG), which extracts from animal fats or vegetable oils (Jozef et al.,
35 2010).The present paper is aimed to explore the effect of two geometrical technologies in terms of emissions
36 from diesel engine.
37 Emissions from diesel engine can be controlled by optimization of spray parameters such as spray
38 angle, number of nozzles, nozzle holes and spray penetration etc. It has proved that; spray parameter affects
39 the emission formation and combustion phenomena inside combustion chamber (Reitz, 1995; Choi et al., 1999;
40 Fuster et al. 2009). One of the potential causes for the emission formation in diesel engine is the presence of
41 diffusion combustion due to variation in spray parameters. The spray parameter improves mixture formation of
42 fuel with air inside cylinder, which leads to enhance combustion performance and produces low emissions
43 from diesel engines (Subramanian et al., 2013). Moreover, researchers are focused to find the further
44 possibility of emission reduction by using biodiesel-diesel blends with improved spray characteristics. Lahane
45 et al., 2015, performed experimental analysis to determine the effects of biodiesel-diesel blend on spray
46 penetration as well as emission formation and observed that spray penetration length was governed by
47 percentage of biodiesel, which increased wall impingement at higher blends. On the other hand, NOx emission
48 increased by 22.8% at blends of high penetration; whereas CO, HC and soot reduced altogether. Furthermore,
49 in the other research, they increased the number of holes in injector from 5 to 6 and found that, NOx mass
50 fraction was reduced from 7.4 g/kWh to 6.6 g/kWh due to reduction in cylinder temperature (Lahane et al.,
51 2014).
52 Spray angle causes the change in impingement point inside the combustion chamber which affects the
53 swirl motion. The improved swirl motion enhances the air - fuel mixing. Gorji et al., 2009, performed
54 simulation analysis with the piston geometry of Caterpillar 3406 truck engine and results showed that, at
55 increased spray angle, NOx increased due to better mixing inside combustion chamber but soot decreased
56 simultaneously. Literature has also shown that the combined effect of improved parameters of spray is able to
57 reduce emissions from diesel engine. Seung et al., 2010, implemented combination of advanced injection timing
58 and multiple injections to examine the effects of narrow spray angle of 70° and 60°. Result showed low NOx
59 and soot emission by using the combination of advanced injection timings and multiple injectors at narrow
60 spray angle. In practice, by changing spray angle in the combustion chamber assures some effect on emission
61 parameters, moreover stated that, emission may depend on internal geometry of the chamber.
ACCEPTED MANUSCRIPT
3

62 One more method of emission reduction which needs to be taken into account is different piston bowl
63 geometry. The air-fuel mixing inside combustion chamber is related to the piston bowl geometry and has a
64 strong impact on emission formation (Dolak et al., 2011; Jaichander et al., 2012). Li et al., 2014, performed
65 numerical simulation on a hemisphere (HCC), shallow (SCC) and omega (OCC) combustion chamber by using
66 the KIVA-4 CFD software at a constant compression ratio of 18.5. Results indicated that, SCC and OCC
67 generate higher NOx emission at low and high engine speed respectively. The combination of spray and
68 improved piston bowl geometry has positive impact over emission reduction. Jesús et al., 2015, carried out
69 experimental analysis with the single and double injection strategy on three different piston bowl geometry,
70 and results have achieved low NOx and soot emission with low and partial load conditions at double injection
71 strategy.
72 It has investigated that; swirl depends much more on piston geometry; whereas re-entrant piston bowl
73 geometry can be a better option in terms of emission reduction. Swirl is used to improve the combustion
74 process by means of a good piston bowl design. Induced swirl - squish interaction results in a tangled turbulent
75 flow field, which is more acute in the re-entrant combustion chamber geometry (Brandl et al. 1979;
76 Arcoumannis et al., 1993). The intensity of swirl and turbulent kinetic energy is high in the re-entrant type
77 combustion chamber, which helps to improve combustion process. It enhances NOx emission, but reduces soot
78 and HC emissions (Payri et al., 2004; Saito et al., 1996). The variation in re-entrant piston bowl shape is also
79 examined to reduce emissions. Prasad et al., 2011, simulated different designs of the re-entrant type piston bowl
80 geometry and stated that, swirl and turbulent kinetic energy is not same for all geometries; however, it may
81 depend on design parameters like the depth and diameter of the bowl. Furthermore, Wei et al., 2014, simulated
82 three spray angles of 146°, 150° and 154°, on a re-entrant type piston geometry and the results described that
83 NOx and Soot mass fraction were lowest at 154° and 146° respectively. Whereas, 150° spray angle was
84 relatively better for stable emission performance.
85 It can be discovered from the above literature survey, that the combination of spray characteristics with
86 piston bowl geometries is possible and additionally, it can improve emission performance. This combination
87 needs to be further discussed in the existing practical model such as hemispherical bowl piston geometry of
88 Kirloskar single cylinder diesel engine (model TV-1). The present work is centered on the same concept of the
89 mixed force of spray angle and piston bowl geometry through AVL FIRE commercial software. This report
90 performs the comparison of Kirloskar single cylinder hemispherical bowl piston geometry with re-entrant
91 piston bowl geometry at different spray angle of 120°, 140° and 160°. Outcomes are described in terms of NO
92 and soot emissions, while compression ratio and bowl volume are keeping constant for whole analysis.
93 2. Model description: sketching, meshing and designing
94 In general, two dissimilar geometry of a piston bowl has
95 considered in the simulation process. The first geometry is commercially available hemispherical bowl piston
ACCEPTED MANUSCRIPT
4

