chapter 2 literature review and...
TRANSCRIPT
6
CHAPTER 2
LITERATURE REVIEW AND OBJECTIVES
2.1 Introduction
In the recent past, the demand for diesel engines has increased rapidly. This is
mainly because of their higher thermal efficiency, better performance and reliability.
In the earlier days diesel engines were considered to be high pollutants than petrol
engines. But with the continuous improvements in the technology, there is a
considerable reduction in the emission levels in diesel engines. Still the research
works are continuing to bring down the levels of emissions; simultaneously the efforts
are continuing in the direction of improving the overall engine performance. In DI
diesel engines the fuel is injected directly into the combustion chamber. Here the
piston crown is a part of the combustion chamber. Fuel atomization, vaporization and
mixing of fuel and air occurs in a rapid sequence within the combustion chamber in
the fraction of a second. This can be achieved with the following: good combustion
chamber, appropriate in-cylinder air motion and fuel injection arrangement. Swirl is
mainly used for getting the adequate fuel- air mixing rates. Air swirl is generated with
the support of a suitable inlet port and it is amplified at the end of the compression
stroke by forcing the air towards the cylinder axis into the bowl-in-piston combustion
chamber. Swirl is basically an organized rotation of air about the cylinder axis.
Though some decay of swirl occurs due to presence of friction during the cycle,
intake generated swirl persists throughout the compression process as well as in the
combustion and expansion processes. The nature of the swirling flow in an actual
engine is extremely difficult to determine. Accordingly, steady state tests are often
used to characterize the swirl. Swirl ratio is defined as the solid-body rotating flow,
7
which has equal angular momentum to the actual flow, divided by crankshaft angular
speed.
The diesel engine combustion is heterogeneous in nature. An improvement in
the combustion process definitely increases its efficiency and reduces the pollutant
formation. Better understanding of the engine in-cylinder fluid dynamics, fuel spray
behaviour, induction generated swirl and the combustion will definitely be helpful in
meeting the great challenges such as fuel economy and pollutant formation. The
sprays and combustion phenomenon in DI diesel engines have been highly influenced
by the induction generated swirl and the number of holes present in the nozzle. Hence,
critical review of the available literature related to the in-cylinder flows, fuel spray
characteristics, induction generated swirls and multi-hole nozzle are presented in the
following section.
2.2 Fluid flows in I.C. engines
Fluid dynamics is a field of science which studies the physical laws governing
the flow of fluids under various conditions. Computational fluid dynamics is used to
generate flow simulations with the help of computers. CFD involves the solution of
the governing laws of fluid dynamics numerically. The complex set of partial
differential equations is solved in geometrical domain divided into small volumes,
commonly known as a mesh or grid. CFD has enabled us to understand the
happenings in the world in new ways. CFD enables analysts to simulate and
understand fluid flows without the help of instruments for measuring various flow
variables at desired locations.
The internal combustion Engine flow modeling is one of the most
challenging fluid dynamics problem, as it is associated with the compressible flow,
density variations, turbulent, cyclic with spatial and temporal perceptions etc.,. The
8
combustion phenomenon is greatly influenced by the fuel-air mixture formation and
the distribution of fuel within the combustion cylinder.
Pearson et al., (1990) [1] have reported computationally efficient simulation
technique. This technique is based on the linearised one-dimensional conservation
equations. These equations are suitable for distributed parameter systems and are
suitable to the requirements of the designer in assessing the relative merits among
different types of manifold configurations. Volumetric efficiencies of measured and
predicted are compared to understand the importance of variable geometry induction
systems.
Aita S et al., (1991) [2] have reported the analysis for the flow in an intake
port-valve-cylinder assembly of a DI diesel engine. The simulation was carried for
both steady state and transient motored situations during the suction and compression
strokes. Generation of angular momentum flux and the induced in-cylinder flow
motion were predicted for a helical port under steady state condition. The predicted
results were correlated and compared with the experimental results. The experimental
results were extracted with the support of oil film visualizations on valve and intake
port surfaces and attached with the local velocity measurements in the cylinder.
The transient flow simulations showed different characteristics of flow motion
in-cylinder and piston bowl during suction and compression processes. It was
reported that the swirl generating capacity of the valve is not the same during the
periods of valve opening and closing. A strong interaction was observed between the
swirling motion and the position & shape of the piston bowl.
Sweetland et al., (1994) [3] have used particle image velocimetry (PIV) to extract
gas velocity and turbulence in a diesel engine. Experiments were conducted on a
single-cylinder caterpillar engine. Optical access was provided for the combustion
9
chamber. The results obtained for the turbulent jet were compared with that of earlier
results obtained through other techniques. The estimates of the turbulence intensity
are obtained from PIV data. The length scales are estimated from the model relations
of k-ε and the turbulent dissipation. The size of vorticity concentrations and the eddies
that are obtained from PIV are reported to be the representatives of the turbulence
integral length scale. The experimental results were also compared with the results
obtained from multidimensional KIVA-3 code. The experimental and the predicted
results are observed to be in good agreement.
Sebastian et al., (1995) [4] have simulated a production engine at part and full
load conditions. Modified engine was analyzed at full load. SPEED CFD code was
used for the analysis. The fuel-air mixing and combustion process are visualized with
the support of the iso-surfaces of stoichiometric mixture. The correlation of this
surface with global quantities such as heat release, pressure, temperature and swirl
ratio was considered. The global properties that are presented here are resolved for
the main chamber and the swirl chamber separately. The formation of thermal NO
and soot are simulated and analyzed.
Kang et al., (1995) [5] have performed in-cylinder flow simulations. KIVA-3
code was used for the analysis. The valve flow conditions that were measured from
the experiments was used as an inflow boundary conditions. The predicted swirl
ratios in a steady flow environment were compared with the swirl ratios of an
operating engine. Some differences were noticed in 3-D flow structures between a
steady flow environment and the operating engine. Axial development of the flow
produced an organized swirl in the case of a steady flow rig. Whereas merging of
multiple swirls and tumble motion were observed in the case of an operating engine.
10
Landress et al., ( 1996) [6] have used the KIVA-3 CFD code to simulate the
flow through the port and the engine cylinder. From the results it was reported that,
mixing of the intake and residual gases is not uniform. Many complex flow structure
developments are happening during intake and they are getting destroyed during
compression. The flow field near TDC exhibits spatial in homogeneities in
temperature, kinetic energy and its dissipation rate.
Taylor et al., (1997) [7] have developed a computational methodology (3D
model) for predicting the losses in the intake regions of IC engines. In order to get
accurate results, the following tasks were implemented in the present methodology,
they are: (i) appropriate modelling of flow physics, (ii) quality of the geometry,
(iii) discretization schemes applied at low viscosity regions and (iv) turbulence of
higher order. This methodology was tested and validated against the data of a variety
of complex 2D and 3D laminar and turbulent flow conditions. The predicted pressure
losses in the intake region of a caterpillar diesel engine are compared with the
experimental data. The analysis was carried out in detail that could describe the
locations of the loss pockets that are associated mechanisms which are contributing
for the losses and the other sources for the losses. Simulations were carried out for
large scale, viscous and for all the turbulent flow situations. From the results it was
observed that there is good agreement between the predicted and the measured values.
Fuchs et al., (1998) [8] have simulated the suction, compression and
combustion processes of a caterpillar diesel engine. KIVA-3 CFD code was used for
the analysis. Seven variations on intake and two injection schemes were considered
during the study. From the results it is observed that, the combustion and emission
behaviour was influenced by one of the following three factors i.e., swirl ratio,
temperature and turbulence.
11
Barths et al., (1998) [9] have studied the comprehensive chemical
mechanisms. This includes all the relevant chemical combustion processes that take
place in a DI diesel engine during auto-ignition, the burnout in the partially premixed
phase, diffusive burning and formation of pollutants. The complete structure of the
flame is preserved without simplifying highly nonlinear dependencies of the
chemistry. Using the representative interactive flamelet model the one-dimensional
unsteady set of partial differential equations is solved online with the 3-D CFD code.
Applying this model pollutant formation in acetane fuelled Volkswagen DI 1900
diesel engine was investigated. It was shown that the soot emissions are primarily
controlled by the mixing process in the cylinder. Numerical simulations for different
injection rates are compared with in house experiments.
Chen et al., (1998) [10] have simulated the engine flow with the support of a
STAR-CD CFD code. Simulation was performed for the inlet port and combustion
chamber flow fields. Volkswagen DI diesel engine having two-valves was used for
the study. The predicted results were compared with the results that are obtained from
laser-doppler anemometer measurements. The results were extracted and compared
for the three periods: valve opening, valve closing and maximum valve lift periods.
The results that are predicted for different engine speeds are validated against the
measured data. The accuracy of the predictions are reported.
Okazaki et al., (1999) [11] have studied exclusively on the design of intake
and combustion systems. Computational fluid dynamics approach was used for the
study. Main attention was on the correlation of the trapping efficiency and the swirl
ratio. The data that was obtained from the engine operation and the steady flow rig
tests were combined in the process of quality improvement. The requirements of the
designers as well as the researchers were satisfied.
12
Baby et al., (1997) [12] have carried out an experimental investigation on in-
cylinder motion, during the intake and compression strokes of a multi-valve engine.
Experiments were conducted on a single cylinder four valve research engine. The
engine was attached with several optical accesses on cylinder liner and cylinder head.
laser doppler velocimetry was used for the measurement of the turbulence and local
velocity in combustion chamber. Effects of different bowl shapes on turbulence, flow
variations, and tumble distortion were analysed. Study was extended to evaluate the
effect of bowl location on the tumble charge angular momentum.
Shenghua et al., (1999) [13] have developed a model based on the Hiroyasu's
multi zone combustion model. Nozzle injection (spray) parameters, induction swirl,
air and fuel composition was considered in the model. Sub models pertaining to zone
velocity, air entrainment rate, droplet evaporation rate, combustion rate etc., were
taken from the latest literature. The model simulation was used to extract the
parameters like cylinder pressure, heat release rate and emissions (NOx and soot). The
predicted results of zone velocity and spray tip penetration are compared with the
predictions that are reported by Hiroyasu. The predicted results showed good
agreement with the experimental data.
Chiu (2000) [14] has consolidated the theoretical accomplishments in droplets
and sprays in the twentieth century, with an emphasis on the evolution of scientific
concepts, paradigms and methodologies. A structural spray theory, which was
developed form an early view of isolated droplets, has evolved in to a new view that
the interaction droplet and micro-scale structures and clusters of many-systems are
also fundamental entities in practical sprays. Outstanding issues and critical
bottlenecks that have prevented further advancement of the existing analytical theory
of droplet physics are examined, and an emerging research trend in a unified theory of
13
droplet phenomena was discussed. Recent accomplishments and future prospects of a
unified theory are presented to coupling the status of this special branch of droplet
science and its future application.
Randall et al., (2005) [15] have studied numerically the quasi-steady
vaporization and combustion of multiple-droplet arrays. Vaporization rates, shapes of
the flame surface and flame locations were found for different fuels and droplet array
configurations. The number of droplets, the droplet arrangement within the arrays
and the droplet spacing within the arrays are varied to determine the effects of these
parameters. It is reported that the droplet interactions, the number of droplets and
relative droplet spacing will affect significantly the vaporization rate of droplets
within the array and consequently the flame shape and its location are also affected.
