Soot Formation in Diesel Engines By Using Cfd

R. Siva Kumar et al. Int. Journal of Engineering Research and Application www.ijera.com
ISSN : 2248-9622, Vol. 6, Issue 6, ( Part -3) June 2016, pp.17-22
www.ijera.com 17 | P a g e
Soot Formation in Diesel Engines By Using Cfd
R. Siva Kumar1
, Dr. K. Tirupathi Reddy2
, P. Madhu Raghava3
, P.Nagaraju4
*( Mechanical Engineering, Santhiram Engineering College , Nandyal-518501, India.)
**( School Of Mechanical Engineering, Rgmcet , Nandyal-518501, India.)
*** ( Mechanical Engineering, Santhiram Engineering College , Nandyal-518501, India.)
**** ( Mechanical Engineering, Santhiram Engineering College , Nandyal-518501, India.)
ABSTRACT
In order to meet the stringent emission standards significant efforts have been imparted to the research and
development of cleaner IC engines. Diesel combustion and the formation of pollutants are directly influenced by
spatial and temporal distribution of the fuel injected. The development and validation of computational fluid
dynamics (CFD) models for diesel engine combustion and emissions is described. The complexity of diesel
combustion requires simulations with many complex interacting sub models in order to have a success in
improving the performance and to reduce the emissions. In the present work an attempt has been made to
develop a multidimensional axe-symmetric model for CI engine combustion and emissions. Later simulations
have been carried out. Commercial validation tool FLUENT was used for simulation. The tool solves basic
governing equations of fluid flow that is continuity, momentum, species transport and energy equation. Using
finite volume method turbulence was modeled by using RNG K-ɛ model. Injection was modeled using La
Grangian approach and reaction was modeled using non premixed combustion which considers the effects of
turbulence and detailed chemical mechanism into account to model the reaction rates. The specific heats were
approximated using piecewise polynomials. Subsequently the simulated results have been validated with the
existing experimental values.
Keywords: IC engines, CFD, FLUENT, RNG K-ɛ model, La Grangian approach.
I. INTRODUCTION
The diesel engine, because of its highest
thermal efficiency among currently available
engines, has been a main power source for over a
hundred years. It’s advantage in thermal
efficiency is due to its combustion characteristics.
Diesel fuel is injected with a high pressure
environment (compression ratio as high as 24:1).
The higher the compression ratio, the more
efficient the cycle [Hey Wood, 1998].
However, for some of the reasons that the
diesel engine is highly efficient, its power density
and exhaust emissions have traditionally been less
desirable than spark ignition (SI) combustion
engines. In a diesel engine, the combustion rate is
controlled by the fuel injection rate and mixing and
diffusion rates, which are usually slower than the
premixed combustion rate in typical gasoline
engines. Diesel engines usually emit more
particulate and NOx than their gasoline court parts
the high temperature and pressure environment in
the diesel engine cylinder because of its high
compression ratio and high combustion
temperatures makes it impossible to completely
prevent NOx form forming. The soot formed in the
fuel – rich regions, although partially oxidized in
the expansion stroke, also remains in considerable
amounts at exhaust valve opening (EVO). The
problem of diesel engine emissions is exacerbated
because of the trade - off feature between NOx and
soot emissions. It is usually impossible to reduce
both kinds of emissions simultaneously, since
factors that tend to decrease one usually increase
the other. For example, retarding the fuel injection
timing is effective to reduce NOx formation by
reducing the peak cylinder temperature and
pressure. However, this method results in an
increase of soot production because more soot
formed due to the lower in - cylinder gas
temperature has shorter time to be oxidized [Lee,
2002]. Increasing the EGR rate can decrease the
NOx emission level, however less oxygen is
available to oxidize soot. Eventually, any change
in these engine parameters will unavoidably affect
other important engine performance measures, like
retarding injection timing causes lower thermal
efficiency and higher Brake Specific fuel
consumption (BSFC) [Hey Wood, 1998].
Increasing environmental concerns and
legislated emission standards have led to the
necessity of considering both conventional and
unconventional means for reducing soot and NOx
emissions in diesel engines which is also the
motivation of the present study. For example,
diesel engine manufacturers are facing the
challenges of the extremely low diesel engine – out
soot emission mandates to be implemented in the
near future. Engine simulation, compared to
RESEARCH ARTICLE OPEN ACCESS