96 geometry and the second one is re-entrant bowl piston geometry. The medium load operating condition has
97 used to simulate the geometries at a constant engine speed of 1500 RPM. The typical sketch of both geometries
98 is shown in fig 1. Moreover, cell numbers, number of cell faces, the number of boundary faces, cylinder
99 volume, compression ratio, connecting rod length and crank radius for both the computational domains are
100 listed in Table 1. The whole combustion chamber domain is divided into symmetrical three parts due to the
101 same number of nozzles used in the experimental and simulation approach. The surface area of both geometry
102 differs according to engine design characteristics, while total volume is kept constant.
103

104
105 (a) Hemispherical bowl piston
106

107
108 (b) Re-entrant bowl piston
109 Fig.1 Outline of the geometries with main dimensions.
110 As expressed in Fig 2, one third mesh has generated by using hexahedral element in AVL
111 FIRE ESE diesel module. As a result of grid independence test, 40868 cells are selected for both geometries. In
112 order to justify the grid quality used in simulation, three mesh densities have compared in terms of pressure
113 results with crank angle as depicted in fig 3. The predictions are well within acceptable limits and simulation
114 time is about 6 h. Thus, the simulation runs with the mesh of 40868 cells.
115
ACCEPTED MANUSCRIPT
5

116 Table 1. Geometrical and meshing quality parameters for bowl designs.
Geometry Cell Number of Number Cylinder Compression Connecting Crank
numbers faces of volume ratio rod length Radius (m)
boundary (m3) (m)
faces
Hemispherical 40868 2022 250 3.2 × 10-5 17.5 0.234 0.055
bowl piston
geometry
Re-entrant 41760 2400 274 3.2 × 10-5 17.5 0.234 0.055
bowl piston
geometry
117

118
119 (a) (b)

120 Fig. 2 Computational grid of (a) Hemispherical and (b) Re-entrant piston bowl geometry at TDC.

29516 40868 50208


50
Pressure (bar)

45
40
35
30
25
20
15
10
5
0
620 670Crank angle
720 (deg) 770 820
121
122 Fig. 3 Grid independence test.

123 3. Numerical simulation:-


124 In the present research, numerical simulation was performed on two different
125 geometries of single cylinder diesel engine i.e. hemispherical and re-entrant bowl piston geometry. Firstly, the
126 physical model (hemispherical bowl piston geometry) of a Kirloskar single cylinder diesel engine was meshed,
127 whereas physical model of re-entrant piston geometry was created by using the same specification of Kirloskar
ACCEPTED MANUSCRIPT
6

128 engine. The specifications of Kirloskar single cylinder diesel engine are given in Table 2 and initial and
129 boundary conditions for simulation are listed in Table 3. The compression ratio, bowl volume, mass injected
130 and speed of the engine were remained constant for the simulation for both the geometry. The 3-D mesh of
131 geometries was generated by using ESE diesel interface code. The AVL FIRE was used to solve the
132 conservation equations. The AVL FIRE is based on the finite volume method and it can perform simulation for
133 each computational cell to preserve conservative properties. Whereas the finite difference method is used for
134 momentum equation.
135 Table 2 Engine specifications. Table 3 Initial and boundary conditions.
136
137 Make Kirloskar engine, Initial pressure 0.65 mPa
138 TV-1
139 Initial temperature 300 k
Number of cylinders 1
140 Piston temperature 550 k
Bore X Stroke 87.5 mm X 110 mm
Liner temperature 425 k
141 Swept volume 661 cc
Head temperature 475 k
142 No. of nozzle holes 3
Fuel spray angle 120° (solid cone)
Connecting rod length 234 mm
143 Injection type Single injection
Rated output 5.2 kw
144 Mass of fuel injected 1.6e-05 kg
Type of combustion chamber Hemispherical open
145 (Piston bowl shape) combustion chamber
Rated speed 1500 rpm
146
147
148 The following conservation equations and hydrocarbon auto ignition mechanism of multi zone
149 model were used in simulation approach.
150 Conservation of mass: -According to the law of conservation of mass for any open system, the sum of mass
151 transfer of any system across the boundaries equal to the rate of change of total mass of the system.
152 Mathematically, it may be written as (Heywood, 1988),
153 𝑚 = ∑ 𝑗𝑚 𝑗 …… (1)

154 Conservation of energy: -According to the first law of thermodynamics, the conservation of energy equation
155 can be expressed as (Heywood, 1988),
156 𝐸 = 𝑄𝑤-W+∑𝑗𝑚𝑗ℎ𝑗 …… (2)

157 Where, right hand side of the equation denotes heat transfer rate, rate of displacement work and enthalpy flux
158 respectively, and left side shows the rate of change of energy.
159 Auto ignition mechanism of hydrocarbon: -AVL fire simulation software uses reaction mechanism to
160 simulate the auto ignition phenomena of diesel fuel. All species are treated as a single entity, those plays
161 similar role in the combustion chemistry. The generalized reactions of the auto ignition model are (AVL FIRE
162 user manual V2013.1).
ACCEPTED MANUSCRIPT
7