For small droplet spacing, the droplet vaporization rate decreases below that obtained
for an isolated droplet by several orders of magnitude. Similarity parameter, which
correlates vaporization rates with array size and spacing was reported. Individual
droplet flames, internal group combustion, and external group combustion were
reported that they are depending on the boundary conditions and droplet geometry.
Anand Kumar et al., (2005) [16] have used KIVA CFD code to analyze the in-
cylinder flow of a four stroke gas engine. The simulation results were compared with
the available experimental results. The effect of different motoring speeds on the
intake-generated turbulence, mass flow rate, velocity, swirl ratio and TKE were
analyzed.
Alfred et al., (2005) [17] have investigated the influence of in-cylinder swirl
patterns along the axis of the cylinder that are obtained during compression on flow
velocities and generated TKE at TDC. In-cylinder flow velocity field is predicted
during induction and compression periods. The analysis was carried out using multi-
14
dimensional modeling. It is observed that the swirl ratio, swirl located near the piston
surface at early compression resulted in higher magnitudes of TKE and swirl inside
the bowl.
Kuleshov (2005) [18] has developed a model that deals with the diesel sprays
and combustion. Submodels that are capable to predict the emissions like NO and soot
have been implemented in the model. The model was used to handle the parameters
like injection strategies, injection direction, swirl intensity, droplet size and the bowl
shape. Inorder to predict the appropriate evaporation rates, nusselt number for the
diffusion process, pressure and temperatures were used. There was observed a good
agreement between the experimental and the predicted results.
Ravindra Aglave (2007) [19] has used KIVA III CFD code. A modified k-
model was used for in-cylinder turbulence. Discrete droplet model was used along
with the sub-models to predict collision, breakup and evaporation. There was
observed establishment of the suitability of intrinsic low-dimensional manifold in
simulations. Their simulation was considering turbulence-chemistry interactions using
a presumed PDF approach with greater accuracy in predicting kinetically controlled
process.
Janakiraman et al., (2007) [20] have presented the importance of combustion
process and stated that the rise in pressure was controlled by combustion process. And
the combustion process in turn was controlled by parameters like injection timing,
compression ratio etc., The parameters like pressure, emissions such as CO, CO2,
NOx, HC and soot against crank angle that are measured from the experiments.
Artificial neural networks technique was used for developing correlations. For this
purpose a feed forward back propagation neural network is used. The network is
15
optimized by varying the number of hidden layers, number of hidden layer neurons,
activation functions and training algorithms.
Pacaud et al., (2009) [21] have worked with HSDI diesel engine on the
following areas: (i) conducting tests at low temperatures (ii) CFD analysis for cold
start operation. The results that are obtained from the tests are conducted at low
compression ratios were compared with the tests of normal compression ratios. From
this comparison it was stated that the delay can be reduced significantly in the case of
a low compression ratio when compared with that of normal compression ratio case.
In addition both the simulation and experimental analysis were carried out in cold
ambient conditions. Correlations between experiments and predictions gave consistent
explanations justifying cold start mechanisms and first order phenomena concerning
to vaporization and combustion.
Errico et al., (2007) [22] have analyzed two different approaches on
complexities associated with diesel combustion. The analysis was carried out with the
support of an open source code called open FOAM. This model was having the sub-
models namely: the eddy dissipation model and the partially stirred reactor model and
models related to emissions. The predicted results especially the cylinder pressure and
the pollutant concentrations were compared with the experimental results.
Liang et al., (2010) [23] have aimed at reducing the gap between the detailed
chemistry associated with the hydrocarbon fuel combustion and the predictions from
computational fluid dynamics. The techniques that were included in the present study
are a surrogate blend optimizer and a guided mechanism reduction methodology. New
methods for coupling the pre-reduced kinetic models accurately with the
multidimensional transport equations were also included in the methodology. The
new methods include the algorithms pertaining to dynamic adaptive chemistry (DAC)
16
and dynamic cell clustering (DCC). These techniques are demonstrated by
determining a multi-component diesel fuel surrogate mechanism, reducing it as
appropriate for the conditions of interest, and then employing the reduced mechanism
in a multidimensional CFD calculation of diesel engine combustion. The CFD
simulation employs the newly developed FORTÉ simulation package, which was
designed to take advantage of the advanced chemistry solver methodologies as well as
advanced spray models. It was proved that the demonstrated techniques are highly
efficient and accurate.
Musu et al., (2010) [24] have introduced a new concept to control the HCCI
combustion. This was called as homogeneous charge progressive combustion
(HCPC). Here while the combustor piston moves in the downward direction the
compressor piston moves in the upward direction. Because of this the air was with the
compressor admitted to the combustor cylinder. At the same time the fuel was
injected into the transfer duct. Fuel evaporates and mixes with the air that brings the
conditions needed for homogeneous combustion. A detailed CFD study was carried
out on a new turbocharged HCPC engine and obtained clean combustion.
Wickman (2003) [25] has optimized the combustion chamber geometry design
of a HSDI diesel engine, for which KIVA-GA code was used. The optimum split-
spray piston design was demonstrated experimentally. This design was proved to be
able to achieve low emissions when using retarded injection timing, high swirl ratio,
high injection pressure, high boost pressure, high EGR rate, and a low compression
ratio.
Hideyuki et al., (2000) [26] have analyzed the fuel spray distribution in a DI
diesel engine. They used pilot and main fuel injections at different piston positions to
prevent the main fuel injection from hitting the pilot flame. From the CFD analysis it
17
is observed that the movement of the piston with a cavity divided by a central lip
along the center of the sidewall effectively separates the cores of the pilot and main
fuel sprays. Experiments showed that an ordinary cavity without the central lip
emitted increased the smoke. whereas smokeless, low NOx operation was realized
with a cavity divided by a central lip even at heavy loads.
Li et al., (2000) [27] have conducted CFD simulations to investigate the
combustion in a direct-inject (DI) diesel engine using the Ricardo engine CFD
program VECTIS and the Ricardo two-zone flamelet (RTZF) combustion model. The
simulation program covered full load and part load operating conditions, each
including 6 to 7 cases. CFD simulation results were compared against engine tests for
the in-cylinder pressures and NOx emissions. The comparison shows that the RTZF
combustion model performs well in all cases studied with no tuning of model
coefficients necessary. The detailed time history of spray, fuel distribution and flame
development obtained from the CFD simulation is proved to be an useful information
towards gaining a better understanding of the features of combustion in DI diesel
engines.
Bo et al., (2009) [28] have presented the CFD simulation that was performed
on a 4-cylinder in-line diesel engine. VECTIS CFD code was used for the study.
The simulation was run through multiple cycles and covered the intake, compression,
spray, combustion and exhaust processes. The simulation was run through multiple
cycles. The converged solution was compared with the engine test data for cylinder
pressure, charge distribution and EGR distribution.
Claywell et al., (2006) [29] have investigated on the intake design to gain
further insight into the nature of the airflow. Intake designs were classified with
reference to the physical layout. Then Ricardo software WAVE (1D) and VECTIS
18
(3D) were used to investigate the performance of the three most common intake
concepts as well as two variations of the base concepts for a naturally aspirated four-
cylinder Formula SAE engine. Simulations were carried out at three rpm using
WAVE and VECTIS coupled at the intake inlet and runner exits of each intake
concept. A 3D simplified CAD geometry of each intake was used for the VECTIS
part of the simulation. An unsteady flow analysis was used instead of a steady flow
analysis due to the nature of the flow within an engine. Intake performance is
determined by several factors: cylinder-to-cylinder volumetric efficiency, time of
choked flow in the restrictor, total pressure loss along the restrictor, sound spectrum
frequency content, and physical packaging characteristics. In addition, intake
manifold and restrictor interaction was discussed. Adverse flow conditions are
visualized. Packaging considerations such as the location of flow bends for minimum
pressure loss were also discussed. It was stated that the conical-spline intake concept
showed the best performance.
Li (2010) [30] has presented the latest development of his work on multiple-
cylinder engine CFD simulation using Ricardo's engine-focused commercial CFD
code, VECTIS. The detailed chemistry-based Ignition progress variable library (IPV-
library) approach was used for combustion modeling. The simulation was carried out
on a 4-cylinder in-line diesel engine. The converged solution was compared to the
engine test data. It was proved that with a detailed chemistry-based combustion
modelling scheme, the approach was having the capability to cover a broader range of
operating conditions and also different types of engine combustion. Additionally, in
terms of computational cost, the use of a combustion reaction library in the
framework of cell-based CFD analysis ensured great efficiency. It has been
demonstrated that multiple-cylinder engine CFD simulation with detailed chemistry-
19
based combustion modeling can be achieved with VECTIS on an affordable time
scale and computational cost.
Kini et al., (2008) [31] have made detailed study on the conventional mesh
generators. The CAD surface data available for mesh generation is far from
satisfactory for volume mesh creation. Dissatisfactions include no node-to-node
matching between mating parts, minute gaps, overlapping surfaces, overlapping parts,
etc. To clean up this kind of data to a level that can be used for volume mesh creation
requires a lot of manual work that could take a couple of weeks or more. The new
model that was developed presented a fast and fully automated, Cartesian cell
dominated projected mesh generation algorithm used in CFD-VisCART that
eliminated the need for CAD data cleaning, thus shaving off weeks worth of time off
the design cycle. This process of mesh generation involved the volume mesh cells
being generated first and then projected onto the surfaces of the CAD model in order
to create the surface mesh automatically. Additionally, the projection method has
intrinsic advantages when dealing with minute gaps and overlapping parts. Shrink-
wrapping (to cover up larger holes or simplify an assembly of parts), automatic hole
detection (to detect any mesh leaks), gap handling (to ensure proper mesh resolution
between neighboring parts) and such other tools together with the novel features of
the projection algorithm mentioned here have reduced the turnaround time for mesh
generation from 3–4 weeks to less than one day.
Golovitchev et al., (2000) [32] have proposed a new methodology that
coupled the generalized partially stirred reactor (PaSR) model with that of highly
efficient numeric to treat detailed oxidation kinetics of hydrocarbon fuels. In this
approach, chemical processes were assumed to proceed in two successive steps they
are: (1) the reaction follows after the micro-mixing is completed on a sub-grid scale,
20
(2) experimental observations on soot formation. KIVA-3 code was used to run the
simulation. The model was used to simulate the turbocharged Volvo DI diesel
engines. It was observed that the soot reduction effect attributed to a shortening of
fuel injection duration.
2.3 Literature on Spray Characteristics
Bo et al., (1997) [33] have used SPEED CFD code for the study. Three-
dimensional model was used for predicting the gas motion and spray characteristics.
Analysis was carried out on a small HSDI diesel engine. The complexities associated
to the engine geometry were included in the model by considering the implicit finite
volume method along with the unstructured mesh. Various sub models are also
incorporated for the model. The predicted results were compared with the results
obtained from LDA measurements, in particular the velocity field during the
induction and compression to the extent of ignition. Quantitative measurements of
spray penetration and local droplet velocities are reported to be in agreement.
Moon et al., (2007) [34] have made an attempt to investigate the spray
characteristics of a swirl injector for a direct injection spark ignition engines. A highly
inclined nozzle was used for the test. The obtained results were compared with the
conventional and the L-stip nozzle. An open hollow cone spray is observed for the 700
taper angle. The taper angle should be optimized to avoid the formation of rich areas
and to increase the spray volume. Improvement in the atomization was observed for a
high fuel temperature injection.