expensive engine experiments, is an efficient way
to investigate various novel ideas to improve
current engine performance, and hence becomes an
essential part of engine research and development.
In addition, simulations can investigate the
transient properties of physical process. e.g., diesel
sprays combustion and emission formations, whose
characteristics are difficult or impossible to obtain
with the current measurement technologies.
II. OBJECTIVE AND APPROACH
The objective of this study was to improve
diesel engine performance using computational
optimization methods. Multiple injection strategies
(split injection), with up to 2, 3 and 4 pulses / cycle
and 1500 bar to 2500 bar injection pressure,
different cone angles, different SOI (start of
injection), different DOI (duration of injection)
were investigated to minimize engine – out NOx,
soot, HC and CO emissions, along with BSFC
using the computational models in FLUENT 6.3,
and to compare the results with the existing
experimental results. The purpose of this
optimization was to gain a detailed understanding
of the mechanisms through winch fuel injection
interacts with other engine parameters and
influences diesel combustion and emissions, and
hence to attempt to generalize the adoption of
multiple injection strategies with regard to
improving HSDI diesel engine performance.
Simulations were performed to understand the
influence on various parameters on engine
performance.
III. METHODOLOGY
The methodology employed in this project was
designed to accomplish the objectives described.
This is shown pictorially in Fig 3.1 and the purpose
behind each of these sections described here in
detail.
IV. SOURCES OF SOOT FORMATION
Fig.4.1 Power Generation
Fig.4.2 Vehicle Exhaust
V. SOOT FORMATION PROCESS
Fig.5.1 Stage-1

Fig.5.2 Stage-2
VI. COMPUTATIONAL FLUID
DYNAMICS
CFD is a sophisticated analysis technique
that the analyst to predict transfer of heat, chemical
reaction, and fluid flow behavior etc. CFD is based
on the fundamental governing equations of fluid
dynamics- the continuity, momentum, and energy
equation. It is a powerful tool to carry out
numerical experiments. This project uses the
Computational Fluid Dynamics –FLUENT
software package.
The process of utilizing FLUENT can be
assumed. Firstly, the geometry and gird is created
using GAMBIT. T Grid can be used to generate 2D
triangular, 3D tetrahedral or 2D and 3D hybrid
volumes mesh from an existing boundary mesh.
Another alternative of creating grids for FLUENT
is using ANSYS or IDEAS (SDRC); or
MSC/ARIES, MSC/PATRAN. Pre BFC and
GeoMesh are the names of FLUENT Pre-
processors that were used before the introduction of
GAMBIT.
Once a grid has been read into FLUENT,
all remaining operations are performed with in the
solver. These include setting the boundary
conditions, defining fluid properties, and material
properties, executing the solution, refining the grid,
viewing and post- processing the results.
VII. MESH GENERATON
The local flow characteristics of high speed
liquids are not necessarily parallel to the flow of
direction. Therefore, it is pertinent to use
unstructured triangular grids. The commercial
mesh generation package GAMBIT is used
.GAMBIT uses the Delaunay triangulation
automatic mesh generation technique, with an
advancing front method. All the meshes are axi-
symmetric in two and three –dimensional. The
advantages of Delaunay triangulation are that no
point is contained in the circum- circle of any
triangle and it maximizes the minimum angle for
all triangular elements, which is a requirement for
good quality finite elements, i.e. the mesh has a
low skewness.
The mesh is generated by applying points
along the boundaries of geometry with certain
stretching function. The geometry made of edges
and faces. It was found that geometries, which
had interiors that were made of multiple faces,
performed better in calculation. This was due to the
increase in versatility of the stretching function
that minimized a large cell volume jumps within
the domain.
The mesh can be modified as a function of
the solution. Post solution adaption is undertaken in
FLUENT. First, the individual cell s is marked for
refinement based on an adaption function, which is
created from the solution data. Next, the cell is
refined based on these adaption marks. The mesh is
adapted by the hanging node technique, whereby
the cells marked for refinement are isotropically
subdivided. Fig 3.2 shows an example of this
technique.
VIII. GOVERNING EQUATION
In the CFD methodology, fluid flows are
stimulated by numerical solving partial differential
equations that governs the transport of flow
quantities also known as flow variables. These
variables include mass, momentum, energy,
turbulent quantities, and species concentrations. In
designing the POME- nozzle, the basic
governing equations that will be used are the
conservation of mass, momentum and energy
equations.
7.1 Continuity Equation
In any fluid flow, the equation for
conservation of mass, or continuity equation, can
be written as follows:
.
( )ju m
t xj


 
 
 
Where ju is the
th
j Cartesian of the
instantaneous velocity, ρ is the density and
.
m
represents the rate of mass generation in the
system. This equation is valid for incompressible as
well as compressible flows.
7.2 Momentum Equation
The law of consecration of momentum
states that

( )j ij
i j i
i i
u
u u f
t xj x x
 
 
  
    
   
Where ρ is the static pressure, ij is the viscous
stress tensor and _ is the body force.
For Newtonian fluids:
2
3
ji k
ij ij
j i k
uu u
x x x
   
    
          
Where  is the fluid dynamic viscosity and ij is
the Kronecker deltas.
Substitute equation 3.4 into 3.3results in the Navier
– Strokes Equation
   
2
3
ji k
i i i ij i
j j j j k
uu
u u j f
t x x x x x

     
      
     
       
7.3 Energy Equation
The energy equation governs the enthalpy
balance. For CFD, the static enthalpy is used
Conservation of energy is described by
 