163
𝑤𝑖
164 Ignition Fu + Ox → 2𝑅 ……. (3)
𝑤𝑝
165 Propagation 𝑅→𝑅 + P ……. (4)
𝑤𝑖
166 𝑅→𝑅 + B ……. (5)
𝑤4
167 𝑅→𝑅 + Q ……. (6)
𝑤2
168 𝑅+Q →𝑅+ B ……. (7)
𝑤𝑏
169 Branching B →2𝑅 ……. (8)
𝑤3
170 Linear termination 𝑅→ I ……. (9)
𝑤𝑖
171 Quadratic termination 2𝑅→ I ……. (10)

172 Where, Fu, Ox, 𝑅, B, Q, P and I denote as hydrocarbon fuel of the structure (CnHmOl), oxygen
173 atom , total radical pool, branching agent, intermediate species, products and inactive ( inert ) species
174 respectively. All subscripts of w indicate the reaction rate of a particular reaction.
175 The PISO (Pressure implicit with the splitting of operator) algorithm is used to solve Navier-
176 Stokes equations by applying pressure- velocity calculation procedure and it is found to be an improved
177 algorithm than SIMPLE (Semi- implicit method for pressure linked equations) algorithm. The GSTB approach
178 handles the set of linear equations (Khaleghi et al., 2003; Versteeg et al., 1995; Ferzigeret al., 2002; Van der
179 Vost et al., 1992).
180 To begin the simulation operation, the initial conditions were chosen according to observational data.
181 The initial pressure and temperature was set equal to 0.65 mPa and 300 K respectively. The initial density was
182 computed by applying an ideal gas equation. The turbulence kinetic energy and length scale were set to 15 m2
183 /s2 and 0.003 m respectively. For boundary conditions, the momentum boundary conditions of cylinder head
184 and liner were specified as fixed wall and boundary condition of the piston was specified as moving wall.
185 Whereas, thermal boundary conditions for piston, liner and cylinder head were set as 550 K, 425 K and 475 K.
186 The present simulation strategy used a k-zeta -f turbulence model to solve energy transport equations leading to
187 enhance the numerical stability of the simulation approach (Hanjalic et al., 2004). This model is more precise
188 than the standard two - equation eddy viscosity k-ε model. The k-zeta -f turbulence model is widely accepted
189 for computational meshes and flow conditions of any dimensionless distance near the wall (Y+). In the
190 presented model, the hybrid wall treatment is integrated near the wall by using standard wall function. The
ACCEPTED MANUSCRIPT
8

191 following wall function equations are employed to represent velocity and temperature profiles near the wall
192 (AVL FIRE user manual V2013.1).
193 In the laminar boundary layer (Y+≤ 11.63)

194 U + = Y+ …… (11)

195 T + = σe U + …… (12)

196 In the turbulent boundary layer (Y+ ˃11.63)

197 U + = ln (Ey+) …… (13)


𝜎𝑒
198 T + = σe, t [U+ + P (𝜎 )] …… (14)
𝑒,𝑡

199 In the above equation, P is a viscous sub layer thermal resistance factor, whereas Y+, U+ and T+ are
200 dimensionless parameters known as distance from the wall, non-dimensional tangential velocity and
201 temperature respectively. These symbols are described as follows:
𝑌𝑈𝑇
202 Y+ = 𝜐

𝑈∗
203 U+ = 𝑈
𝑇

𝜌𝑈𝑇 𝑇𝑤
204 T+ = ( 𝑞 ) Cp T ln ( 𝑇 )
𝑤

𝜏𝑤
205 UT = 𝜌

206 Where, UT is the friction velocity and is local wall shear stress. In the above equation k (Von Karman
207 constant), E (Empirical constant for wall function) and (Turbulent prandtl number) are coefficients and their
208 values are equal to 0.41, 9.0 and 0.90 respectively. Furthermore 𝑇𝑤, 𝑇,𝜌,𝜎𝑒,𝜐,𝑞𝑤, Cp, Y and 𝑈 ∗ have their usual
209 meaning as wall temperature , temperature, density, laminar prandtl number, kinematic viscosity, wall heat
210 flux, constant pressure specific heat, distance from the wall and velocity component parallel to the wall
211 respectively.
212 A coherent flame model was used for simulation of both premixed and non-premixed mixture. This
213 model is based on the laminar flamelet concept. The flamelet models are based on layer separation concept. It
214 presumes that the chemical reaction taking place within the fragile layer, which carved up the unburned gas
215 from fully burned gas. In the present study, ECFM-3Z (Extended Coherent Flamelet Model - Three Zones)
216 was used to couple combustion-spray module to describe direct injection combustion phenomena in diesel
217 engine (Colin et al., 2004; Hasteed et al., 1977). The ECFM-3Z model is widely applied to combine spray
218 model with EGR. This model is applicable for auto ignition; even so, it can be used for auto ignition and spark
ACCEPTED MANUSCRIPT
9