Hannouny et al., (2003) [35] have conducted an experimental and numerical
characterization. It was conducted for high-pressure fuel injection systems. One single
and one double guided multi-hole and valve-covered-orifice type of injector was used
with a common rail system. Mini-sac injectors two in numbers are used for hydraulic
21
electronic unit injection system (HEUI) with different orifice diameters. They have
made an extensive study on the effect of injection system and the operating conditions
on the emissions of a small HSDI diesel engine. A high-speed digital camera was
used to extract the images of the transient sprays. These images were used to obtain
spray tip penetration and cone angles. The droplet sizes averaged over the entire spray
(SMD) were extracted from the images using the light extinction method (LEM). The
KIVA code and the engine emissions measurements are used to correlate the spray
characteristics and the emissions, NOx and soot. The results emphasized the
correlation between the spray behavior and the emissions.
Schmidt et al., (2002) [36] have used the number time counter for droplet
collision algorithm with an embedded collision mesh. It was observed that this change
was able to reduce the grid dependency. The grid dependency was further reduced
with the support of a noval interpolation scheme that transforms the velocities into
polar components before doing interpolation. There was observed a considerable
decrease in grid dependency. All these changes have reduced the uncertainity in
droplet size and improved the spray penetration consistency.
Mohammadi et al., (1998) [37] have developed a single nano-spark back light
photography method in order to record the image of non-evaporating diesel sprays.
Diesel was injected into a high pressure nitrogen gas. From these images a clear
image of fine droplets and spray was obtained. In order to quantify the droplet
characteristics such as, droplet size and shape an image analysis method was
developed and implemented. Double-nano spark photography of diesel sprays was
carried out and relatively clear double exposure images of droplets were obtained on
the same film. Two dimensional size and velocity measurement of droplets were
simultaneously carried out based on these photographs. Distribution of droplets
22
velocity was clarified and used to describe the droplets distribution. The measurement
of two-dimensional velocity of ambient gas was done using a particle tracking
velocimetry (PTV) method.
Zhu et al., (2000) [38] have developed and presented A methodology for
coupling the fuel injection system and its effect on spray characteristics. The method
was applied to a case of a conventional pump-line-nozzle system. Mathematical
models that characterizes the flows from the pump to the nozzle are formulated and
solved using the method of characteristics and finite difference techniques. The nozzle
internal flow was modeled using zero-dimensional flow models, in which the nozzle
cavitation and its effect on the nozzle exit flow were taken into account. The models
are validated with available experimental data. The interaction between the upstream
flow and the nozzle flow, as well as the effects of nozzle flow on the spray
characteristics at the nozzle exit are predicted and analyzed.
Chang et al., (1997) [39] have investigated the effects of fuel viscosity and
also the effects of nozzle inlet configuration on the spray characteristics. Three
different viscosity fuels were used in the study. Two mini-sac six hole nozzles with
different inlet configurations were used. Spray was introduced into a constant volume
chamber using a common rail injector system. A high speed movie camera
synchronized with a pulsed copper vapor laser was used to capture the images of high
pressure transient sprays. The images were analyzed to obtain the spray tip
penetration and the spray cone angle at two regions. Overall SMD was calculated with
light extinction method.
Francisco et al., (1999) [40] have presented a model that deals with the
atomization and evaporation of transient sprays. The model was associated with the
following sub-models namely primary atomization, droplet evaporation and droplet
23
deceleration. The experimental data was extracted from the engine which was
equipped with a two component phase-doppler anemometer. The 3D fluid mechanics
code (PHONICS) was used for predicting the spray.
Jung et al., (2001) [41] have proposed a hybrid model that consists of
modified Taylor analogy break up (TAB) model and discrete vortex method (DVM).
The simulation process was divided into three steps. The first step was analyzing the
breakup of droplet of injected fuel using modified TAB model. The second step was
based on the theory of sieber liquid length. The liquid length analysis of injected fuel
was used for connecting both modified TAB model and DVM. The third step was to
reproduce the ambient gas flow and inner vortex flow injected fuel by using DVM. In
order to examine the hybrid model, experimental study was done on a free
evaporating fuel spray at early injection stage of in-cylinder like conditions. The
numerical results are compared with the experimental ones. The calculated results on
gas jet flow by DVM quantitatively in agreement with the experimental results
especially for the downstream of evaporative spray.
Xiaofeng et al., (2000) [42] have developed a relative velocity correction
model (RVC), combined with the drop tracing system and spherical coordinate
transformation that can work for a 3D case. The new model was implemented in
KIVA-3V code. The calculated results were compared with the experimental data for
both single-hole and three-hole fuel injections, including the spray tip penetrations
and the spray images. The comparison shows that the RVC model performed well for
all the cases.
Patterson et al., (1998) [43] have developed a new spray model to improve the
prediction of diesel engine combustion and emissions. The KIVA-II CFD code was
used for the analysis. The accuracy of modeling the spray breakup process was
24
improved by the inclusion of Rayleigh-Taylor accelerative instabilities, which are
calculated simultaneously with a Kelvin-Helmholtz wave model. This model
improved the prediction of the droplet sizes within a diesel spray and provided a more
accurate initial condition for the evaporation, combustion, and emissions models. An
improved droplet drag model was presented. The improved model accounts for the
increased droplet drag. This was due to the change in the droplet's shape and the
increased frontal area of the droplet. The drag model locally affects the breakup
process and produced a more realistic droplet size distribution. It was reported that it
was highly accurate while performing the vaporization process calculations. This
model was used for the prediction of pressure, heat release, and emissions produced
by single and split injection. New model was capable to predict the results accurately.
Allocca et al., (1992) [44] have studied on a non-evaporating transient high
pressure diesel spray operating under different ambient conditions. Tip penetration
and sauter mean diameter (SMD) were measured using the high speed photography
and the laser light extinction techniques. The simulations were carried out using
KIVA-II code with and without breakup sub-model. The effect of the grid spacing on
the numerical results was also evaluated. The KIVA –II simulations were
underestimating the jet penetrations. The SMDs that are obtained from simulations are
in disagreement with the experiments. The atomization model and the experimental
limitations were showed as the major reasons for the inaccuracies.
Xu et al., (1992) [45] have investigated the discharge coefficients of the
needle-seat opening passage and discharge hole in orifice-type diesel nozzles. Simple
empirical correlations were obtained between these coefficients and needle lift. These
relations were used while calculating the injection pressure. Based on the calculated
transient injection pressure, spray tip penetration was calculated by taking the overall
25
line which covers the trajectories of all fuel elements ejected during the injection
period. The results obtained from the calculations are validated by comparing them
with the results obtained from the wide range of experiments.
Arrègle et al., (1999) [46] have characterized the macro and microscopic
behaviour of diesel sprays generated by a common-rail system. The influence of
injection parameters and boundary conditions was quantified through a detailed
experimental study. The main purpose of this research was to validate and extend the
different correlations available in the literature pertaining to sprays generated by
common-rail systems, i.e., at high injection pressures with small nozzle holes. The
sprays are characterized in an environment which simulates the in-cylinder air density
existing in the real engine when the injection starts. A wide parametric study has
generated the data. This data was used to quantify the influence of the common-rail
pressure, nozzle hole diameter and gas density on the following: (i) spray tip
penetration, (ii) spray cone angle, (iii) geometric volume of the spray, (iv) spatial and
temporal evolution of drop size distribution. The results obtained from the theoretical
analysis are compared with the experimental results.
Stanislav Danov et al., (1999) [47] have developed a mathematical model of
combustion process in a diesel engine based on the theory of the chain reactions for
the higher hydrocarbon compounds. The rates of fuel vaporization and combustion are
defined by the instant values of temperature, pressure, concentration of fuel vapors,
overall diffusion rate, fuel injection rate, and mean fuel droplet size (SMD).
Numerical experiments have been carried out for investigating the interdependencies
between various combustion-related parameters. In specific, the effect of SMD on the
subsequent combustion parameters, such as, pressure, temperature, thermodynamic
properties of air/gas mixture, heat transfer, fuel vaporization, combustion rate, current
26
A/F ratio, gas mixture composition have been investigated. In addition, the integral
indicator parameters of the engine, such as the mean indicated pressure, peak
pressure, compression pressure have been analyzed.
Bianchi et al., (1999) [48] have presented a hybrid model. Their model was
capable to describe both primary and secondary breakup of high-dense high-pressure
sprays. They used the model proposed that was proposed by Huh and Gosman. Huh et
al., model was to compute the atomization of the liquid jet (primary breakup). The
TAB model of O'Rourke and Amsden was used to estimate the secondary breakup.
The atomization model considers the jet turbulence at the nozzle exit. The model also
considers the growth of unstable wave on the jet surface. Hybrid model was validated
against a free non-evaporating high-pressure-driven spray at engine like conditions.
The breakup time evaluation accuracy was improved by tuning the TAB
constant Ck. In the later part, sensitivity analysis of the grid resolution was carried out.
They found that the cell sizes used in common I.C. engine computations are not
sufficient to resolve the transient diesel-spray accurately. Influence of the turbulence
dispersion model on spray structure was discussed and introduced a new one.
Salah Addin et al., (2000) [49] have investigated on the impact of two models
representing two limiting cases of the droplet heat-up process. They are (i) infinite
diffusivity (ID) and (ii) diffusion limit (DL) model. These models work on the
evaporation, self-ignition and subsequent combustion in a diesel spray. The
simulation results showed that, when compared with the DL model, the ID model lead
to an over-prediction of the ignition delay by about 20%. This is attributed mainly to
the under-prediction of the evaporated mass predicted in the case of the ID model,
during the early stages after injection. The time evolution of the combustion processes
in the case of the ID model is observed to lag that was predicted with the DL.
27
Desantes et al., (2007) [50] have developed a model that simulates the diesel
spray dynamic behaviour. As a result of a theoretical reasoning based on momentum
flux conservation in the axial direction of the diesel spray, a mathematical model
which relates the momentum flux with profiles of velocity and concentration was
obtained. The new model was developed by considering the local density variations
and the deduction of the model for a generic schmidt number. A PDPA system and
additional measurements of spray momentum, mass flow rate and spray cone angle
were used in order to validate the model in high density environment and real
injection pressure conditions.
Bykov et al., (2007) [51] have developed a new decomposition technique for a
system of ordinary differential equations. It is based on the geometrical version of the
integral manifold method. The solution was obtained by comparing the values of the
right hand side of these equations, leading to the separation of the equations into ‘fast’
and ‘slow’ variables. The hierarchy of the decomposition was allowed to vary with
time. Equations for fast variables are solved by a stiff ODE system solver with the
slow variables taken at the beginning of the time step. The solution of the equations
for the slow variables was presented in a simplified form, assuming linearised
variation of these variables for the known time evolution of the fast variables. This
can be considered as the first order approximation for the fast manifold. This
technique was applied to analyze the explosion of a polydisperse spray of diesel fuel.
Clear cut advantages are demonstrated from the point of view of accuracy and CPU
efficiency when compared with the conventional approach widely used in CFD codes.
The difference between the solution of the full system of equations and the solution of
the decomposed system of equations was shown to be negligibly small for practical
28
applications. It was observed that in some of the cases the system of fast equations
was reduced to a single equation.