( )
( ) ( )
ij ii i
i j ij a
j j j j j
J hq u
h u h u s
x x x x x

   

   
       
     
Where ijJ is the total diffusive mass flux for
species ai, ih is the enthalpy for species ai, aS
stands for additional sources due to surface
reaction, radiation and liquid spray and jq is
the j component of the heat flux. Fourier_ Law is
employed to model the flux
k
j
T
q K
x
 
     
Where K is the thermal conductivity. This is only
true for incompressible at low Mach number. The
total enthalpy of the energy equation is fully
conservative and is recommended for high –speed
compressible flows. The total enthalpy H is defined
as
2
j ju u
H h 
The governing equation for H can be obtained by
adding the fluid kinematics energy equation to the
static enthalpy equation:
 
( )
( ) ( )
ij i ij i
i a i i
j j j j j j
u J hT
H u H K s f u
t x x x x x x

  
     
      
      
VIII. SIMULATIONS FOR MODEL
VALIDATION
This chapter summarizes the simulation
results performed by other researchers on various
diesel engines, from which the date were drawn for
validating the CFD models developed and used in
this study. Section 4.1 describes the simulation on
the High-Speed Direct – Injection (HSDI) diesel
engines.
The engine is a fully-instrumented single
cylinder version of the caterpillar 3406 heavy duty
truck engine. Simulations were carried at high load
where EGR has been shown to produce
unacceptable increase in particulate emissions. The
fuel system was an electronically controlled,
common rail injector and supporting hardware.
The fuel system was capable to handle four
independent injections per cycle. The spray nozzle
angles used with included angle 1250 and 1400.
8.1 SIMULATION SET-UP
The test engine is a fully instrumented
single cylinder version of a caterpillar 3406 heavy
duty truck engine. The engine is capable of
producing 54 KW at a rated spread of 2100 RPM.
The fuel was an electronically controlled common
rail injector and injection pressure unto 1200 MPa
was possible so as to ensure fuel system reliability.
The injection system is capable of single, double,
triple and quadruple injection schemes. Timing
and duration could be varied independently for
each injection pulse providing a highly flexible
injection system. A 0.6 mm diameter orifice was
used to produce very fast injection rates 2 crank
angles to maximum needle lift at 1600 RPM.
Fig.8.1 Computational Grid for
Simulation

Fig.8.2 Dynamic Mesh of the Grid
Fig.8.3 Pressure Contours-1
Fig.8.4 Pressure Contours-2
Fig.8.5 Temperature Contours-1
Fig.8.6 Temperature Contours-2
IX. GRAPHS
9.1.Validation For Cycle Pressure
Fig.9.1 Graph: P-θ Curve for single injection

9.2. Validation For Nox
Fig.9.2 Graph: NOX Curve for Single
Injection
9.3. Validation For Soot
Fig.9.3 Graph: Soot curve for single Injection
9.4. Validation For Heat Release Rate
Fig.9.4 Graph: Heat release rate curve for single
Injection
X. CONCLUSION
The present investigation is aimed at
studying the simulation for single, using CFD code,
in a direct injection compression ignition engine
and results are compared with the existing
experimental results. The simulation was carries
for single injection for which the comparison with
experimental results were not made due to non
availability of existing experimental results. In
simulation process 2000 particles are injected
during combustion and it was observed that 1890to
1980 particles are only participating in the
combustion.
REFERENCES
[1]. Shimazaki, N., Hatanaka H., Yokota, K., and
Nakahira, T.,“A Study of Diesel Combustion
Process Under the Condition of EGR and
High Pressure Fuel Injection with Gas
Sampling Method”, SAE Paper 960030,
1996.
[2]. Sundoh, S., Kakekawa, T., and Tsujimurwa,
K., “The Effect of Injection Parameters and
Swirl on Diesel Combustion with High
Pressure Fuel Injection”, SAE Paper
910489, 1991.
[3]. Epsey, C. Pinson, J. A., and Litzinger, T.A.,
“Swirl Effects on Mixing and Flame
Evolution in a Research DI Diesel Engine”,
SAE Paper 902076, 1990:1-10
[4]. Ogawa, H., Matsui, Y, Kimura, S., and
Kawashima,J.,“Three- Dimensional
Computation of the Effects of the Swirl
Ratio in Direct- Injection Diesel Engines on
NOX and Soot Emissions”, SAE Paper
961125, 1996.
[5]. Zhang, L., Takatski, T., and Yokota, k., “An
Observation and Analysis of the Combustion
under Supercharging on a DI Diesel
Engine”, SAE Paper 940844, 1994.
[6]. Jafer, D., “Pilot Injection”, Engineering.
October 15, 1937.

Soot Formation in Diesel Engines By Using Cfd

More Related Content

What's hot (19)

Viewers also liked (19)

Similar to Soot Formation in Diesel Engines By Using Cfd (20)

Recently uploaded (20)

Soot Formation in Diesel Engines By Using Cfd