219 ignition as well. The ECFM-3Z model can also be used for multi-component fuel (Hanjalic et al., 2004). The
220 wave breakup model was applied in spray modeling with Weber number constant C2 = 18, which depends
221 upon the physical and dynamic parameters of the injected fuel and the domain fuel. Additionally, it relies
222 mainly on the wavelength of the droplets. This model is used for diesel fuel spray simulation (Reitz, 1987).
223 Spray wall interaction model was used for accounting the effect of non - atomized or non- evaporated fuel
224 particles, which striking on the wall of the combustion chamber. A spray wall interaction model known as
225 'walljet1' was used (Uludogan et al., 1996). In the case of walljet1, the droplets get a rebound or slide over the
226 wall due to the formation of vapor cushion under the droplets. The extended zeldovich mechanism was
227 introduced to calculate the NO emission and this mechanism considers the effect of hydrocarbon radicals,
228 nitrogen and oxygen on NO formation.
229

Experimented Pressure Simulated Pressure Experimented HRR Simulated HRR


25
50
45 20
HRR (J/deg)

40
Pressure (bar)

35 15
30
25
10
20
15
10 5
5
0 0
640 660 680 700 720 740 760 780 800 700 710 720 730 740 750 760
Crank angle (deg) Crank angle (deg)
230
231 (a) (b)
Experimented Pressure Simulated Pressure
Experimented HRR Simulated HRR
60
60
50
50
Pressure (bar)

HRR( J/deg)

40
40

30 30

20 20

10 10

0 0
640 660 680 700 720 740 760 780 800 700 710 720 730 740 750 760
Crank angle (deg) Crank angle (deg)
232
233 (c) (d)
ACCEPTED MANUSCRIPT
10

234 Fig. 4 Comparison of experimental and simulated results of (a) Pressure from diesel fuel at partial load (b) HRR of diesel
235 fuel at partial load (c) Pressure from diesel fuel at full load (d) HRR of diesel fuel at full load.
236
237 The NO formation of nitrogen, oxygen and hydrocarbon radicals are very much depends on
238 combustion temperature (Zeldovich et al., 1947). This model can be coupled with ECFM-3Z combustion
239 model based on equilibrium approach. The kinetic soot model was applied in the present simulation and it can
240 be applied for different fuel classes to identify the behavior of soot formation and oxidation (Appel et al., 200).
241 The kinetic soot model can solve the 1850 gas phase reaction, 186 species and 100 heterogeneous reactions
242 with the participation of micro-heterogeneous particles of different types (AVL FIRE user manual V2013.1).
243 In this study, both Dukowicz model and multi-component evaporation model was applied subject to
244 type of fuel involved in the combustion process (Liu et al., 1993). The utility of spray, NO and Soot models are
245 also present in research to define NO and Soot formation in diesel engine through AVL FIRE simulation code
246 (Petranovic et al., 2015).
247 The exactness of obtained simulation data was compared with observational data of Kirloskar single
248 cylinder diesel engine at partial and full load conditions. Fig 4 shows the validation of the in-cylinder pressure
249 and heat release rate (HRR) against the experimental data of the pressure and heat release rate for the same
250 engine (Soni et al., 2016). It can be corroborated that, the results of simulation and performed experiment
251 match reasonably. Although, qualitative agreement between predicted and experimental results is found to be
252 good. But some discrepancies were observed particularly at TDC. This may attributed to the uncertainty of
253 input parameter such as injection duration.
254
255 4. Result and discussion
256 Mainly the combustion process is non- premixed in the hemispherical and re-entrant combustion
257 chamber. Thus atomization and the diffusion of spray droplets depend on initial air motion inside the
258 combustion chamber. The cavity inside combustion chamber is drawn by aiming the sufficient swirl and better
259 air/fuel mixing. The curved bowl of a piston may not be fully utilized by a fixed spray angle. The change in
260 spray angles may be helpful by using more circumferential momentum, leads to better air/fuel mixing. High
261 circumferential momentum will enhance swirl motion inside a combustion chamber which can be measured in
262 terms of angular swirl velocity, 𝜔𝑠𝑥 (Mattarelli et al., 2004). The swirl angular velocity is described in the
263 following way:
mi {(yi ‒ y0)wi ‒ (zi ‒ z0)vi}
264
n
ωsx=∑i = 1 [ mi{(zi ‒ z0)2 ‒ (yi ‒ y0)2} ] …… (15)

265
266 Where, n and mi is total no of cells and mass inside a cell in a computational design respectively. The Cartesian
267 coordinates are defined as xi, yi and zi while the cylinder axis is represented by x0 and y0. The velocity
ACCEPTED MANUSCRIPT
11