Sung et al., (2006) [52] have made an experimental study on the microscopic
and macroscopic breakup characteristics. Study also deals with the velocity and size
distributions of mono-dispersed droplets in relation to the breakup regimes. The
engine was provided a droplet generator equipped with a piezo stack produced mono-
dispersed droplets. The droplet-breakup phenomenon was captured in microscopic
and macroscopic views by using a spark lamp, a Nd-YAG laser, a long distance
microscope and a CCD camera as a function of the weber number. Along with the
analysis of the images, the droplet size and velocity distributions were measured in
the near nozzle region by a phase doppler particle analyzer system. It was showing the
size and velocity distributions of disintegrated droplets at the bag, stretching and
thinning, and catastrophic breakup regimes. In the bag breakup regime, the droplets
separated into small and large droplets during breakup. Alternatively, the droplets
disintegrated at a shorter duration and formed a cloud, similar to a fuel spray injected
through an injector, in those regimes.
Choongsik Bae et al., (2000) [53] have investigated the spray characteristics
from different holes of VCO (valve covered orifice) nozzles was performed and its
results were compared to standard sac nozzle. The global characteristics of spray,
including spray angle, spray tip penetration, and spray pattern were measured from
the spray images which were frozen by an instantaneous photography. A spark light
source and ICCD were used for the purpose. These spray images were acquired
sequentially from the first injection to fifth injection to investigate injection to
injection variation. For better understanding of spray development and their internal
structures, a long-distance microscope was used to get magnified spray images at the
29
vicinity of the nozzle hole with a laser sheet illumination. Also backward illuminated
images with a spark light source were taken to understand surface structures of dense
spray from VCO nozzle of common-rail injection system.
Laguitton et al., (2002) [54] have described how two high-speed video
cameras were utilised to achieve the pseudo 3-D imaging of the spray and of auto-
ignition sites. The usage of schlieren imaging that enabled vapour phase analysis was
also described. New correlations were established for the liquid and vapour
penetrations and the vapour phases. Attempts were continued by reducing the size of
the injector orifice. An experimental correlation against in-cylinder density for auto-
ignition delay was presented. An increased charge density reduced the penetration
where as it reduced the auto-ignition delay.
Sudhakar Das et al., (2003) [55] have reported the simulation of an outward
opening injector. The analysis was carried out using the modified KIVA-3V code.
The modified injection model was validated with the experimental data. The effect of
computational grid resolution was also reported. Using the validated code, a
parametric study of the effect of nozzle exit diameter, tangential velocity at the nozzle
exit and stream wise velocity on the injector characteristics such as spray penetration,
sauter mean diameter (SMD), and volume distribution (DV90) was carried out. A
strong effect of the tangential component of velocity (swirl) at the nozzle exit was
observed on the scattering (spread) of spray droplets.
Desantes et al., (2006) [56] have conducted research on the diesel spray
injected into chamber having stagnant ambient air. The investigation was continued
with an aim to perform an in-depth analysis on the influence of injection parameters
on the spray dynamics and spray macroscopic characteristics. As a result of
theoretical approach that is based on momentum flux conservation along the spray
30
axis. Model which predicts the spray axis velocity and spray tip penetration was
obtained. Measurements of momentum flux and spray cone angle are needed in order
to predict axis velocity and spray penetration. The chamber density was assumed to be
constant and equal to the density of the pressurized air inside the chamber. A
Gaussian radial profile was assumed for the axial velocity. Experimental results from
a conventional common rail injection system with five axi-symmetric nozzles were
obtained. Tests were carried out for a wide range of injection pressures and densities
to obtain additional information of the model and also for the validation purpose.
These experimental results include a large number of momentum fluxes (impact
force), spray tip penetration and spray cone angle measurements.
2.4 Literature on Swirl
Heywood (1988) [57] has stated that in medium and small DI diesel engines
swirl is used to attain adequate fuel- air mixing rates. Air swirl is generated by
adopting suitable changes in the design of inlet port. Friction at the cylinder wall
surfaces and the turbulent dissipation of the fluid tries to reduce the angular
momentum of the air entering during induction and the same is continued while the
compression process is in progress. Swirl velocity can be increased by adopting
suitable changes in the combustion chamber design.
Ferguson et al., (2004) [58] have tried to present the effect of design changes
on the moment of inertia. They could show that the swirl is proportional to angular
momentum and is inversely proportional to the moment of inertia. The bowl-in-
pistons increases the swirl. Increased swirl is observed to be reducing the moment of
inertia as the piston approaches TDC. It is found that, during compression as the
piston approaches TDC the swirl increases and starts decreasing after TDC more or
less with the same pace.
31
McCracken et al., (2001) [59] have proved that the optimal swirl level changes
as the geometry of the bowl-in-piston changes. Two different bowl-in-piston
combustion chambers were taken into account. Both the bowls were tested for the
same swirl levels. The trends are observed to be the same for both the bowl
geometries. There was observed quantitative differences in the air fuel mixing rate.
The turbulence generated by the jet is also influencing the air-fuel mixing rate.
Fuchs et al., (1998) [60] in their work it has been showed that, the use of bowl
in-piston chamber effects the in-cylinder motions. The high swirl ratios may distribute
the fuel such that it remains in the bowl. This gives a great scope to deplete almost all
of the bowl oxygen during combustion. It has produced stratification of the fuel and
air, and poor late diffusion burn.
Takahashi et al., (1987) [61] have considered different types of combustion
chambers in order to understand the influence of geometry on in-cylinder swirl. It was
observed that the combustion chamber of square type created turbulence at each
corner of the combustion chamber. The turbulence and swirl together contributed for
the enhancement of the air-fuel mixing rate. A flange lip was provided at the entrance
of the square type combustion chamber in order to increase air turbulence. Proper care
was taken to ensure that this turbulence is not permitted to penetrate the area of fuel
spray. It is observed that the lip-type of combustion chamber with high swirl might
affect the cold start. Certain modifications are essential to handle these situations.
They stated that the DI diesel engine performance is influenced by air swirl.
At low engine speeds adopting an air inlet port with high swirl ratio can improve
smoke levels and fuel consumption. But at high engine speeds high swirl port reduces
air flow co-efficient, that lowered the volumetric efficiency but it reduced the
combustion efficiency and the pumping losses were increased. It was suggested that
32
adopting a low swirl inlet port might overcome this situation. But at low engine
speeds and at low inlet swirl increased the smoke levels. In order to maintain the
appropriate fuel air ratios attempts were made by varying the swirl ratio to improve
the fuel consumption and to reduce the smoke levels. This was achieved by
introducing a sub port type variable swirl system in addition to the main inlet port.
Shimada et al., (1986) [62] have made attempts to develop a variable swirl
inlet system with the support of a developed swirl control sub port. Attempts were
made to control the angular momentum of the inlet flow into the cylinder. This
variable swirl system was used to investigate the effects of swirl on direct injection
diesel engine performance and emissions. It was observed that the lower swirl level
reduces the rate of initial stage burning that effects the NOx emission, maximum
cylinder pressure and the rate of cylinder pressure. Same is observed to happen over
the entire range of engine speeds and loads. An improvement in the BSFC and
reduction in the NOx levels were achieved. An intercooled turbocharged engine was
fitted with variable swirl inlet system. There was observed higher low-speed torque,
higher brake horse power and improved cold start. It was also observed that the earlier
conclusions are majorly influenced by the selection of optimum swirl that suits the
engine.
Beard et al., (1998) [63] have conducted experiments on various bowl shapes
such as flat, w-shaped and with or without re-entrant. It is observed that the shape of
the piston bowl is an important parameter to control the turbulence level and fuel air
mixing rates. A small change in the bowl shape has influenced the flow parameters
like swirl numbers and or turbulence intensity. This in turn modifies the combustion
efficiency. It was reported that in the re-entrant region of the w-bowl shape both the
33
swirl and the turbulence level increased at around TDC position when compared with
flat bowl piston.
Takabayashi et al., (2005) [64] have contributed for the development of
Honda accord 1.8 litre lean burn VTECH engine. This engine was analyzed using
Ricardo VECTIS. RTZF combustion model was used for simulating the combustion
process. The results showed that the analysis for both steady and unsteady in-cylinder
calculations is in good agreement with the earlier works. The intake port and piston
crown were optimized to improve the lean burn of air fuel ratio.
Li et al., (2006) [65] have conducted CFD simulations on a DI diesel engine
using Ricardo VECTIS software. It reports that the Ricardo two-zone flamelet
(RTZF) combustion model was used for analysis. Analysis was carried out on a small
HSDI research engine. The simulation continued for 6 cases forming a complete
injection time swing. The results predicted using the RTZF combustion model are in
good agreement with the experimental results.
Stone et al., (1992) [66] have tried to establish the significance of axial swirl
in diesel engines, and proved that the axial swirl is used in medium and small DI
diesel engines for improving the fuel and air mixture rate. Axial swirl contributes for
the generation of turbulence with the aid of interaction of the tangential velocity
because of the inward flow produced by the squish region. The presence of swirl takes
care of the cycle-by-cycle variation in the mean flow. This in turn controls the cycle-
by-cycle variation in combustion.
Dae Choi et al., (2004) [67] have made an attempt to measure the gas velocity
and multi-dimensional numerical simulation. This data was used for the evolution of
the structure of the turbulent flow within the cylinder. The study was carried out on a
bowl-in-piston engine of re-entrant type. The parameters that are contributing to
34
produce the turbulent energy are correlated and compared. The developed model was
assessed using the experimental data. It has been proved that the isotropic portion of
the normal stresses is the main source of turbulent energy for low swirl ratios. This
proved to be significant for all the swirl ratios. At higher swirl ratios the mean swirl
velocity is predominant at TDC. Squish contribution for the generation of turbulence
is due to the swirl velocity distribution. Increased swirl and the spatial distribution of
swirl velocity is appeared to be the effective method to increase the chamber
turbulence. The simulated mean flow fields are in good agreement with the measured.
Miles et al., (2003) [68] have made an attempt to measure both the radial and
the tangential velocity components in a HSDI diesel engine. Three different swirl
ratios and four injection pressures were taken into account. The study was made in the
absence of combustion. Length scales were estimated through Taylor’s hypothesis.
KIVA-3 code with the RNG k-ε turbulence model was used. The simulated and
experimental results indicated that the fuel injection event is the major source of late-
cycle turbulence and the flow structure can be a major source of turbulence at high
swirl ratios.
Magnus Sjoberg (2001) [69] have made an attempt to modify the fuel injector,
such that the fuel may be made to rotate around its axis. This is done to attain
sweeping injections. The sweeping injection enhances the air entrainment into the
spray. This is observed to be the reason for the reduced smoke level. The normal and
non-sweeping injection tends to build up fuel pockets, and the sprays hits the walls of
the piston bowl. Increasing the swirl ratio from 1.65 to 2.47 was not reducing spray
impingement onto the bowl wall impact. Reduction in the smoke levels are attained as
the swirl ratios were increased. It is found that the counter-swirl rotation of the
injector was reducing the impact onto the bowl wall. This has increased the
35
combustion rate and reduced the smoke levels. The fuel close to the bowl wall burns
slowly that raised the soot levels.