268 components are expressed as vi and wi in the vector direction of y and z for a computational cell respectively.
269 Furthermore, swirl ratio is defined as the ratio of swirl angular velocity and engine speed. The quantitative
270 analysis of swirl motion is depicted that, the Re-entrant piston bowl geometry has high swirl motion than the
271 hemispherical piston swirl motion. It is related to the entrapped swirl air by Re-entrant geometry, while
272 hemispherical bowl piston geometry lets trapped air escape from the bowl. Air escaped due to a wide passage
273 of the hemispherical bowl, leads in lower swirl motion.
274 The quality of fuel/air mixing inside the combustion chamber is expressed as “equivalence ratio”. Fig
275 5 shows the variation in fuel/air equivalence ratio for hemispherical and re-entrant bowl piston geometries with
276 the change of crank angle. In case of hemispherical piston, spray distribution of 160° spray angle is consistent
277 compared to other spray angles in the combustion chamber. The spray is rotating clockwise and moving upside
278 in the combustion chamber, when the piston moves from TDC (Top dead center) to BDC (Bottom dead
279 center); meanwhile, equivalence ratio is controlled by swirl and squish. The striking point of sprays varies
280 from mid-section to the upper section (edge) of the chamber while spray angle changes from 120° to 160°
281 respectively. The spray angle with 160° is utilizing more circumferential area to travel towards the bottom of
282 the piston. It results to better use of air inside the chamber and forms homogeneous mixture. It can be seen
283 that, during the piston movement from TDC to 5°CA ATDC (Crank Angle after Top dead center), the
284 accumulation of air fuel mixture at the bottom of the piston is comparatively less at 160° spray angle.
285 Negligible amount of the fuel rich region is observed with 160° spray angle except at the top of the piston
286 (clearance area). The spray penetration of 160° spray angle is less in the upper direction and well distributed
287 throughout the circumference of the piston at 20° CA ATDC; which leads to make sure the best use of the
288 surrounding temperature. The same trend can be observed at 10°CA ATDC, in which accumulation of spray is
289 high at the bottom of the piston for 120° and 140° spray angle, and results in maximum equivalence ratio of
290 3.02 and 2.69 respectively. At the same time, 160° spray angle introduces the uniform charge in the
291 combustion chamber and it is away from the center of the bottom of the piston with a comparatively lower
292 equivalence ratio of 1.98.
293 On the other hand, Re-entrant combustion chamber geometry enhances the swirl speed because of the
294 more circular area at the bottom of the combustion chamber, which is also known as a swirl chamber. The
295 swirl chamber enhances the angular momentum of the charge. This enhancement results in better air fuel
296 mixing inside the swirl chamber. It can be seen from fig 5 that, the high striking point of 160° spray angle
297 results in improved angular momentum and better air fuel mixing, while piston moves from TDC to 5°CA
298 ATDC. While, 120° and 140°spray angles are striking well inside the swirl chamber; it leads to accumulation
299 of spray inside the swirl chamber and possesses less circumferential momentum at 5°CA ATDC. A very fine
300 spray rotation can be observed for 160° spray angle than other spray angles and the same result is observed by
301 Wei for 146° spray angle (Wei et al., 2014), when compare with 150° and 154° spray angle. As crank rotates
ACCEPTED MANUSCRIPT
12

302 from 5°CA ATDC to 10°CA ATDC, 160° spray angle is more homogeneous and most of the charge is close to
303 the guided wall of the swirl chamber. At 10° CA ATDC fuel rich zone of 120°spray angle is collected at the
304 bottom of the swirl chamber and its maximum equivalence ratio is 2.30; whereas spray distribution is uniform
305 in case of 140° and 160° spray angles. The equivalence ratio of 160° spray angle is 2.04, comparatively less
306 than the equivalence ratio of 140° spray angle i.e. 2.12.
307

Hemispherical bowl piston geometry


120° Spray angle 140° Spray angle 160° Spray angle
At TDC

𝟓°ATDC

𝟏𝟎°ATDC

Re-entrant bowl piston geometry


At TDC

𝟓°ATDC
ACCEPTED MANUSCRIPT
13

𝟏𝟎°ATDC

308
309 Fig 5. Comparison of fuel/air Equivalence ratio at different spray angles for both geometries.
310 Fig 6 shows that, the high temperature zone of the hemispherical piston bowl at 160°spray angle is
311 compact and close to the top of the piston whereas temperature regions of other two spray angles are staggered.
312 It indicates the loss of the distribution tendency of spray around the perimeter of the combustion chamber. This
313 effect can be observed at 5° CA ATDC; where, high temperature zones of 120° and 140° spray angles are
314 pointing towards the space of the combustion area and losing the contact from the circumference. This effect is
315 less in case of 160° spray angle than other two spray angle. It has discussed that, the 160° spray angle is
316 travelling more circumferential area compared to other two spray angles. Its high temperature zone is mostly
317 gathered around the circumference of the piston and clearance area. To boot, it has wider temperature region
318 than other spray angles at 10°CA ATDC. The same trend is obtained by Wei for 146° spray angle (Wei et al.,
319 2014). The temperature distribution of spray angles is different while piston moves from 5°CA ATDC to
320 10°CA ATDC. However, high temperature section of different spray angles has same value, but they differ in
321 terms of distribution. Fig 6 demonstrates two high temperature zones of 160° spray angle and remains close to
322 the perimeter of the piston. Whereas, high temperature zones of other spray angles are shifted towards the
323 space of the combustion area. It can be seen that, the higher value of temperature for all spray angle exists near
324 the wall of the combustion chamber. What's more, the spray distribution is shifted to the space of the
325 combustion chamber; which leads to stretching of the high temperature area. This shifting of higher
326 temperature area results in lose their contact from the wall. The Stretching of high temperature area causes
327 reduction in temperature for 120° and 140° spray angles at 10° CA ATDC, though, high temperature zones of
328 160°spray angle stays on the circumference of the piston and leads to high temperature for 160° spray angle.
329 In the case of re-entrant piston bowl geometry, Fig 6 shows the temperature distribution of spray
330 angles at different position of the piston. It is discussed above that; 160° spray angle is utilizing more
331 circumferential area, while striking near TDC. Furthermore, 160° spray angle forms a very fine curve of
332 temperature distribution at 10° CA ATDC because of the even distribution of the charge at the circumference
333 of the swirl chamber. Additionally, most of the part of high temperature zones of 120° and 140° spray angles is
334 collected at the bottom of the swirl chamber as the piston moves from 5° CA ATDC to 10°CA ATDC, this is
335 mainly due to the less circumferential momentum of the charge. It can be observed from fig 8 that at 10° CA
336 ATDC, the strength of high temperature zones of 160° spray angle is changing towards the quad of the
337 combustion chamber and circulating in the clearance area as well. However, high temperature zones of 120°
ACCEPTED MANUSCRIPT
14

338 and 140°spray angle are mostly focused inside the swirl chamber. The impact of high temperature zones of
339 160° spray angle is more on the edge of the distribution curve comparison to the 120° and 140° spray angles. It
340 results in more penetration of the temperature distribution curve in the clearance area of the combustion
341 chamber, thus the temperature of the 160° spray angle is higher compared to the other spray angles.
342
343
344

Hemispherical bowl piston geometry


120° Spray angle 140° Spray angle 160° Spray angle
At TDC

𝟓°ATDC

𝟏𝟎°ATDC

Re-entrant bowl piston geometry


At TDC

𝟓°ATDC
ACCEPTED MANUSCRIPT
15

𝟏𝟎°ATDC

345
346 Fig 6. Temperature distribution of different spray angles for both geometries.
347
348 The velocity field of geometries at different position of crank angle is illustrated in fig 7 for proposed
349 spray angles. As seen in fig 7, the 160° spray angle forms big velocity field and tumble in the combustion
350 chamber, while 120° and 140° spray angle have dense velocity streamlines or vortex. The spread of fuel over
351 the circumference of the piston is more in 160° spray angle compared to other spray angle. This in turn
352 increases velocity of the fuel and forms bigger vortex.

Hemispherical bowl piston geometry


120° Spray angle 140° Spray angle 160° Spray angle
At TDC

𝟓°ATDC

𝟏𝟎°ATDC

Re-entrant bowl piston geometry


At TDC
ACCEPTED MANUSCRIPT
16

𝟓°ATDC

𝟏𝟎°ATDC

353
354 Fig 7. Velocity contours at different spray angles for both geometries.
355
356 Additionally, the rotation of velocity vectors is in the same direction i.e. clockwise. At 5° CA ATDC, 160°
357 spray angle forms homogeneous and strong vortex near the spray block for hemispherical geometry and near
358 the bowl for re-entrant piston geometry. But 120° and 140° spray angle form two weak vortexes inside the
359 combustion chamber and they observed on both sides of the spray block. The rotation of upper vortex is in
360 anticlockwise whereas the lower vortex is in a clockwise direction. This different rotation of vortexes creates
361 two contours of air /fuel mixing in the opposite direction. So, two zones of air/fuel mixing are formed in the
362 combustion chamber. The vortex, which placed above the spray block is responsible for weak air fuel mixing.
363 This weaker vortex is still there at 10° CA ATDC. Thus, it brings incomplete combustion in the cylinder and
364 increases soot formation.
365 Engine performance parameters are depicted in Table 4, and fig 8 shows the NO and soot emission at
366 different spray angles. NO mass fraction of 160° spray angle is higher for both types of the combustion
367 chamber, whereas soot mass fraction is lowest. Higher NO mass fraction is observed at high temperature
368 regions of both geometries and same trend is also found by Wei et al., 2014. The circumferential momentum of
369 120° spray angle is relatively lower in both the geometries. The less circumferential momentum leads to fuel
370 accumulation at the bottom; therefore, fuel makes less contact with air. This insufficient contact forms over
371 rich zones and tends to poor mixing inside cylinder; finally it increases the soot mass fraction. On the other
372 side, equivalence ratio of 160° spray angle is easily spreads and spray impinges at the edge of the both
373 combustion chambers, hence produces higher circumferential momentum. In addition to that, the charge is
374 moving towards the clearance area of the chamber, which results in better utilization of air and low soot mass
375 fraction. The maximum soot mass fraction follows the trend of 120°, 140° and 160° spray angle for both the
376 geometries, when the results are recorded at the last stage of the combustion.
377
ACCEPTED MANUSCRIPT
17

Soot mass fraction


NO mass fraction NO Soot
0.0005 0.00006
0.00005
0.0004
0.00004
0.0003
0.00003
0.0002
0.00002
0.0001
0.00001
0
0
120 140 160 120 140 160
120 140 160 120 140 160
Hemispherical bowl Re-entrant Hemispherical bowl Re-entrant
378 bowl bowl
379 (a) (b)
380 Fig 8. Emissions at different spray angle (a) NO mass fraction (b) Soot mass fraction.
381
382
383 Table 4 Performance parameters at different spray angle of both geometries.
120° Spray angle 140° Spray angle 160° Spray angle
Parameters Hemispherical Reentrant Hemispherical Reentrant Hemispherical Reentrant
bowl piston bowl piston bowl piston bowl piston bowl piston bowl piston
BSFC(g/kWh) 390 381 364 370 342 351
BMEP (bar) 2.21 2.13 2.50 2.41 2.66 2.59
Indicated 18.02 17.90 19.52 19.35 20.38 20
torque (Nm)
Indicated 2.83 2.75 3.07 2.95 3.20 3.02
power (kW)
Indicated 32 30.4 35 33.7 36 35
efficiency (%)
384
385 5. Conclusion