Dembinski et al., (2012) [70] have made an experimental investigation on
spray and mixture formation in a CI engine. An optical engine fitted with a high-
speed digital camera was used. Spray and mixture formation was varied by varying
the swirl, tumble and the injection pressure. Change in the swirl ratio was achieved by
changing the designs of the inlet port, the valve seat masking and the injection
pressure. This was attained by varying the designs of inlet port. In order to evaluate
the combustion process particle image velocimetry software was used. Images of soot
particles were also captured. The experimental results are observed to be well in
agreement with the simulation results. It was observed that the induced swirl was
surviving for the compression and combustion periods, where as the tumble was not
surviving up to the late combustion. The angular velocity is observed to be higher
close to the centre and reduces to the lowest at the outer piston bowl edge. An
increase in the injection pressure increases the deviation from solid body rotation.
Michal Dyer (1979) [71] have conducted experiments to validate computer
models of IC engine combustion. Tests were carried for a set of well characterized
and extensively varied experimental conditions, in order to evaluate the codes
pertaining to kinetics and fluid mechanics. In order to measure the thermodynamic
properties, the engine was attached with laser doppler velocimetry, pressure sensors
and thermocouples. High speed laser shadowgraph filming, laser Rayleigh scattering
and laser refraction techniques were used in the combustion diagnostic process.
Thermodynamic properties of the swirling, uniform pre-combustion mixture
are characterized by measuring pressure, temperature distribution, velocity/turbulence
36
and equivalence ratio. Experiments were carried out on a combustion bomb. The
experimental results were used to validate the computer code.
Gunasekaran et al., (2011) [72] have worked with engines that are arranged
with the shrouds on the intake valves in order to achieve air guided mixture
preparations. In this process, three types of flow structures were considered. They are
low tumble, high tumble and swirl. From the results it is observed that, the high
tumble with late injection provided a better mixture distribution.
Charton et al., (1996) [73] have studied experimentally and numerically on the
combustion of a swirling, stoichiometric and homogeneous mixture. Mixture contains
natural gas and air in a short cylinder. Swirl was induced by rotating a disc. In the
process of varying swirl intensities and turbulence, discs having varied roughnesses
and different disc speeds were considered. By doing so the effect of turbulence and
the induced swirl on combustion were studied. Pressure plot was used to examine the
combustion rates. In order to understand the effect of swirl on flame high speed
schlieren photography was used. It was observed that an increase in swirl intensity
initially reduced the burning span and on further rise in swirl intensity increased the
burning duration.
Vinay Nagaraju et al., (2009) [74] have made an attempt to investigate
experimentally on the effect of swirl ratio and injection pressure of a HSDI diesel
engine. The engine was equipped with a turbo charging facility, a swirl control
mechanism, common rail injection system and an EGR system. A blend of biodiesel
and low sulphur diesel fuel was used for the study. In order to understand the
combustible mixture formation and the auto ignition process the heat release rate was
taken into account.
37
Neal et al., (2012) [75] have made an attempt to study the effect of swirl ratio
on combustion in an optically-accessible diesel engine. Detailed study was made on
extended lift off combustion (ELOC). This is different from traditional combustion.
ELOC resulted in reduced soot levels. In order to visualize the impact of swirl ratio on
combustion, a High-speed imaging of OH chemiluminescence and natural luminosity
were used. Higher swirl ratios have showed longer curved path for the jets.
Ladommatos et al., (1992) [76] have described the three linked computational
models. These models are used for the estimation of the following; (i) swirl generated
during the induction process, (ii) changes in swirl in the case of a bowl-in-piston
combustion chambers during compression while the piston approaches top dead
centre (iii) the interaction of the fuel sprays with swirl, that includes the relative
crosswind velocities between the air and the fuel sprays and also the velocities during
spray impingement. Experimental results were extracted from a single-cylinder DI
diesel engine. Study was carried out by altering both the fuel spray and swirl
parameters systematically. The predicted spray impingement and crosswind velocities
are correlated with the results obtained from experiments. The comparison in
particular was for fuel economy and smoke emission.
Benny Paul and V. Ganesan (2010) [77] have studied on the effect of different
manifold configurations on air motion and turbulence inside a DI diesel engine. The
different manifold congurations such as, helical, spiral, and helical-spiral
combinations are considered here. GAMBIT was used for meshing the 3D model.
STAR-CD CFD code was used to predict the flow characteristics of these models.
The predicted results of swirl velocity at different locations at the end of compression
stroke are compared with the experimental results of the available literature. The
results obtained for helical manifold model showed reasonably good agreement with
38
the measured data given in the literature. RNG k-ε turbulence model was used. The
helical-spiral manifold is observed to give maximum swirl ratio inside the cylinder
than helical manifold.
Shenghua et al., (1999) [78] have developed a new swirl generator to induce
the induction swirl in the cylinder. In this process a short straight port swirl generator
was developed. The swirl generator was having several curvilinear diversion blades
and it was installed at the inlet valve. It is observed that, with the support of different
swirl generators it is possible to generate vary the generated swirl strengths within the
engine. Controlling the airflow rate it is showed that it is possible to generate variable
swirl. Engine tests proved that variably induced swirl has a considerable influence on
diesel engine fuel consumption.
2.5 Literature on Number of Nozzle Orifices
Campanella et al., (1994) [79] have carried out studies on a standard five-hole
V.C.O. nozzle of D.I diesel engine. Several pictures of the spray were taken using the
back-light photography technique. Injected volume, feeding pressure, injection
duration, spray penetration, angle and thickness, and flash delay from the start of
injection were measured both for nozzles of two different manufacturers and for a
group of four injectors before and after a fatigue test of about 1500 hours of running.
Measurements revealed that each spray strongly depends on the needle lift, the back-
pressure and the nozzle geometry. The causes of the most important differences have
been attributed at the different holes inclination, their asymmetric feeding conditions
(due to the eccentricity between needle and nozzle) and the micro defects caused by
drilling operation.
Das et al., (2009) [80] have described a correlation study on fuel spray pattern
recognition of multi-hole injectors for gasoline direct injection (GDI) engines. Spray
39
pattern is characterized by penetration length, which represents the distance of
maximum droplet concentration from the axis of the injector. Five fuel injectors with
different numbers and sizes of nozzle holes were considered in this study.
Experimental data and CFD modelling results were used separately to develop
regression models for spray patternation. These regressions predicted the influence of
a number of injector operating and design parameters, including injection system
operating pressure, valve lift, injector hole length-to-diameter ratio (l/d) and the
orientation of the injector hole. The regression correlations provided a good fit with
both experimental and CFD spray simulation results. Thus CFD offers a good
complement to experimental validation during development efforts to meet a desired
injector spray pattern.
Giorgio et al., (1995) [81] have studied on the spray of a 5-hole, VCO nozzle
for D.I. diesel engines. The data was obtained from each hole and are compared. A
high pressure injection system (elasis patent), electronically controlled, was used. To
investigate the behaviour of each hole a “cap” was used to cover the nozzle in such a
way that only one hole at a time could inject in the spray chamber. The investigation
consisted of a photographic characterization with back-light technique and of the
droplet size distribution through laser diffraction technique via Malvern. Four
pressure values (30, 60, 90, 120 MPa) and three injected flow rates (10, 25, 40
mm3/inj.) were chosen to characterize the behaviour of the nozzle. Each test was
carried out with the nozzle under normal running conditions, measuring the laser
signal background after each acquisition. The influence of the injection frequency, of
the number of the injections needed before running conditions, of the background
measurement and data acquisition were investigated. The investigation revealed that
the presence of an internal part of the spray made of drops of smaller diameter. The
40
different value of droplet size distribution in sprays, obtained from different holes but
working under the same conditions, have been attributed to the different hole
inclinations in respect to the injector axis. More inclined holes produced sprays of
droplet of bigger diameter.
Malbec et al., (2010) [82] have investigated on the air entrainment of multi-
hole Diesel injection using high speed particle image velocimetry (PIV). A multi hole
common rail injector with an injection pressure of 100MPa was used for the study.
The sprays are observed in a high pressure, high temperature cell that reproduces the
thermodynamic conditions which exist in the combustion chamber of a diesel engine
during injection. Typical ambient temperature of 800K and ambient density of
25kg/m3 are chosen. The air entrainment was studied with the PIV technique, giving
access to the velocity fields in the surrounding air and/or in the interior of two
neighboring jets. High acquisition rate of 5000 Hz, corresponding to 200 µs between
two consecutive image pairs was obtained by a high-speed camera coupled with a
high-speed Nd-YLF laser. The effect of neighbouring jets interaction was studied by
comparing four injectors with different numbers of holes (4, 6, 8 and 12) with similar
static mass flow rate per hole. The results show that both the maximum air
entrainment level and the total mass of entrained air are similar for all the injectors,
and therefore are not affected by neighboring jets in the conditions studied. However
the transient behavior of the air entrainment process is affected when the number of
holes is high, hence when two neighboring jets are near (12 holes nozzle): the air
entrainment reaches a maximum at later timings compared to the other nozzles. An
analysis of the velocity fields between the two jets shows that this result might be due
to the air flow inertia generated when the two jets are near. The transient behaviour
after the end of injection (EOI) was also studied using the 4-holes injector. The results
41
showed a very rapid decrease of the mean axial velocity near the nozzle after the EOI.
The air entrainment becomes maximum at a given position and this position
propagates downstream towards the jet tip.
Matthias et al., (2012) [83] have demonstrated collaborative 3D-CFD and
experimental efforts, majorly focused on optimizing the mixture stratification and the
potential for high engine efficiency with low NOx emissions. Performance of the
hydrogen engine was evaluated over a speed range from 1000 to 3000 rpm and a load
range from 1.7 to 14.3 bar BMEP.
Engine maps showed that the hydrogen direct injection engine operating above
35% brake thermal efficiency (BTE) over approximately 80% of the tested operating
range. A more detailed characterization of engine efficiency is done by quantifying
the effects of different loss mechanisms in the engine at relevant points throughout the
engine map. The dominant loss mechanism was heat loss to the combustion chamber
walls and as a function of both engine speed and load. There exists a trade-off
between wall heat losses and other partial losses. As a result, the peak BTE was
observed at 2000 rpm, 13.5 bars BMEP.
A series of engine maps showed efficiency improvements due to optimal
injection timing and also showed efficiency and NOx improvements due to injector
nozzle design. The most promising engine configuration uses a 4-hole nozzle which
showed improvement over the previous 5-hole nozzle. The final engine map showed a
peak BTE of 45.5% and part-load BTE of 33.3%, demonstrating the ability of the
hydrogen direct injection engine to exceed both U.S. DOE light-duty efficiency
targets. The 4-hole nozzle also provides a mixture that is less potent for NOx than the
5-hole nozzle which correlates to a significant decrease in NOx emissions at the peak
42
efficiency operating point. The corresponding map of NOx emissions was dominated
by less than 0.10 g/kwh of NOx.
Scarcelli et al., (2011) [84] have described the validation of a CFD code for
mixture preparation in a direct injection hydrogen-fueled engine. The cylinder
geometry is typical of passenger-car sized spark-ignited engines, with a centrally
located injector. A single-hole and a 13-hole nozzle are used at about 100 bar and 25
bar injection pressure. Numerical results from the commercial code fluent (v6.3.35)
are compared to measurements in an optically accessible engine. Quantitative planar
laser-induced fluorescence provides phase-locked images of the fuel mole-fraction,
while single-cycle visualization of the early jet penetration was achieved by a high-
speed schlieren technique. The characteristics of the computational grids are
discussed, especially for the near-nozzle region, where the jets are under-expanded.