386 The re-entrant swirl chamber is assuring to improve combustion process and emissions at different
387 spray angles compared to the hemispherical bowl chamber. The comparison investigated numerically with
388 different spray angles by using an AVL FIRE commercial CFD code. The equivalence ratio, temperature and
389 velocity distribution of geometries are greatly tempted by the spray angles, which results to change in
390 combustion and emissions accordingly. It can be noted that, at 160° spray angle has better result compare to
391 other spray angles. Executions wise, hemispherical geometry is better than re-entrant geometry at all spray
392 angles, while the comparison is analyzed in terms of force, torque, efficiency, BMEP and BSFC. However,
393 emission parameter represents an opposite tendency. In case of emission parameters, hemispherical geometry
394 shows higher NO emissions than re-entrant geometry. The mathematical value of NO emission is less in case
395 of 120° and 140° spray angle than 160° spray angle. The 160° spray angle prepares homogeneous mixture
396 which leads to complete combustion. Therefore, the soot mass fraction gets reduced for 160° spray angle in
397 both the geometries. The value of the soot mass fraction is almost same for both the geometries at same
ACCEPTED MANUSCRIPT
18

398 operating conditions. Ultimately, it can be concluded that, re-entrant piston geometry is more reliable for the
399 reduction of NO emissions. Additionally; it can be used for soot mass fractions with 160° spray angle, but
400 marginal compromise has to be considered for performance parameters.
401
402
403
404
405
406
407

408

409

410 References:-
411 Tao Li., Zhi-Chao L., Hong-Chao Z., Qiu-Hong J., 2013. Environmental emissions and energy consumptions assessment
412 of a diesel engine from the life cycle perspective. Journal of Cleaner Production, 53: 7-
413 12.http://dx.doi.org/10.1016/j.jclepro.2013.04.034.
414 Jozef M., Ja´n C., Lˇudmila J´., Marek Bˇ., Andrea Kl´., 2010. Second generation diesel fuel from renewable sources.
415 Journal of Cleaner Production, 18:917–926. DOI: 10.1016/j.jclepro.2010.01.018.
416 Reitz R., 1995. Development and testing of diesel engine CFD models. Prog Energy Combust Sci 1995; 21:173–96.
417 http://dx.doi.org/10.1016/0360- 1285(95)00003-Z.
418 Choi CY., Reitz RD., 1999. An experimental study on the effects of oxygenated fuel blends and multiple injection
419 strategies on DI diesel engine emissions. Fuel 1999; 78:1303–17. http://dx.doi.org/10.1016/S0016-2361
420 (99)00058-7.
421 Fuster D., Agbaglah G., Josserand C., Popinet S., Zaleski S., 2009. Numerical simulation of droplets, bubbles and waves:
422 state of the art Fluid Dynamics Research 41; 065001.
423 Subramanian KA., Lahane S., 2013. Comparative evaluations of injection and spray characteristics of a diesel engine
424 using karanja biodiesel–diesel blends. Int J Energy Res, 37:582–97. DOI: 10.1002/er.1947.
425 Lahane S., Subramanian KA., 2015. Effect of different percentages of biodiesel–diesel blends on injection, spray,
426 combustion, performance, and emission characteristics of a diesel engine. Fuel, 139:537–
427 545.DOI:10.1016/j.fuel.2014.09.036.
428 Lahane S., Subramanian K.A., 2014. Impact of nozzle holes configuration on fuel spray, wall impingement and NOx
429 emission of a diesel engine for biodiesel diesel blend (B20). Applied Thermal Engineering 64: 307-
430 314.DOI:10.1016/j.applthermaleng.2013.12.048.
431 Gorji-Bandpy M., Soleimani S., Ganji DD., 2009. The Effect of Different Injection Strategies and Intake Conditions on
432 the Emissions Characteristics in a Diesel Engine. Hindawi Publishing Corporation. International Journal of
433 Vehicular Technology, Article ID 105363, 11 pages. DOI:10.1155/2009/105363.
ACCEPTED MANUSCRIPT
19

434 Seung HY., June PC., Chang SL., 2010. An investigation of the effects of spray angle and injection strategy on dimethyl
435 ether (DME) combustion and exhaust emission characteristics in a common-rail diesel engine. Fuel Processing
436 Technology, 91:1364–1372. http://dx.doi.org/10.1016/j.fuproc.2010.04.017.
437 Dolak J., Reitz R., 2011. Optimization of the piston geometry of a diesel engine using a two-spray-angle nozzle. Proc Inst
438 Mech Eng Pt D J Automobile Eng, 225:406–21.
439 Jaichandar S., Senthil KP., Annamalai K., 2012. Effects of open combustion chamber geometries on the performance of
440 pongamia biodiesel in a DI diesel engine. Fuel 2012; 98: 272–9.
441 Jaichandar S., Senthil KP., Annamalai K., 2012. Combined effect of injection timing and combustion chamber geometry
442 on the performance of a biodiesel fueled diesel engine. Energy; 47:388–94.
443 Li J., Yang WM., An H., Maghbouli A., Chou SK., 2014. Effects of piston bowl geometry on combustion and emission
444 characteristics of biodiesel fueled diesel engines. Fuel, 120:66–73. http://dx.doi.org/10.1016/j.fuel.2013.12.005.
445 Jesús B., José VP., Antonio G., Javier MS., 2015. An experimental investigation on the influence of piston bowl
446 geometry on RCCI performance and emissions in a heavy-duty engine. Energy Conversion and Management,
447 103:1019–1030. http://dx.doi.org/10.1016/j.enconman.2015.07.047.
448 Brandl F., Reverencic., Cartellieri W., Dent JC., 1979. Turbulent air flow in the combustion bowl of a DI diesel engine
449 and its effect on engine performance. SAE Paper 790040.DOI: 10.4271/790040.
450 Arcoumannis C., Bicen AF., Whitelaw JH., 1993.Squish and swirl–squish interaction in motored model engines. ASME J
451 Fluid Mech, 105–12.
452 Payri F., Benajes J., Margeo X., Gil A., 2004. CFD modeling of the in-cylinder flow in direct-injection diesel engine.
453 Comput Fluids, 33:995–1021.DOI:10.1016/j.compfluid.2003.09.003.
454 Saito T., Daisho Y., Uchida N., Ikeya N., 1996.Effects of combustion chamber geometry on diesel combustion. SAE
455 Paper 861186.DOI: 10.4271/861186.
456 Prasad BVVSU., Sharma CS., Anand TNC., Ravikrishna RV., 2011. High swirl-inducing piston bowls in small diesel
457 engines for emission reduction. Applied Energy, 88:2355–2367. DOI:10.1016/j.apenergy.2010.12.068.
458 Wei S., Kunpeng J., Xianyin L., Feihu W., Xin L., 2014. Numerical simulation on effects of spray angle in a swirl
459 chamber combustion system of DI (direct injection) diesel engines. Energy,
460 75:289e294.DOI:10.1016/j.energy.2014.07.076.
461 Heywood JB., 1988. Internal combustion engine fundamentals, NewYork, Mcgrawhill Book Company.
462 AVL FIRE user manual V2013.1.
463 Khaleghi H., Ebrahim F., 2003. Iterative and non-iterative solutions of engine flows using ASM & k–ε
464 turbulence models. In: Proceedings of 11th annual conference of the CFD society of Canada (CFDSC),
465 BC, Canada: Vancouver.
466 Versteeg HK., Malalasekera W., 1995 An introduction to computational fluid dynamics: the finite volume
467 method Essex: Pearson Education.
468 JH Ferziger JH., M Peric M., 2002. Computational methods for fluid dynamics. Berlin: Springer.
469 Van der Vost HA., 1992. Bi-CGSTAB: a fast and smoothy converging variant of Bi-CG for the solution of
470 non-symmetric linear system. SIAM J Sci Stat Comput; 13:631–44.
ACCEPTED MANUSCRIPT
20

471 Hanjalic K., Popovac M., Hadziabdic M., 2004. A robust near-wall elliptic-relaxation eddy-viscosity turbulence model for
472 CFD. Int J Heat Fluid Flow, 25:1047–51.DOI:10.1016/j.ijheatfluidflow.2004.07.005
473 Colin O., Ben KA., 2004. The 3-zones extended coherent flame model (ECFM-3Z) for computing premixed / diffusion
474 combustion. Oil & gas science and combustion, 59 (6), pp593-609.
475 http://dx .doi.org/10.2516/ogst:2004043.
476 Hasteed M., Kirsch L., Quinn C., 1977. The auto ignition of hydrocarbon fueled at high temperature and pressure fitting
477 of mathematical model. Combust flame, 30:45-60. DOI: 10.1016/0010-2180(77)90050-5.
478 Reitz RD., 1987. Modeling atomization processes in high-pressure vaporizing spray. Atomization Spray Techno 3:309–
479 37.
480 Uludogan A., Foster DE., Reitz RD., 1996. Modeling the effect of engine speed on the combustion process and emissions
481 in a DI diesel engine. In: SAE paper 962056.DOI:10.4271/962056.
482 Zeldovich YB., Sadovnikov PY., Frank KDA., 1947. Oxidation of nitrogen in combustion "Translated by M shelef,
483 Academy of science of USSR, Institute of chemical physics, Moscow-Leningrad, 1947.
484 Appel J., Bochorn H., Frenklanch M., 2000. Kinetic modeling of formation with detailed chemistry and physics: Laminar
485 premixed flames of C2 hydrocarbons. Combust flame 121 (1-2), 122-136. DOI: 10.1016/S0010-2180(99)00135-
486 2.
487 Liu AB., Reitz RD., 1993. Modeling the effects of drop and breakup on fuel spray. SAE pap No-930072.
488 DOI: 10.4271/930072.
489 Petronavic Z., Milan V., Neven D., 2015. Towards a more sustainable transport sector by numerical simulating fuel spray
490 and pollutant formation in diesel engines. Journal of Cleaner Production 88:272 – 279.
491 http://dx.doi.org/10.1016/j.jclepro.2014.09.004.
492 Soni DK., Gupta R., 2016. Numerical investigation of emission reduction techniques applied on methanol
493 blended diesel engine, Alexandria Eng. J. (2016), http://dx.doi.org/10.1016/j.aej.2016.02.019
494 Mattarelli E., Borghi M., Balestrazzi D., Fontanesi S., 2004. The influence of swirl control strategies on the
495 intake flow in four valve HSDI diesel engines. SAE paper 2004-01-0112.
496
497

You might also like