Simulation of injection from the single-hole nozzle yields good agreement
between numerical and optical results in terms of jet penetration and overall
evolution. The 13-hole nozzle creates intense jet-to-jet interaction, with all jets
merging into a single effective jet immediately downstream of the under-expanded
region. This phenomenon (usually referred as coanda effect) is more challenging to
the numerical simulation and requires higher level of details in numerical simulation
and grid resolution, with particular regard to the fields near the injector nozzle.
Montgomery et al., (1996) [85] have conducted an experimental study on the
performance and emissions of an engine. Series of experiments were conducted on a
constant volume cold spray chamber. The investigation was carried out with a goal of
exploring the effects of number of holes and size on the emissions and performance of
a DI heavy duty diesel engine. The spray experiments provided insight into the spray
parameters and their role in the engine's combustion processes.
43
The fuel system used for both the engine and spray chamber experiments was
an electronically controlled, common rail injector. The injector nozzle hole size and
number combinations used in the experiments included 225X8 (225 gm diameter
holes with 8 holes in the nozzle), 260X6, 260X8, and 30OX6.
The engine tests were conducted on an instrumented single cylinder version of
the caterpillar 3400 series heavy duty diesel engine. Data was taken with the engine
running at 1600 RPM, 75% load. Engine emissions and performance results include
oxide of nitrogen emissions (NOx), particulate emissions, and brake specific fuel
consumption (BSFC). NOx versus particulate trade-off curves were generated over a
range of injection timings for each nozzle. The pressurized spray chamber was used in
the room temperature spray visualization experiments. Results of the image analysis
that give spray tip penetration length, spray cone angle, and droplet size are presented.
Inorder to understand the effect of nozzle geometry, computer modeling of the engine
was prepared. The KIVA-II code was used to model the combustion with the different
injector nozzle geometries. The results of the modeling effort emphasize the
importance of the details of the spray on diesel engine emissions.
Jung et al., (2011) [86] have implemented the Premixed compression ignition
(PCI) combustion using advanced injection strategy and exhaust gas recirculation in a
DI diesel engine single cylinder. The injection timing swept experiment using a
baseline injector, which had an injection angle of 146° and 8 nozzle holes, obtained
three types of combustion regime: conventional diesel combustion for an injection
timing of 10° CA (crank angle) BTDC (before top dead center), PCI combustion for
an injection timing of 40° CA BTDC and homogeneous charge compression ignition
(HCCI) combustion for an injection timing of 80° CA BTDC. The burn duration,
which was defined as the period from 10% to 90% of the accumulated heat release,
44
was very short in PCI combustion but not in the others. PCI combustion with an
injection timing of 40° CA BTDC was achieved in a range of an exhaust gas
recirculation (EGR) rate from 0% to around 40%.
Two types of different injectors were applied to investigate the effect of
injection angle and the number of nozzle holes on PCI combustion: one had an
injection angle of 70° and 8 nozzle holes, the other had an injection angle of 70° and
14 nozzle holes. These two injectors could implement PCI combustion as well. The
indicated mean effective pressure (IMEP) for both injectors with a narrow injection
angle (70°) was higher than that for the baseline injector because the injected fuel
could be direct toward the piston bowl so that most of the fuel could participate in the
combustion. The IMEP for the injector with 8 nozzle holes was higher than that for
the injector with 14 nozzle holes. On the other hand, when the injection angle was
70°, the injector with 14 nozzle holes had low levels of HC and CO emissions
because of better air utilization compared with the injector with 8 nozzle holes, which
was supported by a spray cone angle analysis. A maximum pressure rise rate (MPRR)
analysis showed that the MPRR for PCI combustion was higher than that for
conventional diesel combustion and HCCI combustion. The MPRR for the injectors
of a narrow injection angle in PCI combustion was higher than that for a baseline
injector. However, for the injectors of a narrow injection angle, there was no
difference in the MPRR of PCI combustion with respect to the number of nozzle
holes.
Kilic et al., (2006) [87] have evaluated the injection nozzles. A measuring
adapter for the determination of the injection quantities for individual spray hole was
developed. They stated that with the measuring adapter it is possible to determine the
jet flow injection quantities that were influenced by the variation in shape and the
45
geometry of the individual nozzle components. The measurements are reported to be
supported by simulation results. The spray hole shape was crucial for flow coefficient
and cavitation in the spray hole.
During fuel atomization the diameter and shape of the injecting holes and the
pattern realized by the nozzle needle in addition to the injection pressure play a
substantial role. Here was asserted that the smaller the spray hole diameter, the larger
the mass flow difference between the individual spray holes due to the increase in
variation in the manufacturing process. Macro- and micro-geometrical structure of the
spray hole surface and the transient areas of the nozzle interior into the spray hole has
a strong influence on the flow conditions within the range (area) of the nozzle needle
seat and within the spray holes and thus on flow values, flow symmetry and
cavitation. The cavitations starts earlier at higher inclination angles (that means the
flow is deflected more) and lower pressure differences.
Design parameters like the angle at which the individual spray hole comes off
the nozzle tip (inclination-angle), as well as manufacturing process driven parameters,
such as the shape of the spray hole inlet edge (grinding radius) influenced the quantity
of fuel flowing through the individual spray hole. The mass flow variability was
reduced by grinding the spray hole either for nozzle to nozzle or for spray hole to
spray hole.
Wehrfritz A et al., (2011) [88] have studied numerically the influence of the
number of fuel sprays in a single-cylinder diesel engine on mixing and combustion.
The CFD simulations were carried out for a heavy-duty diesel engine with an 8 holes
injector in the standard configuration. The fuel spray mass-flow rate was obtained
from 1D-simulations and was adjusted according to the number of nozzle holes to
keep the total injected fuel mass constant. Two cases concerning the modified mass-
46
flow rate are studied. In the first case the injection time was decreased whereas in the
second case the nozzle hole diameter was decreased. In both cases the nozzle holes
(i.e. fuel sprays) was increased in several steps to 18 holes. Quantitative analyses were
performed for the local air-fuel ratio, homogeneity of mixture distribution, heat
release rate and the resulting in-cylinder pressure. The results showed that an
increased number of fuel sprays leads to a more homogeneous fuel distribution, but
also to a more incomplete combustion.
Arcoumanis et al., (2000) [89] have made attempts to provide better
understanding on the cavitating flow characteristics in large scale nozzles. A six-hole
conical sac-type nozzle provided with a quartz window in one of the injection holes
has been used to visualize the flow under cavitating flow conditions. Simultaneous
variation of both the injection and the back chamber pressures were allowed. Images
are obtained at various cavitation and reynolds numbers for two different fixed needle
lifts corresponding to the first- and the second-stage lift of two-stage injectors. The
flow visualization system was based on a fast and high resolution CCD camera
equipped with high magnification lenses which allowed details of the various flow
regimes formed inside the injection hole to be identified. From the obtained images
hole cavitation initiated at the top inlet corner of the hole was identified. The string
cavitation formed inside the sac volume and entering into the hole from the bottom
corner, was also identified. Comparison of the cavitation images obtained in the real
size nozzle with those obtained in an enlarged fully transparent acrylic nozzle replica
has confirmed that similar flow regimes may form inside the injection hole of real size
and large scale injectors. Despite the similarity in the macroscopic cavitation
structures that were found to depend on cavitation intensity, the transient development
of cavitation bubbles was identified to be different inside the two nozzles due to the
47
different residence and life time of the moving bubbles. Based on the flow similarity
between the real size and enlarged multi-hole nozzles and previous extensive
experience on the cavitating flow characteristics in large scale nozzles, better
understanding of the cavitating flow in real size nozzles operating under Diesel engine
conditions has been obtained, which can guide the development of relevant computer
models.
Skogsberg et al., (2005) [90] have investigated on the ways of improving the
spray formation from spray-guided multi-hole gasoline direct injection injectors.
Work was carried out both experimentally using laser diagnostic tools and
numerically using computational fluid dynamics. Laser induced exciplex fluorescence
(LIEF) measurements in a constant pressure spray chamber and optical engine
measurements have showed that the injectors with 6-hole nozzles and 50-degree
umbrella angles are unsuitable for stratified combustion because they produce steep
air-fuel ratio gradients and create a spray with overly-deep liquid fuel penetration as
well as presence of liquid fuel around the spark plug. In order to study the injector
performance, numerical calculations using the AVL FIRE™ CFD code was used. The
numerical results indicated that, increasing the injector umbrella angle, the extent of
piston wall wetting can be decreased. Also, changing the pattern of holes in the nozzle
changes the spray pattern, enabling its optimization with respect to ignition and flame
propagation.
Furthermore, PDA and direct imaging experiments showed that increasing the
l/d ratio by reducing the hole diameter resulted in a decrease in the mean droplet sizes
(D32). The spray angle was found to increase with decreasing l/d ratios. It was
reported that by choosing a suitable l/d ratio, it is possible to control the local AFR
and cross-flow velocity at the spark plug.
48
Blessing et al., (2003) [91] have developed methods for analysing the effect of
individual nozzle configuration parameters on the fuel atomization and the fuel spray
propagation. These are observed to affect the raw emissions of NOx and soot in
modern DI diesel engines, apart from the fuel injection rate, atomization of the liquid
jet and mixing of the fuel with the combustion air. Optical diagnostics and CFD
methods were developed at the Daimler chrysler research. These methods are
combined with an analysis of the injection system hydraulics and are linked to a
detailed analysis of mixture formation and combustion inside an optically accessible
engine. The first part of the paper, methods for the experimental investigation with
transparent 1 and 6 holes nozzles in real size geometries under high pressure
conditions are described. Special emphasis was put on CFD methods for modeling the
cavitating two phase nozzle flow. In the second part the processes occurring in a sac
hole nozzle of a common rail injector during the complete injection event are
discussed. The influences of the spray hole position, inlet rounding and conical shape
of the spray hole are presented. Finally a comparison between a needle lift controlled
(CR=common rail) and a pressure controlled (PLN=pump line nozzle) injection
system was made.
Arcoumanis et al., (1999) [92] have used a one-dimensional, transient and
compressible flow model to simulate the flow and pressure distribution in advanced
high-pressure fuel injection systems; these include electronic distributor-type pumps
with either axial or radial plungers and a common-rail system. Experimental data for
the line pressure, needle lift, injection rate and total fuel injection quantity obtained
over a wide range of operating conditions (from idle to high speed/full load) were
used to validate the model. The FIE system used for validation comprised an
49
electronic high-pressure pump connected to two-stage injectors of different type
including 6 hole vertical and 5 hole inclined conical-sac and VCO nozzles.
A number of important parameters: (i) the injection pressure at the nozzle tip,
(ii) the effective hole area at the hole exit due to the presence of hole cavitation, (iii)
the fuel injection temperature, (iv) the hole-to-hole variability of inclined nozzles,
were predicted for three high pressure fuel injection systems. These parameters are
varied for different nozzles. It was shown that all important nozzle exit conditions for
the control of the subsequent spray development are strongly dependent on the
detailed geometric and operating characteristics of the injection system. It was
facilitated in the computer models.
Hottenbach et al., (2009) [93] have designed a new injection system for DI
Diesel engines to decrease the soot emissions. A nozzle with clustered holes was
observed to be a promising approach to minimize soot production. The basic idea of
the cluster configuration (CC) nozzles was to prevent a fuel rich area in the center of
the flame where most of the soot was produced, and to minimize the overall soot
formation. For this purpose each hole of a standard nozzle was replaced by two
smaller holes. The diameter of the smaller holes was chosen such that the flow rate of
all nozzles was equal. The basic strategy of the cluster nozzles was to provide a better
primary break up and therefore a better mixture formation caused by the smaller
nozzle holes. Three possible arrangements of the clustered holes were investigated.
Both the cluster angle and the orientation to the injector axis are varied. The common
rail diesel injector was installed in a combustion vessel, in order to provide nearly
quiescent high-pressure and high-temperature conditions.
The combustion and soot formation are analyzed using three different
measurement techniques. The hot reaction zone was visualized using OH*
50
chemiluminescence imaging which occurs during the second stage ignition. The local
soot concentration during combustion was measured semi-quantitatively using laser
induced incandescence (LII). In both measurements the soot luminosity was recorded
simultaneously using a second camera or a double frame camera, respectively. The
soot formation was discussed for all nozzles. The data indicated that soot formation
can be reduced using cluster nozzles under these conditions.
Brands et al., (2012) [94] have attempted to characterize the mixture
formation in the sprays emanating from multi-layer (ML) nozzles under
approximately engine-like conditions by quantitative, spatially, and temporally
resolved fuel-air ratio and temperature measurements. ML nozzles are cluster nozzles
which have more than one circle of orifices. They were introduced previously, in
order to overcome the limitations of conventional nozzles. In particular, the ML
design yield the potential of variable spray interaction, so that mixture formation
could be controlled according to the operating condition. In general, it was also a
primary aim of the cluster-nozzle concepts to combine the enhanced atomization and
pre-mixing of small nozzle holes with the longer spray penetration lengths of large
holes. The applied diagnostic, which was based on 1-d spontaneous Raman scattering,
yields the quantitative stoichiometric ratio and the temperature in the vapor phase.
The measurements are conducted in non-reacting sprays. It was established that the
stoichiometric ratio in the region of flame lift-off significantly affects the soot
formation in diesel sprays. The measurements are conducted in a high-temperature,
high-pressure vessel. N-decane is used as the fuel, because it was a commonly applied
model fuel for diesel. The investigated diesel-like sprays emanate from a state-of-the-
art piezo injector.
51
The results of two ML nozzles with two circles of orifices were compared.
The plane of the two holes in each cluster was parallel to the injector axis. The
clustered nozzle holes are convergent (−4°) for one of the ML nozzles, whereas they
are divergent (+4°) for the other one. Two conventional nozzles with one circle of
orifices are also investigated, one with the same flow number as the ML nozzles and
the other one with halved flow number, corresponding to a single hole of the ML
nozzles. The temporal and spatial evolution of the quantitative stoichiometric ratio
and temperature was determined and discussed. Further, the shot-to-shot variability in
these quantities are analyzed. These measurements showed that the ensemble-
averaged fuel-air-ratio distributions are very similar for both ML nozzles and the
reference nozzle with large orifices, but they are significantly different for the
reference nozzle with small orifices. The shot-to-shot variability in the fuel-air ratio
are generally very similar for the ML nozzles as compared to a previously
investigated cluster nozzle with only one orifice circle, indicating that the particularly
complex in-nozzle flow does not lead to enhanced fluctuations of the outcome of the
mixture formation process. The results also lead to conclusions on soot formation in
comparable combusting sprays emanating from ML nozzles. Apparently, the soot-
reduction potential cannot be improved by enhancing evaporation and penetration of
the free spray simultaneously using an ML nozzle. Thus, previously observed reduced
engine-out soot emissions for ML nozzles could be explained by wall impingement or
differences in flame lift-off.
Bianchi et al., (2000) [95] have designed with a goal to improve the air-fuel
mixing and therefore they were characterized by the adoption of high-swirl ports and
re-entrant bowls. Experiments have shown that the high injection velocities induced
by common rail (CR) systems determine an enhancement of the air fuel mixing. CR
52
systems causes a strong wall impingement. The modification was aiming at exploiting
a new configuration of the combustion chamber more suited to CR injection systems
and characterized by low-swirl ports and larger bowl diameter in order to reduce the
wall impingement. The goal was to achieve a higher air flow rate during induction as
well as to reduce the fuel vapor wall impingement without compromising air-fuel
mixing efficiency. This new combustion chamber configuration has been tested
numerically and its performances have been compared to those of a HSDI four valve
diesel engine conventional combustion chamber. The analysis has been carried out by
using a customized version of the CFD code KIVA3. Experimental results of the
conventional engine have been used to validate the numerical models. The influence
of the injection system configuration (i.e, hole numbers, inclination of the spray axis
with respect cylinder head) on pressure cycle and NO x and soot engine-out emissions
has been analyzed too. Computational results seem to indicate that the new
combustion system concept may provide relevant benefits with respect to engine-out
emissions without reducing engine performance.
Gao et al., (2007) [96] have experimentally investigated the spray and
mixture properties of group-hole nozzle with close, parallel or a small included angle
orifices using the ultraviolet-visible laser absorption-scattering imaging technique, in
comparison with the conventional single-hole nozzle. The group-hole nozzle concept
is regarded as a promising approach to facilitate better fuel atomization and
evaporation for direct injection diesel engine applications. Three series of group-hole
nozzles were designed to investigate the effect of group-hole nozzle specification
while varying the included angle and interval between the orifices.
The results suggested that: 1) Group-hole nozzle with very close, parallel
orifices presents the similar spray characteristics with those of the single-hole nozzle.
53
However, with the increase of injection pressure, the group-hole nozzle shows the
potential to produce, to some extent, the better fuel atomization and evaporation
without adversely affecting the spray spatial distributions. 2) Increasing the diverging
angle between the orifices results in significant increase in mass of fuel vapor and
reduction in overall sauter mean diameter (SMD), indicating the evaporation
improvement. 3) In the case of group-hole nozzle with large interval between orifices
in this study, the restrained tip penetration, significant increase in the SMD, along
with large liquid phase mass appearance around the spray tip can be found, which is
considered to be a comprehensive result of the direct droplets collision in the
overlapping part of the two jets and the droplets overtaken and coalescence inside the
fuel spray due to the inferior spray tip penetration and ambient gas entrainment. 4)
The liquid mass proportion and the SMD of the fuel spray from all test group-hole
nozzles are greater than the estimated value by the LAS data of single-hole nozzle
whose orifice diameter is the same as one of the group orifices. Thus, the interaction
between the two jets has the effect of suppression of the ambient gas entrainment and
fuel evaporation, though it has the effect of enhancement of the spray tip penetration.
Tamaki et al., (2010) [97] have performed this study with an aim of improving
atomization characteristics and to obtain excellent spray characteristics with shorter
breakup length, larger spray angle and smaller droplet diameter. A high-efficiency
atomization enhancement nozzle, with large spray angle, short liquid core length and
small droplet diameter was obtained. Investigation was carried out on the atomization
of spray of the multi-hole atomization enhancement nozzle. The effects of dimensions
of the atomization enhancement nozzle such as hole number, the hole diameter,
position of the nozzle hole on atomization of the spray and atomization characteristics
were investigated. As a result, it was clarified that in case of the multi-hole
54
atomization enhancement nozzle with hole number of N=4, breakup length becomes
short about 70 p.c. and spray angle becomes large about two times, droplets of the
spray become considerably small compared with the single hole atomization
enhancement nozzle. Atomization characteristics were improved considerably and
uniform spray mass flux distributions are obtained by using the multi-hole
atomization enhancement nozzle with hole number of N=4.
Xuelong et al., (2011) [98] have proposed the study on a new low-temperature
premixed combustion mode to achieve the simultaneous reduction of NOx and soot
emissions in a volume production diesel engine of CA6DF by reconstructing key
systems. Some developments of this diesel engine are as follows. A straight port and
large diameter combustion chamber of a low compression ratio was developed. Inlet
ports of a high induction swirl ratio were developed. A cooled EGR was developed.
Especially, an ultra-multi hole (UMH) nozzle was developed. It has two layers of
injection holes and a large flow area. Two sprays of the upper and under layers meet
in the space of the combustion chamber. The results showed that the operation range
of this diesel engine to achieve the better low-temperature premixed combustion is as
follows. The speed can cover from the idle speed to the rated speed. The load can
reach to 50% of the full load of the corresponding external characteristics speed. The
NOx and soot emissions of this operation range are simultaneously largely reduced,
even by 80%–90% at most test cases, while keeping the brake-specific fuel
consumption (BSFC) from being significantly deteriorated.
Miao et al., (2009) [99] have stated that the lean premixed compression
ignition combustion provides the potential for simultaneous reduction of NO x and
PM, while imposing moderate penalties on CO and HC emissions. These drawbacks
were overcome in the existing premixed combustion modes of diesel engines using a
55
developed ultra multi hole nozzle. The UMH nozzle has two layers of injection holes
and a large flow area. Two sprays of the upper and lower layers meet in the space of
the combustion chamber. A high-pressure common-rail fuel injection system was
used in this experiment. The fuel injection rate of the UMH nozzle was measured
using the constant volume method, and its spray pattern was recorded using high-
speed digital photography. Combustion and performance experiments with the UMH
nozzle were conducted on a turbocharged intercooled diesel engine. The results
showed that the UMH nozzle exhibited a higher injection rate, shorter injection
duration, shorter spray penetration, and bigger spray angle than those of the
conventional nozzle. These characteristics facilitate better mixing of fuel and air prior
to ignition, and thus NO x and PM emissions were simultaneously reduced with low
CO and HC emissions by combining the UMH nozzle with EGR.
Dahlander et al., (2008) [100] have investigated the spray formation and spray
induced air movement associated with rotational symmetrical and asymmetrical
nozzle hole configurations. Four different nozzles with different hole configurations
and umbrella angle were investigated both experimentally and numerically in a
heated/ pressurized spray chamber. Their influence on spray formation, spray induced
air motion, cross-flow velocity, fuel/air ratio, turbulence and cycle-to-cycle variations
were studied. It was found that rotational symmetrical configurations produce non-
coherent isolated clouds of fuel. It was also found that an asymmetrical configuration
produces a coherent fuel cloud. All the nozzles tested produced partially premixed
vapor clouds, with cycle-to-cycle variations. Stated that these variations may be an
important issue for ignition stability in a closed-spaced combustion system.
Kim et al., (2009) [101] have investigated the spray characteristics of multi-
hole injectors with two different nozzle holes configurations for a gasoline direct
56
injection (GDI) engine according to the various injection conditions. Two GDI multi-
hole injectors with a different location and injection angle of six symmetric holes
located around the nozzle axis were used for this investigation on the spray
characteristics injected six individual plumes. The spray behaviors such as the spray
development process, the spray tip penetration from the nozzle tip, and spray cone
angle were analyzed from the spray images obtained by using the high speed camera.
Also, the local sauter mean diameter (SMD) according to the axial distance from the
nozzle tip and the overall SMD were measured by the droplet measuring system for
the comparison between atomization performances of two injectors. It was found that
the spray tip penetration and cone angle increase as the injection pressure increases. In
the comparison of results between two injectors, test injector with symmetric holes
positioned on the periphery of a larger imaginary circle than the other injector, shows
small values in the spray tip penetration and cone angle. In the atomization
characteristics, the local SMD makes difference according to the axial distance. The
effect of injection pressure on atomization of both sprays at the test injectors showed
that as the injection pressure increased, local and overall SMD show the decreasing
patterns.
George et al., (1998) [102] have adopted RNG model to perform computations
in realistic engine geometries. The parameters explored include the effects of piston
crown shape, injector targeting, glow-plug presence, injection velocity, injection
timing, number of injector holes and initial swirl ratio on mixing. It was concluded
that the mexican-hat design for piston enhances mixing due to the tumbling motion of
the mixture caused by that particular piston shape. The optimum injector hole angle
was found to be 450 with respect to the horizontal.
57
2.6 Literature on Injection Timing
Ergenc et al., (2012) [103] have stated that the quality of combustion majorly
depends on the formation of air-fuel mixture. They also stated that the mixture
formation primarily depends on various injection parameters. They have investigated
on the optimum injection advance angle for DI diesel engine. A single cylinder PLC
controlled DI diesel engine was used for this purpose. Diesel and diesel-ester blends
were used in their study. It was reported that at a particular injection advance peak
pressure occurred in less time, for ester blends.
Zhu Jianjun (2010) [104] has made an experimental analysis on a diesel
engine. Experiments were carried out for various conditions of fuel supply viz., angle
advancement and injection pressures. He has reported that NOx levels were decreased
in both the cases. He also stated that, there was noticed a small effect on Particulates.
Whereas there was no considerable influence on either CO or HC was observed.
Nwafor et al., (2008) [105] have made an experimental investigation on a
diesel engine. They have made an attempt to study the combustion behaviour by
advancing the injection. The tests were carried out on the same fashion for the fuels
viz., diesel and vegetable oil. The performance of the engine is reported to be smooth
at lower loads. They have also noticed that there was some reduction in the ignition
delay. Finally they could recommend the injection advancement for both the fuels.
Cenk Sayin et al., (2009) [106] have carried out experimental investigations
on a diesel engine using diesel fuel blended with ethanol. The experiments were
conducted at different injection timings, two of them fall into under advanced
injection timings and the other two fall into retarded injection timings. For all the
cases, ethanol blend percentage in the diesel fuel was up to 15%. From their study
they could conclude the following:(i)for the retarded injection timings: NOx, CO2 and
58
BSFC are increasing and BTE, CO & HC are decreasing, (ii) for the advanced
injection timings: NOx, CO2 have increased and CO, HC decreased, (iii) with regard
to the BSFC and BTE, better results were reported at the original injection timing.
Nwafor (1999) [107] has made experimental studies on a diesel engine using
natural gas. In this study, the injection timing was advanced and its effects were
observed. Because of this there was noticed a change in the ignition delay. They
proposed an appropriate injection timing at which the engine can run smoothly. They
also reported an increase in the fuel consumption at higher loads.
Kouremenos, et al., (2001) [108] have made both the theoretical and
experimental investigations on a DI diesel engine. Injection advancement was
considered as a major issue to analyse its impact on combustion and emissions. In
addition to that, EGR technique was implemented to control the levels of emissions.
From the study, a reasonable increase in peak pressure with the acceptable levels of
NOx was attained.
Jesus Benajes et al., (2001) [109] have studied on the changes that take place
in combustion and emissions while adopting pre and post injections. Single cylinder
diesel engine was considered for the study. In this process the quantity of fuel being
injected per cycle was also varied to understand its impact on combustion and levels
of emissions esp., NOx and Soot. Following are the major conclusions from their
study: (i) pre-injecting strategy - reduction in the fuel consumption with a penalty of
NOx and soot levels and (ii) post-injection strategy – reduction in the Soot levels with
a penalty of fuel consumption.
Hassan et al., (2011) [110] have experimentally investigated the effects of
advanced injection on an engine. A super charged dual fuel diesel engine was used for
this study. The main focus was on the engine performance and emissions. Injection
59
pressure was maintained to be same for all the cases under consideration. But the
injection flow rates of producer gas as well as the air were varied. The test was carried
out for different loading conditions against various speeds. They have finally
reported that, a considerable reduction in the levels of CO emission, Specific energy
consumption and increased Brake thermal efficiency were noticed.
Hassan et al., (2011) [111] have experimentally investigated on the engine
performance and emissions. They have tried to use the producer gas-diesel and
vegetable oil blends. Air and generated producer gas were injected into the intake
manifold. Injection timing of the pilot fuel was advanced. Experiments were
conducted for the following parameters viz., quantity of fuel, speed and load. It was
concluded that the performance of the engine with producer gas-diesel and vegetable
oil blends have come down. At rated loads the emissions are found to be increasing.
They have recommended the usage of vegetable oil as an alternative fuel for a
supercharged dual-fuel producer gas engine.
Zhixia et al., (2012) [112] carried out theoretical investigation on diesel
engine. They tried to vary the following parameters to study their influence on noise
and emissions: (i) pilot injection timing, (ii) quantity of fuel being injected and (iii)
the timing of main injection.
Yusheng Zhang et al., (2008) [113] have investigated the usage of DME as an
alternate fuel for a DI diesel engine. The following parameters were considered: (i)
engine load, (ii) fuel injection advancement, (iii) nozzle pressure, (iv) crank angle of
EVO and (v) DME. The influence of these parameters on the engine performance and
emissions were studied. DME fuelled engine was proved to be superior at an
appropriate nozzle pressure and fuel injection angle. By retarding the opening of
exhaust it was stated that NOx emissions have decreased.
60
Gorji et al., (2009) [114] have numerically investigated the combustion and
emission characteristics of a diesel engine. Their study majorly deals with the
injection strategies and intake conditions. Following are the parameters that were
taken into account: (i) spray cone angle, (ii) injection timing, (iii) intake temperatures
and (iv) multiple injections. They have noticed considerable changes in the Soot and
NOx formations among the considered cases. Following conclusions were reported:
(i) an increase in the injection angle increased the NO formation but decreased the
Soot levels, (ii) an increase in the cone angle increased the NO formation but no
change in the levels of Soot, (iii) as the injection timing was advanced resulted an
increase in the in-cylinder temperature, pressure and NO formation and reduced levels
of soot, (iv) an increase in the intake temperature resulted a rise in the NO formation
and (v) multiple injection reduced the formation of NO and Soot.
Usman Asad et al., (2009) [115] have carried out experimental investigations
for which a CR 4 cylinder diesel engine was used. The study was on the effects of fuel
injection strategies on the performance and emissions of a CR diesel engine. They
have concluded that, reduction in the levels of NOx and Soot are attainable. This
might lead to an increase in the levels of HC and CO.
Gunabalan et al., (2009) [116] have made attempts to investigate the effect of
early injection and EGR on engine performance and emission. Star-CD code was used
for this purpose. Following are the major conclusions: (i) with early injection and
without EGR – the occurrence of peak pressure has advanced and its value increased
and NOx levels have increased and Soot levels are decreased, (ii) with early
injection and with EGR – reduced the peak pressure, NOx and soot.
Sanghoon et al., (2008) [117] have made experimental investigations on a CI
constant volume vessel. Conditions that were maintained are more or less similar to
61
HCCI. Their attempt was mainly to identify the effect of fuel injection time on the
following: (i) air-fuel mixing, (ii) combustion and (iii) emissions. Electronic fuel
injection system was used to regulate the injection timings. They have noticed some
increase in the ignition delay. The contribution of this in the reduction of NOx is not
that considerable against the considered compression ratios and the engine speed. HC
and CO emissions were decreased noticeably. Whereas NOx has decreased. They
have also concluded that, to enhance the air-fuel mixture, the diesel spray should be
made to impinge near the bowl-lip area.
Suryawanshi et al., (2005) [118] have carried out experimental investigations
on a DI diesel engine. They have tried with an alternate fuel having the blends of
pongamia methyl ester with diesel. The tests were conducted at the standard and
retarded injection timings as well. Reduced levels of CO, HC and smoke was reported
with pongamia. Whereas very small rise in the levels of NOx was reported for normal
injection time. They reported similar behaviour with regard to brake thermal
efficiency and exhaust temperature.
Zhu et al., (2003) [119] have made a theoretical investigation on a CRDI.
They tried to find the extent by which injection timing, EGR etc., would influence the
engine performance and emission. KIVA-3V code was used for their study and
reported the possibility of reducing NOx and soot.
62
Kamal Kishore et al., (2010) [120] have investigated experimentally on the
possibility of using karanj oil blended with petro-diesel in a diesel engine. An
additional set up was attached to the engine using which viscosity of the blend can be
reduced. Finding the optimum injection timing was their major concern. Engine
performance was taken into account while optimizing the injection time.
2.7 Summary
In the literature survey, contributions of eminent researchers in the field of CI
engines have been reviewed. The review of literature revealed the progress made in
the field of modelling of in-cylinder flows of diesel engines. It is clear that the
research in this direction has started early in 1980’s. The progress in the 3D modelling
is witnessed form the 1990’s. Accurate prediction of engine processes such as in-
cylinder flows, spray, combustion and pollutant formation is also witnessed in the
recent period. From the literature survey it is observed that the initial air swirl,
number of nozzle holes and the fuel injection time are controlling the quality of
combustion.
From the literature it is found that the engine performance depends on many
interdependent parameters. They are engine efficiency, fuel economy, injection
pressure, operating pressure, operating temperature, engine design, engine geometry,
combustion process, compression ratio, nozzle geometry, injection orientation,
injector location, injector geometry, injection strategies, geometry of the ports, bowl
geometry, distribution of air fuel mixture within the chamber. In order to improve the
engine performance, it is necessary to optimize all the above interdependent
parameters.
From the literature it is found that researchers have attempted with many
combinations of the above stated parameters but not many with the combination of
63
swirl, number of injector holes and the injection time. Influences of these variables on
the performance of an engine are reported below:
• Swirl in diesel engines is an important parameter that affects the mixing of the
air fuel ratio and the quality of combustion, in turn heat release, emissions and
overall engine performance.
• There is an optimal level of swirl ratio for a particular combustion chamber.
Swirl level can be varied with appropriate design changes of the intake
system.
• Number of nozzle holes of a fuel injector in diesel engines is another
important parameter that affects the mixing of the air and fuel, in turn the
quality of combustion.
• There is an optimal number of nozzle holes for a particular combustion
chamber geometry.
• Fuel injection time in diesel engines is yet another important parameter that
affects the mixing of the air and fuel in turn the quality of combustion.
• There is an optimal injection time for particular combustion chamber
geometry.
Hence the present work aims at studying the influence of Swirl, Number of
nozzle holes of fuel injector and injection time in improving the quality of
combustion.
2.8 Scope and Objectives of the Present Work
From the literature survey the prominence of Swirl, Number of nozzle holes of
fuel injector and injection time in improving the combustion is observed. Scope for
optimising the three parameters viz., swirl, number of nozzle holes and injection time
is also observed. It is also observed that though attempts are made in optimising the
64
performance of IC engine with several combinations of various parameters no major
attempts are made with the combination of the said three parameters. In the present
work, an attempt is made with an objective of optimizing the swirl, number of nozzle
holes of fuel injector and the injection time. A numerical study was carried out on a
DI diesel engine to study the effect of these parameters on spray and combustion.
Recardo-Vectis, a 3 dimensional computational fluid dynamics code has been
used for this purpose. Chapter 3 covers the methodology and details of the models
that are used in the present work. This objective was fulfilled by following these
steps, (1) Identifying the optimum swirl ratio for the engine. (2) Identifying the
optimum number of nozzle holes against swirl ratio. (3) Identifying the optimum start
of injection for the optimum one that was identified from step (2).